The Methodology of Constitutional Theory 9781509933846, 9781509933877, 9781509933860

Citation preview

THE METHODOLOGY OF CONSTITUTIONAL THEORY What sort of methods are best suited to understanding constitutional doctrines and practices? Should we look to lawyers and legal methods alone, or should we draw upon other disciplines such as history, sociology, political theory and moral philosophy? Should we study constitutions in isolation or in a comparative context? To what extent must constitutional methods be sensitive to empirical data about the functioning of legal practice? Can ideal theory aid our understanding of real constitutions? This volume brings together constitutional experts from around the world to address these types of questions through topical events and challenges such as Brexit, administrative law reforms, and the increasing polarisations in law, politics and constitutional scholarship. Importantly, it investigates the ways in which we can ensure that constitutional scholars do not talk past each other despite their persistent – and often fierce – disagreements. In so doing, it aims systematically to re-examine the methodology of constitutional theory. Hart Studies in Constitutional Theory: Volume 2

Hart Studies in Constitutional Theory Series Editors Charles Barzun, University of Virginia, USA Maartje De Visser, Singapore Management University Matthias Klatt, University of Graz, Austria The Hart Studies in Constitutional Theory series publishes thought-provoking works of scholarship addressing diverse aspects of constitutional theory in a concise and crystalline manner. Authors writing for this series cover a wide range of perspectives, methods, and regions, to enhance our understanding of constitutions as central institutions of modern public life. Taken together, the books in this series aim to challenge established wisdom and advance original ideas. This series is a natural home for books interrogating the concepts and structures of constitutions on the national, the supranational and the international level. Its guiding philosophy is that the task of constitutional theory is not only to d ­ elineate the basic structures of government and to protect human rights, but also more broadly to offer methods for grappling with the social, political, and economic problems societies face today. The series is open to theoretical, normative, analytical, empirical and comparative approaches, stemming from legal studies as well as from political philosophy and political science. In its ambition to become a global forum for debate about constitutional theory, the series editors welcome submissions for monographs as well as edited volumes from all parts of the world. Recent titles in this series: Proportionality and Facts in Constitutional Adjudication by Anne Carter

The Methodology of Constitutional Theory Edited by

Dimitrios Kyritsis and

Stuart Lakin

HART PUBLISHING Bloomsbury Publishing Plc Kemp House, Chawley Park, Cumnor Hill, Oxford, OX2 9PH, UK 1385 Broadway, New York, NY 10018, USA 29 Earlsfort Terrace, Dublin 2, Ireland HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2022 Copyright © The editors and contributors severally 2022 The editors and contributors have asserted their right under the Copyright, Designs and Patents Act 1988 to be identified as Authors of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. While every care has been taken to ensure the accuracy of this work, no responsibility for loss or damage occasioned to any person acting or refraining from action as a result of any statement in it can be accepted by the authors, editors or publishers. All UK Government legislation and other public sector information used in the work is Crown Copyright ©. All House of Lords and House of Commons information used in the work is Parliamentary Copyright ©. This information is reused under the terms of the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/ open-government-licence/version/3) except where otherwise stated. All Eur-lex material used in the work is © European Union, http://eur-lex.europa.eu/, 1998–2022. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication data Names: Methodology of Constitutional Theory (Conference) (2017 : University of Reading)   |  Kyritsis, Dimitrios, 1978- editor.   |  Lakin, Stuart, editor.   |  University of Reading, host institution. Title: The methodology of constitutional theory / edited by Dimitrios Kyritsis and Stuart Lakin. Description: Oxford ; New York : Hart, 2022.   |  Series: Hart studies in constitutional theory ; volume 2   |  “The collection of chapters in this volume arises out of a workshop held at the University of Reading in April 2017, funded by a Modern Law Review Seminar Funding Scheme”— ECIP acknowledgments.   |  Includes bibliographical references and index. Identifiers: LCCN 2021053505 (print)   |  LCCN 2021053506 (ebook)  |  ISBN 9781509933846 (hardback)   |  ISBN 9781509957699 (paperback)   |  ISBN 9781509933860 (pdf)   |  ISBN 9781509933853 (Epub) Subjects: LCSH: Constitutional law—Great Britain—Congresses. Classification: LCC KD3989.A2 M48 2017 (print)   |  LCC KD3989.A2 (ebook)   |  DDC 342.41—dc23/eng/20211108 LC record available at https://lccn.loc.gov/2021053505 LC ebook record available at https://lccn.loc.gov/2021053506 ISBN: HB: 978-1-50993-384-6 ePDF: 978-1-50993-386-0 ePub: 978-1-50993-385-3 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.

We dedicate this volume to our families, Denise and Oliver Kyritsis and Beth and Ned Lakin.

vi

Foreword

T

his book marks an important step in the development of a more sophisticated constitutional theory in the United Kingdom. Legal theory here has been rich, but it has concentrated on the nature of law, or of good law, and legal method in general. The intense attention to constitutional and public law theorising found in countries elsewhere has been relatively scarce. This is no doubt in part due to the absence of a codified constitution which too often is said to indicate that we have no constitution at all, or that it is not worth the paper it is not written on. Or due to the absurd notion that our only constitutional rule is that the Queen in Parliament is sovereign. Dicey’s shadow may have loomed too long, but the attacks on his concept of the rule of law by Jennings and others were so effective as to make us forget that Dicey powerfully held that, whether or not parliament could do anything it wished anywhere, untrammelled legislative or executive power ought to be constrained by legality, and respect civil liberties and freedoms. These he felt essentially distinguished a democracy from what he called ‘despotism’. From the 1960s an innovative judiciary started to read principles of accountability into the exercise of even the broadest discretionary power and then, from the 1990s, to identify constitutional principles inherent in any democracy properly so-called. It is remarkable that the rapid acceleration of this kind (summarised under the ‘grounds’ of judicial review), proceeded so long before the academy started considering how the development could be theoretically justified and, if so, whether by the notion of ultra vires or by the tenets of the common law. Ronald Dworkin described judicial discretion as ‘the hole in the doughnut’. The flexible parts of our unwritten constitution might be similarly described. The chapters in this book show how the opening is now being filled by profound and interesting thought on the theoretical basis of our constitutional development and its methodology. Their range and originality is brilliantly categorised and summarised by the editors, who are to be congratulated on a contribution which is bound to have significant influence on our constitution – codified or not. Jeffrey Jowell October 2021

viii

Series Foreword

C

onducting debates over scholarly methodology is a risky affair. As the intellectual historian JGA Pocock once observed, at their worst such debates devolve into scholars telling one another, ‘you should not be doing your job; you should be doing mine.’1 Sometimes, however, the clash of perspectives that methodological dispute entails can produce surprising discoveries, whether of unexpected areas of agreement or of deeper differences in philosophical outlook. The chapters that comprise The Methodology of Constitutional Theory exemplify this latter kind of debate. They vary across constitutional topics – from questions of political legitimacy to theories of adjudication to understandings of administrative law – and across the methods of inquiry employed and endorsed – from philosophical analysis to comparative study to social theory. But they are united in seeing constitutional theory as a distinctive scholarly endeavour framed around a common set of questions about how fundamental norms, rules and understandings structure political life. Both the common set of questions and the diverse approaches employed to answer them make this collection an ideal early volume in the Hart Studies in Constitutional Theory series. The series is premised on the idea that in spite of the variation the world’s constitutions exhibit, constitutional law is capable of being studied in a general way that transcends national borders. This volume not only attests to that truth, but also demonstrates that meaningful debate about how such study is best conducted can cut across methodological borders as well.



1 JGA

Pocock, Political Thought and History: Essays on Theory and Method (Cambridge, 2009) 51.

x

Acknowledgements

T

he collection of chapters in this volume arises out of a workshop held at the University of Reading in April 2017, funded by a Modern Law Review Seminar Funding Scheme award. We would like to acknowledge the generous support of the Modern Law Review. We are very grateful to all contributors to the collection, many of whom have written chapters from scratch with all the attendant challenges of the pandemic. We would also like to thank all participants in the workshop from which this volume springs for their valuable papers, discussions and panel chairing duties: Professor Trevor Allan, Professor Nick Barber, Professor Peter Cane, Professor David Dyzenhaus, Professor Graham Gee, Professor Jeffrey Goldsworthy, Professor Michael Gordon, Professor David Howarth, Professor Aileen Kavanagh, Professor Jeff King, and Dr Sarah Nason. Special thanks go to Professor Dawn Oliver, Professor Maurice Sunkin and Professor Aleardo Zanghellini, who kindly acted as guest chairs of workshop panels. We are also grateful to all other non-presenting attendees. We would like to thank the University of Reading and the School of Law for their support, financial and otherwise. In particular, we thank Sarah Brown in the Law School for all her excellent administrative assistance, and Claire Fletcher from the University Communication & Engagement Department for her invaluable help with marketing and the production of a professional workshop brochure. Finally, we would like to thank the excellent team at Hart Publishing for their help, professionalism, and patience. We are especially grateful to Sinead Moloney, Kate Whetter and Rosie Mearns for their guidance and encouragement throughout the long gestation of this book.

xii

Table of Contents Foreword������������������������������������������������������������������������������������������������������������ vii Series Foreword��������������������������������������������������������������������������������������������������� ix Acknowledgements���������������������������������������������������������������������������������������������� xi About the Contributors�������������������������������������������������������������������������������������� xix The Methodology of Constitutional Theory – Introduction����������������������������������� 1 Dimitrios Kyritsis and Stuart Lakin I. Background�������������������������������������������������������������������������������������������� 1 II. The Scope of the Volume������������������������������������������������������������������������ 4 III. The Chapters and Thematic Groupings��������������������������������������������������� 4 A. Self-Understandings������������������������������������������������������������������������� 5 B. How Do Facts Matter?�������������������������������������������������������������������� 6 C. Morality������������������������������������������������������������������������������������������ 8 D. Social Theory���������������������������������������������������������������������������������10 E. Comparisons����������������������������������������������������������������������������������12 F. Administrative Law������������������������������������������������������������������������13 SELF-UNDERSTANDINGS 1. The Significance of the Common Understanding in Legal Theory�������������������17 NW Barber I. Adherence to the Common Understanding���������������������������������������������18 A. The Nature of the Common Understanding������������������������������������19 B. The Invocation of the Common Understanding in Legal Theory�������������������������������������������������������������������������������������������21 II. Paying Attention to the Common Understanding������������������������������������25 A. An Argument from Intelligibility�����������������������������������������������������26 B. An Argument from the Constitution of Social Institutions���������������26 C. An Argument from the Wisdom of Crowds�������������������������������������27 III. The Limitations of the Common Understanding������������������������������������29 A. Where the Common Understanding is Inconclusive�������������������������29 B. Where the Common Understanding is Clear������������������������������������32 IV. Conclusion��������������������������������������������������������������������������������������������39 2. In Defence of Traditional Methodologies�������������������������������������������������������41 Jeffrey Goldsworthy I. Introduction������������������������������������������������������������������������������������������41

xiv  Table of Contents II. The Orthodox Understanding���������������������������������������������������������������42 A. Common Law�������������������������������������������������������������������������������42 B. Written Constitutions, Legislative Supremacy and Statutory Interpretation��������������������������������������������������������������������������������43 III. Judicial Pragmatism������������������������������������������������������������������������������48 IV. Confusion between the Common Law and Other Kinds of Law�������������51 V. Constitution-making by Judges�������������������������������������������������������������55 VI. Legal Philosophy����������������������������������������������������������������������������������58 VII. Parliamentary Sovereignty Today�����������������������������������������������������������63 3. Constitutional Methodology and Brexit: Adopting a Model-Theoretic Approach������������������������������������������������������������������������������������������������������71 Alison L Young I. Defining Model-Theoretic Approaches��������������������������������������������������73 II. What is Distinctive about a Model-Theoretic Approach to Constitutional Theory?���������������������������������������������������������������������78 III. Why Adopt a Model-Theoretic Approach to Constitutional Theory?�����81 IV. Brexit and Parliamentary Sovereignty����������������������������������������������������89 V. Conclusion�������������������������������������������������������������������������������������������98 HOW DO FACTS MATTER? 4. Slaying the Misshapen Monster: The Case for Constitutional Heuristics����������������������������������������������������������������������������������������������������103 TT Arvind and Lindsay Stirton I. Introduction���������������������������������������������������������������������������������������103 II. Facts, Theories and Traditions: Making the Constitutional World��������106 A. The Social Epistemology of Constitutional Theory����������������������106 B. Taking Facts Seriously�����������������������������������������������������������������107 C. Continuity and the Ascription of Deontic Power��������������������������108 D. Core and Canon in Constitutional Theory�����������������������������������111 III. A Methodology for Constitutional Theory������������������������������������������114 A. A Taxonomy of Constitutional World-Views��������������������������������114 B. The Constitution as a Shield��������������������������������������������������������116 C. The Constitution as the Cornerstone of Social and Political Life��������������������������������������������������������������������������������117 D. The Constitution as a Rulebook���������������������������������������������������118 E. The Constitution as a Truce���������������������������������������������������������121 IV. Heuristics and the Limits of Rhetoric��������������������������������������������������122 A. Understanding the Impasse����������������������������������������������������������123 B. From Rhetoric to Heuristics: Towards a Resolution����������������������125 V. Conclusion�����������������������������������������������������������������������������������������128

Table of Contents  xv 5. Why Common Law Constitutionalism is Correct (If It Is)�����������������������������131 Stuart Lakin I. Introduction����������������������������������������������������������������������������������������131 II. Two Accounts of the British Constitution���������������������������������������������132 A. Goldsworthy’s Orthodoxy (GO)��������������������������������������������������132 B. Allan’s Common Law Constitutionalism (CLC)���������������������������133 C. Are GO and CLC Commensurable?���������������������������������������������134 D. The Motivating Question������������������������������������������������������������140 III. What Makes GO or CLC Correct?�������������������������������������������������������142 IV. GO and CLC as Rival Interpretations of British Constitutional Practice�������������������������������������������������������������������������147 V. Conclusion������������������������������������������������������������������������������������������152 6. Methodological Pluralism and Modern Administrative Law�������������������������153 Sarah Nason I. Subordinating Administrative Law to Constitutional Law���������������������154 II. New Methods of Administrative Law Theory���������������������������������������159 A. Administrative Law and Socio-Legal Methods�������������������������������159 B. The Social Construction of Law: Facts and Values�������������������������162 C. The Plurality of Administrative Law Values and Methods�������������166 D. Doctrine and Theory��������������������������������������������������������������������168 E. Specialist and Generalist Administrative Law, Functionalism and Content Analysis�������������������������������������������������������������������171 F. Computer Science Methods����������������������������������������������������������173 III. Challenges and Opportunities of Methodological Pluralism in Administrative Law������������������������������������������������������������������������������175 MORALITY 7. The Constitution of Legal Authority������������������������������������������������������������179 David Dyzenhaus

I. II. III. IV.

Hart on the Constitution of Authority��������������������������������������������������181 Approaching Natural Law?������������������������������������������������������������������185 The Legal Man vs. the Legal Subject����������������������������������������������������191 Acceptance, Legitimacy, and the Social Contract����������������������������������203

8. Constitutional Law as Legitimacy-Enhancer������������������������������������������������211 Dimitrios Kyritsis I. Introduction����������������������������������������������������������������������������������������211 II. Moral Force and Settlement�����������������������������������������������������������������213 III. Legitimacy vs. Justice���������������������������������������������������������������������������216 A. Legitimacy as Assurance���������������������������������������������������������������216 B. Assurance and Disagreement��������������������������������������������������������220 C. Constitutionalism of Fear?������������������������������������������������������������222

xvi  Table of Contents IV. Two Moralised Methodologies for Constitutional Theory������������������225 A. Constitutions as the Moral Footprints of Settlement�������������������225 B. Assurance vs Moral Construction�����������������������������������������������228 V. Conclusion����������������������������������������������������������������������������������������231 9. A Positivist and Political Approach to Public Law�����������������������������������������233 Michael Gordon I. Introduction��������������������������������������������������������������������������������������233 II. A Basis for Positivist and Political Public Law�������������������������������������234 A. Loughlin’s Theory of Public Law as Political Jurisprudence���������235 B. Political Jurisprudence and Normative Legal Positivism��������������237 III. The Nature of Positivist and Political Public Law�������������������������������241 A. Defining a Positivist and Political Approach��������������������������������241 B. The Coherence of a Positivist and Political Approach – Some Potential Objections����������������������������������������������������������243 IV. The Value of a Positivist and Political Approach to Public Law�����������248 A. A Positivist and Political Analytical Framework��������������������������248 B. Explanation, Justification and Critique of Constitutional Structures����������������������������������������������������������������������������������249 C. Explanation, Justification and Critique of Constitutional Practice��������������������������������������������������������������������������������������251 D. The Utility of Positivist and Political Public Law�������������������������253 V. Conclusion����������������������������������������������������������������������������������������257 SOCIAL THEORY 10. The Material Study of the Constitutional Order�������������������������������������������261 Marco Goldoni I. The Legal Theory of the Material Study��������������������������������������������261 II. The Political Theory of the Material Study����������������������������������������267 III. Thematising the Constitutional Order as Legal Organisation�������������273 IV. Case Study: Constitutional Change����������������������������������������������������278 V. Conclusion����������������������������������������������������������������������������������������281 11. The British Constitution as an Improvised Order������������������������������������������283 David Howarth I. Introduction��������������������������������������������������������������������������������������283 II. Spontaneous Order, Improvisation and Design�����������������������������������285 III. Theoretical Implications��������������������������������������������������������������������286 IV. Interaction between Improvisation and Design�����������������������������������288 V. The Conflictual Side of Improvisation������������������������������������������������289 VI. Assessing Improvisations�������������������������������������������������������������������290 VII. Distinguishing Improvisation from Non-improvisation�����������������������290 VIII. Constitutional Improvisations������������������������������������������������������������292 A. Reforms Proposed Very Early in the Life of a Government�����������293 B. Reforms Proposed in Chaotic Circumstances������������������������������296

Table of Contents  xvii IX. Good or Bad Improvisations?���������������������������������������������������������������298 A. Fixed-term Parliaments Act����������������������������������������������������������299 B. EU Referendum����������������������������������������������������������������������������300 C. The 2005 Act: The Lord Chancellor, Judicial Appointments, and the Supreme Court�����������������������������������������������������������������302 D. The 2010 Act: Prerogatives, the Civil Service and Treaties��������������303 X. Improvising Better�������������������������������������������������������������������������������305 XI. Improvisation and Constitutional Theory���������������������������������������������306 COMPARISONS 12. A Proposal for Defining and Classifying Systems of Constitutional Government������������������������������������������������������������������������������������������������311 Paul Yowell I. Introduction����������������������������������������������������������������������������������������311 II. On Constitutional Government and its History������������������������������������313 III. The Characteristics of Constitutional Government�������������������������������320 A. The Elements of Constitutional Government��������������������������������321 B. On Judicial Enforcement and Bills of Rights���������������������������������324 C. Grimm’s and Raz’s Definitions of a Constitution���������������������������327 IV. Types of Constitutional and Non-constitutional Government���������������329 A. Types of Non-constitutional Government�������������������������������������329 B. Types of Constitutional Government��������������������������������������������333 V. Conclusion������������������������������������������������������������������������������������������337 13. The View from Nowhere in Constitutional Theory: A Methodological Inquiry��������������������������������������������������������������������������������������������������������341 Silvia Suteu I. Introduction����������������������������������������������������������������������������������������341 II. The Comparative Turn in Constitutional Theory����������������������������������342 A. The Nature and Legitimacy of Constitutional Review�������������������343 B. The Use of Comparative Material in Constitutional Adjudication��������������������������������������������������������������������������������346 III. Constitutional Theory and Comparative Constitutional Change�����������348 A. The Revival of Constituent Power Theory�������������������������������������349 B. Constituent Power Theory and Unconstitutional Constitutional Amendments��������������������������������������������������������������������������������353 IV. Conclusion������������������������������������������������������������������������������������������357 ADMINISTRATIVE LAW 14. Common Understandings of Administrative Law�����������������������������������������361 Matthew Lewans I. Introduction����������������������������������������������������������������������������������������361 II. The Puzzle of Administrative Law��������������������������������������������������������362

xviii  Table of Contents III. Legality and Constitutional Formalism�������������������������������������������������364 IV. Dicey on the Absence of Administrative Law����������������������������������������368 V. Common Understandings of Administrative Law���������������������������������372 A. Constitutional Formalism and Administrative Lawlessness������������373 B. Liberty from Administrative Law��������������������������������������������������375 C. Formal Equality Through Judicial Review�������������������������������������377 D. Natural Justice and Administrative Law����������������������������������������379 VI. Conclusion������������������������������������������������������������������������������������������382 15. Methodology in Constitutional Theory: The Case of the Administrative State������������������������������������������������������������������������������������������������������������383 Kristen Rundle I. Introduction����������������������������������������������������������������������������������������383 II. Constitutionalising the Administrative State: Delineating the Object of Inquiry���������������������������������������������������������������������������386 III. Constitutionalising the Administrative State: The ‘Status’ Intervention�����������������������������������������������������������������������������������������388 IV. Constitutionalism and the Administrative State: Reflections from Australia�������������������������������������������������������������������������������������393 V. Constitutional Theory Revisited? The Provocation of the Administrative State�����������������������������������������������������������������������������398 VI. Conclusion������������������������������������������������������������������������������������������401 Index�����������������������������������������������������������������������������������������������������������������405

About the Contributors TT Arvind is Professor of Law and Head of Department at York Law School, University of York, UK. Nick Barber is Professor of Constitutional Law and Theory at the University of Oxford and Fellow of Trinity College, Oxford. David Dyzenhaus is university professor of law and philosophy at the University of Toronto. His most recent book is The Long Arc of Legality: Hobbes, Kelsen, Hart (Cambridge, 2022). Marco Goldoni is senior lecturer in Legal Theory at the School of Law of the University of Glasgow. His research is focused on legal and constitutional theory, Law & the political economy, EU public law. Jeffrey Goldsworthy is an Emeritus Professor of Law at Monash University, a Professorial Fellow at The University of Melbourne, and an Adjunct Professor of Law at The University of Adelaide. A constitutional lawyer and legal theorist, he has written extensively about parliamentary sovereignty, statutory and constitutional interpretation, and Australian constitutional law. Michael Gordon is Professor of Constitutional Law at the University of Liverpool, United Kingdom. He is the author of Parliamentary Sovereignty in the UK Constitution: Process, Politics and Democracy and the co-editor of the UK Constitutional Law Association Blog. David Howarth is Professor of Law and Public Policy at the University of Cambridge (UK), where he is also Head of the Department of Land Economy and a Fellow of Clare College. He is the author of Law as Engineering: Thinking about What Lawyers Do (2013). His public service includes two terms as a UK Electoral Commissioner and one term as a member of the House of Commons, during which time he was a member of the Constitutional Affairs and Justice Committees and of the Committee on Reform of the House of Commons (the Wright Committee). Dimitrios Kyritsis is Reader in Law at the University of Essex. He writes in legal philosophy and constitutional theory. He is the author of Shared Authority: Courts and Legislatures in Legal Theory (Hart Publishing, 2015) and Where Our Protection Lies: Separation of Powers and Constitutional Review (OUP, 2017). Stuart Lakin is Associate Professor in Law at the University of Reading. Matthew Lewans is the TransCanada Chair of Administrative and Regulatory Law at the University of Alberta Faculty of Law.

xx  About the Contributors Sarah Nason is Senior Lecturer in Law at Bangor University. Kristen Rundle is a Professor of Law at Melbourne Law School, University of Melbourne, Australia. She teaches and researches in the fields of legal theory and administrative law. Her book, Forms Liberate: Reclaiming the Jurisprudence of Lon L Fuller (Hart Publishing, 2012) was awarded the University of Melbourne Woodward Medal in the Humanities and Social Sciences (2017), and second prize, UK Society of Legal Scholars Peter Birks Book Prize for Outstanding Legal Scholarship (2012). Lindsay Stirton is Professor of Public Law in the School of Law, Politics and Sociology, University of Sussex; he is also adjunct researcher in the School of Law, University of Tasmania. Silvia Suteu is Associate Professor in Public Law at University College London, Faculty of Laws. Her research interests are in comparative constitutional law and constitutional theory, in particular the theory and practice of constitutional change, deliberative constitutionalism, and gender and constitutionalism. Her monograph, Eternity Clauses in Democratic Constitutionalism (OUP, 2021). Alison L Young is the Sir David Williams Professor of Public Law at the University of Cambridge and a Fellow of Robinson College, Cambridge. She is also a legal advisor to the House of Lords Constitution Committee. Paul Yowell is the Benn Fellow and Tutor in Law at Oriel College and Associate Professor, Faculty of Law, University of Oxford. He is the author of Constitutional Rights and Constitutional Design: Moral and Empirical Reasoning in Judicial Review and co-author of Legislated Rights: Securing Human Rights Through Legislation. He researches broadly in public law and legal theory, with particular interests in the separation of powers, constitutional theory, comparative constitutional law, and human rights.

The Methodology of Constitutional Theory – Introduction DIMITRIOS KYRITSIS AND STUART LAKIN I. BACKGROUND

W

e live in an age of conflict and polarisation in politics, law and constitutional practice. Brexit, the potential breakup of the Union, the pandemic, Black Lives Matter, climate change and more are placing stress on the relationship between states, branches of government, and individuals and the state. The customary British constitutional response to such challenges is one of p ­ ragmatism – small, incremental adjustments aimed at achieving comity between unhappy ­parties.1 But this approach is under strain. The executive branch of government is at loggerheads with the other branches, bent on corrective or retaliatory (depending on how one looks at it) action to rein in the powers of courts, and reclaim the prerogative power to call general elections. The public too are deeply divided. Protests and counter-protests, sometimes violent, are occurring more frequently, fuelled by evermore partisan sections of the media. Even a former UK Supreme Court Justice has advocated civil disobedience in response to COVID restrictions.2 In academic books and journals about the constitution – whose raison d’être is arguably to provide a cool, dispassionate forum away from the heat of real-life conflicts – temperatures are no lower.3 Scholars view the same legal and constitutional events and decisions in radically different ways. It is part-truth, part-parody to say that a political constitutionalist will boo a Supreme Court decision while a common law constitutionalist will cheer (or vice versa). Positivist and interpretivist legal theorists seem to be at a permanent impasse over the precise role of morality in the law. Public lawyers continue to disagree about the nature of their discipline, whether, for instance, it is primarily one of political theory, political science or a combination of the two. To a great extent these disagreements stem from the peculiar methodological ­challenges posed by the object of our study. For instance, constitutional law has a

1 See, for instance, D Oliver, ‘Parliament and the Courts: A Pragmatic (or Principled) Defence of the Sovereignty of Parliament’ in A Horne, G Drewry and D Oliver (eds), Parliament and the Law (Oxford, Hart Publishing, 2013, 309. 2 J Sumption, Law in a Time of Crisis (Profile Books, 2021) ch 12. 3 See M Loughlin, The Idea of Public Law (Oxford, Oxford University Press, 2003) 28.

2  Dimitrios Kyritsis and Stuart Lakin special affinity with politics, which perennially casts its ‘legal’ character in doubt. It is, as they say, ‘political law’. It thus raises the question whether there is more to be learnt from how it is understood and debated by legal practitioners, or whether we should instead pay more attention to how it functions in the broader society and especially how it figures in the deliberation and practice of other actors, such as ­politicians. Constitutional law is also typically thought to embody a measure of moral aspiration. It enshrines weighty moral principles such as democracy, separation of powers and fundamental rights norms. Claims for and against the legitimation of state policy characteristically invoke it. These – and many other – features seem to pull in opposite directions, and it is not at all a given that they can be accounted for in a coherent and theoretically cogent way. At a minimum, they put pressure on the idea that traditional legal doctrine can provide determinate and convincing solutions to the thorny issues for which we seek a solution in a national constitution. Constitutional theory cannot directly resolve the many political, legal and constitutional challenges facing nations; but we feel confident that it has the resources to offer frameworks within which constitutional actors can understand, assess and respond to those challenges. In order to exploit those resources, we need to work hard to acquire suitable ­methods and measures of success.4 Needless to say, no methodological panacea emerges from the chapters in the volume. A 400-page book on how to approach and understand legal, political and constitutional polarisations is bound to replicate many of those same polarisations. But the contributions give us cause for optimism. We feel inclined to conclude, with Dworkin, that law and democracy (and indeed all sorts of other constitutional values, doctrines and practices) are still possible despite our disagreements.5 Indeed, the dominant theme of the chapters speaks precisely to the issues of conflict and polarisation just discussed. It is the possibility of harmony, co-existence or meaningful disagreement between different constitutional methods. This theme manifests itself in diverse ways. Some contributors argue that the key to understanding constitutions is to challenge the traditional disconnect between empirical and normative constitutional analysis. As Nason puts it, to employ one type of analysis and not the other can result in a distorted view of a constitution, one removed from its realities. Similarly, Arvind and Stirton call for theorists to remove entrenched biases and predispositions in their work, by which they privilege one world view and rejects all others. Replacing this non-heuristic approach with a heuristic one will result, they argue, in leaving space for pragmatic politics to unfold. One can draw a line from these chapters to Howarth’s improvisatory model of constitutions. For him, the concept of improvisation is a tool with which to understand the socio-cultural realities of constitutions, including occasional confusion and incoherence. Rather than wish away such ‘messiness’ with idealised moral theory, Howarth says, we do well to understand how constitutional actors improvise their way out of such situations. 4 See the excellent discussion in NW Barber, The Constitutional State (Oxford, Oxford University Press, 2010). 5 See R Dworkin, Is Democracy Possible Here? Principles for a New Political Debate (New Jersey, Princeton University Press, 2006).

The Methodology of Constitutional Theory – Introduction  3 Some contributors argue that a method can yield insights about a constitutional practice without filling the whole space of constitutional theory, and without elbowing out other methods. For instance, Alison Young’s model theoretic method, and Michael Gordon’s positivist and political method both offer perspectives on constitutions which, they argue, sit happily alongside other perspectives. Young is at pains to show how her approach overlaps with, but is distinct from, critical, interpretative and ideal models of constitutions. Gordon’s method is built on a reconciliation between the two apparently opposing theories of Waldron and Loughlin. He, like Young, carefully situates his method among, rather than atop, related and potentially antagonistic approaches such as the critical, interpretative, historical, and political constitutionalist. Even the chapters in the volume that engage the vexed debates between positivist and anti-positivists, and between those who defend legislative supremacy and legislative intent over common law constitutionalism show room for rapprochement. While Lakin argues for a common law constitutionalist understanding of the British Constitution, his main aim is to show that an interpretative method of constitutional theory may recommend any number of different understandings, including the orthodox, positivist one defended by Goldsworthy in this volume. Similarly, while Goldsworthy maintains his robust defence of traditional positivist methodologies, he proposes a version of interpretivism that may be palatable to legal positivists, one that emphasises common understandings of constitutional content rather than controversial moral judgement about such content. From Goldsworthy, we can draw a line to Barber who argues for the importance and limitations of popular understandings of institutions and normative principles. Even if common understandings need to be checked for analytical or moral mistakes, he contends, such understandings aid intelligibility and can raise challenges to debates around institutions and principles that deserve an answer. Here group wisdom can meet the wisdom of experts. Perhaps the most direct and extensive attempt to reconcile positivist and antipositivist legal theories is found in Dyzenhaus’s chapter. Dyzenhaus wants to derive his brand of antipositivism from an interpretive reconstruction of key commitments of two arch-positivists, HLA Hart and Hans Kelsen. A similar reconciliatory ambition fuels Kyritsis’ contribution. Although he puts forward an unabashedly moralised methodology for constitutional theory, he insists that one of the measures of its success is its ability to account for the fact that constitutional law also embodies a contingent political settlement. This is a fact that is commonly taken to give legal positivism an edge over its opponents. Several of the chapters in the volume deal with disagreement and conflict as a crucial parameter that frames the proper methodology for constitutional theory. Perhaps, as Barber suggests, the contested nature of moral concepts reduces the utility of appeals to a supposed common understanding of those concepts. To the extent that constitutional theory makes reference to such concepts and seeks to settle ethical disagreements among constitutional lawyers, Barber’s point is relevant to it as well. Conversely, morality may explain, say Dyzenhaus, Kyritsis and Lakin, why people owe their allegiance to a political regime despite their ethical disagreements. In fact, it may show that the constitution, with its status as higher-order law, has a special role to play in securing their allegiance under these circumstances.

4  Dimitrios Kyritsis and Stuart Lakin II.  THE SCOPE OF THE VOLUME

Before we introduce the chapters and chapter themes, it is necessary to say a few words about the scope of the volume. Our title and focus for this volume is the methodology of constitutional theory, but we have not been prescriptive about any of these terms. In particular, the reference to theory is not intended to carry any connotation of morality, abstraction or editorial approval. It is an empty vessel which our contributors have filled in varied and imaginative ways. A common question in the chapters concerns the relationship between constitutional theory and practice, but it will be seen that the meaning of practice is as disputed as the meaning of theory. Similarly, the title reference to constitutional is enough to put contributors in the same ballpark, but we have left it up to individuals to define their precise territory. Several contributors, for instance, have taken an administrative law focus. We have also not attempted to distinguish constitutional theory from other disciplines, for instance sociology or political science. We instead want to allow that a successful methodology of constitutional theory may draw upon techniques associated with any of these disciplines (if they are indeed different disciplines). This view is reflected in several of the chapters, and across the different organising themes of the book. In terms of the jurisdictional focus of the volume, we did not intend a silent UK in our title, but readers will detect such a focus in many of the chapters. In some cases, the contributors have tailored their arguments to the UK Constitution, but for the most part aspects of UK constitutional practice serve as illustrations or case studies for the general methodological claims advanced. Perhaps there is good theoretical reason for this emphasis. The UK Constitution has been described as ‘indeterminate, indistinct and unentrenched’.6 Whatever the merits of that view, it is one that may attract authors seeking to impose some sort of order on constitutional argument. Be that as it may, the UK is not the only jurisdiction considered here. This is clearest in the chapters dealing with the methods of constitutional comparison but can be seen in other parts of the volume as well. III.  THE CHAPTERS AND THEMATIC GROUPINGS

We have structured the volume according to rough thematic emphases in the chapters. Readers (and indeed authors) may identify more obvious emphases, but we have arrived at the following six: Self-Understandings, How do Facts Matter? Morality, Social Theory, Comparisons and Administrative Law. Such is the richness of the chapters, we could almost have placed any chapter under any theme, or indeed arrived at many different combinations of themes. The themes themselves overlap and inter-link in manifold ways. No matter what view one takes on these questions, we feel confident that the chapters cover very broad terrain in debates about the methodology of constitutional theory. In the space remaining in this introduction, we introduce our chosen themes and the chapters we have assigned to them.

6 SE

Finer, V Bogdanor, and B Rudden, Comparing Constitutions (Oxford, Clarendon Press, 1995) 40.

The Methodology of Constitutional Theory – Introduction  5 A. Self-Understandings In chapter one ‘The Significance of Common Understanding’, Barber discusses one of the central planks of the traditional approach to constitutional theory, the idea that constitutional theory is meant to explicate the concept of the constitution, as it is understood by the community of concept users. Since HLA Hart, this idea has been associated with legal positivism. A methodology for constitutional theory oriented towards the common understanding presents a number of advantages. First, it pushes theorists to produce accounts of constitutional concepts that are intelligible to those who employ them. Relatedly, those accounts will be attuned to the way in which social understandings shape social institutions like the constitution and thus more contextual. Finally, those accounts will benefit from the wisdom of the crowds vested in the common understanding. While broadly sympathetic to this approach, Barber also explores some of its limitations. He contends that the common understanding is not always determinate and may not serve all theoretical purposes as well, for example the radical critique of existing institutions. He also maintains that some important features of social institutions may well elude participants and that the wisdom of the crowds argument applies with diminished force in the case of moral concepts. In chapter two, ‘In Defence of Traditional Methodologies’, Jeffrey Goldsworthy approaches the idea of common understanding in the context of debates between positivist and interpretivist constitutional lawyers. He finds perplexing the idea – which he attributes to interpretivists – that the meaning of a legal and constitutional practice could be anything other than what its participants believe it to be. To claim otherwise, he insists, is to impose a new practice rather than recognise the existing one. Elsewhere in the volume and beyond, interpretivists have their say. While common understandings may determine constitutional content, says Lakin, this cannot be so simply as a matter of descriptive fact. Moral values may recommend such determinants, but this will depend on the peculiar history and practices of a constitutional order. In other work related to his chapter in this volume, Barber lands somewhere between Goldsworthy and Lakin. His argument is that an interpreter must bring her ethical beliefs to the task of interpretation, but that she should also consider ‘which of these features are important to the recipients of her theory’.7 Goldsworthy begins his chapter with the claim that constitutional law requires multiple methodologies, and that the particular methodologies needed to ascertain the law depend on which type of law is in question. He resists the interpretivist view that there is a ‘master methodology’ such that the answer to any legal question depends on the most fundamental principles of the common law, made and developed by judges. While he allows that the interpretivist account may have some plausibility in relation to the common law, he contends that this theory was never plausible in the case of written constitutions and statutes. Judges consistently speak of giving effect to the intentions of the legislature rather than holistic moral interpretation.



7 Barber,

The Constitutional State (n 4) 11.

6  Dimitrios Kyritsis and Stuart Lakin This is so regardless of occasional pragmatic, results-orientated dissembling. Case law and scholarship today and through history in Britain, New Zealand, Australia and America show that the common law is subordinate to written constitutions and – under the doctrine of legislative supremacy or sovereignty – statute. And, as a matter of legal philosophy, a positivist defence of parliamentary sovereignty is true whether one is a positivist or an interpretivist. He warns that there could be severe practical problems if – as interpretivists arguably claim – officials and citizens could adopt different and incompatible moral interpretations of their legal arrangements. Legal systems are more likely to be stable and effective if participants accept a consensus approach (whether positivist or interpretivist) than an interpretivism that depends on the moral judgements of the interpreter. Here, the theme of common understanding takes the form of a distinctive constitutional method. In chapter three, ‘Brexit and Constitutional Theory’, Alison Young explicitly distinguishes her approach to constitutional theory from one based on tracking what most people think. Her own brand of a model-theoretic approach to constitutional theory seeks to optimally combine attentiveness to social facts and normative appeal judged along two dimensions, the model’s fittingness to serve a certain purpose and the independent merit of that purpose. She advocates the use of constructed-type models in constitutional theory that aim to offer a normatively defensible ideal-type of a constitutional concept that is constrained by what is objectively probable or feasible – rather than merely possible – in a specific set of social circumstances. In this sense, constructed-type models differ from interpretative approaches to constitutional theory, whose stated primary goal is to further our understanding of the existing constitution; constructed-type models blend understanding and prescription. Conversely, by being anchored in objective probability, constructed-type models can make critiques of constitutional practices and proposals for reform that are more effective than those made by theories guided solely by an ideal understanding of constitutional concepts. Young demonstrates these virtues through a model-theoretic analysis of the impact of Miller II on the UK Constitution. She argues that the controversy over the Supreme Court’s decision can be fruitfully construed as not about the ‘correct’ understanding of parliamentary sovereignty but about which understanding of this important constitutional concept is sufficiently responsive to empirical facts of constitutional and political practice and normatively appealing. B.  How Do Facts Matter? In order to understand a constitutional practice, we arguably need some criteria as to which facts or aspects of the practice carry which type of significance. We might think that this is especially important if there is widespread disagreement about the content or meaning of the practice. Does our criteria of importance need to be rooted in arguments of political morality as Kyritsis and Dyzenhaus contend? Or is indirect, non-moral evaluation to be preferred, by which we seek to identify the important features of a practice without need of controversial ethical judgements. Are criteria of importance relevant only to legal, doctrinal and theoretical

The Methodology of Constitutional Theory – Introduction  7 scholarship; or must/can sociologists, political scientists, historians and others also employ such criteria? In the second part of the book, we see three perspectives on these types of questions. In chapter four, ‘Slaying the misshapen monster: The case for constitutional heuristics’, Arvind and Stirton are critical of processes of identifying facts that have constitutional significance, and then assigning significance to those selected facts. They detect a similar pattern in the construction and transmission of traditions. In each case, they argue, the methodology suffers from the failure to recognise the extent to which theories are shaped by the predilections and predispositions that underlie them. As if to underline this point, they outline four families of approaches that often drive a constitutional theorist’s selection of facts. The doctrinaire insistence of absolute fidelity to only one, they contend, will inevitably lead to positions in ­constitutional debates becoming entrenched, polarised and unyielding. Indeed, these are the symptoms, they argue, of the implicit non-heuristic epistemology underpinning the work of many constitutional theorists. These difficulties can only be overcome, they argue, by giving a greater role to a heuristic approach to constitutional issues, one that relates more closely to the shifting needs of the polity, and which opens up a more pragmatic space for theoretical debate and doctrine-making in constitutional scholarship. In chapter five, ‘Why Common Law Constitutionalism is Correct (If It Is)’, Lakin argues for a process and conclusion of precisely the type that Arvind and Stirton reject. The correct account of a constitution, he says, depends on an interpretative judgment about which moral values, in combination with which facts, give the most morally attractive understanding of the practice. He takes as the focus of his chapter, two leading accounts of the British Constitution: the positivist, ­orthodox one put forward by Jeffrey Goldsworthy, and the interpretative, common law constitutionalist one propounded by Trevor Allan. To begin with, he explains the sense in which the two accounts represent rival models of the salient moral and empirical facts and features of British constitutional practice. He then turns to the different ways in which theorists have sought to defend their preferred model. The dominant descriptive methods employed by Goldsworthy and other positivists, he argues, suffer from a common defect: they each assume the correctness of something like the orthodox account, rather than independently establish its correctness. The better method, he contends, is an interpretative one, operating at a deeper level of abstraction than Allan’s common law constitutionalism. This method engages questions and concerns of political morality plausibly shared by each account: why does a practice bind, and why is a practice one to which people owe their allegiance (if they do). In so doing, it opens the way for important and genuine disagreement about constitutional practices. It also encourages a conciliatory attitude on the part of scholars. Lakin concludes (tentatively) that the common law constitutionalist model, and the types of moral values that underpin that model, gives the better understanding of key features of British constitutional practice than the orthodox model. In any case, he says, if we accept that there are correct doctrinal propositions of constitutional law and constitutional practice, then we must also accept that there are true accounts of which facts constitute the law and the constitution. There is a problem with this

8  Dimitrios Kyritsis and Stuart Lakin aim for correctness, say Arvind and Stirton. A constitutional theory is not merely an account of an existing practice, but a framework for the future development of a practice. In order for a constitutional theory to be correct, it must therefore successfully predict which future facts will come to pass. The interpretivist cannot answer this challenge, they say, by offering an account of the political values that a good constitution ought to serve. This is precisely what leads to rhetorical clashes between different styles of public law thought. In chapter six, Methodological Pluralism and Modern Administrative Law, Nason joins Arvind and Stirton in warning about methodological bias in the criteria of significance for constitutional understanding. A case in point, she argues, is the government Independent Review on Administrative Law (IRAL). As she sees it, the Review was biased towards a particular position, the one adopted by the Judicial Power Project and other scholars who think that judicial review should not be about rights and constitutional principles. Nason’s remedy for such biases is to shift from the traditional polarised debates between legal and political constitutionalists towards a plurality of methods and values. In particular, more rigorous empirical research, and more attention to research already conducted, is needed, she argues, to show how public bodies actually operate, and how and which constitutional values work in practice over time. The total separation between the outcomes of empirical and normative methods, she warns, can lead to the production of legal theories that are based on misconceptions about the day-to-day experiences of law (as exemplified by the IRAL). However, this works both ways. If socio-legal research designed to uncover empirical evidence about judicial review, and other elements of administrative law, does not openly engage to at least some extent with constitutional and administrative law theory, she says, it risks producing results that are underpinned by particular interpretations of legal, political and administrative values, that are not made clear, and that are therefore not capable of being properly analysed and challenged. In developing these two perspectives and their relationship, Nason highlights a rich variety of philosophical concepts and theories from which empirical researchers will benefit. At the same time, she highlights new directions in administrative law to which administrative lawyers and theorists need to respond: the plurality of contested and evolving administrative law values beyond overtly constitutional values and principles such as separation of powers; the realities of the state as revealed by Bell, Kirkham and O’Loughlin and others; and the emergent contribution of computer science to administrative law analysis. C. Morality For many constitutional lawyers, morality is only relevant to the evaluation of extant constitutional law. Alternatively, they allow that it plays at best an interstitial role in constitutional argument, when other methods of determining the content of constitutional rights and duties have run out or when the constitution, identified in a non-moral way, directs constitutional actors to have resort to it. The chapters in the third part envisage a more robust role for moral considerations in constitutional theory, one that transcends the stark contrast between identification and

The Methodology of Constitutional Theory – Introduction  9 evaluation of the law. This position invites the objection that it collapses what the law is and what it ought to be. It therefore becomes paramount from the perspective of methodology not only to single out the compartment of morality one takes to play this constitutive role but also to articulate the desiderata (other than moral appeal) that govern this choice and explain how a choice that satisfies these desiderata maintains the distance between the ‘inner morality’ of constitutional law and ideal justice. The three chapters in this part propose different responses to these theoretical challenges. A further issue concerns the point at which moral considerations enter constitutional theory. For authors like Gordon moral argument is only operative at the level of establishing the authority of bodies with the power to make or amend the law. Others like Dyzenhaus maintain that the question of law’s authority is also pertinent to adjudication and, thus, moral considerations are constitutive of our constitutional rights and duties at what Gordon aptly calls the ‘retail level’ as well. In chapter seven, ‘The Constitution of Legal Authority’, David Dyzenhaus seeks to find the moral core of constitutional law in the idea of legitimacy, but he associates legitimacy with a regime’s acceptance by those subject to it. For him, though, this connection of law – and especially constitutional law – with legitimacy is incipient even in the positivistic theories of Hart and Kelsen. To this end, Dyzenhaus advances an unorthodox interpretation of Hart’s account of the authority of law as requiring acceptance of subjects as well as officials; and he argues, contrary to Hart and Raz, that de facto authority is always de jure because of the role of legal officials, particularly judges, within the institutional framework of the law in giving answers to the question of the legal subject: ‘But, how can that be law for me?’ It is essential for a theory of law to explain that the majority of those subject to the law understand their subjection as serving their interests even when they disapprove of the content of some of the laws. This is especially so, he says, if officials relegate individuals on the losing side of the struggle to second-class status – ‘out-groups’ rather than ‘in-groups’. Such individuals will not be able to accept the compromise as a sufficient basis to make sense of their legal order as an order of legal right or authority, rather than one of unmediated coercive power. As one would expect, Dyzenhaus applies and illustrates these arguments at times to the question of segregationist laws. In chapter eight, ‘Constitutional Theory as Legitimacy Enhancer’, Dimitrios Kyritsis defends a moralised methodology for constitutional theory that is fuelled by the substantive claim that the purpose of constitutional law is to enhance the moral legitimacy of a political regime in a special way, namely by structuring government such that it will reliably and systematically operate in a morally justifiable way. He calls this dimension of legitimacy assurance legitimacy. He argues that this methodology can successfully account for two salient features of constitutions: a) their heightened binding force; and b) their contingent character as a political settlement. A focus on assurance legitimacy as opposed to other moral principles such as equality or the rule of law can explain why constitutional settlement matters morally such that we have reason sometimes to prefer it to ideal justice. That is because by committing to a constitution of the right sort a polity offers citizens a standing assurance that they will not be placed at the complete mercy of the victorious side. In exchange for

10  Dimitrios Kyritsis and Stuart Lakin this assurance it is reasonable to demand that they give their allegiance to the polity despite its moral imperfections (within certain limits). In chapter nine, ‘A Positivist and Political Approach to Public Law’, Michael Gordon offers a different understanding of the role of morality in constitutional theory. At first blush, it might be surprising to include him in this part of the book at all. In an important respect, the thrust of his claim is to limit the impact of moral considerations in constitutional argument. Even so, he adds an unmistakable normative dimension to positivist constitutional theory. Following Waldron, and in a way that is perhaps redolent of the questions and concerns of Kyritsis and Dyzenhaus, he describes as crucial the need to understand how law operates within political government: how to explain the authority, functions and limitations of law. He defends an approach to public law which is, as he puts it, both positivist and political in its orientation. In so doing, he attempts to reconcile Jeremy Waldron’s theory of normative positivism with Martin Loughlin’s anti-positivist, political jurisprudence. Despite the ostensible competition between the theories, Gordon contends that there are strong affinities between them. This reconciliation yields insights, he says, at both the ‘wholesale’ level of constitutional design and the ‘retail’ level of constitutional practice. At the wholesale level, the positivist and political approach explains, for instance, why we might understand law as a set of generally authoritative and enforced rules which exist and have effect regardless of their substantive compatibility with standards of justice: positive law is the product of a specific set of governmental arrangements, and the authority of law flows from the fact it is located in such a political system, rather than because legal rules exist to serve any preordained set of moral ideals or a particular scheme of justice. At the ‘retail’ level, a positivist and political approach, he says, allows us to understand legal decisions in a political context, while preserving the distinctiveness of legal reasoning; and it functions as a necessary constraint on judicial power. Gordon closes with three examples of the insights his approach can bring to public law: parliamentary sovereignty, the Miller cases, and the status of human rights within constitutional law. D.  Social Theory The chapters of this collection reveal a fundamental distinction of great methodological value between two goals constitutional theory can set for itself. One goal is, we might say, doctrinal. It is to explore how other disciplines such as politics and sociology might be brought to bear on problems of constitutional law doctrine. In other words, it is to investigate whether those other disciplines can help determine what is the correct answer to questions regarding our constitutional law rights and duties. Perhaps the clearest example we have met thus far is Nason’s chapter discussed above. The other goal is explanatory. It is, as Goldoni puts it, to ‘understand the reality of existing constitutional orders’. In this guise, constitutional theory might want to know, for instance, whether there is a distance between ‘the law in the books’ and what political and government actors actually do, explain the social forces that brought about constitutional change or the way in which the constitution structures society. In both cases the content of the law is

The Methodology of Constitutional Theory – Introduction  11 taken largely as fixed, and the focus is on its efficacy or its evolution. The peculiar challenge facing this methodological approach is to avoid reductionism about the practice being investigated, to analyse the patterns that it exhibits, whether or not ­participants are conscious of them, but without betraying the complexity of the practice and its norms. In chapter ten, ‘The Material Study of the Constitutional Order’, Marco Goldoni pursues this explanatory task by drawing on conceptual resources from political sociology and even political theology. The material study of the constitution that he espouses investigates what he calls the internal relation between the constitution and the societal formation of which it is a part. It pays special attention to the way various social forces, such as the organisation of labour and reproduction of society, the conflicting interests of its different segments, and even aspects of its social imaginary shape and are structured by the constitutional order. It also explores how the constitutional order emerges as a system of norms of a special sort characterised by unity, fundamental objectives and bearing subjects. Goldoni illustrates the utility of the material method through a case study that is of particular value to UK constitutional lawyers, namely constitutional change. The material study reveals that such change can sometimes occur outside formal channels of constitutional amendment and, conversely, that it may not be present in cases where the constitution has, as a formal matter, been significantly altered. In chapter eleven, ‘The British Constitution as an Improvised Order’, Howarth joins Goldoni in offering explanations for constitutional change outside of formal constitutional processes. He brings to constitutional theory concepts more usually associated with music and engineering: improvisation and design. He poses the questions: to what extent is UK constitutional reform improvised? And to what extent is an improvisatory model a success or failure? Throughout his chapter, he carefully unpicks the concept of improvisation, contrasting it with ‘the spontaneous’, and ‘the designed’. Improvisers, he tells us, consciously try to bring their actions into harmony with the actions of others, but unlike planners and designers, they are not aiming to bring about a chosen final pattern. Improvisation is also characterised, he says, by using tools for purposes for which they were not originally intended on materials that just happen to be at hand. He illustrates these characteristics using, for instance, common law reasoning, statutory interpretation, the drafting of statutes, and judicial activism. In the latter sections of his chapter, he details some possible instances of constitutional improvisation within the British Constitution, identifying the Fixed-term Parliaments Act 2011 and EU Referendum Bill 2015 as prime examples. He then asks whether these are examples of good or bad improvisations: whether they either set future developments on a path that continued to solve problems, or led to a ‘maze’. In a way that seems to insert a wedge between doctrinal and explanatory approaches to the constitution, Howarth closes with some observations about his social scientific account of constitutions as compared to normative accounts. The latter accounts, he argues, are often highly idealised and implausible, failing to take into account the reality of improvisatory reasoning, confusion and incoherence in the law. This may be right, but, perhaps, as Nason and Bell contend, there need not be a tension between the functionalist concerns of Howarth’s improvisatory model, and

12  Dimitrios Kyritsis and Stuart Lakin the normative concerns of others. An accurate account of a constitution may depend on insights from both of these approaches. E. Comparisons In a globalised world, the facts that constitutional theory may (or must) pay attention to may include facts about the constitutional law of other jurisdictions. This, anyway, is the battle cry of the bustling comparative constitutional law literature. But what can we learn from such comparative insights? For instance, what can these insights tell us about the merit of our own Constitution or the best way to interpret its abstract provisions? Paul Yowell and Silvia Suteu try to answer these questions and in so doing offer much needed terms of engagement for our exploration of foreign constitutions. Importantly, they seek to neutralise the ever-present threat of parochialism. They thus confront the question to what extent we ought to thin down our own moral convictions in exchange for being more ecumenical, as Yowell urges, and how we can make our comparisons sensitive to context, as explained by Suteu. In chapter twelve, ‘A Proposal for Defining and Classifying the Types of Constitutional Government’, Yowell offers the essentials of a classificatory scheme of types of constitution. He steers a middle course between a purely descriptive classificatory criterion that equates a constitution with a form of government, on the one hand, and, on the other, a thickly normative criterion that sees the constitution as a necessary element of political legitimacy and thus essentially setting functional as well as substantive benchmarks for the exercise of state power that are drawn from moral theory. He calls his classificatory scheme thinly normative because it employs exclusively functional criteria regarding the structure of government, adherence to which is a means for ensuring the accountability of those in power. Importantly, this scheme does not rely on the idea that constitutions must have a canonical character and give courts the power to enforce constitutional rights against the legislature. Using this scheme, he distinguishes, first, constitutional from non-constitutional government, and, further downstream, presidential from parliamentary systems. Yowell claims that this classification pays important methodological dividends: By not equating constitutionalism with its liberal – or any other parochial – variant, it purports to capture a broader range of political systems and is therefore better suited for comparative constitutional law studies. In chapter thirteen, ‘The View from Nowhere in Constitutional Theory: A Methodological Inquiry’ Silvia Suteu makes a plea for a constitutional theory that is more attentive to context and variation across jurisdictions. Drawing on recent developments in comparative constitutional law, she argues that constitutional comparison can serve both as a useful corrective to sweeping claims about the superiority (especially moral) of one constitutional arrangement over other and as a source of valuable knowledge and inspiration from constitutional experience. The former virtue is illustrated through the debate over the legitimacy of constitutional review, whereas the latter through the judicial use of foreign jurisprudence. However, Suteu warns against the over-hasty reliance on constitutional comparison. In particular, she highlights the importance of robust case selection with a view to avoiding epistemic bias and the

The Methodology of Constitutional Theory – Introduction  13 need to look closely at the practice of constitutional concepts and not only their intellectual history. Suteu demonstrates the potential and pitfalls of comparative analysis in the burgeoning literature on constituent power and unconstitutional constitutional amendments. F.  Administrative Law Most characteristically, constitutional theory deals with the structure of government at the highest level. Does it have anything to say, though, about the legal regulation of the modern, sprawling administrative state? This Part explores the extent to which the concepts and methods of constitutional theory can apply to the study of the administrative state and enrich our understanding of it. However, this rapprochement of the two fields is not without its complexities. Constitutional theory cannot unthinkingly perpetuate the traditional model of legal authorisation built on the idea of statutory delegation that occludes just as much as it enlightens. It must confront administrative law’s nature and the peculiar challenges it poses. For example, it must account for the fact that administrative bodies enjoy vast discretion in many areas of social life. It must also acknowledge the special virtues and benefits that they are supposed to bring to constitutional government. It may, then, be that the best way jointly to satisfy these desiderata is by carving out a distinct source of constitutional authority or a distinct constitutional status for the administrative state. In chapter fourteen, ‘Common Understandings of Administrative Law’, Matthew Lewans identifies a number of conceptions (or ‘common understandings’) of the place of the executive in the constitutional landscape as complementing the place of legislation and the common law. In addition, he uses landmark administrative law cases to show how these conceptions have also animated judicial philosophies. These conceptions have proposed formal power-conferring rules, individual liberty, equality before the law, and natural justice as the touchstone of administrative law and its raison d’être. Lewans shows that these conceptions are underpinned by and combine in different ways enduring ideas and argumentative strategies that he traces back to the constitutional theories of William Blackstone and Alfred Venn Dicey, and the puzzles and difficulties that they face are to a large degree those that Blackstone and Dicey sought to resolve in their constitutional theories. Thus, his approach illustrates the value of bringing intellectual history to bear on the study of administrative law. In chapter fifteen, ‘Methodology in Constitutional Theory: The Case of the Administrative State’, Kristen Rundle frames the issue as one about the constitutional status of the administrative state. Her discussion is motivated by dissatisfaction with traditional accounts situating the administrative state in the classic tripartite separation of powers. This dissatisfaction is due to what she sees as both the descriptive inaccuracy and normative inadequacy of those accounts. To locate what is left out by them she turns to the theories of Gillian Metzger and Kate Glover Berger who have theorised the administrative state not primarily in terms of legislative delegation and judicial supervision, as constitutional orthodoxy would have it, but more directly in terms of either the benefits that the administrative state as a separate unit of legal

14  Dimitrios Kyritsis and Stuart Lakin analysis brings to constitutional government, as in Metzger’s theory, or of its place in the contemporary Canadian constitutional architecture, as in Berger’s. Rundle then applies the same methodological shift to the Australian context. Her primary target are the twin ideas that administrative law has a primarily supervisory rather than facilitative function and that its crux is to determine legally valid administrative action. She concludes by proposing that in all three jurisdictions what unifies our exploration of the administrative state in all its multifarious expressions is a concern with the relationship between the government and the governed.

Self-understandings

16

1 The Significance of the Common Understanding in Legal Theory NW BARBER*

T

hose who debate the nature of social institutions and the content of ­constitutional principles often pay little attention to the character of the arguments they deploy. It is sufficient that these arguments are persuasive – and, it might be thought, there is little to be gained in enquiring into the degree to which they ought to persuade. There are dangers in spending too long pondering questions of methodology; all too often, such reflection achieves little, and becomes introspective and boring. The danger of ignoring methodology, in contrast, is that we fail to examine some of our foundational assumptions about the nature of the project we are engaged in, and, sometimes, these unexamined assumptions determine the accounts we go on to produce. The fear is that once examined these assumptions might prove rather less attractive than we had hoped. I have written on the general question of methodology in legal and constitutional theory elsewhere.1 This chapter is more narrowly drawn and focuses on one issue that is relevant to legal scholarship: the role of the common understanding of an institution or principle in formulating a successful account of that entity. It is often thought, or at least assumed, that what people think about institutions or principles has a role to play in the way we theorise about them. Indeed, accounts of social institutions such as law or the state, and principles such as the separation of powers or sovereignty, are sometimes challenged on this basis. An objector will proffer an instance of one of these institutions, or a version of these principles, that she claims the common understanding would endorse, and use this example as a criticism of the theory that is being advanced. For instance, an account of a legal order that included

* This paper was originally published in the Oxford Journal of Legal Studies and is reprinted with the kind permission of the editor and Oxford University Press (NW Barber, ‘The Significance of the Common Understanding in Legal Theory’ (2015) 35(4) Oxford Journal of Legal Studies 799–823, https:// doi.org/10.1093/ojls/gqv017). 1 I begin the task in NW Barber, The Constitutional State (Oxford, Oxford University Press, 2010) ch 1.

18  NW Barber coercion as a feature of the account – asserting that legal orders are coercive social systems – might be countered with the example of a society of angels. The angels would need law to resolve coordination problems, they would need a legal order, but they would not need these rules to be backed by force. The counter-example appeals to the common understanding of a legal order: this would acknowledge the angelic system as an instance of this entity, and its acknowledgement shows that law need not be coercive. Similarly, an account of the state that included international recognition as a feature of that social institution might be countered with the example of Taiwan. Taiwan, it might be asserted, is captured by the common understanding of the state, yet is not recognised as a state by the international community. Consequently, international recognition cannot be a necessary feature of the state. This chapter will consider the significance of the common understanding in formulating such accounts. It will argue that the common understanding is always relevant, and relevant in a number of different ways, but is not decisive in the manner that some of its strongest supporters may hope. What people think of institutions and principles is only one factor amongst others in the formulation of accounts of these institutions and principles. The chapter divides into three parts. In the first part, the claims of the common understanding are surveyed. Often, its primacy is implied by the claims of theorists rather than argued for directly, but some authors explicitly use the common understanding as the criterion for success of their accounts. In the second part of the chapter the merits of the common understanding will be considered. There are at least three reasons why a theorist should pay regard to the common understanding. First, the account produced needs to be intelligible to the community it addresses. Second, as social institutions are constituted by rules, and shaped by people’s beliefs and dispositions, the common understanding of social institutions will at least partly determine their nature and operation. Finally, the common understanding may provide a pointer towards important features of the institution or principle: an argument based on the ‘wisdom of crowds’ may require us to pay attention to what people think is significant and valuable, even if we struggle to see this ourselves. The third section of the chapter considers the limitations of these arguments. Though they all show a role for the common understanding in accounts of institutions and principles, none are decisive: there is latitude for theorists to depart from the common understanding and, indeed, to produce successful interpretations of these phenomena such departure is often essential. I.  ADHERENCE TO THE COMMON UNDERSTANDING

The common understanding is frequently – though sometimes indirectly – invoked in legal theory as a test of the correctness of an account of an institution or principle. Before surveying the invocation of the common understanding in legal philosophy it is necessary to consider the nature of the common understanding itself; the thing that is being appealed to when arguments of this type are made.

The Significance of the Common Understanding in Legal Theory  19 A.  The Nature of the Common Understanding The common understanding is a function of the understandings held by individuals. The following paragraphs will develop an account of the common understanding in two stages. First, the understandings of individuals will be discussed. Individuals possess and use concepts that relate to the institutions and principles that are the subject of analysis by the theorist. Second, the process by which these disparate concepts combine into the common understanding will be examined. The common understanding is an artefact of a group, an artefact which is both formed by and shapes the concepts held by individuals. Concepts are mental representations of the world, units of thought that pick out and group together features of the world.2 They are quintessentially personal entities but also, as we shall see, are shaped by our interaction with others in our community. The two leading contemporary theories of concepts are the prototype and the exemplar models. Each seeks to map the ways in which we generate, refine, and hold concepts. They make psychological claims about the nature of concepts: asserting that their model captures the ways in which our brains divide up units of thought. The prototype model presents concepts as summary representations of a category of entities. This is a little like the ‘ideal-type’ model in social theory, a model that advances a description of a collection of phenomena which seeks to capture the essential, or the most significant, shared features of those entities.3 It differs from the ideal-type, though, in that the objects included within the concept need not share a set of features – indeed it is possible that the objects do not share any significant features.4 Prototypes include a range of features and not all of these need be present for an entity to be identified as included within the concept. The various features have degrees of significance, and entities can be more or less typical instances of the concept depending on which of the prototype’s features they possess and degree to which these features are present. To simplify, on the prototype model the individual holds a concept of a ‘dog’ when she has a mental set of features that constitute a prototype of a dog. Dogs are animals, have four legs, bark, eat meat, and so forth. This prototype then enables her to identify an entity in the world as being an incident of a dog – even if it lacks four legs and is, doubtless to its regret, fed vegetarian food. Exemplar models present concepts as collections of memories of encounters with entities encompassed by the concept.5 The brain stores these encounters in a web, and new experiences are then categorised by virtue of their similarity to remembered instances. These memories break down into collections of features that connect

2 GL Murphy, The Big Book of Concepts (Boston, MIT Press 2002) 1–5; S Carey, The Origins of Concepts (Oxford, Oxford University Press, 2009) 5–10. 3 P Baert, Philosophy of the Social Sciences (Cambridge, Polity Press, 2005) 46–51. 4 Murphy, Big Book of Concepts (n 2) 44–5. 5 ibid 49–54.

20  NW Barber instances within the concept. Like the prototype model, the instances included within the concept need not have any single significant feature in common, and typicality is, again, a matter of degree, depending how closely a newly encountered entity resembles those memories encompassed by the concept. On this model, rather than the brain producing an abstracted account of a ‘dog’, it is now presented as storing – more or less clearly – memories of the dogs that the individual has encountered. These memories are united by the shared features they possessed – the dogs were animals, had four legs, and so forth. When a person encounters a new entity that may amount to a dog, this creature is categorised by reference to these past experiences. For our purposes it is unnecessary to choose between the prototype and exemplar models of concepts. They share a number of common features, though, that are worth noting. In each model, concepts are aspects of individuals’ mental worlds. We each have our own collection of concepts that are – normally – similar to, but not identical with, those of others. People may draw conceptual boundary lines in different places, they may differ in terms of the features included within the concept and the weight ascribed to those features. Sometimes this presents problems: people can find that conceptual differences make comprehension and communication difficult. Sometimes, though, this can be an advantage. Variations between similar concepts might be determined by the use people make of those concepts; they may be shaped by the needs of the tasks those individuals engage in. Psychologists have shown, for example, that biologists and landscapers tend to have distinctively different sets of concepts relating to tree-types, and also show differences between similar concepts within those sets.6 When the common understanding is referenced in legal theory it is not – or, at least, it is not normally – treated as the sum of all of the related individual concepts in the group. If the common understanding were this broad it would be almost meaningless, including all manner of odd outliers. The common understanding has a normative dimension; it seeks to constrain diversity between the various related concepts held by individuals. Indeed, the common understanding could be described as the folk-theory of a concept; the community’s understanding of what it means to possess a correct version of that concept.7 When presented as an argument against an account in legal theory, the common understanding is given an explicitly critical edge: it is invoked to show that some concepts held by individuals are flawed. Whilst concepts are possessed by individuals, they are partly shaped and refined by the interaction of individuals within a community. They are set in a dynamic relationship with the community’s understanding – the common understanding – of phenomena. The first part of the community’s role in the crafting of concepts stretches back into the distant past. It is likely that we have evolved to be sensitive to certain distinctions, and it is probable that some, at least, of the mechanisms

6 DL Medin, EB Lynch, and JD Coley, ‘Categorization and Reasoning Among Tree Experts: Do All Roads Lead to Rome?’ (1997) 32 Cognitive Psychology 49. See also Murphy, Big Book of Concepts (n 2) 232–234 and G Deutscher, Through the Language Glass (London, William Heinemann, 2010) ch 4. 7 See the discussion in B Leiter, ‘The Methodology Problem in Jurisprudence’ in B Leiter, Naturalizing Jurisprudence (Oxford, Oxford University Press, 2007) 175–80.

The Significance of the Common Understanding in Legal Theory  21 that we use to develop and refine concepts are innate.8 Second, and connectedly, the community may shape and refine the concepts held by individuals more directly through teaching and feedback on usage, invoking the common understanding.9 A child who included cats within her concept of dogs (a different error from simply calling a cat a dog) would need to have the differences between these two groups of creatures sketched out for her.10 Such feedback continues into adulthood, as when we are told that, for example, a tomato is ‘really’ a fruit, or that a whale is a mammal and not a fish. When I write of the common understanding in this chapter, then, I am referring to the community’s understanding of a concept, the deviation from which will often result in the correction of the concept-holder. It is not just that the holder of the concept has a different concept to similar concepts held by others in the group – that much is unremarkable – but this difference is understood by those in that community as amounting to a flaw in the way in which the holder of the concept divides up the world. The reasons for adherence to the common understanding, the supposed justification for correction, are often left unarticulated – some of the possible reasons for correction are considered in the next section of this chapter. B.  The Invocation of the Common Understanding in Legal Theory When the common understanding is invoked in argument, the claim is that the account of an institution or principle presented by a theorist contains or implies the holding of a version of a concept that falls outside of the common understanding and which is – because it falls outside of the common understanding – flawed. The common understanding has been invoked in legal theory as a litmus test for accounts of both institutions and principles. In the context of institutions, the argument tends to be made directly: a successful account of a social body is one that expounds and clarifies our common understanding of that institution. In the context of ethical principles the argument is more ambitious, and takes a metaphysical turn. Perhaps all that can be done in expounding principles is to explicate the common understanding of their content. Perhaps such principles – like democracy, the separation of powers, or the rule of law – exist only as parts of our shared moral language, and the task of the theorist is confined to explicating the meaning they have within that language. i. Institutions It is common to argue that the task of legal theory is to help us understand how our society understands the institutions that operate within it. It is significant that the starting point of much modern jurisprudence, Herbert Hart’s classic work 8 Carey, The Origins of Concepts (n 2) ch 12; S Mithen, The Prehistory of the Mind (London, Thames and Hudson, 1996). 9 There is disagreement as to how significant this formal learning process is in concept acquisition: Carey, The Origins of Concepts (n 2) 250–254 and E Margolis and S Laurence, ‘Learning Matters: The Role of Learning in Concept Acquisition’ (2011) 26 Mind & Language 507. 10 Murphy, Big Book of Concepts (n 2) 350–2.

22  NW Barber The Concept of Law, contains an apparent commitment to an elucidation of a form of the common understanding in its title.11 Hart’s inclusion of the phrase ‘the concept of’ in the title of his work could mean one of at least three things. First, and most innocuously, Hart could be reminding us that the book will contain a concept of law; the concept of law he has formulated to illuminate the practice of law. This seems unlikely: any book about law will contain such a concept, and the reminder would be unnecessary. Furthermore, the use of ‘the’ rather than ‘a’ suggests something beyond this truism. Second, and more solipsistically, the book could be advanced as Hart’s personal concept of law, the concept of law from his perspective. If this were the case, if all the book aspired to do was to set forth Hart’s understanding, the ambitions of the book would be very limited. The only grounds on which its correctness could be challenged would be one of sincerity: that Hart had not honestly set down his personal concept of law. Third, and most probably, Hart could be understood as seeking to capture our collective understanding of that social institution, the common understanding of law that is a function of the concepts of law found within our community. Once Hart begins the main body of his book, he shows little or no interest in the concept of law, and instead focuses on the nature of law as a social institution12 – suggesting that whilst his methodological commitment to the priority of the common understanding was prominent, it was not very deep.13 Joseph Raz, writing in Hart’s tradition, has been more methodologically explicit and the common understanding may play a greater role in his work than in Hart’s. Raz argues that it is ‘a major task of legal theory to advance our understanding of society by helping us to understand how people understand themselves’.14 Raz writes that he follows Hart in ‘equating complete mastery of a concept with knowledge and understanding of all the necessary features of the objects to which it applies’15 and that, broadly speaking, ‘the explanation of a concept is the explanation of that which it is a concept of’.16 By mastering the concept – which I take Raz to mean a refined form of the common understanding – we reach a superior understanding of the phenomena to which the concept relates. Others writing in the positivist tradition have also endorsed this methodology.17 11 See further N Stavropoulos, ‘Hart’s Semantics’ in J Coleman (ed), Hart’s Postscript: Essays on the Postscript to the Concept of Law (Oxford, Oxford University Press, 2001). 12 See the discussion in B Bix, ‘Joseph Raz and Conceptual Analysis’ (2007) 6(2) American Philosophical Association Newsletter on Law and Philosophy 1 and B Leiter, ‘Legal Realism, Hard Positivism, and the Limits of Conceptual Analysis’ in B Leiter, Naturalizing Jurisprudence (Oxford, Oxford University Press, 2007) 123–5. 13 See further, J Finnis, ‘Hart as a Political Philosopher’ in J Finnis, Philosophy of Law (Collected Essays vol. IV) (Oxford, Oxford University Press, 2011) 259–264 and J Gardner, ‘Law in General’ in J Gardner, Law as a Leap of Faith (Oxford, Oxford University Press, 2012) 274–9. 14 J Raz, ‘Authority, Law and Morality’ in J Raz, Ethics in the Public Domain (Rev ed, Oxford, Oxford University Press, 1995) 237. See also J Raz, ‘Can There Be a Theory of Law’ in J Raz, Between Authority and Interpretation (Oxford, Oxford University Press, 2009) 31: ‘In large measure what we study when we study the nature of law is the nature of our own self-understanding.’ 15 J Raz, ‘Can There Be a Theory of Law?’ in J Raz, Between Authority and Interpretation (n 14) 21. 16 J Raz, ‘Two Views on the Nature of the Theory of Law: A Partial Comparison’ in J Raz, Between Authority and Interpretation (n 14) 55. 17 See, eg, J Coleman, ‘Incorporationism, Conventionality, and the Practical Difference Thesis’ in J Coleman, Hart’s Postscript: Essays on the Postscript to The Concept of Law (Oxford, Oxford

The Significance of the Common Understanding in Legal Theory  23 There are at least two ways in which this approach to legal philosophy could be understood. First, it could amount to a claim that the important features, or the ‘necessary’ features, of an entity should shape the common understanding, and, in turn, shape the concepts we hold that refer to that entity. Now the role of the theorist is to help shape the common understanding, rather than to study the common understanding directly. If this is the case, the common understanding of, and our concepts of, the entity examined are of greatly reduced significance: almost all of the methodological work is being performed by the test of what is ‘important’ or ‘necessary’,18 and we need a further account of the content and application of these criteria. Understood in this way, conceptual analysis has become a distraction, rather than a central part of the process of understanding phenomena such as law and the state. A second interpretation of this methodology – and one that may be more faithful to the apparent priority given to the common understanding in the work of theorists – would treat it as arguing for the primacy of an account of a refined version of the common understanding which acts as a proxy for an understanding of the entities to which that understanding relates. So, we examine the common understanding of law to gain a better understanding of law as a social institution. These statements catch the intuitions that ground an adherence to the common understanding. Perhaps it is, or should be, the business of legal philosophy to reflect on the popular understandings of institutions and the principles that underpin them. And, moreover, perhaps this reflection is the best way of coming to grips with the phenomena to which the popular understanding refers. The reasons why we might wish to study the common understanding of an entity are considered in the middle part of this chapter. But at least one of the attractions of this approach might be that it provides a standard by which the success of an account of law can be gauged. Brian Bix has argued that conceptual claims are not falsifiable – or, he quickly qualifies – are not falsifiable in any obvious way.19 This depends on how we understand conceptual analysis, but if we take it – as Hart and Raz appear to do – as the examination and refinement of the common understanding of phenomena, one of its attractions is that it does open the possibility of falsification. Claims about the common understanding are falsifiable: the assertions they contain may be mistaken. When we are considering individual concepts (my concept of law) the claim may be false if it is insincerely or mistakenly made. When we are considering the community’s understanding (the common understanding of law, a function of individual

University Press, 2001) 106–8 and G Lamond, ‘Coercion and the Nature of Law’ (2001) 7 Legal Theory 35, 38. 18 It is worth noting that the understanding of concepts that regards them as definitions, as resembling sets of rules specifying the necessary and sufficient elements an entity must possess in order to gain inclusion in the category, has largely been abandoned by psychologists working in the area: Murphy, The Big Book of Concepts (n 2) ch 2; E Margolis and S Lawrence, ‘Concepts and Cognitive Science’ in E Margolis and S Lawrence, Concepts: Core Readings (Boston, MIT Press, 1999), 14–27; see also B Leiter, ‘The Methodology Problem in Jurisprudence’ in B Leiter, Naturalizing Jurisprudence (Oxford, Oxford University Press, 2007), 175–81 and D Priel, ‘Jurisprudence and Necessity’ (2007) 20 Canadian Journal of Law and Jurisprudence 173. 19 B Bix, ‘Conceptual Questions and Jurisprudence’ (1995) 1 Legal Theory 465, 467.

24  NW Barber concepts of law) the claim is false if as a matter of fact – and it is a matter of fact – the account does not square with our common understanding. The common understanding is an artefact of a group: it may be vague or disputed in some areas – and then it would be false to claim it is plain or uncontroversial – but in other instances it will be clear and widely understood. The content of the common understanding can sometimes be ascertained through empirical study, through opinion polls and the like. Whilst neither Hart nor his followers have ever been tempted to conduct surveys of this type, he may have been drawn to the claim embodied in the title of his book precisely because it raised the possibility of falsification: perhaps this explains his long-discussed commitment to ‘descriptive sociology.’20 ii. Principles In addition to institutions, legal scholars often spend time reflecting on those principles that animate, or should animate, these entities. The rule of law, the separation of powers, state sovereignty – amongst others – are principles that have a special relationship with the legal institutions of the state.21 We might ask, then, what role the common understanding should play in formulating such principles. When, if ever, is it correct to criticise an account of, for example, the separation of powers for failing to match the common understanding of what that principle requires? Michael Oakeshott provided what might be the most extreme defence of the common understanding’s role in shaping ethical principles – but it is an argument that, in more moderate terms, might tempt many theorists. Starting from the plausible claim, that will be considered further later in this chapter, that an account of ethical principles must be sufficiently close to the common understanding to be intelligible, Oakeshott developed the far less plausible claim that the common understanding is determinative; that an articulation of the common understanding of the principle is the only task open to the scholar.22 For Oakeshott, political ideologies were the product of, rather than the producers of, political activity;23 to generate an account of a political principle it is necessary to examine how a particular community invokes these principles. Changes to political practice are, or should be, generated by reflection on the strands of activity and thought already found within that tradition.24 On this view, the point of reflecting on a political or constitutional principle is to better understand the traditions of a community. Oakeshott draws an analogy with learning a language;25 it is only once we have grasped (and, perhaps, have accepted) the principles subscribed to by a political community that we can engage with them.

20 HLA Hart, The Concept of Law 2nd edn (Oxford, Oxford University Press, 1994) vi. 21 I discuss the nature of principles in Barber, The Constitutional State (n 1) 85–87, and NW Barber, The Principles of Constitutionalism (Oxford, Oxford University Press, 2018) ch 1. 22 See the discussion of Oakeshott in M Loughlin, Public Law and Political Theory (Oxford, Oxford University Press, 1992) 64–83 and G Webber and G Gee, ‘Rationalism in Public Law’ (2013) 76 MLR 708. 23 M Oakeshott, ‘Political Education’ in M Oakeshott, Rationalism in Politics and Other Essays (Indianapolis, Liberty Press, 1991) 61–4. 24 ibid 60. 25 ibid 62; M Oakeshott, ‘Political Discourse’ in M Oakeshott, Rationalism in Politics and Other Essays (n 23) 74–5.

The Significance of the Common Understanding in Legal Theory  25 If Oakeshott’s approach were endorsed, the popular understanding of a normative principle would be of crucial importance. The extent to which the theorist’s account of the principle mapped on to the popular understanding would be one of the most significant measures by which the success of the account could be gauged. iii.  Fear of the Common Understanding Of course, one problem with using the common understanding as the test for the success of an account of an institution or principle is that the common understanding may contain disagreement: people may, and often do, argue about the nature of institutions or the content of principles. The significance of disagreement will be considered later, but sometimes disagreement over the common understanding might be the motivation behind a tempting academic strategy: those papers that survey the uses of a term, illustrate the disagreements over its meaning, and which conclude that the term is too disputed or ambiguous to be used in a ­scholarly way. So, for instance, ‘society’,26 ‘sovereignty’,27 ‘separation of powers’28 and ­‘constitutionalism’29 have all been nominated for academic proscription. Even such apparently innocuous terms such as ‘group’,30 ‘revolution’,31 and ‘intelligence’32 have their challengers. The conclusion that the term should be abandoned allows the theorist to include all possible versions of the principle, or understandings of the institution, without having to choose between them – and without exposing herself to the uncertainties of the common understanding.33 II.  PAYING ATTENTION TO THE COMMON UNDERSTANDING

The previous sections sought to survey the influence of the common understanding on the formulation of accounts of institutions and principles, concluding by discussing the despair that such adherence can cause when the common understanding is disputed or unclear. Those who accord significance to the common understanding are right to do so: there are at least three reasons why we should give weight to the common understanding of a social institution or principle.

26 B Tamanaha, Jurisprudence of Law and Society (Oxford, Oxford University Press, 2001) 206. 27 SI Benn, ‘The Uses of “Sovereignty”’ (1955) 3 Political Studies 109. 28 G Marshall, Constitutional Theory (Oxford, Oxford University Press, 1971) 124. 29 Jo Murkens suggests this as one option in J Murkens, ‘The Quest for Constitutionalism in UK Public Law Discourse’ (2009) 29 OJLS 427, and see also J Waldron, ‘Constitutionalism – A Sceptical View’ in T Christiano and J Christman (eds), Contemporary Debates in Political Philosophy (New York, Wiley, 2009). 30 F Allport, Social Psychology (Boston, Houghton Mifflin Co, 1924) 5–8. 31 See the discussion of RG Collingwood in A Hatto, ‘“Revolution”: An Inquiry into the use of a Historical Term’ (1949) 58 Mind 495, 516. 32 J Flynn, What is Intelligence? Beyond the Flynn Effect (Cambridge, Cambridge University Press, 2007) 20. 33 The claim that a theorist has discovered an ‘essentially contested concept’ relating to a principle or, sometimes, an institution might also be motivated by this fear. On essentially contested concepts, see S Besson, The Morality of Conflict (Oxford, Hart Publishing, 2005) 70–8.

26  NW Barber A.  An Argument from Intelligibility First, all writers must speak in the language of a community; as many have pointed out, language is a shared institution, with common rules and shared beliefs determining correct usage. A theorist who is providing an account of the state or the separation of powers must ensure that her interpretation of these things is sufficiently close to the common understanding to be comprehensible to her audience.34 An account of the state that concluded it was a small mammal with fur would fail for this reason: the theorist has her wires crossed; this is an account of a mouse, not a state. Similarly, though less dramatically, a theorist who argued that the separation of powers was directed towards the question of the division of power between legislatures at different levels in a federation could be criticised for making a simple error: the account is of subsidiarity, not the separation of powers. B.  An Argument from the Constitution of Social Institutions The argument from intelligibility shades into a related, but more interesting, observation about the nature of social institutions.35 Social institutions are the creations of people. It is people who adopt and act upon the rules that constitute these institutions,36 and their beliefs about institutions will affect their nature and operation. A community’s understanding of the nature and point of institutions may shape the relationship of individuals to that institution, the functioning of the institution, and the relationships between individuals within the community. These are all features that might be included within an account of that entity. For instance, a community that regarded law as an institutionalised expression of force may have a different type of legal order to a community that regarded law as a formalised subset of ethical reasoning. Amongst other things, these different understandings of law would affect the way judges exercised their function, and how ordinary citizens responded to legal rules. It would also affect the ways in which people in the community viewed rule-breakers and understood their own legal obligations. Of course, the significance of the common understanding in this example does not just go in one direction: there will also be those in the community who reject this understanding of law. The effect that these dissenters have on the institution will also be framed by the common understanding, and their challenge to the institution may also affect its operation and nature. These differing understandings of institutions may also influence the creation of rules that govern their operation. A society that understood the legal order as

34 Though sometimes surprising assertions can be made for rhetorical effect: see T Endicott, ‘Herbert Hart and the Semantic Sting’ in J Coleman (ed), Hart’s Postscript: Essays on the Postscript to The Concept of Law (Oxford, Oxford University Press, 2001) 51–5. 35 C Taylor, ‘Social Theory as Practice’ in C Taylor, Philosophy and the Human Sciences: Philosophical Papers Volume 2 (Cambridge, Cambridge University Press, 1985) 98–104. 36 For a good discussion, see JR Searle, The Construction of Social Reality (London, Penguin, 1996).

The Significance of the Common Understanding in Legal Theory  27 a form of institutionalised moral reasoning might give judges a wider discretion to modify and develop law than that found in a society which understood it as a set of commands backed by threats. More broadly, the common understanding may also affect the non-legal rules that constitute institutions. The community may apply standards to the conduct of officials that go beyond the law: a judge may have acted in conformity to the law, but may still be regarded by the community as having acted in an inappropriate manner. There may be non-legal social rules defining the proper conduct of the judge, in addition to her legal duties, and these non-legal rules may, in part, have emerged as a result of the common understanding. As I have argued at length elsewhere, it is impossible to provide a satisfying account of a social institution without being attentive to these socially-created rules.37 In short, the common understanding of institutions in the abstract will often have implications for the operation of the local manifestation of that institution: a good account of an institution may need to include an account of how the communities in which it operates understand it. This argument from the constitution of institutions also has implications for the formulation of principles. Whilst it might be possible to comprehend some moral principles – like justice or charity – outside of the context of social institutions, the type of principles that tend to be of direct interest to legal theorists – like sovereignty or the separation of powers – are intimately connected to state institutions and will play out in different ways in different societies. For example, if we understand the separation of powers as a principle of constitutionalism,38 rather than as a description of a category of constitutions or a rule found within a constitutional order, it is unlikely that there is a single, uniquely correct, model of constitution required by that principle.39 The best constitution for a state will inevitably depend, in part, on that state’s current constitution, its history, and its political community. If the separation of powers speaks to all states, it must be flexible: it is likely that, for instance, parliamentary systems and presidential systems are alternative ways of working through its demands. How institutions and communities currently exist – an existence that is partly constituted and shaped by people’s understanding of those institutions – will shape the implications of those principles that apply to those institutions. C.  An Argument from the Wisdom of Crowds A third reason for paying attention to the common understanding of institutions and principles is that it is reasonable to assume that the conceptual divisions drawn by a community over time map differences of significance in the world. By reflecting on our common understandings, we may access a pooled wisdom about the phenomena and ideas to which these understandings refer. This explanation of the importance 37 Barber, The Constitutional State (n 1). 38 See Barber, The Principles of Constitutionalism (n 21) ch 3. 39 I discuss the separation of powers further in Barber, ibid and NW Barber, ‘Self-Defense For Institutions’ (2013) 72 CLJ 558.

28  NW Barber of the common understanding connects with a set of arguments currently popular in political science: those drawing on the ‘wisdom of crowds’.40 One of the most interesting of the many arguments found within writings on the wisdom of crowds is the Condorcet Jury Theorem. This shows that, in some situations, the decisions of large groups can provide more reliable indicators of facts than the decisions of individuals. Condorcet demonstrated that when each individual has a better than 50 per cent chance of giving the correct answer to a question, the more people who are asked a question, the greater the likelihood that the majority of the group surveyed will provide the right answer.41 Cass Sunstein illustrates this with the example of a group of three people each of whom has a 67 per cent chance of giving the right answer. The chance that a majority of that group will choose the correct answer is 74 per cent.42 The more people who are added to this group – provided they beat the 50 per cent mark – the higher the chance the majority will pick the correct answer. If the group is sufficiently large the likelihood of it choosing the right answer can be a practical certainty, even if the chance of any given member getting the wrong answer remains quite high. The Condorcet Jury Theorem could ground an argument for the importance of the common understanding. The common understanding of institutions and principles held by a community has developed over a long period of time. These understandings draw lines: they divide up the world into sets of things and collections of ideas. The divisions are the product of an unknowably large number of decisions made by an unknowably large number of people, more or less consciously. These thoughts take us back to our earlier discussion of the formation of concepts. If concepts are acquired by, and refined through, a mix of our inherent capacities and feedback from others, it is at least likely that the conceptual lines we draw will follow distinctions of importance. The inherent capacities we utilise are likely to have evolved in such a way as to allow us to pick out distinctions that are of significance to us and, relatedly, the feedback from others – perhaps itself motivated by evolved traits – will often be animated by a concern to bring forward significant features. When we critique another’s concept it will often be because we believe it to obscure or elide matters of importance: their conceptual structure is unhelpful, less useful than it could otherwise be. A defence of the evolutionary explanation for the significance of conceptual divisions is far beyond the ambitions of this chapter, but it is, at a minimum, reasonable to assume that when people draw these lines they are seeking to draw distinctions that are of importance to them; that they seek to make their community’s set of concepts as useful as possible, catching distinctions of significance. This might help explain why some sets of concepts are pretty much universal. For instance, people within just 40 On these arguments, see C Sunstein, A Constitution of Many Minds (Princeton NJ, Princeton University Press, 2009) and A Vermeule, Law and the Limits of Reason (Oxford, Oxford University Press, 2009), both of which are discussed in NW Barber, ‘Two Meditations on the Thoughts of Many Minds’ (2009-2010) 88 Texas Law Review 807. The Freudian analyst Bruno Bettelheim advanced a similar ­argument to justify his claim that fairy tales embody complex psychological truths: see B Bettelheim, The Uses of Enchantment (London, Penguin, 1991) 26. 41 See PH Edelman, ‘On Legal Interpretations of the Condorcet Jury Theorem’ (2002) 31 Journal of Legal Studies 327 for a careful discussion of the possible variations on the Theorem. 42 Sunstein, A Constitution of Many Minds (n 40) 25.

The Significance of the Common Understanding in Legal Theory  29 about all communities distinguish between night and day, between the edible and the inedible. These distinctions track important features in the world; all of us need to possess concepts that capture these divisions. If it is accepted that people generally try to draw conceptual boundaries along lines of significance, Condorcet may provide a reason for us to pay attention to these decisions. Assuming that each member of the community has a better than 50 per cent chance of making the correct determination about such matters – or, even if this is not the case, that people’s errors tend to balance each other out – the conceptual lines drawn by a community will give us an insight into distinctions of importance in the world. The study of concepts could then act as a useful proxy for the study of the world: if the community’s determinations are more reliable than the determinations of any given individual, we are more likely to successfully identify matters of importance through this proxy than by observing the world directly. III.  THE LIMITATIONS OF THE COMMON UNDERSTANDING

When reflecting on the limitations of the common understanding, there are two collections of situations that need to be considered. First, there are those situations in which the common understanding turns out to be unclear or disputed. Second, there are those situations in which the common understanding is clear. A.  Where the Common Understanding is Inconclusive The first, and most obvious, difficulty with invoking the common understanding as a test for the success of a theory is that the common understanding is often disputed or unclear. The common understanding is a function of the concepts held by members of the community. These related concepts, captured by the common understanding, will vary in a wide range of respects. Amongst many other possible differences, one person’s concept may extend to examples not included by others, or people may ascribe differing amounts of significance to various features of the concept. Whilst the common understanding does not encompass the totality of the set of similar individual concepts – part of the point of the common understanding is to reign in variation by critiquing the concepts individuals adopt – it will also, ordinarily, leave some latitude for vagueness and disagreement. There will, normally, be a great deal of shared ground between similar concepts held by different people – just about everyone will agree that an apple is an instance of a fruit, for example. But outside these core cases people may become much less certain. Sometimes people’s concepts will have different reaches – and sometimes, indeed, the reach of a concept will vary over time: psychologists who asked a group of people whether an olive was a fruit discovered that over a fifth of the group gave a different response when surveyed a couple of weeks later.43 When theorists appeal to the common understanding it will sometimes be the case that most people have never thought about these questions, and do not

43 Murphy

Big Book of Concepts (n 2) 20.

30  NW Barber have a settled answer to give; there may not be agreement, or disagreement, about whether the common understanding extends to these incidents. A number of disputes in legal philosophy may fall within the realm of uncertainty. Most people will not have an opinion about many of the questions that tax legal philosophers and, indeed, there is little reason for them to formulate one. When Joseph Raz, in the course of an argument for the sources thesis, invokes ‘fundamental features of our understanding of a certain social institution …’44 he may be expecting rather a lot of ‘our understanding’ – though he may be quite right in his analysis of those social institutions referenced by that understanding.45 When hypothetical examples are used in the plea to the common understanding, the problem of popular apathy may be especially strong. One of the standard debates in legal theory surrounds the question of whether law is necessarily coercive. We know that in our world law is always backed by coercion, but Raz conjures the example of a society of angels.46 These folk would require law, to solve the coordination problems that arise on the celestial level, but, being of perfect good nature, they would not require this legal order to be backed by force. Raz bases his claim on conceptual necessity; it is, he claims, ‘logically possible’ that such a legal system could exist, and this is enough to show that law is not necessarily coercive. Kenneth Einar Himma, in sharp contrast, contends that it is ‘a conceptual truth that legal obligation involves coercive state enforcement’ and that the society of angels would have ‘transcended law’.47 Both Raz and Himma rest their arguments on appeals to our collective understanding of law. Raz thinks that we will all recognise the angelic system of coordination as an instance of law, Himma thinks we will all recognise it as a different form of social ordering. But the likelihood is that they will both be disappointed.48 Such an appeal to the common understanding is likely to produce bemusement: most people would struggle to imagine such a world, much less have a view about whether or not it would require a legal order.49 Occasionally the claim that the task of the theorist is the elucidation of the common understanding slides into a slightly more ambitious claim: that it is the task of the theorist to tidy up concepts by making them logical and coherent.50 This

44 J Raz, ‘Legal Positivism and the Sources of Law’ in J Raz, The Authority of Law (Oxford, Oxford University Press, 1979) 50. See the discussion in WJ Waluchow, Inclusive Legal Positivism (Oxford, Oxford University Press, 1994) 113–4. 45 See further R Dworkin, ‘Hart’s Postscript and the Point of Political Philosophy’ in R Dworkin, Justice in Robes (Cambridge MA, Harvard University Press, 2006) 151–2. 46 J Raz, Practical Reason and Norms 2nd edn (Princeton NJ, Princeton University Press, 1990) 158–9. 47 KE Himma, ‘Law’s Claim of Legitimate Authority’ in J Coleman (ed) Hart’s Postscript: Essays on the Postscript to The Concept of Law (Oxford, Oxford University Press, 2001) 308, fn 42. 48 See R Dworkin, ‘Thirty Years On’ in R Dworkin, Justice in Robes (Cambridge MA, Harvard University Press 2006) 201. Raz discusses the problem in the context of authority: J Raz, ‘The Problem of Authority: Revisiting the Service Conception’ in J Raz, Between Authority and Interpretation (Oxford, Oxford University Press, 2009) 130–1. 49 See further, R Sorensen, Thought Experiments (Oxford, Oxford University Press, 1999) 43–6. Brian Leiter makes a similar point about the debate between ‘hard’ and ‘soft’ positivists: B Leiter, ‘Legal Realism, Hard Positivism, and the Limits of Conceptual Analysis’ in B Leiter, Naturalizing Jurisprudence (Oxford, Oxford University Press, 2007) 131–5. 50 J Coleman, ‘Methodology’ in J Coleman and S Shapiro (eds), The Oxford Handbook of Jurisprudence and the Philosophy of Law (Oxford, Oxford University Press, 2002) 313, and J Raz, ‘Can There Be a

The Significance of the Common Understanding in Legal Theory  31 goes beyond the critical edge already applied to individual concepts by the common understanding. Even if the common understanding is vague, and people would generally be unsure whether an example counted as an instance of an institution or version of a principle, and would not regard the holder’s concept as flawed, perhaps the theorist should help the community refine and clarify its conceptual apparatus by developing clearer, better, concepts. To a certain extent, this argument is correct. A good account of a constitutional institution or principle will be grounded in the common understanding of that institution or principle, and will work outwards from this point. The account that is subsequently developed should improve upon, and ideally replace, the common understanding. But it is virtually impossible to undertake this task simply by making use of the resources that are found within the common understanding. It will be rare that the common understanding is internally incoherent or logically flawed. This is not to say that it may not suffer from other sorts of problems. It may include factual or moral errors, errors that it is the task of the theorist to expose. The common understanding of the state might include the belief that a state must be, or must strive to become, coterminous with a national group if it is to remain stable. This is a factual error: there are many stable, successful, and longlasting states that lack such a connection. The common understanding of ‘respect’ might include the notion that people of noble birth are due ‘respect’ for this reason. This is a moral mistake: a privileged birth does not make someone more worthy of respect than those of less privileged origins. In each case, the theorist has work to do – the conscious or unconscious elements of these common understandings need to be exposed and critiqued – but in neither case can this critique be undertaken from within the resources provided by the common understanding itself. A similar problem arises where the community disagrees about the content of the common understanding. The debate between that very loose collection of academics that march under the banner of legal positivism and the equally broad group of ‘natural lawyers’ that supposedly oppose them might be reflected in the community as a whole. The divide between those who are inclined to think of law as a mechanism for regulating behaviour, a mechanism characterised by its technique and not by its point, and those who tend to think of law as a formalised subset of moral reasoning, characterised both by its technique and by its point, is as old as philosophy itself, and can be found in many (maybe most) cultures that possess law. It would not be surprising if this disagreement also manifested in the community that lies beyond the academy. Whether because the philosophical debate has influenced the popular understanding of law, or because the popular understanding of law has framed the philosophical debate, it is quite possible that if the common understanding of the nature of law were surveyed, we would discover that it is split between those who tend towards the positivist side and those who are more tempted by the natural law account. In a bid to tighten up the group of people from whom the common understanding is extracted, some writers have argued that we should focus our attention on

Theory of Law’ in J Raz, Between Authority and Interpretation (Oxford, Oxford University Press, 2009) 31.

32  NW Barber specialist subgroups within the community, subgroups who might be thought to have a special appreciation of the institution referenced by the common understanding. So, for example, Scott Shapiro argues that theorists trying to understand law should examine the understandings of legal professionals and legal scholars.51 This may help avoid some disagreement and uncertainty found in the community at large, but the refinement raises its own problems. The first part of this chapter introduced a study conducted into the tree-related concepts held by biologists and landscapers.52 Though there were many similarities between the concepts held by those within these groups, there were also differences in the ways the groups tended to draw conceptual lines and the features their concepts tended to pick out as significant. The same issue might arise around concepts relating to law. The attention of legal philosophers has often focused on the role of the courts and the role of the judge. This might reflect the interests and concerns of lawyers and academics, who tend to be involved in resolving legal disputes around existing laws. If the understandings of others involved in the legal world were tested, a different account might emerge. Legislators, civil servants, and parliamentary draftsmen might regard legislation as at the core of a legal order, with courts as a secondary feature. Indeed, for ordinary people, who are not involved in the legal world, statutes and legislatures might be at the forefront of their understanding of law – with the court a rare hazard to be steered around, whenever possible.53 Furthermore, whilst those seeking to understand the law have a group of ‘­professionals’ whose beliefs they can study, how should this methodology be applied to other social institutions, such as the state or the nation? An account of the state that just focused on the understanding of politicians or state officials might be profoundly different from the understandings of ordinary state members; perhaps overlooking the less attractive aspects of the institution. And finally, even if we concentrate on the understanding of the institution possessed by a specialised sub-group, it is quite possible that many of the ambiguities and disagreements found in the wider community will be reflected in their understanding. These last paragraphs have explored at length what is, in essence, quite a simple point. If a theorist makes an appeal to the common understanding and the common understanding is vague or is the subject of controversy within the community, the appeal fails for this reason.54 B.  Where the Common Understanding is Clear Sometimes, the common understanding is fairly clear – or, at least, is clear enough in the context of the particular dispute. Even in these cases, though, it is far from

51 S Shapiro, Legality (Cambridge MA, Harvard University Press, 2011) 15–6. Hart makes use of the understanding he imagines to be held by ‘any educated man’: HLA Hart, The Concept of Law 2nd edn (Oxford, Oxford University Press, 1994) 3. 52 Medin, Lynch, and Coley, ‘Categorization and Reasoning’ (n 6). 53 See F Pirie, The Anthropology of Law (Oxford, Oxford University Press, 2013), especially chs 2 and 3, for a study of a range of different understandings of law found in different branches of scholarship. 54 See further the discussion in Priel, ‘Jurisprudence and Necessity’ (n 18) 178–84.

The Significance of the Common Understanding in Legal Theory  33 obvious why this should determine the success of an account of an institution or principle. Occasionally, of course, it is the common understanding that is the focus of study. The question of how a community conceives of law, or understands the demands of democracy, is an inherently interesting one. If a social scientist is seeking to provide an account of how a community understands these things, it is always a fair criticism to argue that her account misses or misstates elements of that understanding – though normally we would expect that such criticism would be grounded in some evidence of the community’s views, and not rest entirely on the critic’s intuitions. Generally, though, our interest is not primarily in the common understanding, but rather in the institutions or principles that the common understanding is directed towards.55 As John Finnis waspishly writes, ‘I think theories of law are concepts of law, not theories of concepts.’56 Even when the common understanding is clear, the arguments for its significance – set out in the first part of this chapter – are far from conclusive. i.  The Argument from Intelligibility Revisited The first argument for the significance of the common understanding invoked the need to render an account intelligible: if the theorist’s account diverges too sharply from the common understanding there is a suspicion, at least, that she is not really talking about the same entities as those referenced by the community. But the simple demand that the account be intelligible is not all that exacting. People hold differing versions of concepts that relate to similar entities without rendering communication between them impossible. Sometimes, the distinctions may be compatible. The understanding of the state held by a lawyer and a cartographer may identify different features of that institution as important, and yet be capable of being combined into a single, richer, account. Sometimes, the theorist will seek to provide a radical interpretation of social phenomena, an interpretation that purports to bring out important features of the practice of which members of their community were unaware. For example, Lenin claimed that an important feature of the state is that it acts as a mechanism for class oppression; this is part of its true nature.57 People within his community, he asserted, did not realise this because the powerful cunningly hid the truth from them. Lenin’s account – if correct – necessarily differed from the popular understanding of the state. But it is unlikely that it would raise problems of intelligibility. Those hearing the account would recognise that Lenin was providing an interpretation of a set of phenomena they would identify as states, even if he were making a claim about those entities that differed from their understanding.

55 Sorensen, Thought Experiments (n 49) 31–2. 56 J Finnis, ‘Law and What I Truly Should Decide’ (2003) 48 American Journal of Jurisprudence 108, 125 and J Finnis, Natural Law and Natural Rights 2nd edn (Oxford, Oxford University Press, 2011) 426–7. Brian Leiter makes a similar observation from a profoundly different starting point: B Leiter, ‘Science and Methodology in Legal Theory’ in B Leiter, Naturalizing Jurisprudence (Oxford, Oxford University Press, 2007) 189–90. 57 V Lenin, The State and Revolution, transl. R Service (London, Penguin, 1992) 7–9. See further K Dyson, The State Tradition in Western Europe: A Study of an Idea and Institution (Oxford, Oxford University Press, 1979) 104–7.

34  NW Barber The demands of intelligibility are weaker still when applied to normative principles, such as the separation of powers, or democracy. To an extent these principles are what they ought to be; people routinely argue about their content whilst recognising that they are arguing about the content of a single principle. Principles like the rule of law, the separation of powers, or democracy, are collections of moral reasons that direct us to act in certain ways. As Ronald Dworkin argued, we agree that such principles relate to a particular value – or collection of values – and we agree, furthermore, that their correct application will turn on a correct understanding of these values.58 Those who claim, for example, that democracy requires the vindication of the will of the majority, or – in contrast – requires the protection of certain minority group rights, are making rival claims about our obligations to others within our community. To be intelligible, any understanding of democracy must relate to the connection between the individual and collective decision-making. But within these broad boundaries, the best understanding of democracy will depend on the truth and weight of the moral reasons that apply to us in these situations. Each of the competing camps agrees, or should agree, that the best understanding of democracy is the one that makes it the most morally attractive version of the principle it can be. They disagree, though, over which interpretation that is. In the earlier part of this chapter, Michael Oakeshott’s claim that ethical principles should be treated like social institutions, as the creatures of their cultures, was introduced. This approach to the interpretation of principles stands in stark contrast to that given in the previous paragraph. Of course, there is some value in Oakeshott’s approach: we often find incarnations of ethical principles within cultures, a community’s understanding of the principle that is specific to a particular time and place. Consequently, it is possible to provide histories of constitutional principles that chart their origins and development. M. J. C. Vile, for example, provides an illuminating account of the separation of powers, showing how it emerged from the ancient idea of mixed government and tracing the forms the principle has taken in the work of scholars and the constitutions of states.59 But reflection on ethical principles should not be abandoned to the intellectual historians. For most of the time we treat ethical principles as dynamic things, which can be argued over and improved, and not just as mere historical artefacts; indeed, it is the product of this activity that provides the material for the historian to study. We are debating the best interpretation of these ethical principles, and this debate may take us a long way from the understanding that prevails in our community. To say that, for example, John Rawls’ A Theory of Justice departed from the utilitarian consensus that held sway when it was published would not, normally, be thought to be either a point of praise or point of criticism of the book. It differed from the common understanding – but if Rawls could convince his audience that this was a more morally attractive account of justice, then the common understanding should shift towards Rawls’ understanding. Oakeshott might reply to this by contending that people are just mistaken. They are deluded if they fail to perceive ethical principles as cultural artefacts; that novel 58 Dworkin, ‘Hart’s Postscript’ (n 45) 145–55. 59 MJC Vile, Constitutionalism and the Separation of Powers 2nd edn (Indianapolis, Liberty Fund, 1998).

The Significance of the Common Understanding in Legal Theory  35 interpretations of such principles are errors, and that the only proper way of understanding them is through historical or cultural study. If Oakeshott were to run this argument he would be forced to make the rather uncomfortable claim that whilst the bulk of people cannot be mistaken about the content of ethical principles they can be, and almost universally are, mistaken about their nature. But, to be fair to Oakeshott, there is no escape from arguments of this type. If we accept the premise that ethical principles are merely social artefacts, the creations of particular cultures at particular times, the form of study Oakeshott advocates is the only form of analysis possible. But it is not a premise many of us would be inclined to accept. ii.  The Argument from the Social Construction of Institutions Revisited The second set of arguments for the significance of the common understanding argued that a community’s understanding of an institution normally played a part in the constitution and operation of that body. This claim was always a limited one. It asserted that some aspects of social institutions – and, connectedly, principles – were determined by people’s beliefs and attitudes. It is possible, even probable, that there are also important features of social institutions of which people are unaware. Some of the rules that constitute the institution may be so deeply internalised that people are generally unaware of their operation.60 In these cases even those following the rule might not realise that their behaviour was governed by it. And it is possible that social institutions could possess important characteristics that people systematically fail to recognise. Lenin may not have been right about the state as an instrument of, and being animated by, class oppression – but it is perfectly possible that some social institutions exist and prosper for reasons that those taking part in them fail to comprehend.61 iii.  The Argument from the Wisdom of Crowds Revisited The third set of arguments in support of the common understanding invoked the supposed wisdom of crowds; the capacity of a group to correctly identify facts more reliably than any single individual. The Condorcet Jury Theorem, presented in the first half of this chapter, is certainly attractive, but there are a number of serious limitations on its usefulness – limitations which are so profound as to render the Theorem almost worthless as an argument for the priority of the common understanding in most cases. The first, and most obvious, constraint on the utility of the Condorcet Jury Theorem is that the members of the group surveyed must have a greater than 50 per cent likelihood of picking the correct answer.62 If, in contrast, each member of the 60 Barber, The Constitutional State (n 1) 110–3. 61 The classic example is the rain-dance: a social practice that its participants believe they undertake because it creates rain, but which anthropologists argue is practised because it brings the community together in times of crisis. See CG Hempel, ‘The Logic of Functional Analysis’ in M Martin and L McIntyre (eds), Readings in the Philosophy of Social Science (Boston, MIT Press, 1994). 62 Sunstein, A Constitution of Many Minds (n 40) 28–9. See also J Waldron, ‘Unintentional Legislation’ in J Waldron, Law and Disagreement (Oxford, Oxford University Press, 1999) ch 6.

36  NW Barber group has a less than 50 per cent chance of choosing correctly, the Theorem shows that the more members who vote on an issue, the smaller the chances of the group choosing successfully. A sufficiently large group consisting of individuals each of whom has a 49.9 per cent chance of reaching the correct answer is almost guaranteed to reach the wrong outcome, even though each member has a reasonable chance of choosing correctly. This hazard with Condorcet is mitigated if we can check the previous decisions of the group or can assess the reliability of people in general. If we know that a particular group has done well in answering a particular question, or we are sure that people in general have over a 50 per cent chance of making the right decision on a particular issue, Condorcet provides a solid argument for preferring the judgement of the group to our own. A plausible case can be made for the Condorcet Theorem operating at the broadest level of a community’s conceptual structure. But there is far less reason to assume that our community’s common understanding of the detail of social institutions is constituted by decisions that pass this standard. As Cass Sunstein and Adrian Vermeule have shown, there is a collection of common scenarios in which people can herd towards error. Amongst the most important of these are those caused by information cascades and those caused by undeserved reputation. The simplest example of an information cascade occurs when members of a group are requested to deliver their opinion on a question publicly and sequentially.63 The first person to deliver her opinion does so free of influence, but the remaining members will be affected by the verdicts of their fellows. If they believe that the earlier voters had information they lacked – or even that the earlier voters were equally wellinformed, but had reflected on the question with more care – the later voters might resolve to follow the pronouncements of their earlier colleagues. If the earlier voters were mistaken, their error will then cascade through the group. Sometimes information cascades are fragile things – they just require a couple of independent-minded spirits to bring them to an end – but sometimes they are very powerful indeed. At some point an information cascade can develop into a form of groupthink, where to question the correctness of the asserted fact challenges a rule of the group. It is no longer the case that the later voters in the group concur with the earlier voters because they believe the earlier voters knew something they did not, but, rather, to disagree would now challenge the coherence and strength of the group. In this situation the phenomenon of ‘groupthink’ may arise.64 It may become an aspect of being a good group member that the member agrees with the pronouncements of those who have gone before her. Might groupthink distort the common understanding of a social institution and, in consequence, give those who focus on this understanding a distorted impression of the world? It is certainly possible. It is often asserted, for instance, that states require a special connection to a national group or religion. Sometimes, the people making

63 Sunstein, A Constitution of Many Minds (n 40) 88–91; C Sunstein, Why Societies Need Dissent (Cambridge MA, Harvard University Press, 2003), ch 3; Vermeule, Law and the Limits of Reason (n 40) 73–7. 64 I Janis, Groupthink 2nd edn (Boston, Houghton Mifflin, 1982) ch 8. Discussed further in Barber, ‘Two Meditations’ (n 40) 109–13.

The Significance of the Common Understanding in Legal Theory  37 this assertion have a vested interest in hoping that this is the case: they seek to defend the national or religious ties of their own state by arguing that such ties are unavoidable. Similarly, some groups have an interest in presenting social institutions in a bad light, insisting that they are unavoidably abusive or oppressive. The common understanding of law found in Marxist groups or the Critical Legal Studies movement presents it as an inevitably oppressive system, shoring up class division and gender discrimination. Our discussion of groupthink shades into a further, and similar, danger: the risk of bias caused by reputation.65 On some occasions the bias caused by reputation will amount to a form of information cascade: people make decisions about the content of the common understanding on the assumption that a particular person has special knowledge of an area, and so defer to her judgment. On other occasions the bias caused by reputation is distinguishable from an information cascade – for instance, where people are led to agree with another because of a desire to please that person or because that person is (wrongly) believed to have the authority to determine a particular issue. For theorists who use the common understanding as determinative of an account of an institution, the problems caused by reputation are intriguing, to say the least. A scholar whose account of an institution is widely read and popular may push the common understanding towards her own account. So, for example, by this criterion, a very successful book such as The Concept of Law may be rendered correct simply by virtue of its popularity. Having read the book, people may come to endorse Hart’s elucidation of the common understanding (that is, the concept of) law, and, in so doing, make his book a more accurate account of what it purports to describe. In this context, indeed, nothing succeeds like success.66 Use of ‘wisdom of crowds’ arguments becomes particularly problematic when the question addressed to the group relates to moral facts.67 Applying the Condorcet Jury Theorem to moral facts, assuming it is appropriate to talk of ‘facts’ in this context, raises a number of problems. Some of these relate to the nature of morality and some to our capacity to possess knowledge of moral truths. First, some moral philosophers would reject the claim that there are unique moral truths to be found. For relativists, the best that can be said of a moral claim is that it is true relative to an ethical system. For subjectivists, the best that can be said of a moral claim is that the person advancing the claim believes it to be true; that is, she sincerely believes the claim she makes. If either of these metaphysical accounts is correct, the Condorcet Jury Theorem could, at best, help us assess whether a purported moral fact really did fit a given ethical framework, or whether the person advancing the claim was sincere. The Theorem could not help us decide which of several rival moral claims should be preferred: a number of conflicting or even contradictory moral beliefs could all be ‘correct’ given these understandings of the nature of morality. 65 Sunstein, A Constitution of Many Minds (n 40) 91–2; C Sunstein, Why Societies Need Dissent (Cambridge, Mass, Harvard University Press, 2003) ch 4. 66 As Joseph Raz writes, ‘To the extent legal theorists acquire influence their views tend to be selfverifying’ in J Raz, ‘On the Nature of Law’ in J Raz, Between Authority and Interpretation (Oxford, Oxford University Press, 2009) 98. 67 Both Vermeule and Sunstein contend that Condorcet can be applied to moral facts: Vermeule, Law and the Limits of Reason (n 40) 8; Sunstein, A Constitution of Many Minds (n 40) 98–9.

38  NW Barber The fundamental question of the plausibility of these positions, that is, whether relativism or subjectivism is correct, cannot be addressed in this chapter. But these grand metaphysical questions lead us to a second, more modest, but also more powerful, objection to the use of the Theorem in the context of moral facts. Even if it is accepted that morality is an objective affair, that there are truths to be found about morality, what it means to have knowledge of a moral fact differs from what it means to have knowledge of a physical fact. When a physical fact is asserted there are steps by which this proposition can be proved or disproved. When an assertion is made about the length of a river, the weight of an ox, or the number of beans in a jar, processes exist which allow us to test the claims: the river can be measured, the ox can be weighed, the beans can be counted.68 An assertion of a moral fact, in contrast, cannot be proved in the same way. The proposition that torturing the innocent is wrong, for example, is true, but if it is disputed there is nothing that can be weighed or measured to resolve the dispute. Moral facts cannot be demonstrated to be true in the same way that physical facts can be proven. Finally, groups can make systematic moral mistakes.69 Groups are prone to certain types of moral error. They are often prejudiced against those who are not members of the group or those who are members of the group but deviate from their prototypical model of a group member.70 Perhaps this prejudice helps explain, in part, the common understanding of democracy as unconstrained majoritarianism. As we have seen, groups can adopt irrational beliefs, whether as the result of an information cascade, a misplaced sense of the reputation of a leader or for some other reason. Groups can have moral ‘blind-spots’. An otherwise virtuous group might have strange prejudices against particular life-styles, or be slow to identify the abuses undertaken by their state as torture. It would be a serious error to subordinate our own beliefs about right and wrong to those of the group. When our moral beliefs diverge from the consensus we should certainly be willing to reconsider, but sometimes we will be correct to stick to our convictions: the group is wrong, and we are right. The common understanding of the group cannot tell us when that common understanding is mistaken. Arguments that invoke the common understanding of an ethical principle consequently receive little support from the ‘wisdom of crowds’. Even if we assume that when people make decisions about the content of principles like democracy and the separation of powers they are trying to shape these principles to fit the moral reasons that should animate them, there is no guarantee that they will make the right ­decision – or that their errors will balance out. When we are providing an account of these principles that aims to guide those working in the political sphere, the common understanding of the principle provides, at best, a place to start from and a weak constraint on our interpretation.

68 J Surowiecki, The Wisdom of Crowds (New York, Little, Brown, 2004) ch 1. 69 Sunstein, A Constitution of Many Minds (n 40) 169–71; Vermeule, Law and the Limits of Reason (n 40) 47. 70 For discussion of this phenomenon, see MA Hogg, ‘Social Categorisation and Group Behaviour’ in MA Hogg and S Tindale (eds), Blackwell Handbook of Social Psychology: Group Processes (Oxford, Blackwell, 2001).

The Significance of the Common Understanding in Legal Theory  39 IV. CONCLUSION

This chapter has pursued one strand of argument within the debates surrounding the methodology of legal theory. The significance of the common understanding is argued for expressly by some, but is assumed by a great many more writers, some of whom regard it as the main, perhaps even the only, criterion for a successful account of an institution or principle. This chapter has not claimed that the common understanding is unimportant. Most superficially, when the theorist is writing for the community she is studying, locating the account within the common understanding will ensure that it is intelligible to that community. Furthermore, the common understanding will also shape the institutions studied, affecting the constitution and operation of these entities, and, as the requirements of many of the principles that legal theorists study are partly determined by these institutions, the common understanding will affect these principles, too. Finally, it is at least reasonable to suppose that the lines that are drawn by the common understanding sometimes track matters of importance in the world. For these reasons, those who invoke the common understanding in debates around institutions and principles may be raising challenges that deserve an answer: if your account of a state excludes Taiwan, you need to be able to provide a good reason for its exclusion. The common understanding is an appropriate place to begin an account of an institution or principle, and does impose some constraints on the account produced. However, this chapter has also argued that each of the reasons for adherence to the common understanding is limited, and it may often fail to possess the analytical weight some of its supporters ascribe to it. The constraints of intelligibility are relatively weak: people are easily able to comprehend radical or surprising accounts of social institutions. The role of rules and social dispositions in shaping institutions is significant, but cannot provide a complete account of their nature. And whilst it may be reasonable to suppose the conceptual lines drawn by the community track matters of importance, it would be rash to treat the lines drawn as conclusive. There are many reasons why the community might be wrong. This chapter raises and leaves open an obvious question: if the common understanding is not the sole and definitive test of the success of a theory, what further criteria must a successful account meet? This broad, and vitally important, question cannot be answered here.71 This chapter has focused on one aspect of the methodology that should be applied to the production of an account of an institution or principle, and discussed one of the criterion that test the success of such theories. The common understanding may be a good place to start an account of an institution, but it is a bad place to conclude such an account – and whilst it does constrain the account that is produced, these constraints give a great deal of latitude to the theorist.



71 Though

I do explore it at some length in Barber (n 1).

40

2 In Defence of Traditional Methodologies JEFFREY GOLDSWORTHY* I. INTRODUCTION

T

here is no single methodology that must be used by constitutional lawyers and theorists. They may pursue questions that are not purely legal, but historical, comparative, philosophical, economic, evaluative and so on, using different methodologies as appropriate. To answer even purely legal questions, about what the law is in relation to some issue, different methodologies are appropriate, depending on what kind of law is involved. Constitutional law in Anglophone legal systems includes fundamental norms that are customary (in Britain and New  Zealand, establishing parliamentary sovereignty, parliamentary privileges and executive prerogatives), written constitutions (except in those two countries), statutes and common law, and the operation of all of these is intertwined with constitutional conventions. The orthodox understanding is that there is no single methodology for establishing the contents of customary norms, written constitutions, statutes and the common law, let alone constitutional conventions. The methodologies needed to ascertain the law clearly depend on the nature of law: on whether law is entirely or ultimately a matter of fact alone, as legal positivists argue, or of some combination of fact and value. That is the most basic, and contentious, question in legal philosophy. But answering it does not guarantee a single, unitary methodology; for example, it might be that customary norms, written constitutions and statutes are best understood in legal positivist terms, whereas the common law is best understood in non-positivist terms. Even if legal positivism is true in general, and all law is ultimately a matter of fact, customary norms, written constitutions, statutes and common law are arguably constituted by such different kinds of facts that different methodologies are required in each case. Some theorists claim that the contents of all four types of law can be ascertained and clarified only through a single non-positivist methodology, a process they call

* Thanks to Stuart Lakin for comments on an earlier version of this chapter.

42  Jeffrey Goldsworthy ‘interpretation’ that is ultimately guided by moral judgments.1 In Britain, so-called ‘interpretivists’ tend to equate this master methodology with the methodology of the common law, which they regard as the most fundamental body of law.2 They say that it is the source of the authority of customary constitutional norms, written constitutions, and statutes, and also limits their authority, thereby helping to determine their contents. Answering any legal question therefore depends partly on identifying the most fundamental principles of the common law, which underpin the legal system as a whole. Hence the term ‘common law constitutionalism’. This chapter will defend the orthodox understanding that constitutional law requires multiple methodologies, which it will briefly describe, against the extravagant claims for morally driven interpretivism made by common law constitutionalists. But an important qualification should be noted. What I call the orthodox understanding was orthodox until relatively recently, but is now at risk of being eroded due to various challenges whose nature and consequences will be discussed in sections V, VI and VII. II.  THE ORTHODOX UNDERSTANDING

A.  Common Law The orthodox understanding of common law is relatively simple and straightforward. It regards the common law as a body of doctrines, principles and rules made and developed by the judiciary, which is subordinate to the constitutional doctrine of parliamentary sovereignty, written constitutions and statutes.3 The methodology employed by judges in ascertaining, clarifying and developing the common law includes consideration of general principles identified as underlying particular rules and doctrines. But Ronald Dworkin’s interpretivist thesis that judges aspire (in principle) to identify and apply a grand, all-encompassing theory, providing the best possible moral justification of the entire vast body of the law – including written constitutions and statutes – is implausibly far-fetched.4 So is his related idea that judges never need to make new law, because if they strive hard enough to develop that grand theory, they can always discover existing legal principles – even if previously unrecognised – capable of settling any legal controversy.5 This perpetuates the old declaratory theory of law, that at least since the 1970s has rightly been rejected as a ‘fairy tale’.6 It disguises judicial law-making, which might be perfectly legitimate or even necessary, as law-finding. 1 N Stavropoulos, ‘Legal Interpretivism’, Stanford Encyclopedia of Philosophy, https://plato.stanford. edu/entries/law-interpretivist/. 2 T R S Allan, The Sovereignty of Law; Freedom, Constitution and Common Law (Oxford, Oxford University Press, 2013), 3, 32, 121, 132, 135. 3 The doctrine of parliamentary sovereignty is not part of the common law in the usual modern sense of judge-made law: see the text to and following n 80 below. 4 R Dworkin, Law’s Empire (London, Fontana, 1986). 5 R Dworkin, Taking Rights Seriously (Cambridge, Harvard University Press, 1977), ch 4. 6 Lord Reid, ‘The Judge as Lawmaker’ (1972) 12 Journal of the Society of Public Teachers of Law (New Series) 22.

In Defence of Traditional Methodologies  43 Dworkin relied heavily on how he thought judges actually decide ‘hard cases’, in which the law is prima facie obscure or under-determinate.7 His interpretive theory has some plausibility in relation to the common law, because judges do infer general principles from particular decisions and rules, and use those principles in deciding novel cases, overruling incompatible precedents, and generally developing the common law.8 But the theory was never plausible in the case of written constitutions and statutes. As we will see, it is inconsistent with the way judges themselves have traditionally described their primary objective in interpreting statutes.9 They have consistently spoken of giving effect to the intention of the legislature, rather than to a holistic interpretation that makes the law as a whole morally the best that it can be. It is true that they often take into consideration other legal norms, including common law principles; but as we will also see, they have almost always justified this by appealing to likely legislative intentions or purposes.10 Dworkin himself, in the years after writing Law’s Empire, adopted a kind of originalist approach to constitutional law.11 Even Nicos Stavropoulos, also an interpretivist, concedes that interpretivism is not the orthodox view of statute law.12 Because the common law is judge-made, and subject to continuing judicial ­development, the judiciary has extensive discretion with respect to how rapidly and radically to develop it. Judges must weigh up the relative importance of stability and predictability on the one hand, and of justice and good governance on the other. They usually prefer to develop the law cautiously, incrementally, and – to make developments more predictable – consistently with principles explicitly or implicitly relied on in earlier judgments, or mentioned in obiter dicta, albeit often in expansionary or novel ways. The distinction between rationes decidendi and obiter dicta is of some importance, but often difficult to draw in practice. Some areas of law – modern administrative law being a primary example – have been developed much more rapidly and sometimes radically than other areas. Because this is largely a matter of judicial discretion and temperament, not much more can usefully be said in a brief summary of common law methodology. But as we will see, when there is overlap between the common law and the other forms of law, or confusion about the boundaries between them, judicial law-making can expand beyond its proper limits. B.  Written Constitutions, Legislative Supremacy and Statutory Interpretation The orthodox understanding regards the common law, including its deepest principles, as subordinate to written constitutions because they are the supreme law, which 7 Dworkin, Law’s Empire (n 4) ch. 1; Dworkin, Taking Rights Seriously (n 5) ch 4. 8 See S Perry, ‘Judicial Obligation, Precedent, and the Common Law’ (1987) 7 Oxford J Legal Studies 215; M Walters, ‘The Unwritten Constitution as a Legal Concept’, in D Dyzenhaus and M Thorburn, The Philosophical Foundations of Constitutionalism (Oxford, Oxford University Press, 2016), 33. 9 Section II.B, below. 10 ibid. 11 See J Goldsworthy, ‘Dworkin as an Originalist’ (2000) 17 Constitutional Commentary 4; L Alexander, ‘Was Dworkin an Originalist?’, in W Waluchow and S Sciaraffa (eds), The Legacy of Ronald Dworkin (Oxford, Oxford University Press, 2016), 299. 12 Stavropoulos, ‘Legal Interpretivism’ (n 1) section 2, ‘The Orthodox View’, last viewed 13 March 2020.

44  Jeffrey Goldsworthy judges along with other officials and citizens are legally bound to obey. Common law norms, just like statutory provisions, cannot be legally valid or enforceable if they are inconsistent with the written constitution.13 As three Justices of the Australian High Court have said: ‘If the Constitution, expressly or by implication, is at variance with a doctrine of the common law, the latter must yield to the former … it is clear that the Constitution must prevail.’14 The common law is also subordinate to statute law, because of the doctrine of legislative supremacy (or sovereignty, as it is called in Britain), which is part of the constitutional law – whether unwritten or written – of all Anglophone countries. In Britain, the doctrine of parliamentary sovereignty maintains that every statute enacted by Parliament is legally valid, regardless of its content.15 Statutes are superior to and override inconsistent common law, not vice versa. As eight members of the UK Supreme Court affirmed in Miller No 1, ‘it is not open to judges to apply or develop the common law in a way which is inconsistent with the law as laid down in or under statutes, i.e. by Acts of Parliament’.16 The doctrine of parliamentary sovereignty is accepted in its fullest sense only in Britain and New Zealand. But even in Australia, which has written constitutions, it is accepted that there are no limits to the laws that Parliaments can validly enact other than those expressly or impliedly imposed by those constitutions. In other words, the legislative authority of Australian Parliaments is not legally constrained by extra-constitutional norms, including the common law.17 The same is true in the US. American judges and scholars have often observed that the judiciary has no authority to hold a statute invalid except on the ground that it violates some norm expressed or implied by one of their written constitutions.18 Some have explicitly concluded that, when acting within constitutional boundaries, their legislatures are as sovereign as the British Parliament.19 The doctrine of legislative supremacy heavily influences how statutes must be i­nterpreted. It has played a pivotal role in recent American debates,20 with

13 For the position in Australia, see J Goldsworthy, ‘The Constitution and its Common Law Background’ (2014) 25 Public Law Review 265, 269–70. 14 Theophanous v Herald & Weekly Times Ltd (1994) 182 CLR 104, 126 (Mason, Gaudron and Toohey JJ). 15 See J Goldsworthy, The Sovereignty of Parliament, History and Philosophy (Oxford, Clarendon Press, 1999), and J Goldsworthy, Parliamentary Sovereignty, Contemporary Debates (Cambridge, Cambridge University Press, 2010). 16 R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5, para 42, per Lord Neuberger, Lady Hale, Lord Mance, Lord Kerr, Lord Clarke, Lord Wilson, Lord Sumption and Lord Hodge. For examples of statutes overriding the common law after controversial judicial decisions, see M Gordon, Parliamentary Sovereignty in the UK Constitution (Oxford, Hart Publishing, 2015), 23, n 49. 17 Goldsworthy, The Sovereignty of Parliament (n 15) 1. 18 J Goldsworthy, ‘The Real Standard Picture, and How Facts Make it Law; a Response to Mark Greenberg’ (2019) 64 The American Journal of Jurisprudence 163, 170–71. 19 J Kent, Commentaries on American Law 10th edn (Boston, Little, Brown & Co., 1860), 503; R Pound, ‘Common Law and Legislation’ (1908) 21 Harv Law Review 383, 392. See also T M Cooley, A Treatise on the Constitutional Limitations Which Rest Upon the Legislative Power of the States of the American Union (Boston, Little, Brown & Co., 1868), 87. 20 See, eg, D A Farber, ‘Statutory Interpretation and Legislative Supremacy’ (1989) 78 Georgia Law Review 281, esp 281 n 3; W Eskridge, ‘Spinning Legislative Supremacy’ (1989) 78 Georgia Law Review 319 (1989); E O Correia, ‘A Legislative Conception of Legislative Supremacy’ (1992) 42 Case Western

In Defence of Traditional Methodologies  45 textualists and purposivists both claiming to respect the doctrine more reliably than the other.21 This claim is often expressed in terms of courts being ‘faithful agents’ of the legislature,22 which Cass Sunstein describes as ‘the most prominent [American] conception of the role of courts in statutory construction.’23 As John Manning explains: In a constitutional system predicated upon legislative supremacy (within constitutional boundaries), judges – as Congress’s faithful agents – must try to ascertain as accurately as possible what Congress meant by the words it used.24

This interpretive rule has for many centuries been generally upheld by Anglophone courts, which have repeatedly asserted that the primary object of statutory interpretation is ‘to give effect to the intention of the [law-maker] as that intention is to be gathered from the language employed having regard to the context in connection with which it is employed’.25 It was recently restated by Lady Hale, with the agreement of four other members of the UK Supreme Court: The goal of all statutory interpretation is to discover the intention of the legislation. That intention is to be gathered from the words used by Parliament, considered in the light of their context and their purpose … The question [here] is whether, in the light of the words used, their context and the purpose of the legislation, Parliament must have meant the Crown to be bound.26

This general interpretive rule has been described as ‘the only rule’, ‘the paramount rule’, ‘the cardinal rule’ and ‘the fundamental rule of interpretation, to which all others are subordinate’.27 It is itself constitutional in character, because it concerns the relationship between the authority of the legislature and that of the judiciary.

Reserve Law Review 1129; E M Maltz, ‘Rhetoric and Reality in the Theory of Statutory Interpretation: Underenforcement, Overenforcement, and the Problem of Legislative Supremacy’ (1981) 71 Boston University Law Review 767. 21 T A Aleinikoff, ‘Updating Statutory Interpretation’ (1988) 87 Michigan Law Review 20, 23; A C Barrett, ‘Substantive Canons and Faithful Agency’ (2010) 90 Boston University Law Review 109, 112–13 and 157–58; J F Manning, ‘Without the Pretence of Legislative Intent’ (2017) 130 Harvard Law Review 2397, 2425–27. 22 See, eg, Barrett, ‘Substantive Canons and Faithful Agency’ (n 21) esp 112–17; J F Manning, ‘Textualism and the Equity of the Statute’ (2001) 101 Columbia Law Review 1, 6–7, 9–22; R H Fallon Jr, ‘Three Symmetries Between Textualist and Purposivist Theories of Statutory Interpretation – and the Irreducible Roles of Values and Judgment Within Both’ (2014) 99 Cornell Law Review 685, 685–86. 23 C R Sunstein, ‘Interpreting Statutes in the Regulatory State’ (1989) 103 Harvard Law Review 405, 415; see also Fallon, ‘Three Symmetries’ (n 22) 685–86. 24 J F Manning, ‘Textualism and Legislative Intent’ (2005) 91 Virginia Law Review 419, 419 (2005). See also Sir P B Maxwell, On The Interpretation of Statutes 2nd edn (London, Maxwell & Son, 1883), 1: ‘Statute law is the will of the Legislature; and the object of all judicial interpretation of it is to determine what intention is conveyed, either expressly or by implication, by the language used.’ 25 Attorney-General v Carlton Bank [1899] 2 QB 158, 164 (Lord Russell). 26 R (on the application of Black) v Secretary of State for Justice [2017] UKSC 81, paras 36 and 37. The final sentence bars any argument that by ‘the intention of the legislation’ Lady Hale meant something different from ‘the intention of the legislature’. See also the observations of the Court of Appeal in R (Miller) v The Prime Minister [2019] EWHC 2381 (QB), para 61, quoted at n 139 below. 27 Respectively, Sussex Peerage Case (1844) 8 ER 1034, 1057 (Tindall CJ); Attorney-General (Canada) v Hallet & Carey Ltd. [1952] AC 427 (Lord Diplock); Mills v Meeking (1990) 169 CLR 214, 234 (Dawson J); Amalgamated Society of Engineers v Adelaide Steamship Co Ltd (1920) 28 CLR 129, 161 (Higgins J).

46  Jeffrey Goldsworthy The law provides an array of further principles, maxims or canons of statutory interpretation. The legal meaning of a statute could reasonably be defined as the result of applying all relevant interpretive principles. But there is theoretical debate about the nature of many of them. Some regulate the kinds of evidence of legislative intentions that is admissible in a court; the admissibility of ‘legislative history’ has been of particular concern, and has varied over time and across common law jurisdictions.28 Other principles, such as the so-called ‘linguistic canons’, directly assist in clarifying the legislature’s probable communicative intentions.29 Yet others require a higher than usual standard of proof of certain kinds of legislative intentions, especially intentions to encroach upon traditional common law doctrines and rights.30 Sometimes the meaning that the lawmaker intended to communicate remains obscure even after the judge has exhausted all admissible evidence of it, due to ambiguity, vagueness, self-contradiction or gaps. Some interpretive principles arguably function as tie-breakers, to resolve indeterminacies in intended meaning in favour of common law doctrines and rights. These principles help to supplement the meanings of statutes that would otherwise be insufficiently determinate to resolve a legal dispute.31 If indeterminacies remain after they have been resorted to, judges must exercise lawmaking discretion, guided by their sense of justice and the public interest. Other interpretive principles allow, in exceptional circumstances, the partial correction or rectification of linguistic meaning. Examples in which this is (at least arguably) legally permissible in common law jurisdictions include: (i) complying with a legislative directive, in a subsequent statute, to attribute to earlier statutes meanings that differ from their original ones;32 (ii) correcting obvious drafting errors, including inconsistencies;33 (iii) ‘reading down’ overly broad terms to avoid constitutional invalidity;34 (iv) incrementally adjusting the meanings of provisions to enable them to achieve their intended purposes, which unanticipated social or technological developments may have made otherwise impossible;35 (v) adding qualifications to avoid undesirable results that the legislature is unlikely to have intended, and

28 Goldsworthy, Parliamentary Sovereignty (n 15) 247–51. 29 See Barrett, ‘Substantive Canons and Faithful Agency’ (n 21) esp 117 n 27. 30 J Goldsworthy, ‘The Principle of Legality and Legislative Intention’, in D Meagher and M Groves, eds, The Principle of Legality in Australia and New Zealand (Federation Press, Sydney, 2017), 46; A Perry, ‘Strained Interpretations’ (2019) 39 Oxford J Legal Studies 316. 31 Barrett, ‘Substantive Canons and Faithful Agency’ (n 21) 109; A M Samaha, ‘On Law’s Tiebreakers’ (2010) 77 University of Chicago Law Review 1661, 1708–17. 32 For an example, s 3 of the Human Rights Act 1998 (UK). Dale Smith calls such statutory provisions ‘modifier laws’, and discusses their contribution to legal content in ‘Is the High Court Mistaken About the Aim of Statutory Interpretation?’ (2016) 44 Federal Law Review 228. 33 A Scalia and B Garner, Reading Law, The Interpretation of Legal Texts (Thomson West, 2012), 234–39. 34 Severance (ie, excision) of terms in order to ensure constitutional validity surely goes beyond ­interpretation, and amounts to judicial amendment. 35 J Goldsworthy, ‘The Case for Originalism’, in G Huscroft and B Miller, eds, The Challenge of Originalism: Theories of Constitutional Interpretation (Cambridge, Cambridge University Press, 2011) 42, 62–63; J Goldsworthy, ‘Interpreting the Constitution in Its Second Century’ (2000) 24 Melbourne University Law Review 677, 699–701.

In Defence of Traditional Methodologies  47 would probably have preferred to avoid, in unusual circumstances that it apparently did not anticipate (this used to be called ‘equitable interpretation’);36 and (vi) adding (or ‘reading in’) so-called ‘implied terms’ on the ground that they are practically necessary for the text to achieve its intended purpose.37 None of this alters the fundamental principle that statutory interpretation is concerned first and foremost with inferring from all admissible evidence the communicative intentions of the legislature. Rectifying interpretation admittedly involves altering the meaning communicated by a statute, but in all the examples just listed this is to advance the legislature’s apparent purposes. As ‘faithful agents’, to use the American term, they sometimes depart from strict adherence to the instructions of their principals, in order better to advance their principals’ apparent intentions and purposes.38 This is consistent with the doctrine of legislative supremacy, and also with the rule of law and fairness to those subject to the law, if in the circumstances the likelihood of a discrepancy between what the statute, strictly construed, actually communicates, and what it was intended to communicate or achieve, is sufficiently obvious. The doctrine of legislative supremacy is challenged by the objection that many of the purposes attributed to statutes are independent of the legislature’s actual intentions.39 These purposes include promoting justice, public welfare, and the legitimacy of law; making legislating more democratic and legislation more publicly accessible; promoting coherence in the legal system; and upholding principles of the common law.40 These are purposes of the legal system, or the community, as a whole, and might on occasion conflict with the immediate purposes of the legislature.41 To a large extent this is consistent with the orthodox view. First, even when courts are attempting to clarify a legislature’s actual communicative intention, they may be justified in understanding the meaning of a provision in the light of unstated tacit assumptions that protect some traditional common law right or principle. Second, courts when supplementing the meaning of an under-determinate provision may have no choice but to rely on extra-statutory principles, including common law or even moral principles (including rights). Third, when engaging in rectifying interpretation, courts may be justified in ­adjusting the meaning of a statutory provision in order to protect such a ­principle, on the ground that Parliament can reasonably be presumed to have a standing

36 See, eg, R Ekins, The Nature of Legislative Intent (Oxford, Oxford University Press, 2012), 275–84; J Evans, ‘A Brief History of Equitable Interpretation in the Common Law System’, in T Campbell and J Goldsworthy, eds, Legal Interpretation in Democratic States (Ashgate, 2002), 67; J Evans, ‘Reading Down Statutes’, in R Bigwood (ed), The Statute: Making and Meaning (Wellington, LexisNexis, 2004), 123. 37 See, eg, L Hand, The Bill of Rights (Cambridge, Harvard University Press, 1958), 29; see also id. 14. See also J Goldsworthy, ‘The Implicit and the Implied in a Written Constitution’, in R Dixon and A Stone, eds, The Invisible Constitution in Comparative Perspective (Cambridge, Cambridge University Press, 2018) 109, 135–140. 38 See nn 22–24, above. 39 M Greenberg, ‘Legislation as Communication? Legal Interpretation and the Study of Linguistic Communication’, in A Marmor and S Soames (eds), Philosophical Foundations of Language in the Law (Oxford, Oxford University Press, 2011) 217, 252, summarised at 255. 40 ibid 252. 41 ibid 253.

48  Jeffrey Goldsworthy commitment to respect it, and therefore must have overlooked the provision’s impact on it. There is an obvious danger of disingenuous judicial misuse of this capacity. But clearly a legislature can have standing commitments. Consider section 3 of the Human Rights Act 1998 (UK), which requires courts to ‘read and give effect’ to ­legislation, as far as possible, so as to be compatible with enumerated rights. This in effect declares Parliament’s standing commitment to respect those rights and charges the courts with helping it do so, through their interpretation of other legislation. The idea of a legislature having standing commitments can plausibly be used to explain common law presumptions of its intentions. It can be reasonable to infer standing commitments from past practices, as well as from express declarations, even if the former are more questionable. Traditional, long-established presumptions, well known to parliamentary counsel, are no longer questionable, even if they were originally the artificial products of judicial pragmatism.42 If I know that others attribute standing commitments to me, and do nothing to disavow them, I corroborate the attribution and dispel misgivings about it. The standing commitment eventually becomes a tacit assumption, which then forms part of the legislature’s communicative intention. The presumption that new criminal offences include an element of mens rea can be used as an example. Our everyday, common sense notion of a moral offence implicitly includes a mens rea component. It is not implausible that when a new criminal offence is created, legislators have such a notion in mind; moreover, experienced drafters take it for granted that the familiar mens rea requirement will apply, especially given that this presumption is so well known. The legislature is reasonably regarded as relying on the drafters’ expert knowledge in helping to communicate its intentions. If there is genuine doubt about this, such a presumption might operate as a sensible epistemic hurdle, the courts in effect requiring a high degree of certainty before they will conclude that the statute was intended to exclude the usual mens rea requirement. Even if the preponderance of the evidence suggests that the legislature did intend to exclude the requirement, the courts might not regard that as sufficient to satisfy the degree of certainty required. In this way the courts would give effect to a standing intention that is reasonably attributable to the legislature. This does not make actual legislative intention irrelevant; it merely imposes an evidentiary hurdle to protect a reasonable default presumption. Many if not most of the presumptions applied in administrative law may be justifiable in much the same way. III.  JUDICIAL PRAGMATISM

Judges indulge in occasional furtive acts of result-oriented pragmatism, by which I mean more or less deliberate departures from the law for consequentialist reasons. 42 For discussion of the propriety of courts continuing to apply canons or ‘presumptions’ that were originally foisted on legislatures by judges acting pragmatically, and contrary to the doctrine of legislative supremacy, see Barrett, ‘Substantive Canons and Faithful Agency’ (n 21) 160–163. On pragmatism generally, see section III below.

In Defence of Traditional Methodologies  49 Although judges proclaim their fidelity to legislative supremacy, and its interpretive corollary, the primacy of legislative intention and purpose, they are sometimes surreptitiously unfaithful to both. This is understandable; judges are human, and prefer not to facilitate injustice or bad governance. Indeed, they may sometimes be morally justified in covertly disobeying the law in order to avoid extreme injustice.43 Consider the development of what is now called the ‘principle of legality’, which provides enhanced protection of an updated collection of interpretive presumptions of legislative intention. It is often alleged that the courts are not always sincere when they interpret statutes in the light of such presumptions.44 On this view, judges either occasionally or often use the presumptions to camouflage their de facto amendment of statutory provisions, in order to protect ‘common law values’ long cherished by the judiciary.45 In other words, they pretend to be clarifying the pre-existing meaning of a provision while in reality rectifying it, but without legal warrant. The traditional justification for these presumptions was entirely consistent with legislative supremacy.46 Aileen Kavanagh, who disapproves of parliamentary sovereignty, has conceded that the orthodox justification for applying the statutory presumptions is the fact that, in general, legislators know, or can be taken to know, that their legislation will be interpreted and understood in light of them. They are part of the known background against which Parliament legislates and of which it should be aware.47

But lately this traditional justification has increasingly been dismissed as an artificial rationalisation or fig leaf. Former Australian Chief Justice Sir Anthony Mason once described strong presumptive rules as having an ‘evident fictional character’ because ‘they do not reflect actual legislative intent’.48 Others have claimed that the presumptions ‘no longer have anything to do with the intent of the Legislature; they are a means of controlling that intent’.49 In reality, it is said, the courts have stubbornly protected common law values from legislative interference, while acknowledging political constraints on their ability to do so.50 The long-standing judicial resistance to privative clauses has involved dissembling. Many commentators have claimed that in Anisminic,51 the House of Lords used specious reasoning to circumvent Parliament’s command that decisions of

43 See J Goldsworthy, ‘The Limits of Judicial Fidelity to Law: the Coxford Lecture’ (2011) 24 Canadian Journal of Law and Jurisprudence 305. 44 For a typical American example, see S F Ross, ‘Where Have You Gone, Karl Llewellyn? Should Congress Turn Its Lonely Eyes To You?’ (1992) 45 Vanderbilt Law Review 561, 563. 45 See Goldsworthy, Parliamentary Sovereignty (n 15) 242, 304–312. 46 Bropho v Western Australia (1990) 93 ALR 207, 215. 47 A Kavanagh, Constitutional Review Under the UK Human Rights Act (Cambridge, Cambridge University Press 2009) 99. 48 Sir A Mason, ‘Commentary’, (2002) 27 Australian Journal of Legal Philosophy 172, 175. 49 L Tremblay, ‘Section 7 of the Charter: Substantive Due Process’ (1984) 18 University of British Columbia Law Review 201, 242; see also Kavanagh, Constitutional Review (n 47) 335. 50 J Burrows, ‘The Changing Approach to the Interpretation of Statutes’ (2002) 33 Victoria University of Wellington Law Review 981, 982–3, 990–5, 997–8; J Evans, ‘Controlling the Use of Parliamentary History’ (1998) 18 New Zealand University Law Review 1, 44. 51 Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147, [1969] 2 WLR 163.

50  Jeffrey Goldsworthy a statutory authority could not be judicially reviewed.52 New Zealand Justice E.W. Thomas recommended that we candidly admit what judges do in such cases: ‘I know of no rule of law or logic which would make judicial disobedience more ­palatable simply because it is done covertly.’53 Britain’s then leading administrative lawyer, Sir William Wade, observed that ‘the judges have proved willing to turn a blind eye to constitutional impediments for the sake of their historic policy of refusing to tolerate uncontrollable power’, and added that ‘I support the judges in resorting to every possible argument, convincing or otherwise’ in pursuing this policy.54 The so-called ‘ultra vires’ doctrine in administrative law holds that the grounds on which judges review the decisions of the executive government are imposed by statute, implicitly if not explicitly. Sir John Laws famously described this doctrine as a fig-leaf ‘serving to provide a façade of constitutional decency, with lip-service to the sovereign Parliament, while being out of touch with reality’.55 He thought that the judges created many or most of the grounds of review, but tried to conceal this because they believed it to be inconsistent with parliamentary sovereignty. Lord Steyn of the House of Lords also described the ultra vires doctrine as a ‘dispensable fiction’.56 I attribute some of these claims of infidelity to legislative intention to unjustified scepticism about the reality of implicit assumptions and standing commitments. As explained in the previous section, the traditional justification of the presumptions often remains plausible. The ‘principle of legality’ can be applied in a spirit that is faithful to Parliament’s intentions, provided that they are sufficiently clear; indeed, I have argued elsewhere that the principle’s operation is logically dependent on such fidelity.57 Nevertheless, I acknowledge that some judges have dissembled.58 They have pretended to be faithful to Parliaments’ presumed intentions, instead of frankly admitting that they were really protecting common law values. In other words, they have engaged in what Roscoe Pound called ‘spurious interpretation’, which amounts to ‘put[ting] a meaning into the text as a juggler puts coins, or what not, into a dummy’s hair, to be pulled forth presently with an air of discovery’.59 But it is important to ask why judges have engaged in this subterfuge. Surely it is to maintain a facade of legal orthodoxy, which is committed to legislative supremacy and the interpretive primacy of legislative intention and purpose. According 52 For example, HWR Wade and CF Forsyth, Administrative Law 7th edn (Oxford, Oxford University Press, 1994), 737; E W Thomas, ‘The Relationship of Parliament and the Courts’ (2000) 5 Victoria University of Wellington Law Review 9, 27; Kavanagh, Constitutional Review (n 47) 98 nn 39 and 105. 53 Thomas, ‘The Relationship of Parliament and the Courts’ (n 52) 27. 54 Sir W Wade, ‘Beyond the Law: a British Innovation in Judicial Review’ (1991) 43 Administrative Law Review 559, 564 and 566; see also his Constitutional Fundamentals rev’d edn (Boston, Stevens & Sons, 1989) 86. 55 Quoted by H W R Wade and C F Forsyth, Administrative Law 9th edn (Oxford, Oxford University Press, 2004) 39. 56 J. Steyn, Democracy Through Law (Aldershot, Ashgate, 2004) 131. 57 Goldsworthy, ‘The Principle of Legality’ (n 30) 46; P Sales, ‘Legislative Intention, Interpretation, and the Principle of Legality’ (2019) 40 Statute Law Review 53; P Sales, ‘In Defence of Legislative Intention’ (2019) 48 Australian Bar Review 6. 58 For the 19th Century, see, eg, Lord Devlin, ‘Judges and Lawmakers’ (1976) 39 Modern Law Review 1, 13–14. 59 R Pound, ‘Spurious Interpretation’, (1907) 6 Columbia Law Review 379, 382.

In Defence of Traditional Methodologies  51 to Kavanagh, the ‘politics of judicial lawmaking in the UK’ required the judges to do this: given the fact that respecting the will of Parliament is thought to be one of the most fundamental principles of statutory interpretation … we should not be surprised if judges routinely say that their judgments give effect to the will of Parliament.60

Hence, the judges have appealed to ‘legislative intentions’ that are really ‘constructed’ by them for use as ‘a rhetorical device’ providing a ‘cloak of respectability’.61 In other words, these judges have always known that they would flout constitutional orthodoxy and risk conflict with the legislature if they were to admit that they sometimes tamper with legislation to protect moral and political values of their choosing. The crucial point is that if there is a conflict between what the courts say they are doing (in this instance, acting on a genuine presumption of legislative intention), and what they are really doing (protecting common law values regardless of the legislature’s intentions), we should conclude that the law mandates what they say they are doing rather than what they are actually doing. As I have explained elsewhere: what judges say they do is much better evidence of the scope of their lawful authority than what they might sometimes actually do. Any claim about their lawful authority must be able to pass the test of public candour: if that claim is not and cannot be candidly asserted in public by them, it is surely false. How can judges possess a legal authority that they dare not publicly assert, and feel compelled to keep secret?62

It follows that the legal content of a statute is what orthodox interpretive principles, and the constitutional doctrines underlying them, authorise judges to identify or construct, regardless of occasional judicial dissembling. Such dissembling suggests that the judiciary has sometimes engaged in result-oriented pragmatism, camouflaged to avoid the appearance of impropriety and the risk of inter-institutional conflict, rather than interpretation and application of an imaginary common law constitution. The judiciary might, if the process continues unchecked, succeed in constructing a common law constitution, but that is a different matter.63 IV.  CONFUSION BETWEEN THE COMMON LAW AND OTHER KINDS OF LAW

I observed previously that when there is an overlap between the common law and other forms of law, or some confusion about the boundaries between them, then the judicial power to change the common law may be mistakenly extended to those other forms of law.64

60 Kavanagh, Constitutional Review (n 47) 335, 81. 61 ibid 98, nn 38, 82, 115; see also Barrett, ‘Substantive Canons and Faithful Agency’ (n 21) 120–21. 62 Goldsworthy, ‘The Limits of Judicial Fidelity’ (n 43) 308. For a related discussion, see M Barczentewicz, ‘The Illuminati Problem and Rules of Recognition’ (2018) 38 Oxford Journal of Legal Studies 500. 63 See sections V and VII below. 64 See conclusion of section II.A, above.

52  Jeffrey Goldsworthy It can be difficult to distinguish clearly between enacted law and common law, given that in interpreting enacted law judges also legitimately develop it, by supplementing the meanings of provisions that would otherwise be under-determinate, and occasionally rectifying them.65 Moreover, the principles that judges use to interpret statutes are often called ‘common law principles’, because they were developed over centuries by the judiciary in somewhat the same way as ordinary common law principles. In both Canada and Australia, whose Constitutions were originally statutes enacted by the UK (and Imperial) Parliament, apex courts initially chose to apply traditional principles of statutory interpretation.66 Hence, Australian High Court Justice Sir Owen Dixon once said that ‘the common law as antecedent to the Constitution … supplies … principles in aid of its interpretation and operation’.67 But the label ‘common law’ should not be allowed to mislead us. The vast bulk of the common law consists of substantive rules and principles, governing property, contracts, torts and so on, that have been made and developed by the judiciary over centuries. Judges are recognised as having authority unilaterally to change them, by extending, contracting or repudiating them. By contrast, principles of statutory and constitutional interpretation concern the interpretation of laws, created by other institutions possessing superior lawmaking authority, that the judges do not have authority to change (except, as previously explained, in strictly limited ways to prevent the lawmakers’ intentions or purposes being frustrated).68 It follows that the judges do not possess the same extensive authority to change these interpretive principles according to their assessment of justice and the public interest, as they have to change ordinary common law. As Australian High Court Justice (later Chief Justice) Sir Gerard Brennan observed: The authority of the courts to change the common law rules of statutory construction must … be extremely limited for the courts are duty bound to the legislature to give effect to the words of the legislature according to the rules which the courts themselves have prescribed for the communication of the legislature’s intentions.69

There is some scope for the judiciary to modify the principles of statutory and constitutional interpretation, but not so as to subvert or usurp the authority of the original lawmakers or those with the power of amendment. Were the judiciary to modify those principles so as to enable it, in effect, to amend statutes or constitutions contrary to the intentions and purposes of those who made them, the judiciary would flout the constitutional principles of legislative supremacy, the separation of powers and the rule of law, which are unwritten constitutional norms in Britain and New Zealand, and either expressed or implicit in the written constitutions of other Anglophone legal systems. 65 See section II.B above. 66 J Goldsworthy and G Huscroft, ‘Originalism in Canada and Australia: Why the Divergence?’, in R Albert and D Cameron (eds), Canada in the World: Comparative Perspectives on the Canadian Constitution (New York, Cambridge University Press, 2018) 183. 67 Sir O Dixon, ‘Marshall and the Australian Constitution’, in his Jesting Pilate and Other Papers and Addresses (Prymont, Law Book Co, 1965) 166, 174. 68 Section II.B above. 69 Corporate Affairs Commission (NSW) v Yuill (1991) 172 CLR 319, 322.

In Defence of Traditional Methodologies  53 In this vital respect, statutes and constitutions differ from the common law, which is not built on a substantial legal foundation enacted by institutions with lawmaking authority superior to the judges’. But the coexistence of common law and enacted (statutory and constitutional) law might make it psychologically difficult for judges to respect this difference. Judges whose role includes the ongoing review and development of the common law possibly succumb more often than their counterparts in civil law systems to the temptation to change the meanings of statutes and constitutions. That is an empirical hypothesis that cannot be tested without extensive comparative research, which is beyond the scope of this chapter. But Australian High Court Justice Sir Victor Windeyer once suggested that this was both inevitable and desirable: ‘[i]n any country where the spirit of the common law holds sway the enunciation by courts of constitutional principles based on the interpretation of a written constitution may vary and develop in response to changing circumstances’;70 even ‘the concepts that a written constitution embodies’ may ‘grow and develop as the needs of men change’.71 In the US, this view has been developed into a theory of constitutional interpretation called ‘common law constitutionalism’,72 and other theories of ‘living constitutionalism’. Justice Windeyer’s suggestion is impossible to reconcile with other more orthodox statements he made, such as that the Court may not change the meaning of the Constitution’s words, and may only reveal implications that are actually in the Constitution, not add new ones to it.73 As Sir Gerard Brennan also said: [I]t is clear that judicial development of the common law is a function different from judicial interpretation of statutes and of the Constitution … The Court, owing its existence and its jurisdiction ultimately to the Constitution, can do no more than interpret and apply its text, uncovering implications where they exist … The notion of ‘developing’ the law of the Constitution is inconsistent with the judicial power it confers. Clearly the Court cannot change the Constitution …74

I have often argued that for these and many other reasons, the orthodox approach to constitutional interpretation in Australia is a form of moderate originalism, as it used to be and arguably still is in Canada and the US.75 Moderate originalism is essentially the same methodology, applied to written constitutions, that was traditionally applied to statutes, so it includes scope for judicial supplementation and rectification in strictly limited circumstances. In both cases, the methodology is based on deference to the authority of the law-maker, whether it be a constitution-maker or

70 Victoria v Commonwealth (1971) 122 CLR 353, 396. 71 Sir Victor Windeyer, Some Aspects of Australian Constitutional Law (Speech delivered at the JA Weir Memorial Lecture, Edmonton, 13–14 March 1972) 36–37. 72 D Strauss, The Living Constitution (Oxford, Oxford University Press, 2010). 73 R v Commonwealth Conciliation and Arbitration Commission; Ex parte Association of Professional Engineers (1959) 107 CLR 208, 267; Victoria v Commonwealth (1971) 122 CLR 353, 402. For full ­discussion, see J Goldsworthy, ‘Justice Windeyer on the Engineers’ Case’ (2009) 37 Federal Law Review 363. 74 Theophanous v Herald & Weekly Times Ltd (1994) 182 CLR 104, 143–44. 75 For Australia, J Goldsworthy, ‘Originalism in Constitutional Interpretation’ (1997) 1 Federal Law Review 1; for Canada, Goldsworthy and Huscroft, ‘Originalism in Canada and Australia’ (n 66); B J Oliphant and B Sirota, ‘Has the Supreme Court of Canada Rejected ‘Originalism’?’ (2016) 42 Queen’s Law Journal 107; for the US, W Baude, ‘Is Originalism Our Law?’ (2015) 115 Columbia Law Review 2349.

54  Jeffrey Goldsworthy a legislature, and fidelity to its intentions and purposes, insofar as they have been communicated or manifested by the making of the law in question. That deference and fidelity are implicit in the constitutional separation of the powers to make, and to interpret, the law. A second example of the sometimes confusing overlap between common law and statute law can be observed in the development of modern administrative law. Judicial review has always required very close reading and careful interpretation of statutes in order to identify and clarify limits to the administrative powers they confer. A distinction was traditionally drawn between enforcing the legal boundaries within which the administration exercises discretionary powers, and reviewing the merits of the administration’s discretionary decisions inside those boundaries. The role of the judiciary was to hold firm the legal boundaries, but not to intrude into the merits of a decision: the administration was responsible only to Parliament and the electorate for decisions that were lawful, even if arguably unfair or unwise. However, judges were able to shift the legal boundaries within which decisions could be made, and they did so regularly so as to expand their own powers of review. Eventually, the orthodox theory that they were merely enforcing legal limits imposed (expressly or impliedly) by statute came to be seen as implausible, and was challenged by the theory that they were in fact enforcing limits imposed by the common law – which means, by the judges themselves, since they make the common law.76 This led to debate between proponents of the ‘ultra vires’ theory, and of the ‘common law’ theory, of judicial review.77 Given that some of the communicated content of enacted laws can be genuinely implied or implicit, and also that orthodox interpretation can include rectification by way of ‘reading in’ so-called ‘implications’ necessary to achieve the law-maker’s immediate purposes or protect its standing commitments, it can be very difficult to distinguish between genuine and fabricated implications. I have argued, elsewhere, that while it is quite plausible to regard many of the traditional grounds of judicial review as concerning limits genuinely found in statutes, others may be judicial constructs that are nevertheless legally legitimate.78 Those who prefer the ‘common law’ theory of judicial review to the old ‘ultra vires’ theory are able to regard modern administrative law as largely judge-made: a result of judicial inventiveness rather than the application of principles discovered in the depths of a common law constitution.79 But does it follow from the courts’ success in developing administrative law, that they legally may, or morally ought to, engage in a similar process of developing constitutional law, by gradually and stealthily subordinating statutes to a form of constitutional judicial review in the absence of a written constitution? The development of administrative law was always subject to

76 See the essays collected in C Forsyth (ed), Judicial Review and the Constitution (Oxford, Hart Publishing, 2000), and for a defence of the common law theory, especially that of Paul Craig. 77 ibid. 78 Goldsworthy, Parliamentary Sovereignty (n 15) 281–85. 79 See the story told in JNE Varuhas, ‘Administrative Law and Rights in the UK House of Lords and Supreme Court’, in P Daly (ed), Apex Courts and the Common Law (Toronto, University of Toronto Press, 2019) 225.

In Defence of Traditional Methodologies  55 oversight and potential override by Parliament, but if its sovereignty were limited, the development of a common law constitution might not be. V.  CONSTITUTION-MAKING BY JUDGES

A third example of overlap or confusion, and perhaps the most difficult to sort out, concerns the boundary between, on the one hand, the common law, and on the other, customary constitutional norms that establish Parliament’s legislative sovereignty and privileges, and executive prerogatives. This brings us to the issue mentioned but ­postponed at the end of the Introduction: the extent to which the orthodox understanding of constitutional law is at risk of being eroded due to various recent challenges posed by judicial reasoning and academic theorising. I have long argued, with some success, that the doctrine of parliamentary sovereignty is not a matter of common law in the modern sense of judge-made law.80 It is in a different category because it was not originally made by the judges and may not be changed by them unilaterally; it is, instead, constituted by a widespread consensus among senior legal officials in all branches of government.81 If parliamentary sovereignty is to be labelled a doctrine of common law, this must be in a much older sense of the term, such as ‘long-standing custom’ recognised but not unilaterally made by the judges. But this creates a considerable risk of confusion between this sense of common law, and the modern sense of judge-made law. The success of British judges in developing modern administrative law over the last half-century, to widespread acclaim, may have emboldened some of them to attempt a similar development of constitutional law. In the two Miller cases, the Supreme Court is arguably repositioning itself to become a ‘constitutional court’ that can adjudicate the major political conflicts of the day.82 The courts have also been busily recognising – or creating – ‘constitutional principles’, ‘constitutional rights’ and ‘constitutional statutes’, and using these to interpret statutes in ways that arguably flout the related doctrines of legislative supremacy and the interpretive primacy of legislative intention and purpose. If judges interpret legislation according to ‘constitutional principles’ that they themselves have developed, contrary to Parliament’s apparent intentions, then they elevate themselves to co-authors of the law that results from their interpretations. Parliament is then no longer the sole author of the statutes it enacts; no matter what words it uses, or with what apparent intentions, their meanings will be altered to give effect to values preferred by the judges. This is applauded by the academic opponents of legislative supremacy, who claim that the meaning of any statute is ‘the joint responsibility of Parliament and 80 For one example of success, see Lord Bingham, The Rule of Law (London, Penguin, 2011) ch 12. 81 Goldsworthy, The Sovereignty of Parliament (n 15) ch 10; Parliamentary Sovereignty (n 15) ch 2; J Goldsworthy, ‘Parliamentary Sovereignty and Constitutional Change in the United Kingdom’, in P Leyland, R Rawlings and A Young (eds), Sovereignty and the Law: Domestic, European and International Perspectives (Oxford, Oxford University Press, 2013) 50. 82 J Murkens and R Masterman, ‘The New Constitutional Role of the Judiciary’ (LSE Law, Policy Briefing, July 2014).

56  Jeffrey Goldsworthy the courts’ and therefore its construction a ‘collaborative enterprise’.83 As we have seen, when supplementing or rectifying statutory meaning, courts do frequently contribute to the meanings of statutes, but they should do so as Parliament’s faithful agents, seeking where possible to implement its objectives or protect its standing commitments. If they depart from that role, then as Richard Ekins has argued, it is difficult to see how their interpretations can be reconciled with the fundamental constitutional norm that it is Parliament, not the courts, that has the authority to make statute law. If making statute law were a collaborative enterprise in that radical sense, Parliament would merely provide raw material, in the form of a text, which the judges would then combine with their own material to make the law.84 The separation of legislative and adjudicative powers would plainly be breached. The increasing judicial tendency to describe some common law rights and principles, and some statutes, as ‘constitutional’ might develop into a radical challenge to the doctrine of parliamentary sovereignty.85 That description is not in itself contrary to orthodoxy. As for statutes, it has always been accepted that the British Constitution consists partly of various important statutes dealing with matters of government, and some are so important that Parliament is very unlikely to intend to meddle with them indirectly, as a side-effect of provisions dealing with lesser matters. That is why Lord Wilberforce said in 1967 that he felt ‘some reluctance to holding that an Act of such constitutional significance as the union with Ireland Act is subject to the doctrine of implied repeal’.86 The current trend to deem ‘constitutional statutes’ immune to o ­ rdinary implied repeal, launched by Laws J in Thoburn and approved by the Supreme Court in HS2, is consistent with parliamentary sovereignty provided that it remains sincerely based on the orthodox reasoning described previously, which gives effect to a standing intention that is reasonably attributable to the legislature.87 Laws J erred in attributing this trend to judicial development of ‘the common law’.88 The Supreme Court in HS2 has correctly attributed it to recognition that Parliament is unlikely to either contemplate or authorise the abrogation of fundamental principles contained in constitutional instruments or recognised by common law.89 Describing important principles as ‘constitutional’ is a long-standing feature of British constitutional discourse, and in itself, entirely consistent with parliamentary 83 TRS Allan, ‘Legislative Supremacy and Legislative Intention: Interpretation, Meaning, and Authority’ [2004] Cambridge Law Journal 685, 689 n 13; also P Joseph, ‘Parliament, the Courts, and the Collaborative Enterprise’ (2004) 15 King’s College Law Journal 321, discussed in Goldsworthy, Parliamentary Sovereignty (n 15) 275–80; and D Kyritsis, Shared Authority: Courts and Legislatures in Legal Theory (Oxford, Hart Publishing, 2014). 84 R Ekins, ‘The Relevance of the Rule of Recognition’ (2006) 31 Australian Journal of Legal Philosophy 95, 100. 85 The following paragraphs borrow from Goldsworthy, Parliamentary Sovereignty (n 15) 314–18. 86 Earl of Antrim’s Petition (House of Lords) [1967] AC 691. See further Goldsworthy, Parliamentary Sovereignty (n 15) 312–14. 87 See the end of section II.B, above; the word ‘ordinary’ is used to acknowledge that repeal by necessary implication must still be possible. 88 Goldsworthy, Parliamentary Sovereignty (n 15) 312–14. 89 R (HS2 Action Alliance Ltd) v Secretary of State for Transport [2014] UKSC 3, para 207. This is consistent with Paul Craig’s argument in ‘Constitutionalising Constitutional law: HS2’ [2014] Public Law 373, 385–86.

In Defence of Traditional Methodologies  57 sovereignty. Sir Robert Chambers, Blackstone’s successor in the Vinerian Chair at Oxford, embraced parliamentary sovereignty whole-heartedly, but condemned a particular statute on the ground that ‘though not illegal, for the enaction of the supreme power is the definition of legality, [it] was yet unconstitutional’, because it was ‘contrary to the principles of the English government’.90 This distinction between law and constitutional principle was perpetuated by other writers in the eighteenth and nineteenth centuries, and survives today in the distinction between law and constitutional convention.91 Dicey, for example, referred to ‘the fundamental principles of the constitution and the conventions in which these principles are expressed’.92 He did also describe as ‘constitutional’ certain principles that were extrapolated from judicial decisions, including principles that protected personal liberty.93 But of course all these principles were subject to Parliament’s legislative authority. Today, the subtle distinctions encoded in this traditional terminology are increasingly liable to be misunderstood or obfuscated. It is already being claimed that ‘the incremental development by the courts of a body of “constitutional rights”… ha[s] rendered our traditional understandings of the subordinate role of courts in relation to Parliament obsolete’.94 This fashionable way of thinking might eventually tempt the judiciary to claim the power to protect rights from legislative interference. Having laid the foundation for doing so, in a substantial body of dicta, judges may eventually feel emboldened to declare that: ‘of course, if these principles are constitutional, they must by definition control even the power of Parliament’. That would be illogical, because there is nothing magical about the word ‘constitutional’: it is not a kind of fairy dust, which when sprinkled on a select group of statutes and common law norms makes them immune to legislative revision. Critics of parliamentary sovereignty, including some judges, openly talk about the courts ‘chipping away at the rock of parliamentary supremacy’,95 ‘inching forwards with ever stronger expressions when treating some common law rights as constitutional’,96 and as ‘[s]tep by step, gradually but surely’ qualifying the principle of parliamentary sovereignty.97 One senior judge has said that ‘the common law has come to recognise and endorse the notion of constitutional, or fundamental, rights’,98 and although Parliament currently remains sovereign, this may change through ‘the

90 R Chambers, A Course of Lectures on the English Law Delivered at the University of Oxford 1767–1773 (TM Curley, ed, University of Wisconsin Press, 1986) vol I, 141. 91 J Austin, The Province of Jurisprudence Determined (HLA Hart, ed, Weidenfeld & Nicholson, 1954) 257–58; Madzimbamuto v Lardner-Burke [1969] 1 AC 645, 723 per Lord Reid. 92 A V Dicey, Introduction to the Study of the Law of the Constitution 10th edn (Indianapolis, MacMillan, 1964) 445; see also 434. See also Goldsworthy, The Sovereignty of Parliament (n 15) 190–92. 93 Dicey, Introduction to the Study of the Law (n 92) 197, and more generally, 195–98, 201–202. 94 Murkens and Masterman, ‘The New Constitutional Role of the Judiciary’ (n 82) 3. 95 J Jowell, ‘Parliamentary Sovereignty Under the New Constitutional Hypothesis’ [2006] Public Law 562, 575. 96 R Cooke, ‘The Road Ahead for the Common Law’ (2004) 53 International and Comparative Law Quarterly 273, 277. 97 Jackson v Attorney-General [2006] 1 AC 262, [104] (Lord Hope). 98 International Transport Roth GmbH v Secretary of State for the Home Department [2002] EWCA Civ 158; [2002] 3 WLR 344 [71] (Laws LJ).

58  Jeffrey Goldsworthy tranquil development of the common law’.99 Some senior judges have already, in dubious obiter dicta, questioned whether Parliament is still sovereign.100 Support from legal theorists for these developments can be divided into two kinds. Common law constitutionalists maintain that the judges are using the method of interpretivism to faithfully identify and clarify constitutional principles already immanent in the depths of the common law.101 But others, whom I will call common law pragmatists, are not committed to interpretivism. They acknowledge that the judges are endeavouring to change the Constitution, one step at a time; they accept this as entirely legitimate, and urge the judges to continue. They advocate the construction of a common law constitution, rather than the enforcement of an existing one.102 I suspect, but cannot prove, that the common law constitutionalists are in a minority. I will discuss the merits of both kinds of support after a digression into legal philosophy. VI.  LEGAL PHILOSOPHY

For many years Trevor Allan and I have debated whether or not the doctrine of parliamentary sovereignty is part of the British Constitution.103 Allan is a common law constitutionalist, heavily influenced by Ronald Dworkin’s interpretivist theory of law. But in recent work he chides Dworkin for failing to take his own insights to their logical conclusions, which are similar to those of strong natural law. Any government action inconsistent with fundamental rights ‘lacks legal force or validity’,104 so ‘a gravely unjust provision … cannot be law’,105 ‘even if [it is] asserted in the name of Parliamentary sovereignty’.106 But before holding that a statute is invalid, judges should strive to give it a meaning consistent with justice. ‘[A] statutory provision has no power to inflict serious injustice; it means what it ought to mean when all the pertinent moral considerations are taken into account.’107 99 Sir John Laws, ‘Illegality and the Problem of jurisdiction’, in M Supperstone and J Goldie (eds), Judicial Review 2nd edn (London, Butterworths, 1997) 4.17. 100 R (Jackson) v Attorney General [2005] UKHL 56, [2006] 1 AC 262 [102] (Lord Steyn), [104], [107], [126] (Lord Hope), [159] (Baroness Hale); Thoburn v Sunderland City Council [2002] EWHC 195 (Admin), [2003] QB 151 [60] (Laws LJ); Axa General Insurance Ltd v Lord Advocate [2011] UKSC 46; 3 WLR 871, [49-51] (Lord Hope) and [153-54] (Lord Reed); Moohan v Lord Advocate [2014] UKSC 67; [2015] AC 901 [35] (Lords Hodge, Neuberger, Clarke and Reed, and Lady Hale); R (Public Law Project) v Lord Chancellor [2016] UKSC 39, [2016] AC 1531, [20] (Lords Neuberger, Lady Hale, Lords Mance, Reed, Carnwath, Hughes and Toulson). See n 141 below for criticism of the dicta in Jackson. 101 See nn 1 and 2 above. 102 For one of many examples, see R Masterman and Se-Shauna Wheatle, ‘Unity, Disunity and Vacuity: Constitutional Adjudication and the Common Law’, in M Elliott, JNE Varuhas and SW Stark (eds), The Unity of Public Law?: Doctrinal, Theoretical and Comparative Perspectives (Oxford, Hart Publishing, 2018) 123. 103 For the latest exchanges see Allan, The Sovereignty of Law (n 2) (consult ‘Goldsworthy’ in Index), and LB Crawford, P Emerton and D Smith (eds), Law Under a Democratic Constitution; Essays in Honour of Jeffrey Goldsworthy (Oxford, Hart Publishing, 2019) ch 12 (see esp n 1) and 297–300. 104 TRS Allan, ‘Political Obligation and Public Law’, in Crawford, Emerton and Smith, Law Under a Democratic Constitution (n 103) 252; see also 256. 105 ibid 260. 106 ibid 268. 107 ibid 255.

In Defence of Traditional Methodologies  59 I take these claims to diverge radically from constitutional orthodoxy in Britain. The doctrine of parliamentary sovereignty embodies a kind of parochial legal ­positivism in relation to statute law: any statute enacted (posited) by Parliament is legally valid, regardless of its moral quality.108 Allan acknowledges that the positivist view is more commonly held than his: ‘Discussions of political obligation usually proceed on the assumption that the law’s requirements can be identified independently of questions about their moral bindingness … as a matter of social fact.’109 He later refers to this as ‘the standard view’.110 But my disagreement with Allan is not due solely to him being an ultra-Dworkinian and me a legal positivist. My positivist defence of parliamentary sovereignty can be – indeed, already has been – reworked into an interpretive defence along Dworkinian lines.111 Disagreements between competing Dworkinian interpretations are very hard to resolve because they ultimately rest on contentious moral judgments, which are often impossible either to prove or disprove. Dworkinian interpretivism requires that even apparently settled legal norms be rejected as mistakes if they are inconsistent with the deeper moral principles that are taken to provide the best fit and moral justification of the legal system overall. The criteria of fit and of justification, and the balance between them, all require moral judgment; Dworkin says that they represent ‘different aspects of a single overall judgment of political morality’.112 That judgment may be the subject of irresolvable disagreements. Brian Leiter has argued that Dworkin’s theory is one of ‘esoteric law’, which makes it possible that no-one can truly know what the law is.113 Because all the key elements of Dworkinian interpretation require moral judgment, we can never be sure that we have established which interpretation of the law is the best, and so we can never truly know (although we may believe) that a legal proposition is true. The legal positivists’ theory makes knowledge about the law more attainable, because it is ­ultimately concerned with matters of fact.114 The dependence of Dworkinian interpretations on the moral judgments of each interpreter is likely to create severe problems at a practical level, given the risk of officials and citizens adopting different and incompatible interpretations of their legal arrangements due to divergent moral judgments. Allan denies that the interpretation of the highest court has special, binding authority.115 On his view, all individuals 108 I realise, of course, that the doctrine itself must rest on a deeper foundation whose nature is itself open to jurisprudential debate. 109 Allan, ‘Political Obligation and Public Law’ (n 104) 249. 110 ibid 253 and 267. 111 See, eg, the final chapter of Goldsworthy, The Sovereignty of Parliament (n 15) esp 254 and 271. 112 R Dworkin, Justice in Robes (Cambridge Mass, Belknap Press, 2006) 15. 113 See B Leiter, ‘Book Review’ (2006) 56 Journal of Legal Education 675, 675; ‘Explaining Theoretical Disagreement’ (2009) 76 The University of Chicago Law Review 1215, 1248. 114 See, eg, Andrei Marmor, The Language of Law (Oxford, Oxford University Press, 2014) ch 3 (‘Truth in Law’). Stuart Lakin argues that empirical arguments are incapable of settling disagreement about legal practices: S Lakin, ‘The Manner and Form Theory of Parliamentary Sovereignty: A Nelson’s Eye View of the UK Constitution?’ (2018) 38 Oxford Journal of Legal Studies 168, 179–80. But legal positivists would maintain that where empirical enquiry is unable to identify a consensus among senior officials, there is no law. In that situation, judges may be forced to make moral judgments, and hope that other senior officials accept them. The issue of ‘manner and form’ that Lakin discusses is an example: see Goldsworthy, The Sovereignty of Parliament (n 15) 259. 115 As reported in Goldsworthy, The Sovereignty of Parliament (n 15) 264–65.

60  Jeffrey Goldsworthy are responsible for their own interpretive judgements about the law,116 even though their reasoning and conclusions may be controversial.117 An observation of Jeremy Bentham seems apposite: ‘from a set of data … law is to be extracted by every man who can fancy that he is able: by each man, perhaps a different law: and these then are the monades [basic metaphysical entities] which meeting together constitute the rules’ of law.118 As Gerald Postema commented, this ‘transforms a matter of public rules into a matter of essentially and unavoidably private conjecture’.119 The practical consequences of interpretive disagreement among senior legal officials about constitutional fundamentals could be disastrous. If Parliament and the judiciary were to adopt conflicting interpretations of Britain’s unwritten Constitution, Parliament affirming its sovereignty and the judiciary denying it, the authority of both the interpreters and their interpretations might itself be in dispute.120 It hard to see how that dispute could, in practice, be resolved by Dworkinian interpretation. To proclaim that the legally correct interpretation would be whichever of the contenders (Parliament’s or the judiciary’s) is truly the best would hardly be helpful. Both could plausibly claim that their interpretation was truly the best. As Jeremy Waldron has put it, ‘the prescription most commonly advanced by the [moral] realist, ‘Choose the objectively right answer’, is, though metaphysically admirable, politically impotent’.121 Waldron’s point is corroborated by the fact that when judges disagree among themselves about the best interpretation of ordinary law, for practical purposes their disagreement is not, and could not be, resolved by the criterion of objective rightness or truth. That is precisely what is in dispute. Their disagreement is settled, instead, by a decision procedure – long established by consensus – whereby the majority’s interpretation prevails. Unfortunately, there would be no accepted decision procedure should Parliament and the judiciary disagree about parliamentary sovereignty: the disagreement would include advocacy of incompatible decision procedures. The disagreement would have to be settled by capitulation, compromise or coercion. Even if one of these two conflicting interpretations of the Constitution were ‘correct’, according to Dworkin’s criteria, this could not be proved. Appeals to moral truth cannot provide the stable foundation that legal systems need to function effectively or even to endure. What is needed (among other things) is widespread official consensus about fundamental decision procedures – about the location and scope of law-making authority, or about the location of authority to resolve disagreements about such matters, which would sooner or later amount to the same thing. Any assumption or stipulation that judges necessarily possess superior authority of the second kind would beg the question.

116 Allan, The Sovereignty of Law (n 2) 39–40 and 160–61; and Allan, ‘Political Obligation and Public Law’ (n 104) 260–61 and 267–68. 117 Allan, ‘Political Obligation and Public Law’ (n 104), 252. 118 J Bentham, Of Laws in General (HLA Hart ed, 1970), 192, quoted by G Postema, Legal Philosophy in the Twentieth Century: the Common Law World (Dordrecht, Springer, 2011) 541. 119 Postema, Legal Philosophy (n 118) quoting his ‘The Normativity of Law’, in R Gavison (ed), Issues In Contemporary Legal. Philosophy: The Influence of H.L.A. Hart (Oxford, Clarendon Press, 1987) 81, 103. 120 See Goldsworthy, ‘Parliamentary Sovereignty and Constitutional Change’ (n 81) 50. 121 J Waldron, ‘Moral Truth and Judicial Review’ (1998) 43 American Journal of Jurisprudence 75, 81.

In Defence of Traditional Methodologies  61 The possibility of dangerous disputes between the political and judicial branches of government about their constitutional authority can never be prevented, no matter what legal philosophy is generally accepted. But intuitively, such disputes seem more likely to arise and less likely to be resolvable peacefully if a theory of law were widely accepted that holds moral truth to be the ultimate criterion of legal truth.122 If so, legal systems are more likely to be stable and effective if participants accept either some version of legal positivism, which holds that law ultimately rests partly on a widespread consensus among senior officials about the most fundamental norms, or some kind of consensus-seeking interpretivism that does not depend on the moral judgments of the interpreter.123 There are good reasons to prefer some kind of non-moral interpretivism to the Dworkinian kind. Law precedes philosophical reflection about it. It can exist in the absence of careful philosophical reflection on the part of participants, or indeed anyone within the broader community. Even in mature legal systems, the busy ­practical people who help to create law, apply it, and resolve disputes about it rarely have the time or expertise to devote to deep philosophical reflection. Of course, their deliberations and actions may rely partly on working assumptions of a jurisprudential nature, perhaps gleaned from philosophers, and it is part of the job of philosophers to help practitioners clarify those assumptions, make them consistent, and fill in gaps in their self-understanding. But apart from such tidying-up work, legal philosophers must for the most part take as given the practice as it is, including widespread assumptions on which it rests, and seek to understand and explain it. They must avoid wishful thinking, which projects onto the law their personal aspirations and visions. What the law is necessarily depends far more on its practitioners’ understandings and practices than on those of legal philosophers; the former constitutes the subject-matter that the philosophers study. Philosophers are helpmates to practitioners, not vice versa. If widespread, routine assumptions and understandings of practitioners are repudiated or radically revised by philosophers, then the latter are not seeking to understand and explain the subject-matter, but instead, to reconstitute it. Interpretivist theories, therefore, cannot stray too far from the shared self-understanding of participants – or, in other words, from tradition and orthodoxy – even if some aspects of that understanding need revision for the sake of overall coherence. The task of the interpreter who seeks to understand a practice must be to clarify and harmonise the reasons or principles that actually motivate its participants, in ways that they are capable of recognising as enhancing their own understanding of their practice.124 Dworkin himself has explained that [j]udges must not read their own convictions into the Constitution … They must regard themselves as partners with other officials, past and future, who together elaborate a 122 Cf D Smith, ‘Agreement and Disagreement in Law’ (2015) Canadian Journal of Law and Jurisprudence 183, 190–94. 123 See Goldsworthy, ‘Parliamentary Sovereignty’ (n 81), and for a non-Dworkinian interpretive method, K Toh, ‘Jurisprudential Theories and First-Order Legal Judgments’ (2013) 8/5 Philosophy Compass 457, 465–66; F Schroeter, L Schroeter and K Toh, ‘A New Interpretivist Metasemantics for Fundamental Legal Disagreements’ (2020) 26 Legal Theory 62. 124 G Postema, ‘“Protestant” Interpretation and Social Practices’ (1987) 6 Law and Philosophy 283, 288–89 and 308–19; Schroeter, Schroeter and Toh, ‘A New Interpretivist Metasemantics’ (n 123) 94.

62  Jeffrey Goldsworthy coherent constitutional morality, and they must take care to see that what they contribute fits with the rest.125

If this is true of judges, then other interpreters must also avoid reading their own moral convictions into the law; instead, they must aim to clarify and harmonise the ‘positive’ constitutional morality that is already explicitly or implicitly accepted by practitioners, and especially senior officials.126 I am frankly baffled by the suggestion that if I were to travel to Mongolia, and seek to understand its legal system, my own moral judgments would be paramount, rather than being at best mere heuristics to help me understand the judgments of Mongolian officials.127 Moreover, Dworkin’s reference to ‘other officials’ must be taken seriously. The task is not to provide the best interpretation of only judicial assumptions and understandings, but those of senior officials in general, whose co-ordinated practices are jointly responsible for maintaining an efficacious legal system.128 The foundations of legal systems are typically established more through the settlement or avoidance of violent conflict – whether by capitulation, compromise or coercion – than by philosophical reflection. Philosophical disagreements can create or intensify conflicts, but the victors are not necessarily the better or more persuasive philosophers; they may simply be more powerful. Compromise may help to avoid or settle such conflicts, requiring both sides to accept what each regards as a less-than-ideal modus vivendi. Even written constitutions established through debate, negotiation and voting typically embody many compromises. Often what are compromised are conflicting opinions about the trustworthiness of different kinds of officials and methods of decision-making – in other words, about what Scott Shapiro calls the ‘economy of trust’.129 A compromise may strike a mutually acceptable balance of the differing risks of official misbehaviour feared by contending interest groups. Law operates on the foundation of such compromises or settlements, and any genuine understanding or interpretation of it must account for them. An interpretation resting on noble moral ideals that repudiates them will risk destabilising the legal system they help maintain. This is particularly so when those compromises are centuries old. Dworkin himself acknowledged that it is ‘uncontroversial’ that widespread and unchallenged understandings among lawyers and judges play an important role in familiar legal systems – that the American legal system would not function as effectively as it does if American lawyers did not agree that the requirements of the Constitution are part of American law, for example ….130 125 Dworkin, Law’s Empire (n 4) 10; and see also Goldsworthy, The Sovereignty of Parliament (n 15) 253–54 and 271–72 nn 158–159. This is also required by an important objective of Dworkinian interpretation, which is to enable legal officials to act with ‘integrity’, by applying the same set of principles to all members of their community. If judges were to interpret the law according to their own personal moral convictions, rather than those of the legal community to which they belong, their interpretations would be more likely to differ from one another, and litigants to be subjected to different principles. 126 As Stuart Lakin, an interpretivist, states, ‘The personal moral convictions of a theorist must give way to the moral theory reflected in the practice’: Lakin, ‘The Manner and Form’ (n 114) 182. 127 Either way subject, of course, to ‘fitting’ the history of Mongolian legal practices. 128 Lakin, ‘The Manner and Form’ (n 114) 182 n 105, citing Kyritsis, ‘Shared Authority’ (n 83) ch 4. 129 S Shapiro, Legality (Cambridge Mass, Belknap Press, 2013) 331–77. 130 Dworkin, Justice in Robes (n 112) 189.

In Defence of Traditional Methodologies  63 He denied the legal positivist claim that legal reasoning, strictly speaking, is limited to implementing such understandings; he maintained that it also includes interpretive reasoning, guided by moral judgement, to decide issues that such understandings leave unresolved. Presumably legal reasoning, as he saw it, can also include interpretive reasoning that challenges previously unchallenged understandings. But considerable moral weight must surely be accorded to the value of not destabilising a generally satisfactory and functioning legal system. That is how Dworkinians can cope with most objections: by including considerations shown to be relevant by objectors among the institutional data that any interpretation must accommodate, and adjusting their interpretations accordingly. But in doing so, they risk having to adopt interpretations – such as mine – that they would prefer to avoid. VII.  PARLIAMENTARY SOVEREIGNTY TODAY

We can now return to the recent developments in judicial reasoning, summarised in section V, that either already do or may soon challenge the doctrine of parliamentary sovereignty. As noted then, support from legal theorists for these developments can be divided into two kinds. Common law constitutionalists maintain that the judges are faithfully identifying, clarifying and applying pre-existing constitutional principles immanent at the deepest levels of the common law.131 Common law pragmatists, on the other hand, acknowledge that the judges are endeavouring to change the constitution, applaud them for doing so, and urge them to continue.132 These two positions have different implications for the kinds of criticisms that can be made of these developments. As Dworkin has said: It is important, practically and politically, to determine what judges may and must do in the exercise of their responsibility to enforce the law, and to distinguish that from other judicial acts and decisions that must rely on a different and more controversial kind of justification.133

Common law constitutionalists regard the judges as merely identifying and enforcing existing law, whereas the pragmatists acknowledge that the judges are endeavouring to change the law, whether or not they have legal authority to do so. Common law constitutionalism has the rhetorical advantage of pre-empting certain kinds of criticisms of judicial law-making. When judges determine existing law, they act within the core of their judicial function, and even critics who regard a decision as erroneous should acknowledge this. Very robust criticism, especially by politicians, is then open to the objection that it threatens the judges’ independent exercise of their judicial function. But when judges make new law, such criticism may be appropriate; making new law turns more on policy evaluation than legal acumen, and is subject to criticism on policy grounds. A fortiori, if judges attempt to change laws that they lack authority to change – such as through the de facto amendment of statutes or written



131 See 132 See

n 2, above. n 102 above. Justice in Robes (n 112) 199.

133 Dworkin,

64  Jeffrey Goldsworthy constitutions, or the repudiation of long-standing customary rules of an unwritten constitution – they may be subject to even more robust criticism. The declaratory ‘fairy tale’ embraced by some common law constitutionalists attempts to delegitimise such criticisms, by obfuscating orthodox distinctions between declaring existing law and making new law, between legal judgment and policy choice, and between the kinds of laws that judges may and may not change unilaterally. Common law constitutionalists, as methodological interpretivists, welcome – indeed, celebrate – judicial reliance on contestable value judgments in deciding constitutional law cases.134 But I would expect them to oppose proposals that the appointment of judges should depend partly on a more explicit assessment of their moral and political philosophies. Yet, if their version of interpretivism were to become orthodox judicial methodology in constitutional cases, that would seem not only inevitable but entirely proper. Appointment to any position requires scrutiny of the candidates’ ability to carry out its responsibilities; if a position requires the frequent making of contestable value judgments, the ability to do so wisely must surely be a criterion for appointment. That is one reason why it is dangerous for judges to intrude too far into political decision-making. Do the recent developments that threaten parliamentary sovereignty vindicate Trevor Allan’s common law constitutionalism? His theory confronts many obstacles, in addition to the theoretical objections to Dworkinian interpretivism discussed in the previous section. Any plausible interpretation must, as we have seen, accommodate the understandings and practices of senior legal officials in general, not just those of the judiciary.135 Given Parliament’s strong commitment to preserving its sovereignty, that must make it difficult to discredit the doctrine of parliamentary sovereignty. But even if we focus solely on the understandings and practices of judges, the two recent Miller cases now make this even more difficult. When Miller No 1 was decided by the High Court, the Chief Justice, the Master of the Rolls and Sir Philip Sales (who is now on the Supreme Court) noted that: It is common ground that the most fundamental rule of UK constitutional law is that the Crown in Parliament is sovereign and that legislation enacted by the Crown with the consent of both Houses of Parliament is supreme …. Parliament can, by enactment of primary legislation, change the law of the land in any way it chooses. There is no superior form of law than primary legislation, save only where Parliament has itself made provision to allow that to happen.136

On appeal, an eight-member majority of the Supreme Court affirmed that the doctrine, understood explicitly in Diceyan terms, is ‘a fundamental principle of the UK Constitution’.137 The judgments also employed the positivist concept of ‘the 134 Eg, T Fairclough, ‘Evans v Attorney General [2015]: The Underlying Normativity of Constitutional Disagreement’, in S Juss and M Sunkin (eds), Landmark Cases in Public Law (Oxford, Hart Publishing, 2017) 285. 135 Goldsworthy, ‘Parliamentary Sovereignty and Constitutional Change’ (n 81) and section VI above. 136 R (Miller) v Secretary of State for Exiting the European Union [2016] EWHC 2768 (Admin), para 21; see also paras 22, 23 and 79. 137 R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5, para 43; and see also paras 61 and 67 (Lord Neuberger, Lady Hale, Lords Mance, Kerr, Clarke, Wilson, Sumption and Hodge). It was also affirmed by Lord Reed (paras 183 and 334) and Lord Carnwath (para 274).

In Defence of Traditional Methodologies  65 rule of recognition’, and confirmed that the common law cannot validly contradict statute law.138 When Miller No 2 was decided by the High Court, a new Chief Justice, the Master of the Rolls and the President of the Queen’s Bench Division again re-affirmed the orthodox view: The concept of Parliamentary Sovereignty recognises that the Queen in Parliament is able to make law by primary legislation without legal restraint, save such restraint as it has imposed on itself for the time being …. The interpretation of legislation is for the courts which seek to give effect to the intention of Parliament divined from the statutory language, examined in accordance with established principles of statutory interpretation.139

On appeal, all 11 members of the Supreme Court unanimously affirmed that a ­fundamental principle of the Constitution is the principle of Parliamentary sovereignty: that laws enacted by the Crown in Parliament are the supreme form of law in our legal system, with which everyone, including the Government, [and logically, therefore, the courts] must comply.140

These clear statements by so many senior judges are a major obstacle for those who made so much of the dicta in Jackson’s case that challenged parliamentary sovereignty, despite those dicta being heavily based on demonstrable errors of fact and law.141 It is simply not plausible to discount the statements in the Miller cases as mere dicta that can be rejected as dispensable ‘mistakes’ due to their being inconsistent with the morally best interpretation of the constitution as a whole. The common law constitutionalists must rely on two other recent and overlapping developments. The first is the occasional reliance on ‘constitutional’ rights or principles in order to interpret statutes contrary both to Parliament’s words and to its obvious intention in uttering them, which contradicts parliamentary sovereignty in substance if not in form; the second are dicta suggesting that judges will refuse to obey privative or ouster clauses that diminish their jurisdiction to review the decisions of administrators, tribunals and inferior courts.142 138 R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5, paras 42, 43, 60, 61, 67 (the majority), 173, 223–27 and 183 (Lord Reed), and 274 (Lord Carnwath). 139 R (Miller) v The Prime Minister [2019] EWHC 2381 (QB), para 61. 140 R (Miller) v The Prime Minister; Cherry v Advocate General for Scotland [2019] UKSC 41, para 41. 141 R (Jackson) v Attorney General [2005] UKHL 56, [2006] 1 AC 262, [102] (Lord Steyn), [104], [107], [126] (Lord Hope), [159] (Baroness Hale). Lord Steyn’s assertion at [102] that the judges created the principle of parliamentary sovereignty is demonstrably false, as is Lord Hope’s related assertion at [126] that the principle was ‘created by the common law’, if he means judge-made law. See Goldsworthy, The Sovereignty of Parliament (n 15) ch 10; Goldsworthy, Parliamentary Sovereignty (n 15) ch 2. Lord Hope’s suggestion at [104]–[106] that the Acts of Union 1707, the European Communities Act 1972, and the Human Rights Act 1998 all qualified or limited parliamentary sovereignty are highly questionable: Goldsworthy, The Sovereignty of Parliament (n 15) 165–73; Goldsworthy, Parliamentary Sovereignty (n 15) 287–304. The strongest case concerned the European Communities Act, but Brexit has surely put that to rest: see in particular R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5, [60] (Lord Neuberger, Lady Hale, Lords Mance, Kerr, Clarke, Wilson, Sumption and Hodge). See also Gordon, Parliamentary Sovereignty in the UK Constitution (n 16) 204–206. 142 It is not surprising that judges are determined to defend and even expand their own authority, through whatever creative or implausible arguments might seem useful; this has happened around the world. See J Goldsworthy, Interpreting Constitutions; A Comparative Study (Oxford, Oxford University Press, 2006) 345.

66  Jeffrey Goldsworthy But given the clear and strong affirmation of parliamentary sovereignty in the Miller cases, these other developments are much more consistent with legal pragmatism than common law constitutionalism.143 Implausible interpretations of statutes, which flout Parliament’s authority surreptitiously, are hardly ringing endorsements of a common law constitution purporting to limit that authority.144 Since Anisminic this has been the preferred strategy in dealing with statutes designed to restrict the courts’ judicial review jurisdiction, although in Privacy International members of the Supreme Court flirted with two other strategies. One of these strategies, endorsed by Lord Carnwarth, Lady Hale and Lord Kerr, is consistent with common law constitutionalism. They asserted that: the rule of law is the most basic principle of the British system of law, explicitly recognised by Parliament itself;145 ‘there are certain fundamental requirements of the rule of law which no form of ouster clause (however “clear and explicit”) could exclude from the supervision of the courts’;146 and ‘a natural application of the constitutional principle of the rule of law’ is that ‘it is ultimately for the courts, not the legislature, to determine the limits set by the rule of law to the power to exclude judicial review’.147 The latter two statements are obiter dicta,148 and were in effect repudiated by Lords Sumption, Reed and Wilson.149 But Lord Carnwath, Lady Hale and Lord Kerr also endorsed a second, different strategy, which also found favour with Lords Lloyd-Jones, Sumption and Reed.150 Adopting an argument first made by Sir John Laws,151 these six judges proposed that parliamentary sovereignty itself requires that independent courts have jurisdiction to review actions of administrators, tribunals and inferior courts, to ensure that they comply with Parliament’s statutes and not thwart its lawmaking authority.152 On this view, Parliament’s inability to diminish that jurisdiction is akin to its inability to bind itself in the future: both these limits to its authority are, in a sense, merely apparent, because they are necessary to maintain that same authority. Both limits are conceptual and internal to Parliament’s sovereignty, rather than normative and ­external to it.153 This strategy is unconvincing. A sovereign Parliament is able to protect its own authority, and decide for itself on the best methods for doing so. It is somewhat ironic that judges should claim that the courts’ power to interpret statutes is essential

143 See section III, above. 144 See text to n 62, above, on the significance of a lack of judicial candour. 145 R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22, para 119, 120 and 121. 146 Paras 122, 123 and 144. 147 Paras 131, 132 and 144. This statement is inconsistent with the endorsement of parliamentary sovereignty by Lady Hale and Lord Kerr in both Miller cases, and by Lord Carnwath in Miller No 2: see nn 137 and 140, above. 148 Para 119: ‘We are not therefore concerned with the difficult constitutional issues which might arise if Parliament were to pass legislation purporting to abrogate or derogate from those accepted principles.’ 149 Paras 208 and 209. 150 Lord Sumption clearly distinguishes between the two strategies at para 208. 151 R (Cart) v The Upper Tribunal [2011] QB 120, paras 36–40. 152 Privacy International, n 145, para 132 (‘as an essential counterpart to the power of Parliament to make law’), read with paras 114 and 116, per Lord Carnwath, Lady Hale and Lod Kerr. 153 ibid, para 169 (Lord Lloyd-Jones), 208 and 210 (Lords Sumption and Reed).

In Defence of Traditional Methodologies  67 to maintain Parliament’s sovereignty, when sometimes courts have conspicuously refused to interpret statutes with fidelity to Parliament’s obvious intention.154 That track record is one reason why, in exceptional circumstances, Parliament might reluctantly conclude that it has good reason to enact a privative clause. If it is truly sovereign, it can choose to vest interpretive and enforcement powers in an agency other than the regular courts. To say that it cannot do so is to say that it is not sovereign. I acknowledge that Parliament would be prudent not to press this point through legislation, given the judges’ adamant opposition to it. This is one occasion when a mutually respectful modus vivendi is sensible. But for present purposes, a more important point is that this second strategy is difficult to reconcile with common law constitutionalism. Given the Supreme Court’s express commitment to parliamentary sovereignty, Lord Lloyd Jones, Sumption and Reed are clearly seeking to avoid any impression that they advocate limiting Parliament’s authority in order to protect ‘the rule of law’, except insofar as this is essential to maintaining parliamentary sovereignty itself. This is not an endorsement of common law constitutionalism. It might be argued that by declaring parliamentary sovereignty to be ‘a’ fundamental principle of the Constitution, the Supreme Court in the Miller cases committed itself only to accepting parliamentary sovereignty as one such principle along with others, so that in certain circumstances it might be outweighed by one or more of them, and in particular, the rule of law.155 The difficulty is that this would be inconsistent with Parliament being, truly, sovereign. It is one thing to maintain that there are two or more constitutional principles; even Dicey acknowledged that the rule of law is one of them.156 But the doctrine of parliamentary sovereignty holds that all statutes enacted according to lawful manner and form are legally valid; if statutes could sometimes be declared invalid or simply disapplied because they violate some other constitutional principle, then Parliament would not be sovereign. It is sheer self-contradiction to maintain both that Parliament is sovereign, and that its sovereignty may sometimes be outweighed by competing constitutional principles. If that were the Supreme Court’s position, it would be another indication of an essentially pragmatic approach to the issue. If the Court is to embrace common law constitutionalism, it should frankly declare that Parliament is not sovereign, and face the inevitable and fierce political backlash. This brings us to the common law pragmatists, who realise that in attempting to limit parliamentary sovereignty the courts would be attempting to change the Constitution – in effect, to construct a common law constitution, rather than enforce an existing one.157 Their position is open to inter-related jurisprudential and political objections. 154 A prominent example is Evans v Attorney-General [2015] UKSC 21; 2 WLR 813. It is ironic that Lord Neuberger, who wrote one of the majority’s leading judgments in that case, had previously endorsed the doctrine of parliamentary sovereignty: ‘Who Are the Masters Now? Second Lord Alexander of Weedon Lecture’, 6 April 2011. 155 A thesis defended by J McGarry, in ‘The Principle of Parliamentary Sovereignty’ (2012) 32 Legal Studies 577, and Intention, Supremacy and the Theories of Judicial Review (London, Routledge, 2016) ch 6. 156 Dicey, Introduction to the Study of the Law of the Constitution (n 92) ch 4. 157 See n 102 above.

68  Jeffrey Goldsworthy What authority does the judiciary have to change the Constitution in such a fundamental respect, according to common law pragmatists? They are not Dworkinian interpretivists. And they can no longer put forward the simplistic thesis that the doctrine of parliamentary sovereignty is one of judge-made common law, which the judiciary has legal authority to change unilaterally. That thesis has been refuted.158 Of course, they could nevertheless urge the judiciary to claim, or at least purport to exercise, a legal authority it does not possess, and hope to get away with it. They could rely on the legal positivist thesis that the doctrine of parliamentary sovereignty is part of the rule of recognition underpinning the British Constitution, and is therefore constituted by widespread consensus among senior legal officials.159 The judges could therefore stage a partial withdrawal from the consensus they previously supported, while attempting to modify it. The problem is that while the doctrine could be subverted by the judges withdrawing their acceptance of it, they could not replace it with an alternative of their choosing without securing the agreement of other senior officials. Those officials would not be legally or morally bound meekly to accept that the judiciary has authority to change such a fundamental and longstanding element of the Constitution. So judicial repudiation of the doctrine might lead to uncertainty, disagreement, instability or even dangerous conflict at the highest levels of government. Common law pragmatists might be prepared to take that risk. Many pragmatists downplay legal formalities and legal logic; that the judiciary lacks legal authority unilaterally to change certain kinds of laws would not weigh heavily with them. They applaud judicial creativity provided it is ‘progressive’ and successful. There are undoubtedly historical examples of judicial creativity that changed the Constitution in positive ways, despite its flouting legal orthodoxy at the time. It can be successful if other senior officials are persuaded of its merits, or are unwilling to resist it, perhaps because they are taken unawares and decide that it is too late or politically imprudent to resist. Judicial pragmatism is in practice often surreptitious rather than open, concealed by ‘fig leaves’ professing fidelity to orthodoxy.160 Mark Aronson has observed that during the early development of judicial review of administrative decisions, the courts ‘may have disguised some of their more creative moves behind a veil of technicality and dubious precedents’.161 Masterman and Wheatle observe that ‘[f]urtive steps are being taken’ to promote ‘a constitutional common law’, through the recent recognition of ‘constitutional common law rights’ and ‘constitutional statutes’.162 Common law pragmatists might urge judges to continue to proceed slowly and cautiously, sometimes dissembling as to the true reasons for their approach, by developing a growing body of decisions and dicta that can eventually (if not already) be declared to have ushered in ‘a new constitutional settlement’.163 158 See Goldsworthy, The Sovereignty of Parliament (n 15) ch 10; Goldsworthy, Parliamentary Sovereignty (n 15) ch 2. 159 Goldsworthy, ‘Parliamentary Sovereignty and Constitutional Change’ (n 81) and section VI, above. 160 See section III, above. 161 M Aronson, ‘Retreating to the History of Judicial Review?’ (2019) 47 Federal Law Review 179, 180. 162 Masterman and Wheatle, ‘Unity, Disunity and Vacuity’ (n 102) 147. 163 Privacy International (n 145), para 131 (Lords Carnwarth, Hale and Kerr).

In Defence of Traditional Methodologies  69 The political objection to this course of action is that the judiciary lacks moral as well as legal authority to attempt to impose restraints or limits on Parliament’s authority without any democratic input or warrant. Lord Neuberger shrewdly observed that ‘it is a feature of all constitutional courts that that they generously interpret the constitution and tend to bestow power on themselves.’164 But as Lord Millett cautioned, ‘any change in a fundamental constitutional principle should be the consequence of deliberate legislative action and not judicial activism, however well meaning’.165 In Miller No 1, eight members of the Supreme Court stated that: We cannot accept that a major change to UK constitutional arrangements can be achieved by ministers alone; it must be effected in the only way that the UK constitution recognises, namely by Parliamentary legislation. This conclusion appears to us to follow from the ­ordinary application of basic concepts of constitutional law to the present issue.166

This is consistent with a referendum being morally required for truly fundamental constitutional change, because Parliamentary legislation would be needed. How ironic if the Supreme Court itself were to claim legal authority to bring about – or to have already brought about – such a change, without the assent of Parliament expressed in legislation, or of the people expressed in a referendum.

164 Lord Neuberger, ‘The Role of Judges in Human Rights Jurisprudence: a Comparison of the Australian and UK experience’ (Speech at Supreme Court of Victoria Conference, 8 August 2014), 19. See also n 142 above. 165 Ghaidan v Godin-Mendoza [2004] UKHL 30; 2 AC 557 [57]; see also Bingham, The Rule of Law (n 80), 168–170. 166 R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5, para 82; see also para 100 (Lord Neuberger, Lady Hale, Lords Mance, Kerr, Clarke, Wilson, Sumption, and Hodge). See also para 173 (Lord Reed): ‘Since that will be a fundamental alteration in the UK’s constitution, it can only be effected by Parliamentary legislation.’

70

3 Constitutional Methodology and Brexit: Adopting a Model-Theoretic Approach ALISON L YOUNG*

B

rexit has dominated the legal, political and constitutional landscape since the referendum. It set the stage for, at the time of writing, the only two cases in which the Supreme Court sat in its maximum composition of 11 Justices.1 It also gave rise to the first reference concerning the validity of an Act of the Scottish Parliament,2 and a reference under Article 267 of the Treaty on the Functioning of the European Union (TFEU) concerning whether EU law permitted states to ­unilaterally revoke a notification to the European Council to withdraw from the EU Treaties.3 In terms of its political impact, Brexit has, at least indirectly, caused: the resignation of two Prime Ministers; the calling of two early parliamentary general elections; the enactment of two Private Members Bills which required the Prime Minister to act contrary to governmental policy;4 novel interpretations of parliamentary Standing Orders, particular Standing Order No 24; a greater role for the legislature in the scrutiny of international Treaties under the provisions of the so-called meaningful vote;5 and evidence both of the legislature providing detailed and exacting scrutiny of primary legislation designed to facilitate Brexit,6 and of legislation enacted swiftly with little, if any parliamentary scrutiny to implement the Withdrawal Agreement,7

* The author wishes to thank Hayley J Hooper, Leah Trueblood and the editors of this collection for their comments on earlier drafts. All errors remain those of the author alone. 1 R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5, [2018] AC 61 and R (Miller) v Prime Minister; Cherry v Advocate General for Scotland [2019] UKSC 41, [2020] AC 373. 2 In re UK Withdrawal from the European Union (Legal Continuity)(Scotland) Bill [2018] UKSC 64 [2019] AC 1022. 3 C-621/18 Wightman v Secretary of State for Exiting the European Union [2019] QB 199. 4 The European Union (Withdrawal) Act 2019 and the European Union (Withdrawal) (No 2) Act 2019. 5 European Union (Withdrawal) Act 2018, s 13. 6 See further M Russell, ‘Brexit and Parliament: The Anatomy of a Perfect Storm’ [2020] Parliamentary Affairs 1 and A McConnell and S Tormey, ‘Explanations for the Brexit policy fiasco: near-impossible ­challenge, leadership failure or Westminster pathology?’ (2020) 27 Journal of European Public Policy 685. 7 European Union (Withdrawal Agreement) Act 2020, which was enacted in three weeks.

72  Alison L Young and the Cooperation and Trade Agreement between the UK and the European Union.8 Its impact on the UK Constitution is harder to assess – not least because those adopting different background constitutional theories have, in turn, used these ­theories to shape their interpretation of the impact of Brexit on the UK Constitution. Brexit has been analysed through the lenses of political constitutionalism9 and common law constitutionalism,10 as well as providing both a constitutional moment in which the UK should move towards the adoption of a new constitutional settlement with a written, codified constitution,11 and an illustration and application of the long-standing traditions of the UK Constitution.12 It is hard to deny that Brexit has posed a series of constitutional challenges. It has placed the Union of the United Kingdom under tension. The use of a referendum also challenges parliamentary sovereignty, particularly as the will of the people challenged the will of Parliament. Whilst these may seem interesting issues in terms of law, politics, and the UK Constitution, why, if at all, should they be of interest to constitutional theorists, or a chapter examining the methodologies of constitutional theory? Such events, at best, would appear to be peripheral, or, at worst, distractions from the elevated, abstract thoughts of a constitutional theorist. This chapter will argue that Brexit provides a backdrop against which to illustrate the relative strengths, weaknesses, and limitations of different methodologies of constitutional theory. This is precisely because it provides a series of examples which questioned the empirical data which inform generalisations made about the UK Constitution. It will first set out the methodology I applied when developing my theory of democratic dialogue – a model-theoretic approach to constitutional theory.13 As well as providing an account of the model-theoretic approach, the chapter will explore how this approach to constitutional theory differs from other approaches. It will do so by explaining the specific way in which a model-theoretic approach develops models of constitutions, as well as exploring the advantages of adopting a model-theoretic approach to provide an account of the differences in methodology between modeltheoretic and other approaches to constitutional theory. Finally, it will provide a further example of the distinct nature of the model-theoretic approach by comparing and contrasting different theoretical accounts of the impact of Brexit on parliamentary sovereignty. 8 European Union (Future Relationship Act) 2020, enacted in one day through both Houses of Parliament, receiving just 4 minutes of legislative scrutiny by the House of Commons at the committee stage. 9 See, eg, K Ewing, ‘Brexit and Parliamentary Sovereignty’ (2017) 80 MLR 685, M Gordon, ‘Parliamentary Sovereignty and the Political Constitution(s): From Griffith to Brexit’ (2019) 30 King’s Law Journal 125 and ‘Referendums in the UK Constitution: Authority, Sovereignty and Democracy after Brexit’ (2020) 16 European Constitutional Law Review 1. 10 See, eg, D Campbell, ‘Marbury v Madison in the UK: Brexit and the Creation of Judicial Supremacy’ (2018) 39 Cardozo Law Review 921 and G Anthony ‘Brexit and the Common Law Constitution’ (2018) 24 European Public Law 673. 11 B Ackerman, ‘Why Britain Needs a Written Constitution – and Can’t Wait for Parliament to Write One’ (2018) 89 Political Quarterly 584 and V Bogdanor Beyond Brexit: Towards a British Constitution (New York, IB Taurius, Bloomsbury, 2019). 12 M Taylor, ‘Brexit and the British Constitution: A Long View’ (2019) 90 Political Quarterly 719. 13 A L Young Democratic Dialogue and the Constitution (Oxford, Oxford University Press, 2017) ch 6.

Constitutional Methodology and Brexit  73 I.  DEFINING MODEL-THEORETIC APPROACHES

The model-theoretic approach finds its origins in the semantic approach to scientific theories, developed by Suppe, van Fraassen, Giere and Suppes.14 The central tenet of a model-theoretic approach to scientific theories argues that scientific theories are not linguistic entities but set-theoretic entities. In other words, theories are not collections of linguistic propositions, explored through an analysis of their meaning. Instead, theories are ‘extralinguistic entities which may be described or characterised by a number of different linguistic formulations’.15 These extralinguistic entities are referred to as models. According to model-theoretic approaches, scientific theories are theories about models. Scientists develop abstract models that aim to replicate, in a more simplified manner, complex realities. These models are then used to predict behaviour. The results from the model are tested against results from experiments. Models are evaluated both descriptively and normatively. A good model needs to fit well with reality. Consequently, empirical data is used to test the model to determine how well it fits with reality. In addition, the model is evaluated to determine how well it serves its purpose, as well as evaluating the importance of its purpose. It may seem odd to suggest that a methodology that was originally developed to explain scientific theories would have any application to law, let alone to constitutional theory. However, others have proposed similar arguments in favour of a model-theoretic approach to law. Roederer, for example, advocates adopting a modeltheoretic approach to legal theory,16 suggesting also that this approach is preferable for areas of applied legal theory, such as constitutional theory.17 Chapter eleven in this volume also illustrates an adoption of a model-theoretic approach, examining how far models of improvisation can provide a means of explaining recent UK ­constitutional developments and the consequences of adopting this approach.18 Moreover, model-theoretic approaches require theorists to adopt models of constitutions. It is not hard to find specific reference to the use of models in works examining aspects of constitutional theory, or of constitutions. Stephen Gardbaum, for example, refers to his theory as the ‘new commonwealth model’ of constitutionalism.19 Gee and

14 See F Suppe, The Structure of Scientific Theories (Urbana-Champagne, University of Illinois Press, 1974) and The Semantic Conception of Theories and Scientific Realism (Urbana-Champagne, University of Illinois Press, 1989), P Suppes ‘A Comparison of the meaning and use of models in mathematics and the empirical sciences’ in P Suppes (ed), Studies in the Methodology and Foundations of Science (Reidel, Dordrecht, 1961), BC van Fraassen ‘On the extension of Beth’s semantics of physical theories’ (1970) 37 Philosophy of Science 325 and ‘The semantic approach to scientific theories’ in N Nersessian (ed), The Process of Science (Kluwer, Dordrecht, 1987) and R Giere Explaining Science: A Cognitive Approach (Chicago, University of Chicago Press, 1988). 15 Suppe, Structure (n 14) 221. 16 C Roederer, ‘Negotiating the Jurisprudential Terrain: A Model Theoretic Approach to Legal Theory’ (2003) 37 Seattle University Law Review 385. 17 ibid 44–450. 18 D Howarth, ‘The British Constitution as an Improvised Order’, ch 11 in this volume. 19 S Gardbaum, The New Commonwealth Model of Constitutionalism (Cambridge, Cambridge University Press, 2016), ‘Reassessing the New Commonwealth Model of Constitutionalism’ (2010) 8 ICON 167 and ‘The New Commonwealth Model of Constitutionalism’ (2001) 49 American Journal of Comparative Law 707.

74  Alison L Young Webber also note that accounts of political constitutionalism often refer to models.20 Common law constitutionalism, political constitutionalism and legal constitutionalism are often presented as rival models of the UK Constitution. However, mere examples of theorists advocating or using model-theoretic approaches does not provide sufficient support for adopting a model-theoretic approach to constitutional theory – though it may suggest that adopting a model-theoretic approach is possible. In order to investigate the model-theoretic approach further and assess its suitability as a methodology for constitutional theory, we need to explore what is meant by a model, whether it is possible to develop models of constitutions and, if so, whether there is any value in doing so. A model is an idealised abstraction of reality. It can operate both as a map and as a simplification, choosing to highlight particular characteristics. When we think of models, we often imagine a scale model – for example, a scale model of Ely Cathedral. The purpose of a scale model is to replicate reality. We would evaluate different scale models of Ely Cathedral according to the accuracy with which they reproduced the cathedral on a smaller scale. When providing an account of a model in constitutional theory, however, we are clearly not referring to a scale model. Although we can appreciate scale models of the Houses of Parliament, or of the Supreme Court, these are not the work we expect from constitutional theorists. A constitutional theorist adopting a model-theoretic approach would both adopt a different type of model than a scale model, as well as devising the model for a different purpose. In addition to scale models, there are also analogue and theoretical models. Analogue models focus on the relationships between different elements of the subject of the model. Unlike a scale model designed to replicate Ely Cathedral, an analogue model applied to the cathedral could be used to explain the patterns of visitors to the cathedral – for example how long they spent visiting the cathedral, which features of the cathedral were the most popular, how weather patterns or the holding of events impacted on the number of visitors to the cathedral, and whether visitor numbers increased following the use of the cathedral tower as a nesting site by a pair of peregrine falcons. Analogue models would not just focus on providing a daily tally of footfall in Ely Cathedral, but would aim to relate this footfall to other events. The purpose of an analogue model can be either to provide a better explanation of a complex series of relationships, to help further understanding, or to provide a basis from which to make predictions of future behaviour. Those wishing to increase visitors to Ely Cathedral might use an analogue model to determine which aspects of the Cathedral to renovate, or to feature in their publicity campaigns. Theoretical models are used to provide a novel means of investigating and potentially understanding a complex subject. Like analogue models, they can be used as heuristic devices to help further understanding. However, they can also be developed as ideal-types. Theoretical models can sometimes deploy metaphors, using more familiar concepts or patterns applied to a new field of investigation to help further understanding of this new field. The use of metaphors and models may help to provide information or explanations that would not otherwise be achieved,

20 G

Gee and GCN Webber ‘What is a Political Constitution?’ (2010) 20 OJLS 273, 290–291.

Constitutional Methodology and Brexit  75 helping theorists to explore fields of investigation in a new manner. Like analogue models, theoretical models provide explanations and hypotheses whose accuracy needs to be checked in comparison with the primary empirical data.21 Both analogue and theoretical models can be used when adopting a model-theoretic approach to constitutional theory. However, whilst a model-theoretic approach uses models, it is not the only methodological approach to do so. To understand the distinct nature of a model-theoretic approach to constitutional theory, we need to draw further distinctions between different types of model. When providing an account of different models of sovereignty, David Held describes models as ideal types or heuristic devices which order a field of inquiry. They assist in clarifying the primary elements or constitutive characteristics of a domain, although in so doing they bring them into sharper relief and delineation than can be found within the ‘messier’ world of everyday law and politics.22

Similarly, Gee and Webber argue that theorists use models to provide a simplified account of the constitution, whilst appealing to a particular ideal.23 The unique feature of a model-theoretic approach is that it recognises the distinct features of heuristic models and ideal-types, accepting that all models have both a normative and a descriptive dimension. Most constitutional theorists who refer to models use them as heuristic devices. When acting as a heuristic device, a model aims to reflect real experiences and provide a simplified account of essential elements of a particular aspect of a constitution. The purpose of a heuristic device is to explain or describe. The dual purpose of a heuristic model to describe and explain means that heuristic models require a balance to be made between simplification and accuracy. A completely accurate model may be too complicated to further understanding, whereas an over-simplified model may further understanding at the cost of accuracy. Gardbaum’s ‘new Commonwealth model’ of constitutionalism, for example, is best understood as a heuristic model.24 His work provides an explanation of a novel form of human rights protection found in Canada, New Zealand, Australia and the UK by providing an account of the common features of the Canadian Charter of Fundamental Rights and Freedoms, the New Zealand Bill of Rights Act 1990, the Victorian Charter of Human Rights and Responsibilities Act 2006 and the Australian Capital Territories Human Rights Act 2004 and the Human Rights Act 1998. These features are contrasted with the form of human rights protections found in systems such as the US or Germany, where a constitutional court has the power to strike down legislation which contravenes constitutionally protected human rights. Gardbaum’s model identifies four key features of the new Commonwealth model of constitutionalism: (i) a codified charter of rights found in a legal document;

21 For further discussion, see M Black, Models and Metaphors (Ithaca, Cornell University Press, 1962) 219–243. 22 D Held, ‘Law of States, Law of Peoples: Three Models of Sovereignty’ (2002) 8 Legal Theory 1, 1–2. 23 Gee and Webber, ‘What is a Political Constitution?’ (n 20) 291. 24 See the references in n 19.

76  Alison L Young (ii) a requirement for political branches of the Constitution to review legislation for its compatibility with human rights protections during the enactment of legislation and prior to the legislation coming into force; (iii) a provision for courts to review legislation for its compatibility with human rights that goes beyond a requirement to interpret legislation in a manner that respects human rights; and (iv) the ability of the legislature to have the final say on the content of legislation by enacting laws using the ordinary procedure for enacting legislation.25 However, these key features of the new Commonwealth model of rights protections do not provide a fully ­accurate account of any one specific example of this model of rights protections. The model also glosses over significant differences in approach between the protection of human rights in Canada, New Zealand, Australia and the UK. This is not to criticise Gardbaum’s model. Rather, it is to recognise that the lack of accuracy is the price paid for developing a heuristic model that is able to set out and explain the salient features of an emerging novel form of human rights protection. Models may also provide an ideal-type; an abstract concept which presents an ideal derived from an assessment of important features.26 Ideal-type models are similar to heuristic models as they too simplify and focus upon key features when developing a model. Ideal-type models, however, have a different purpose from heuristic models. The aim of an ideal-type is to provide a ‘pure case’, abstracted from empirical data but not dependent on empirical data for its validity.27 The model does not merely provide a description of a particular concept, but also provides an idealised version of that concept, often based on a predominantly normative account. An ideal-type may be theoretically possible but need not be reflected or exemplified in empirical data. Ideal-types are often used not just to describe, but to evaluate particular constitutional principles, concepts, laws, or events. As a model, it too is simplified and may overlook complications. In addition, when contrasted to reality, empirical data often fails to comply with any one particular ideal-type but will nevertheless incorporate aspects of a range of ideal-types. Ideal-types are often used to assess current constitutional arrangements, contrasting the reality with a theoretically feasible, if hypothetical, ideal. For example, Ian Cram criticises the control of emergency powers in the UK by analysing the UK’s arrangements according to the ideal-types of ‘democratic formalist’ and ‘liberal common law’ schools. Both of these are models that Cram adapts from Schauerman.28 Cram uses these models to analyse primary legal materials and prevailing academic commentary. His aim is to illustrate a shift in UK public law scholarship exemplified by a change in model. The types of model developed in a model-theoretic approach to constitutional theory are neither purely heuristic nor purely ideal-type models falling, instead, between the two. They are best understood as a constructed-type models. To understand what is meant by a constructed-type model, we need to revisit our account of 25 Gardbaum, New Commonwealth Model (n 19) 29–31. 26 F Weinert, ‘Weber’s Ideal Types as Models in Social Sciences’ (1996) 41 Royal Institute of Philosophy Supplement 73. 27 ibid 74. 28 I Cram ‘Beyond Lockean Majoritarianism – Emergency Institutional Failure and the UK Constitution’ (2010) 10 Human Rights Law Review 461, 475–479.

Constitutional Methodology and Brexit  77 ideal-types. We described ideal-types as providing a more abstract, idealised version of reality than heuristic models. Heuristic models aim to explain. Whilst they may simplify reality, there is still a need for heuristic models to fit sufficiently to empirical data to ensure that they can perform their purpose of explaining aspects of real constitutions. Ideal-types do not require the same relationship to empirical data. However, Weber’s account of ideal-types refers to these models both as purely abstract entities, designed to provide an idealised model that may not need to describe or explain real data and to models that have at least some relationship to empirical data. Weinert describes this as a tension between ideal-types which provide an account of objective possibilities and ideal-types which provide an account of objective probabilities.29 When ideal-types provide an account of objective possibilities, they focus more on providing an idealised or normative model. Their aim is to provide an account of the essential components of a particular concept that may be possible to achieve, but which need not reflect the real world. When understood in terms of objective probabilities, ideal-types have more of a connection to real data. There is no requirement that the ideal-type describe or reflect empirical data. Rather, ideal types focus on principles. However, rather than aiming to provide as pure an account as possible of the essence of a particular principle or concept, the ideal-type aims to provide an account that it is probable to achieve in reality. Having drawn this distinction, Weinert provides an account of constructed-types, drawing on the work of McKinney and Becker, which he regards as providing a form of ideal-type understood in terms of objective probability as opposed to objective possibility. Weinert concludes that these models are best suited for use by those adopting a model-theoretic approach to social sciences.30 In a similar manner, constructedtype models provide the best model for a model-theoretic approach to constitutional theory. Constructed-type models in constitutional theory aim to provide ideals of constitutional government, or of specific constitutional principles – eg, parliamentary sovereignty or the rule of law. These constructed-type models need not map reality but do need to provide an account that it would be probable to achieve in the real world, either generally, or in a specific constitutional system. The aim of a constructed-type model is neither purely heuristic, nor purely theoretical. However, it aims both to provide an account of a normatively defensible ideal of constitutionalism and of a model that it may be feasible or probable to achieve in a particular legal system. The more feasible it is to achieve the constructed-type model, the more the model serves as a heuristic device. I would argue that the success, or otherwise, of a constructed-type model can be determined according to three factors – although not all three may be needed depending on the purpose of the model and of the perspective from which the model was being evaluated. First, there is a need for a model to fit reality, not in terms of providing an accurate account of empirical data, but in terms of resting on assumptions that are at least feasible or possible to achieve in a particular constitutional system. Second, the model itself has to have internal consistency, even if this is to recognise that the constitution it



29 Weinert, 30 JC

‘Weber’s Ideal Types’ (n 26) 79. McKinney, Constructive Typology and Social Types (New York, Appleton-Century-Crofts, 1966).

78  Alison L Young fits is one in which there can be ad hoc constitutional reform with resulting haphazard and inconsistent outcomes. Whilst a constitution can be haphazard, a model of that constitution that was internally inconsistent – ie, one that suggested both that constitutional change was always consistent or always haphazard – would fail as a constitutional model. This is particularly true when developing a constructed-type analogue model, which focuses on explaining the relationships between different institutions of the constitution. Third, the model can be assessed according to its normative credentials, evaluating the norms and ideals underpinning a particular model of the constitution. This criticism is context-dependent. We may criticise a model that aims to achieve normative values with which we do not agree, however, this criticism may not mean that the model itself is invalid. It may merely be an undesirable model. II.  WHAT IS DISTINCTIVE ABOUT A MODEL-THEORETIC APPROACH TO CONSTITUTIONAL THEORY?

Adopting a model-theoretic approach to constitutional theory enables a theorist to develop a normative constitutional theory, or to focus on producing models that may facilitate the achievement of constitutionally desirable outcomes (eg a model of human rights that is more likely to produce the ‘right’ outcome in a human rights case) or facilitate the achievement of principles considered by a particular theorist to be desirable (eg the facilitation of a particular conception of the rule of law). However, it does not entail the claim that these normative values are required in order for a particular set of institutional arrangements to be regarded as ‘constitutional’. Rather, the theorist is merely setting out a feasible model of a particular constitution, particular types of constitution, or constitutions writ large, designed to achieve particular normative goals. This model may well be compatible with other models of the constitution, particularly as models simplify and may, for example, only focus on one particular way of analysing the relationship between different governmental institutions. However, despite a model-theoretic approach including both descriptive and normative elements, it does not confuse descriptions of the constitution with assertions of normative value – ie, confusing an ‘is’ with an ‘ought’. Rather, the model proposes a possible framework from which to analyse constitutions. The model also aims to explain constitutions, and modifications to the model might need to be made in the light of empirical data. Nevertheless, questions as to the extent to which a model fits a particular legal system, or as to the internal logic of the model, or the desirability of the values it promotes, are separated. They provide different means of evaluating a particular model. In order to understand this in more detail, we need to compare and contrast a model-theoretic approach with other approaches to constitutional theory. There are two ways in which we can distinguish a model-theoretic approach from other general approaches to constitutional theory. First, we can investigate whether other approaches to constitutional theory use models and, if so, the type of model they use. Second, we can evaluate whether there are methodological differences between model-theoretic

Constitutional Methodology and Brexit  79 and other approaches to constitutional theory. This section will focus predominantly on the differences between the types of model used in a model-theoretic approach and other approaches to constitutional theory. The second distinction will be examined in more detail in the following section. As has been noted by others working in constitutional theory, it can be difficult to provide an account of distinct methodological approaches to constitutional theory as those working in the field rarely discuss methodology.31 Nick Barber provides an account of three possible approaches to constitutional theory: the historically-minded; the critical constitutional approach; and the interpretative approach. Although he regards these as different and compatible paths, it is clear from his account both that Barber adopts an interpretative approach and that he regards this approach as foundational.32 Barber’s account of different approaches provides a starting point from which to distinguish a model-theoretic from other approaches to constitutional theory. Barber describes a historically-minded approach as one which aims to provide an account of the historical origins of a particular constitution. For example, a theorist adopting a historically-minded approach to the UK Constitution would examine historical events or legal sources that played a key role in the establishment of the UK Constitution. The aim of a historical approach is to provide key evidence of the origins of a constitution in order to provide a historically accurate account of a constitution or a specific constitutional principle. Those who adopt a critical approach to constitutional theory devise a normative framework, often relying on ethical or moral principles, through which to evaluate the UK Constitution. An interpretative approach to constitutional theory also aims to provide an account of constitutional principles. However, it does not do so from a purely historical perspective. It focuses on setting out an account of the key features of these principles, providing an account of a central or standard interpretation of these principles as opposed to an account of the historical origins of a particular principle of the constitution. These principles may then be used to critically evaluate a particular constitution according to how well it promotes these constitutional values. The extent to which this is possible depends upon the level of abstraction of the interpretative approach. An abstract account of the key features of the rule of law may be used to criticise the account of the rule of law in a particular legal system. An interpretative approach to the rule of law in the UK Constitution could be used to criticise a specific legal case that used the rule of law in a manner which diverged from these key features. Those adopting a historically-minded approach to constitutional theory tend not to use models when providing their account. According to Barber, the approach of those adopting a historically-minded approach is similar to that adopted by naturalists in the natural sciences. Most historically-minded approaches to constitutions are descriptive. They aim to describe the origins of the constitution. To the extent that

31 N W Barber, The Constitutional State (Oxford, Oxford University Press, 2010) ch 1. 32 See also M Lewans, ‘Common Understandings of Administrative Law’, ch 14 in this volume. Dimitrios Kyritsis also appears to take a similar approach, providing an account of constitutional law that is best able to capture its salient features. See D Kyritsis, ‘Constitutional Law as Legitimacy Enhancer’ ch 8 in this volume.

80  Alison L Young this is an accurate account of historically-minded approaches to constitutions, they differ from model-theoretic approaches in that these accounts do not rely on models. They are not providing an account of a non-linguistic entity. Nor do they aim to provide an account of the salient features of a constitution, other than to explain the historical origins of these features. Both critical and interpretative approaches to constitutional theory lend themselves to the use of models. Barber’s approach to interpretative theory, for example, develops common understandings. Barber argues that common understandings are similar to concepts which, in turn, are similar to models. He refers in particular to prototype and exemplar models of concepts or common understandings. Barber refers to a prototype as similar to an ideal-type. It provides an account of the essential features of a constitution or a constitutional principle. These are features that are shared by the principle, but it is not the case that all examples of the principle need to share these features. He refers to exemplar models as a collection of memories or encounters with the principle. The purpose of common understandings is to aid understanding. They provide a means of judging accounts of constitutional principles, in terms of their accuracy. Any account of a constitutional principle that does not sufficiently fit the common understanding can be rejected as an inaccurate account of that constitutional principle.33 Both prototypes and exemplars are best understood as heuristic models. Their aim is to further understanding by providing an account of common understandings of a constitution, aspects of a constitution, or of constitutional principles. Common understandings can then underpin other forms of constitutional theory. Historically-minded theorists use common understandings to help provide an account of a constitutional principle before then exploring the origins of this particular ­principle. Critical constitutional theorists use a common understanding to determine the content of a particular constitutional principle, before evaluating the common understanding of that principle according to a particular normative theory. Regardless of whether a theorist sets out to develop a common understanding, or to use a common understanding when adopting other approaches to constitutional theory, a common understanding operates as a heuristic model. Its success depends upon the extent to which it captures what are believed to be the key features of a concept, or what can be regarded as standard examples of a particular concept. It differs from a constructed-type model used in a model-theoretic approach to constitutional theory as a constructed-type model need only provide an account that is objectively possible. This would not be sufficient for a common understanding, which seeks to explain how a concept is used or understood. Common understandings require a more specific fit to the essential features, or what we would regard as standard examples, of a particular concept – although this fit need not be perfect.34 Those adopting a critical approach to constitutions could also use models. These models are closer to an ideal-type model understood as an objective possibility as opposed to an objective probability. Theorists adopting this approach are 33 N W Barber, ‘The Significance of the Common Understanding in Legal Theory’ (2015) 35 OJLS 799. 34 This is clearly illustrated in Kyritsis’s defence of a moralised constitutional theory in ch 8 in this volume.

Constitutional Methodology and Brexit  81 more concerned with providing idealised moral or ethical frameworks from which to evaluate current constitutional principles or case law. These frameworks may be devised from a range of approaches. Theorists may deploy intuition to determine what they perceive to be an ideal of a constitutional principle. Gee and Webber point to a growing use of rationalism in public law, including in constitutional theory, where reason is used to identify orderly solutions to practical problems. Their work draws on the account of Oakeshott, who criticised this use of reason and the use of theoretical knowledge, contrasting this with a pragmatic approach.35 Rationalism can be used to develop heuristic models, providing an account of common understandings. However, Oakeshott also regarded approaches that developed principles and used these principles as an ideology as rationalist. He defined these approaches as establishing a set of abstract principles.36 Understood in this sense, a rationalist approach is used in a similar manner to the approach of a critical constitutionalist approach, as opposed to an interpretative approach. Those adopting a critical approach to constitutions arguably use models as if they were ideologies. There is more of a concern as to what the constitution ought to be, with models used as a standard of criticism from which to evaluate current constitutional arrangements. Model-theoretic approaches use different models from those adopted by theorists using a critical approach to constitutions. Constructed-type models need more of an anchor to reality than pure ideal-types or ideologies that are then used to critically evaluate particular constitutions. Whilst there is an evaluative dimension to constructed-types models, the models nevertheless recognise that these values are based on assumptions. Constructed-type models can only apply when there is a sufficient correlation between assumptions underpinning values or normative principles and the empirical data about the constitution to which they are applied. It is this connection that aims to ensure that a constructed-type model is objectively probable. There are, therefore, differences in terms of the type of model used by a model-theoretic approach when compared and contrasted to other approaches to constitutional theory. However, the more important difference concerns the purpose of developing models and the methodology used when applying models to particular constitutions. This can be best explained through analysing the advantages of adopting a model-theoretic approach to constitutional theory. III.  WHY ADOPT A MODEL-THEORETIC APPROACH TO CONSTITUTIONAL THEORY?

Model-theoretic approaches were originally developed in the physical sciences as a means of responding to perceived problems with the received view of science. Under the received view, scientists develop theories that provide a generalised law – eg,



35 G

Gee and G Webber, ‘Rationalism in Public Law’ (2013) 76 MLR 708. 716.

36 ibid

82  Alison L Young Newton’s laws of motion – and then devise experiments to prove or disprove these theories. The received view of science provides rules by which to distinguish between statements about theories and statements about empirical data, and which regulate how empirical data can prove or disprove a particular theory. These rules are required to ensure that any empirical data remains neutral. Neutrality ensures that empirical date is untainted by assumptions relating to a particular theory. Without this neutrality, empirical data cannot prove or disprove a particular general theory. It requires a distinction to be drawn between what is, as set out in the empirical data, and what ought to be, as set out in the general theory.37 When analysing the extent to which model-theoretic approaches to science can apply to legal theory, Roederer argues that criticisms of the received view of science can apply in a similar manner to approaches to legal theory. Legal and constitutional theorists devise theories about the law or constitutions. These theories are, in turn, tested for their accuracy according to empirical data. However, this approach to legal and constitutional theory also requires a clear delineation between statements about theories and statements about empirical data, as well as accounts as to how the accuracy of legal and constitutional theories can be proved or disproved with reference to empirical data. Criticisms about the received view of science, therefore, may apply in a similar manner to legal and constitutional theory. Having pointed out these similarities, Roederer recounts the main criticisms of the received view of science, focusing in particular on correspondence rules, which determine when and how empirical data proves or disproves a particular theory. These rules require a strict separation between descriptive and evaluative language. However, it is argued that it is difficult, if not impossible, to maintain this separation. Scientific experiments devised to prove or disprove particular theories frequently failed to provide definitive prove and, instead, displayed the weaknesses of using the received view. For example, earlier, we referred to Newton’s laws as an example of a scientific theory. Newton’s third law of motion states that for every action, there is an equal and opposite reaction. This theory is expressed as a law as it provides an account of what ought to be. It also provides an account of what is. The law has been derived from a series of observations and experiments where different forces are applied to a range of objects. Difficulties would arise were a scientific experiment to be devised to test this theory that demonstrated that, despite force being applied to a particular object, the object did not move, or produced a reaction that was greater than or less than the force applied to that object. This would disprove the empirical claim. But would it require a rejection of, or a modification to Newton’s third law, recognising either circumstances or elements or compounds to which it did not apply? Also, how would we know if this experiment disproved Newton’s theory, or was, in some sense, a failed experiment precisely because it did not produce the outcome we would expect when applying Newton’s third law of motion?

37 For a discussion of this and its application to legal theory, see Roederer, ‘Negotiating the Jurisprudential Terrain’ (n 16) 404.

Constitutional Methodology and Brexit  83 Model-theoretic approaches were devised to overcome these difficulties as they do not rely on correspondence rules to determine whether empirical data proves or disproves a particular general theory. Instead, model-theoretic approaches adopt constructed-type models; models that need not fully correspond with empirical data in order to be proved or disproved. Rather, constructed-type models provide an account of an objective-probability. Empirical data is not used to prove a particular theory, but to evaluate the extent to which a particular model is based on realistic or unrealistic assumptions. In addition, a model-theoretic approach accepts that it is not possible to provide a clear distinction between accounts of theories and accounts of empirical data. Constructed-type models recognise that any model contains normative evaluations. Constructed-type models used in model-theoretic approaches provide a clear account of the normative theories or assumptions on which they are based. Models can be evaluated in terms of their normative credentials, their own internal logical consistency, and the accuracy of the assumptions on which they are based. By keeping these forms of evaluation separate, models can provide a means of recognising the difficulty of distinguishing between descriptive and normative statements, whilst ensuring that information about what is and what ought to be are sufficiently distinct. Constitutional theory is not the same as scientific experiments. Nevertheless, similar issues of the need for, and criticisms of, correspondence rules can be found in constitutional theory. In particular, difficulties arise when trying to determine which theories of constitutional law, or distinct accounts of constitutional principles are better, or more valid, or more accurate than others. The problem of assessing legal theories, particularly jurisprudential theories, has been explored by Brian Bix when investigating how to evaluate different conceptual accounts of law in order to determine which account is better than other accounts. Bix provides an illustration of these difficulties by setting out a conceptual claim, similar to the claims made by legal theories: the conceptual claim ‘swans are white’, which he distinguishes from the empirical claim ‘all swans are white’. In a similar manner to the received theory in science, we would regard the conceptual claim that ‘swans are white’ as proved to be accurate if we could prove the empirical claim that ‘all swans are white’. Difficulties would arise if, following a trip to Australia, we were to observe an Australian black swan – a bird that appears in every regard to resemble a swan, except for the fact that it has black as opposed to white feathers and a red as opposed to an orange beak. The discovery of a black swan disproves the empirical claim that all swans are white. It is not true that all swans are white if you discover a swan that is black. Bix argues, however, that the discovery of a black swan does not disprove the truth of the conceptual claim ‘swans are white’ in the same manner. Upon the discovery of a black ‘swan’, a proponent of the conceptual claim that ‘swans are white’ can reply that the alleged black swan is clearly not a ‘swan’. By definition ‘swans’ are white. Therefore, the black bird in front of him is clearly not a ‘swan’ – it is perhaps a swan-like bird, but it is not a ‘swan’ as it is black and not white.38 If this is the case, how do we prove and disprove conceptual claims, and what, if anything, is the relationship between an empirical and a conceptual claim?

38 B

Bix, Jurisprudence: Theory and Context 6th edn (London, Sweet & Maxwell, 2012) 18.

84  Alison L Young Having set out this specific problem for legal theory, Bix argues that conceptual claims should be analysed not according to whether they are true, but according to whether they provide a good analysis for a particular practical purpose. Does a conception explain something that is interesting or important, or establish a good evaluative test for a particular purpose?39 The discovery of a black swan may lead a proponent of the conceptual claim ‘swans are white’ to question the value or importance of his conceptual claim. Is it important to continue to maintain that ‘swans are white’, so that the swan-like black bird is not a ‘swan’, or are there other similar features between the swan-like black bird and other birds such that the concept of a ‘swan’ should be changed, focusing on other features of a ‘swan’ – such as its size, that it is a waterfowl, or that it has a very long neck, rather than the colour of its feathers?40 A model-theoretic approach to legal theory responds to these concerns. Model-theoretic approaches do not make conceptual claims like the claim that ‘swans are white’. Rather, a theorist would propose a model of a ‘swan’, where swans are white. The discovery of a swan-like bird with black feathers would provide the theorist with the opportunity to evaluate his model of a ‘swan’. The model would no longer be based on assumptions that provided a perfect fit with empirical data. As such, the theorist would question the accuracy of her model and whether, despite its inaccuracy, it nevertheless had value, depending on the purpose of the model. If the purpose of the model was to provide an accurate classification of waterfowl, there would be more of an incentive to change the model than if its purpose were to provide an evaluation of birds, where ‘whiteness’ was a virtue, or if the model only provided an account of wildfowl living in the UK. If white-feathered birds are more valuable than others, a model that wished to classify birds according to their value may prefer to keep a model of swans as white, arguing that a black swan was best understood as a swan-like bird with black feathers as it did not have the same value as a white-feathered bird. If the model was focused on providing an account of wildfowl behaviour, and swans with white and black feathers acted in different ways, there may be good reasons for keeping the model of a swan as a bird with white feathers. However, if they acted in the same way, then there are good reasons to change the model of a swan to now include swans with black as well as white feathers. When applied to swans this assertion may seem extremely odd. But, when we apply it to the terms used in constitutional theory, for example to a model of a ‘liberal democracy’, the model-theoretic approach makes more sense. A constitutional theorist may use a model of liberal democracy when analysing the UK Constitution. However, there may be aspects of the UK that do not match the model of a liberal democracy. The theorist then faces a choice; to decide that a model of liberal democracy is unsuited for application to the UK Constitution, or to modify the model of a liberal democracy so that it can fit the UK Constitution. If the constitutional theorist merely found examples of UK court decisions, legislation, or of institutions that did



39 ibid

32.

40 Roederer

‘Negotiating the Jurisprudential Terrain’ (n 16) 422–425.

Constitutional Methodology and Brexit  85 not always promote the values found in her model of a liberal democracy, she would be more likely to modify her model so as to ensure the UK still fits within the model of a liberal democracy. Were these differences such that the UK was no longer able to further the values her model of liberal democracy aimed to promote, she would be more likely to change her model of a liberal democracy, classifying the UK as a purported example of a liberal democracy as it failed to be a liberal democracy in an important sense.41 This account of the justification for adopting a model-theoretic approach in science and in legal theory helps to provide a further means of distinguishing a ­ model-theoretic approach from other approaches to constitutional theory. As discussed above, historically-minded theorists focus on providing an accurate account of the origins of constitutions or of constitutional principles. Barber describes this approach as one that adopts the approach of natural sciences in the social sciences.42 Understood in this manner, the historically-minded approach not only distinguishes itself from a model-theoretic approach as it tends not to use models, but also because it adopts a different methodology. If Barber is correct in his classification, historicallyminded approaches to constitutional theory resemble the received view in science, which is rejected by the model-theoretic view. Historically-minded theorists develop accounts of the origins of a constitution or a constitutional principle. This account combines a conceptual and an empirical claim. The conceptual claim is proved or disproved by examining empirical data. However, when empirical data is incomplete, or can be interpreted in different ways to provide distinct historical accounts, there may be no accepted methodology through which to choose one account of the origins of a constitutional principle over another. We may just have to accept that there are competing historical accounts. Those adopting a critical approach to constitutional theory, if they provide models, provide ideal-types that need not match reality. The model is used to provide a normative framework through which to evaluate constitutions. These normative models can be developed in a variety of ways – through applying rational principles, through reasoning in a more intuitive manner, or through developing an ideology. As critical approaches to constitutional theory do not seek to provide an accurate account of empirical data, providing instead a normative model against which to evaluate empirical data to determine how well this corresponds to a normative framework, it is not subject to the same criticisms as those that apply to the received view of science. Nevertheless, difficulties can arise when distinguishing between different normative accounts. Why should a theorist adopt one particular normative account of the constitution over another? Moreover, this can often mean that different theorists talk at cross-purposes, providing competing accounts of constitutions whilst failing to engage with other theories, or failing to explain why a particular theory should be preferred over another. A model-theoretic approach differs from an approach that is designed solely to criticise particular constitutions. Model-theoretic approaches have a different aim,



41 ibid.

42 Barber,

The Constitutional State (n 31) ch 1.

86  Alison L Young wishing to provide models that may help analyse and predict behaviour and not merely provide a critical ideal from which to evaluate current constitutional arrangements. Moreover, adopting a model-theoretic approach provides a means through which we can evaluate competing models of the constitution. A model-theoretic approach provides a clearer account of the normative values a particular model wishes to promote, setting out the empirical assumptions on which the achievement of these normative values rests. There is no claim made in a model-theoretic approach that a particular model provides the best account of constitutionalism, only that it provides a valuable account of constitutions, with the model setting out the particular values the model aims to facilitate. In addition, a model-theoretic account is not adopted in order to provide a clear means through which to criticise specific constitutions, constitutional principles, or constitutional case law according to the extent to which they meet or adhere to a particular set of normative goals or values. A model-theoretic approach may be used to evaluate constitutions in a similar manner to those adopting a critical approach to constitutions. However, modeltheoretic approaches also recognise that different normative values compete and that the achievement of one particular normative goal may mean that another cannot be perfectly achieved. In addition, evidence that a particular goal is not being achieved is not used as a standard of criticism of a particular constitution. Rather, it may provide evidence of a need to modify the model, either because the model rests on assumptions that are not sufficiently objectively probable, or because the model provides an account of a relationship between different institutions or principles that is inaccurate, or because there is a logical inconsistency in the model. The interpretative approach to constitutional theory and the model-theoretic approach use similar models, with the interpretative approach classifying these as common understandings. However, common understandings are best understood as heuristic as opposed to constructed-type models. Barber argues that adopting an interpretative approach to constitutional theory also provides a means of responding to Bix’s criticisms of conceptual theories and the problem of determining which conceptual theory is better than another. Barber argues that adopting common understandings resolves this problem as it provides a measure against which different interpretative theories may be evaluated. An interpretative theory which provides an account which fits the common understanding is better than a different interpretative theory which fails to fit the common understanding of a concept. A particular account of a theory of law can be judged empirically by how well it fits a particular community’s common understanding of that concept. For example, an account of the separation of powers in the UK could be tested according to how well this account fitted the facts of the common understanding of the separation of powers in the UK.43 Barber is right to recognise that this can provide a means of evaluating different specific conceptions of constitutional principles. Nevertheless, Barber’s approach



43 ibid

806–7.

Constitutional Methodology and Brexit  87 differs from a model-theoretic approach in three ways. First, Barber’s approach evaluates different conceptions according to how well they fit a particular common understanding; it is a means of evaluating specific conceptions according to how well they fit another, more general, conception. Model-theoretic accounts assess models by looking at similar issues of fit, but this is to determine whether there is sufficient fit between the empirical data and the assumptions on which a particular model is based. It is not used to determine whether one model of the constitution is better than the other. Rather, it is used to evaluate whether the model can apply to a particular constitution. Second, model-theoretic approaches do not evaluate specific models by assessing how well they fit a particular common-understanding. Rather, they evaluate particular models by assessing whether the model really does facilitate the values the model aims to promote, focusing on whether the model is internally consistent. Moreover, model-theoretic approaches also need to provide an account of why the model wishes to promote particular values. Model-theoretic approaches also assess whether a particular model is based on assumptions that are met to a sufficient degree in the political constitutional system to which the model is applied. All three of these issues are used to evaluate a particular model of a constitution, or of constitutions in general. Third, model-theoretic approaches provide a different solution to Bix’s conundrum. This can best be explained by means of an example. If we return to Bix’s conceptual claim that ‘swans are white’, Barber’s means of evaluating the conceptual claim would be to assess how far the claim ‘swans are white’ complied with the common understanding of a ‘swan’; this common understanding being derived either as a prototype or an exemplar model of a ‘swan’. Whilst this may provide a means of assessing specific conceptions of a ‘swan’, it does not provide a means of evaluating the common understanding of a swan. If there were two competing common understandings – one that claimed that swans were white and others that did not include such a requirement – how would we determine which common understanding of a ‘swan’ was the most accurate? If this were done merely by assessing how far these competing common understandings of a ‘swan’ matched empirical data, then this would merely replicate Bix’s original criticism of conceptual claims, discussed above. Model-theoretic approaches are designed to answer this specific criticism. A model-theoretic approach does not examine competing common understandings but competing models of a swan. Each model is assessed according to the extent to which the model rests on factual assumptions that are sufficiently met by empirical data, whether the values promoted by a particular model really did have value, and the extent to which the model achieves these values. In doing so, the model-theoretic approach provides a better means of evaluating models. It does not focus on whether there is sufficient empirical data corresponding with the common understanding to support that particular common understanding. Rather, it provides an alternative means of evaluating models. Empirical data need not correspond perfectly with the model, rather there merely needs to be a sufficient fit for the model to be feasibly applied either generally or specifically. Models can also be evaluated from different perspectives, according to their internal coherence and the values they are designed to facilitate.

88  Alison L Young It could be argued that the different use of empirical data and different perspectives from which a theory can be evaluated do not provide a good enough means of avoiding Bix’s critique of conceptual claims. The arguments criticising correspondence rules can apply equally to model-theoretic approaches given that ‘fit to empirical data’ is a means, if not the only means, of evaluating competing models. Surely an assessment of ‘fit’ will also require the development of correspondence rules? If so, all a model-theoretic approach may do is appear to avoid a problem as opposed to resolve it.44 It is easier to understand how this criticism can apply to an evaluation of competing common understandings when applying the interpretative approach to constitutional theory. One of the means of devising a common understanding is through developing an exemplar model. This works by examining how a specific term – eg, the separation of powers – is used by, for example, officials in a particular legal system, or members of the judiciary in that legal system, or the population writ large. In order to be accurate, a common understanding of the separation of powers would require sufficient fit to how one or more of these groups use the term ‘the separation of powers’. It is hard to see how this could be determined other than through a form of correspondence rule, ensuring sufficient fit between the conception of the common understanding of the separation of powers and the empirical data about how this group used the term the separation of powers. A model-theoretic approach, however, differs from the development of an exemplar model of a common understanding. A model of the separation of powers, for example, would not be judged purely on the extent to which the model provided an accurate account of how the term ‘separation of powers’ was used in a particular community. As a constructed-type model, the model need not provide a perfect fit to empirical data in order to be a ‘good’ model. It need merely provide a model that is objectively possible to achieve. This minimises the difficulties that arise from correspondence theories. Empirical data is not used to determine the ‘true’ model of the separation of powers, or the best common understanding of the separation of powers in terms of its fit to empirical data. Rather, empirical data is used to assess whether a particular model of the separation of powers can be feasibly applied to a specific constitution, or constitutions more generally. When a model of the separation of powers fails to sufficiently fit empirical data, a theorist need not reject the model as inaccurate in the way that a theorist adopting an interpretive approach to constitutional theory would need to reject a specific common understanding of the separation of powers as failing to provide an accurate account. Rather, a theorist can choose whether to modify the model of the separation of powers to provide a better account of empirical data, or to conclude that the model proposed does not apply to a specific constitution.

44 See A Chakravarrty, ‘The Semantic or Model-Theoretic View of Theories and Scientific Reason’ (2001) 127 Synthese 325, NCA da Costa and S French, ‘The Model-Theoretic Approach in the Philosophy of Science’ (1990) 57 Philosophy of Science 248, S French and J Ladyman, ‘Reinflating the Semantic Approach’ (1999) 13 International Studies in the Philosophy of Science 102, S French and J Saatsi, ‘Realism about Structure: the Semantic View and Non-Linguistic Representations’ (2005) Philosophy of Science: Supplement Proceedings of PSA 2004.

Constitutional Methodology and Brexit  89 This choice will be influenced by the specific aim of the model of the separation of powers. Models can be used to help predict behaviour, or to facilitate a specific aim or value. When the aim of a model is used to predict outcomes in a particular constitution, disparities between predicted outcomes and reality will lead to a theorist changing her model. When the model is used to facilitate the achievement of particular values, the theorist will examine whether this is caused by too great a divergence between the factual assumptions on which the model is based and the empirical data about the separation of powers in a particular constitution. When the divergence is too great, the theorist will conclude that this particular model of the separation of powers is not suited to that constitution. In addition, when adopting a model-theoretic approach, a model of the separation of powers would also be evaluated for other purposes. First, the model would need to be internally consistent. A model of the separation of powers would fail if it the model relied on flawed reasoning. For example, a model of the separation of powers should be rejected if it is based on contradictory assumptions, for example if it required both that there was a strict separation between the executive and the legislature and also that members of the executive were also members of the legislature. A model of the separation of powers should also be rejected if it is based on circular reasoning. Second, a model of the separation of powers may be devised in order to achieve particular normative values – eg, to prevent the abuse of power or to promote the efficient distribution of power between governmental institutions. The model could be criticised by others who believed that the values promoted by this model of the separation of powers were the wrong values, or if those values were undesirable. This analysis would not provide a reason for concluding that the model of the separation of powers was wrong. It would, however, be a criticism of the model in terms of the values it wishes to promote, or the normative assumptions on which it was based. A model-theoretic approach, therefore, is more immune to the criticisms of correspondence theories. It does not rely on correspondence with empirical data to determine whether a model is true. Rather, it uses empirical data to assess whether the factual assumptions on which a model is based fit particular examples of a constitution, modifying the model or its application accordingly. Second, it relies on other means through which to evaluate the model than its ‘truth’ or ‘fit’ to empirical data. IV.  BREXIT AND PARLIAMENTARY SOVEREIGNTY

The previous sections described a model-theoretic approach to constitutional theory, explaining how this differs from other approaches both as regards to the type of model adopted by a model-theoretic approach and also in terms of its specific methodology. In doing so, it expounded the reasons why the model-theoretic approach had been adopted in scientific research, why this may be suited to legal and constitutional theory, and whether the proposed advantages of adopting a model-theoretic approach could also be achieved by applying this methodology to constitutional theory. However, it can be difficult to understand methodologies when they are

90  Alison L Young presented in the abstract. This section aims to provide further clarification by applying a model-theoretic approach to one aspect – the definition of parliamentary sovereignty. In the second Miller decision, the Supreme Court provided an account of parliamentary sovereignty which, arguably, was broader than current understandings.45 Parliamentary sovereignty has been traditionally interpreted by explaining how UK law resolves the conflict between two Acts of the Westminster Parliament, where later legislation overrides earlier legislation either expressly or by implication. In Miller; Cherry, the Supreme Court preferred a broader conception of parliamentary sovereignty, recognising this principle as one that promotes democracy more generally, including a requirement that the legislature meets regularly to enact legislation and to hold the executive to account.46 The preference for this broader interpretation of parliamentary sovereignty can be seen clearly in the following extract from the judgment of the court, delivered by Lady Hale and Lord Reed: the effect which the courts have given to Parliamentary sovereignty is not confined to recognising the status of the legislation enacted by the Crown in Parliament as our highest form of law. Time and again, in a series of cases since the 17th century, the courts have protected Parliamentary sovereignty from threats posed to it by the use of prerogative powers, and in doing so have demonstrated that prerogative powers are limited by the principle of Parliamentary sovereignty …. The sovereignty of Parliament would, however, be undermined as the foundational principle of our constitution if the executive could, through the use of the prerogative, prevent Parliament from exercising its legislative authority for as long as it pleased. That, however, would be the position if there was no legal limit upon the power to prorogue Parliament (subject to a few exceptional circumstances in which, under statute, Parliament can meet while it stands prorogued). An unlimited power of prorogation would therefore be incompatible with the legal principle of Parliamentary sovereignty.47

This judgment changes the empirical data about the UK Constitution. Parliamentary sovereignty, as a concept, now includes a requirement that Parliament meets regularly, at least to the extent that there can be no unlimited power of the executive to prorogue Parliament. It is no longer feasible to argue that the legal principle of parliamentary sovereignty, empirically, is merely about conflicts between statutes. As an example of a change in empirical data, it provides a good means through which to analyse how this change would be analysed by constitutional theorists adopting a historical-minded, a critical, an interpretative, or a model-theoretical approach to constitutional theory. A historical-minded theorist would examine historical documents to determine the content of parliamentary sovereignty. In doing so, she would be able to assess the

45 R (Miller) v Prime Minister; Cherry and Advocate General for Scotland [2019] UKSC 41, [2020] AC 373. 46 See, generally, P Craig, ‘The Supreme Court, Prorogation and Constitutional Principle’ [2020] PL 248, M Loughlin, ‘A Note on Craig on Miller; Cherry’ [2020] PL 278 and P Craig, ‘Response to Loughlin’s Note on Miller; Cherry’ [2020] PL 282 and the essays in [2019–2020] UK Supreme Court Yearbook forthcoming. 47 Miller; Cherry (n 45) [41]–[42].

Constitutional Methodology and Brexit  91 extent to which the decision of the Supreme Court departed from her understanding of the content of parliamentary sovereignty drawn from these historical documents. It may be that different theorists had distinct accounts of the best conception of parliamentary sovereignty, understood from a historical perspective. However, if we accept the criticisms of correspondence theories, it may be that all we can do is accept these competing accounts, explaining how well we think the decision of the Supreme Court matches or departs from these competing historical accounts. A constitutional theorist adopting a critical approach would rely on her normative account of parliamentary sovereignty to assess the extent to which the decision of the Supreme Court complied with her account. A theorist who relied on a normative account of the constitution which facilitated democracy, for example, might welcome the decision of the Supreme Court. Its account of parliamentary sovereignty focuses on ensuring that the government, who are indirectly elected, are held to account by Members of Parliament who are directly elected. In doing so, democracy is facilitated as this account of parliamentary sovereignty aims to ensure that the people can hold MPs and the government to account for their actions. However, a constitutional theorist whose critical approach preferred a constitution which facilitated the establishment of a strong government would reject the decision, precisely because it failed to enable the government to prorogue Parliament when it wished for as long as it wished. The difficulty with both the historical and the critical approaches is that of determining which account of the decision in Miller; Cherry, if any, is better than the others. We can see this when we analyse some of the academic reactions to this decision. Ekins and Finnis, for example, argue that parliamentary sovereignty means that sovereignty is found in legislation enacted by the Queen-in-Parliament.48 They reach this conclusion through drawing on accounts of the historical origins of parliamentary sovereignty. We could, therefore, interpret their argument as merely suggesting that the Supreme Court has failed to comply with the historically correct interpretation of parliamentary sovereignty. However, there are also critical elements to their theory. Both prefer an account of parliamentary sovereignty that restricts its understanding to resolving conflicts between two statutes, with the later overriding the earlier statute. This is because they argue in favour of a constitution where there is a strong government, where the political party from which the government is drawn has a strong majority in the House of Commons. A strong government has the backing of the people from the previous general election and is able to govern so long as it holds the confidence of the House of Commons. An alternative theory is that of Hasan Dindjer, who argues that parliamentary sovereignty recognises legislation as the highest form of law.49 Consequently,

48 J Finnis, The Law of the Constitution before the Court: Supplementary Notes on The Unconstitutionality of the Supreme Court’s prorogation judgment’, available at https://policyexchange.org.uk/wp-content/ uploads/The-Law-of-the-Constitution-before-the-Court.pdf and R Ekins, Parliamentary Sovereignty and the Politics of Prorogation, available at https://policyexchange.org.uk/wp-content/uploads/2019/09/ Parliamentary-Sovereignty-and-the-Politics-of-Prorogation3.pdf. 49 See H Dindjer, ‘Prorogation as a Breach of Parliamentary Sovereignty’, UK Const L Blog (16 Sept. 2019), available at https://ukconstitutionallaw.org/.

92  Alison L Young parliamentary sovereignty requires that Parliament must be able to legislate in order to enact statutes that then become the hierarchically superior form of law in the UK. This entails a reading of parliamentary sovereignty which matches that adopted by the Supreme Court. Dindjer’s approach is less historically-minded and more critical. Dindjer argues that this is a better understanding of parliamentary sovereignty. How are we to determine whether Finnis, Ekins or Dindjer provide a better account of parliamentary sovereignty? Moreover, how are we to use the data from the decision in Miller; Cherry? Do we see this as providing empirical evidence that Dindjer provides a better account of parliamentary sovereignty, or do we criticise the judgment for failing to adhere to the historically correct, or normatively better interpretation of parliamentary sovereignty? Barber would argue that any account of parliamentary sovereignty can only be gained through adopting an interpretive approach to constitutional theory. In adopting this approach, a constitutional theorist needs to determine the common understanding of parliamentary sovereignty. However, it is hard to see how this helps to progress an assessment of whether we should adopt the approach of Finnis, Ekins or Dindjer, and whether we should criticise or welcome the decision of the Supreme Court. This is because the decision of the Supreme Court is, at least arguably, now part of the common understanding. It is evidence of how ‘parliamentary sovereignty’ is used by the 11 Justices of the Supreme Court who decided Miller; Cherry. However, it is also open to those adopting an interpretive approach to constitutional theory to argue that the decision of the Supreme Court is not part of the common understanding of parliamentary sovereignty. Any common understanding of parliamentary understanding is best drawn from a series of historical sources, not just from the most recent decision of the Supreme Court. If you adopt the first common understanding, then Dindjer’s argument is the most convincing. If we adopt the latter, then the arguments of Finnis and Ekins are more convincing. The problem that arises is that adopting an interpretive approach to constitutional theory does not tell us which common understanding of parliamentary sovereignty is correct. It may merely leave us with the conclusion that there are competing common understandings. Does this mean that we are merely free to choose the one we find the most appealing? A model-theoretic approach takes a different approach. It classifies different conceptions of parliamentary sovereignty as competing models of parliamentary sovereignty. It would investigate the assumptions that underpin these distinct models of parliamentary sovereignty, as well as the normative values that are promoted by adopting these distinct models of sovereignty. Model-theoretic approaches also test whether distinct models of parliamentary sovereignty achieve the normative values they are meant to achieve, as well as questioning their value. Whilst this may not provide a conclusion that the views of Finnis, Ekins or Dindjer are ‘correct’, or help determine the best way of determining a common understanding of parliamentary sovereignty, it provides an alternative means of evaluating competing models of parliamentary sovereignty. This, in turn, helps to better facilitate discussion than merely arguing about which account better fits a particular common understanding or trying to determine whether one common understanding is superior to another.

Constitutional Methodology and Brexit  93 To illustrate a model-theoretic approach in more detail, we will focus on three possible models of parliamentary sovereignty: Paul Craig’s interpretation of Dicey, based on a model of self-correcting unitary democracy; Michael Gordon’s manner and form model of sovereignty based on democracy; and my own model of parliamentary sovereignty based on democratic dialogue.50 This approach uses evidence of the decision in Miller; Cherry in a different manner. It does not ask whether there was a change in direction, or whether the conception of parliamentary sovereignty accepted by the Supreme Court fits a common understanding of parliamentary sovereignty. Rather, it assesses the factual assumptions on which a particular model of parliamentary sovereignty is based. It asks whether the decision in Miller; Cherry means that these factual assumptions on which a particular model is based are no longer tenable. If this is the case, a model-theoretic account does not require that this model of parliamentary sovereignty should be abandoned. Rather, it raises a further question as to whether the aims of a particular model will still be achieved given this possible divergence from previous empirical data concerning the nature of parliamentary sovereignty in the UK. Craig’s interpretation of Dicey’s model of the UK Constitution starts with Dicey’s account of parliamentary sovereignty: that Parliament, as the supreme law-making institution, can enact legislation on any subject matter that it wishes and that its legislation cannot be questioned in any court or any other place outside of Parliament. However, Dicey’s theory recognises one limit on the sovereign law-making power of Parliament; Parliament cannot enact legislation that binds future parliaments. To do so would limit the future law-making power of Parliament, restraining it from enacting legislation on a particular subject matter unless it adopted a law-making procedure different from, and more onerous than, the law-making procedure used to enact ordinary legislation. Craig argues that this account of parliamentary sovereignty rests on a theory of a self-correcting unitary democracy. Dicey’s theory distinguishes between legal and political sovereignty. Parliament holds legal sovereignty, with political sovereignty vesting in the people. This is exercised through their role in general elections. Dicey argued that there were both internal and external limits on the actions of elected Members of Parliament. The internal limits stem from the consciences of MPs and from the limits placed on the actions of MPs by other MPs in the House of Commons. The external limits stem from the electorate. Craig explains how Dicey’s theory of parliamentary legislative supremacy ensures that the internal and the external limits on MPs are the same because each MP is aware of the need to face the electorate – now at least once every five years. To place legal limits on legislation would upset this balance. This, in turn, could undermine the political sovereignty of the people which underpins the legal sovereignty of Parliament.51 50 These are not the only models of parliamentary sovereignty, though they do help provide a good trio through which to explain adopting a model-theoretic approach. For an alternative analysis of p ­ arliamentary sovereignty, and the impact of both Miller I and Miller; Cherry on the continued accuracy of parliamentary sovereignty as a principle of the UK Constitution, see J Goldsworthy, ‘In Defence of Traditional Methologies’ ch 2 in this volume. 51 P Craig, ‘Dicey: Unitary Self-Correcting Democracy and Public Law’ (1990) 106 LQR 105.

94  Alison L Young Gordon provides an alternative model of parliamentary sovereignty.52 He argues for a manner and form theory of parliamentary sovereignty, where any one Parliament should be able to bind future parliaments as to the manner and form in which future legislation should be enacted, thereby empowering Parliament to entrench legislation. However, this should only be the case when the manner and form requirement is one that enhances democratic decision-making. Gordon explains this through the example of a possible interpretation of the Parliaments Acts 1911–1949. To the extent that these provisions bind future parliaments, they do so in a manner that enhances democracy, limiting the powers of one of the unelected components of the Queen-in-Parliament – the House of Lords – and empowering the House of Commons, the democratically-composed component of the Queen-in-Parliament. Gordon’s theory would also permit entrenchment through referendum requirements. Referendums enhance democratic decision-making as an example of direct as opposed to representative democratic decision-making. Gordon’s model is based upon an account of democracy which favours democratic decision-making. Parliament is sovereign as it is the institution with the most democratic credentials in the UK Constitution, given that the House of Commons is democratically elected, and the government is drawn from members of the House of Commons. My model of parliamentary sovereignty is similar to that of Dicey in that it too argues that Parliament ought not to be able to legally bind its successors. However, it does so for different reasons. My model of parliamentary sovereignty forms part of my model of democratic dialogue, a constructed-type analogue model that investigates interactions between the institutions of the Constitution. My model of democratic dialogue focuses on fostering relations between institutions designed to facilitate the achievement of a range of values in the Constitution, including a checks and balance model of the separation of powers, collaborative as opposed to majoritarian democracy and the protection of human rights.53 My model of parliamentary sovereignty also distinguishes between parliamentary legislative supremacy and parliamentary sovereignty. Parliamentary legislative supremacy is an assessment of the relative law-making powers of the Westminster Parliament. I adopt Dicey’s conception of parliamentary legislative supremacy, meaning that Parliament cannot legally bind its successors. Later legislation that contradicts earlier legislation impliedly repeals earlier legislation, although I also argue that it is rare for later legislation to contradict earlier legislation as most apparent conflicts can be resolved through compatible interpretation. I adopt an account of sovereignty that is similar to that of Martin Loughlin; the power that vests in the institutions of the state that empowers the state to act as a state.54 Whilst this too vests in Parliament, I do not argue that Parliament’s sovereignty in this sense is unlimited. Parliament can

52 M Gordon, Parliamentary Sovereignty in the UK Constitution: Process, Politics and Democracy (Oxford, Hart Publishing, 2015) and M Gordon, ‘Parliamentary Sovereignty and the Political Constitution(s): From Griffith to Brexit’ (2019) 30 King’s Law Journal 125. 53 Young, Democratic Dialogue (n 13). 54 M Loughlin, ‘Why Sovereignty’ in R Rawlings, P Leyland and A L Young (eds), Sovereignty and the Law: Domestic, European and International Perspectives (Oxford, Oxford University Press, 2013) and M Loughlin and S Tierney, ‘The Shibboleth of Sovereignty’ (2018) 81 MLR 989.

Constitutional Methodology and Brexit  95 place limits on its sovereignty by entering into international treaties, for example. It is also possible for a state to be sovereign and yet to place constitutional limits on the exercise of its law-making power, although, clearly, this is not the case for the Westminster Parliament. My model of parliamentary sovereignty also distinguishes between the power to enact legislation, and the power to enact and modify constitutions. Dicey’s account of parliamentary sovereignty denies any conceptual distinction between constitutional and ordinary legislation. Parliament is sovereign and cannot bind its successors as to either the form or content of legislation. I argue that the power to create or modify fundamental aspects of the Constitution should not belong solely to the Westminster legislature. Rather, in a system like the UK that has no codified or entrenched constitution, the power to create or modify fundamental aspects of the constitution requires acceptance by both the courts and the legislature, which incorporates the government. In this sense, sovereignty is shared between courts and Parliament. Officials in the legal system must accept changes to the Constitution as valid. This is because both the courts and the legislature have distinct strengths and weaknesses when determining constitutional issues given their distinct composition, expertise, and their decision-making processes. Whilst legislatures bring democratic accountability and a better ability to balance a range of interests, courts use legal reasoning, focusing on protecting long-standing normative principles of the common law, helping to ensure the constitutional values developed over time in the UK are not eroded. Parliamentary legislative supremacy is adhered to in order to ensure a more justifiable means of modifying the UK Constitution, understood both in terms of process and outcome. Without it, Parliament alone would be able to entrench legislation, thereby modifying the Constitution. My theory argues that modifications of the Constitution are more legitimate and more likely to promote valuable outcomes if they involve both the legislature and the courts. Each of these competing models of parliamentary sovereignty rests on the facilitation of different values. An assessment of their relative merit, therefore, is not solely concerned with the extent to which each theory can account for a potential change in empirical data. However, any change to this data may provide a reason for rejecting the application of a particular model of sovereignty to the UK Constitution if this change in data undermines the factual assumptions on which the model is based. Dicey’s model of a self-correcting unitary democracy has already been criticised as to whether the assumptions on which it is based sufficiently fit the UK Constitution, given the changing circumstances since Dicey’s theory was first proposed. Devolution means that we can no longer regard the UK as a unitary system. Moreover, as Dicey himself recognised, the increasing role of political parties in the UK, and the everincreasing number of MPs on the ‘payroll’ who are thereby covered by collective ministerial responsibility, undermine the extent to which the external restrictions on MPs match the internal restrictions. The decision in Miller; Cherry does not provide evidence in and of itself for rejecting the self-correcting unitary democracy model or for questioning its application to the UK Constitution. However, when added to further evidence, it questions the extent to which this model, and its account of parliamentary sovereignty, can serve as an analogue model in order to predict behaviour.

96  Alison L Young There is no need in Dicey’s theory to distinguish between the legislature and the executive, or to require the legislature to be able to check on the executive in order to enhance parliamentary accountability. Nor is there a need for Parliament, rather than the executive, to take important constitutional decisions. Applying Dicey’s model, therefore, would not have predicted the outcome in Miller; Cherry. The change in empirical data may also question the extent to which Dicey’s self-correcting model of sovereignty achieves its aims. Is it really the case that the views expressed in the House of Commons, and the restraint on the actions of the government, match those of the electorate if the executive is able to act so as to override the wishes of the legislature? A different conclusion is reached when assessing the extent to which the decision in Miller; Cherry applies to Gordon’s model of parliamentary sovereignty. The change in Miller; Cherry does not provide a reason to question the application of Gordon’s model to the UK Constitution per se. However, the means through which a change in the content of parliamentary sovereignty was achieved may question the suitability of Gordon’s model as an account of the UK Constitution. It questions the extent to which the application of Gordon’s model to the UK Constitution facilitates democratic decision-making when changes to the UK Constitution can be made with neither a direct nor an indirect mandate, which, in turn, questions the extent to which an application of Gordon’s model to the UK Constitution facilitates the achievement of the values on which it is based.55 Gordon’s model advocates a manner and form account of parliamentary sovereignty. The decision in Miller; Cherry does not provide any further evidence to support the acceptance of a manner and form in the UK Constitution. Nor does it provide any evidence to reject it. The decision in Miller; Cherry also promotes democratic decision-making. The Supreme Court stepped in to facilitate democratic decisionmaking by recognising the need for the House of Commons to enact legislation and for the legislature to continue to hold the government to account. However, the way in which this was achieved was in a manner that failed to facilitate the aim of democratic decision-making. The change came about through the courts modifying the common law, rather than through a democratic-decision making process involving either Parliament or the people. The ability of the courts to modify the common law in this manner questions the extent to which, when applied to the UK Constitution, Gordon’s model will ensure a broader facilitation of democratic decision making. It also requires Gordon to assess how far his model may need to be changed to prevent decisions of the court which may specifically facilitate democratic decision-making but do so in a manner that is undemocratic. My model of democratic dialogue rests on an assumption of continuing parliamentary legislative supremacy. This also is not challenged by the decision in Miller; Cherry. Consequently, in a manner similar to our analysis of Gordon’s model, the decision does not challenge the existence of the factual assumptions on which my model is based in the UK Constitution. Therefore, the model can still be applied to

55 See J Goldsworthy, ‘In Defence of Traditional Metholodogies’, ch 2 in this volume, for a similar criticism.

Constitutional Methodology and Brexit  97 the UK Constitution both as an analogue model to predict behaviour and as a means of achieving the values on which the democratic dialogue model is based. One of the values the democratic dialogue model aims to facilitate is a system of checks and balances between the different institutions of the Constitution. This is facilitated through inter-institutional interactions that can be used by one institution to prevent another from trespassing into or limiting the proper constitutional role on another institution. Continuing parliamentary legislative supremacy is required to facilitate these checks and balances, counter-balanced by the role of the courts to develop the common law. The conception of parliamentary sovereignty adopted in Miller; Cherry also helps to facilitate checks and balances by recognising the importance of the legislature holding the executive to account. Miller; Cherry provides a further example of inter-institutional interaction designed to facilitate checks and balances. The Supreme Court quashed the prorogation of Parliament as this was an example of the government going beyond its proper constitutional role, using the power of prorogation to undermine the proper constitutional role of the legislature.56 Whilst the decision provided an example of how Gordon’s model, when applied to the UK Constitution, may not achieve the values it wishes to facilitate, the opposite is true when we analyse my model of democratic dialogue. This suggests that my model of democratic dialogue would be more accurate as an analogue model of predicting behaviour in the UK Constitution than Gordon’s model of democratic dialogue. It also suggests that, when applied to the UK Constitution, my model of democratic dialogue is more likely to facilitate the values on which it is based than if we apply Gordon’s model to the UK Constitution. Empirical data is used differently when adopting a model-theoretic approach to constitutional theory. It assesses the suitability of a particular model to apply to a specific constitution. The data is not used to question the truth of a model, in the sense of assessing how well it applies to the facts. Rather, it is used to determine the suitability of a model’s application to a specific constitution. Miller; Cherry, for example, may provide further evidence for questioning the suitability of Dicey’s model of a self-correcting unitary democracy in the UK Constitution. The UK Constitution no longer fully matches the factual assumptions on which the model is based, calling into question its utility as a means of predicting outcomes or as a means of facilitating the achievement of the values on which it is based when applied to the UK Constitution. A change in empirical data does not mean that a model should be changed. It may also be possible to argue that the empirical data should be challenged. This depends upon the aim of a model of the constitution. We argued above that Miller; Cherry provides a potential means of questioning the extent to which Gordon’s model is suited to the UK Constitution in that it provides an example of constitutional change in a manner that does not facilitate democratic decision-making. If Gordon’s model aims to facilitate democratic decision-making, then Miller; Cherry 56 AL Young, ‘The Prorogation Case: Re-inventing the Constitution or Re-imagining Constitutional Scholarship’ in D Clarry (ed), The UK Supreme Court Yearbook, Volume 10 2018–19 Legal Year (London, Wiley, 2021) 357.

98  Alison L Young may provide a justification for the model to be modified, such that when applied to the UK Constitution it prevents this potential erosion of democratic decision-making. If Gordon’s model aims to provide a set of values from which to criticise constitutions, then Miller; Cherry would provide an example of a decision that merited criticism. It is important to recognise also that fit with empirical data is not the only means through which models can be evaluated. Each of these models can also be evaluated in terms of their internal logic and the aims that they promote. There may be a reason, for example, for rejecting Dicey’s model of a unitary self-correcting democracy in preference for Gordon’s model of parliamentary sovereignty if we are persuaded that the account of democracy facilitated by Gordon’s model is better than that facilitated by Dicey. A preference for a strong system of checks and balances, albeit one that may come at the price of some aspects of democratic decision-making, may provide a justification for preferring my model of democratic dialogue to Gordon’s model of parliamentary sovereignty. A model theoretic approach recognises that these are different issues from an assessment of fit with empirical data, enabling an approach that can be both normative and descriptive without that approach confusing questions as to what is from questions of what ought to be. Whilst this may not provide a definitive justification for adopting one model over another, adopting a model-theoretic approach to constitutional theory may minimise the extent to which theorists are talking at cross-purposes and help to clarify different ways in which constitutional theories, and their relationship to empirical data, can be assessed. V. CONCLUSION

The circumstances surrounding Brexit challenged a number of the factual assumptions and, potentially, the values underpinning or promoted by the UK Constitution. This chapter is not the place to fully explore the consequences of Brexit on the UK Constitution. However, in presenting an account of a model-theoretic approach to constitutional theory, I hope to provide an effective tool through which to analyse constitutions more generally and the impact of Brexit on the UK Constitution more specifically. A model-theoretic approach helps to resolve some of the difficulties that arise when comparing competing conceptions of the UK Constitution. Rather than focusing on assessing which conception provides a more accurate account of constitutional history, or better reflects how a constitutional term is used in society or a particular sub-group of society, a model-theoretic approach develops models of constitutions. These models are best understood as constructed-types. They do not need to provide a fully accurate account of empirical data but do need to be based on empirical data that is objectively probable if the model is to achieve the normative values the model is designed to promote. Models of the Constitution are evaluated in terms of the accuracy of the assumptions on which they are based and the values they aim to achieve. The model also needs to be tested to ensure that it is able to achieve the normative values it wishes to achieve, focusing on whether there are any logical inconsistencies in the model. A model can also be used to criticise constitutional developments.

Constitutional Methodology and Brexit  99 Although model-theoretic approaches may appear novel in constitutional theory, they are not novel to social sciences or scientific theory. Indeed, although we may not talk of using a model-theoretic approach in constitutional theory, constitutional theorists often refer to models. This chapter has tried to explain how a model-theoretic approach may be suited to constitutional theory and also how this may help to resolve the difficulties that arise when empirical evidence may call into question conceptual claims. In doing so, it hoped to clarify what theorists may mean when they refer to models of the constitution, as well as providing a tool that can be used by others to develop alternative models of the UK and other constitutions.

100

How Do Facts Matter?

102

4 Slaying the Misshapen Monster: The Case for Constitutional Heuristics TT ARVIND AND LINDSAY STIRTON* I. INTRODUCTION

O

n 9 February 1784, Adriaan Kluit, rector magnificus at the University of Leiden, ascended the podium to deliver his valedictory address: ‘On the abuse of constitutional law’.1 Kluit was speaking at a time of fierce contestation in Dutch politics between the Orangists (with whom Kluit identified) and the anti-Orangist ‘Patriots’. The language of constitutional theory had become a key weapon in their battle, and it was to that language that Kluit’s address was directed. Its contents were sharp and uncompromising. The Patriots, Kluit said, had created ‘a misshapen monster in constitutional law’.2 The excellent provisions of the Dutch state’s settled constitutional institutions had established civil liberty on the firmest foundations, but these accomplishments were threatened by those who, posing as the commonwealth’s physicians, were taking it down a path of ruin.3 The Patriots’ constitutional doctrines, Kluit warned, were a destructive plague bringing catastrophes upon commonwealths.4 This chapter is motivated by our sense that little has changed since Kluit’s day in the methodology of constitutional theory and how it deals with disagreement. The debate between Kluit and the Patriots was at one level a theoretical debate as to the meaning of liberty, and whether civil liberty mattered more than political liberty. Yet the argument about these concepts was suffused with claims about facts: about the Netherlands’ Batavian past and commercial present, about the regicides, oppressions,

* We are grateful to the University of York for funding the translation of non-English source documents used in this chapter, to Firat Cengiz for invaluable assistance with locating Turkish sources, and to Elizabeth Craig, Simon Halliday, Alex Latham-Gambi, Will McCready, Lawrence McNamara, and the volume editors for comments on an earlier draft. 1 A Kluit, Academische Redevoering over het Misbruik van ’t Algemeen Staatsrecht (Leiden, Luzac en van Damme, 1787). 2 ibid 25. 3 ibid 2–3. 4 ibid 97.

104  TT Arvind and Lindsay Stirton and tumults that afflicted polities at different points of time, about which political treatises were useful guides and which should be disregarded, and so on. And, crucially, the parties’ disagreement was marked by an inability to agree on which of these facts were of significance in understanding the true nature of the eighteenthcentury Dutch constitution.5 And so it remains in the present day. Constitutional theorists, unlike political theorists, are concerned with the specifics of a particular constitutional system or set of systems. The goal of constitutional theory is to identify the institutions and principles that underpin the public life of a specific state, rather than consider how all states ought to be organised. Of their nature, they therefore necessarily (if implicitly) claim to be not merely dialectically effective (likely to be persuasive), but also epistemically effective (likely to produce a true understanding of the Constitution).6 Notwithstanding their claim to merely be advancing arguments about how a particular constitution ought to be understood, they are beneath their surface concerned with facts precisely as the theories of Kluit’s day were. Where, then, does this epistemic confidence come from? This question is rarely discussed explicitly. Theorists rarely discuss the process through which their theories were formed, and their writing usually seeks only to persuade the reader of its correctness rather than (as, for example, in science) to explain why the process by which the theory was constructed or tested meets criteria for justified belief. The rightness or wrongness of a theory is treated as a matter to be uncovered through argumentation and persuasion, rather than through an assessment of the processes that underpinned its construction and the biases if any that they bequeathed to the theory.7 Kluit’s language in terming opposing theories of the eighteenth century Dutch Constitution a misshapen monster and a destructive plague may well be more extreme than the terms a modern-day theorist would use, but this is a matter of degree rather than kind. The underpinning assumptions remain the same: each theory presents itself as a correct understanding of the principled and institutional underpinnings of a constitution’s excellence that safeguards its accomplishments, unlike its rivals. The impact of this on constitutional theory has not been a happy one. The reliance on persuasion means that constitutional theory lacks any method beyond argumentation for judging the truth or falsity of a given theory of the Constitution, reducing it to a cycle of disputation between different ‘styles of public law thought’.8 But even more fundamentally, rhetorical or philosophical attractiveness is by itself an inadequate

5 WRE Velema, ‘The concept of liberty in the Dutch Republic, 1780-1787’ in WRE Velema (ed), Republicans: Essays on Eighteenth-Century Dutch Political Thought (Leiden, Brill, 2007) provides a concise and accessible overview of the broader debate and Kluit’s position within it. 6 On the role of epistemic and dialectical effectiveness in disagreement, see E Sosa, ‘The epistemology of disagreement’ in A Haddock, A Millar, and D Pritchard (eds), Social Epistemology (Oxford, Oxford University Press, 2010) 293–297. 7 A similar point was made by Herbert Simon about public administration theory in the mid-twentieth century. See HA Simon, ‘The proverbs of Administration’ (1946) 6 Public Administration Review 53. 8 M Loughlin, Public Law and Political Theory (Oxford, Clarendon Press, 1992) 62.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  105 basis for evaluating whether a particular constitutional theory ought to be accepted by a polity. Constitutional theories have non-trivial, real-world consequences because they influence the conduct of public affairs, and the history of constitutional thought is littered with constitutional theories whose rhetorical power made them appear normatively attractive in their day but whose practical consequences were deeply troubling. Examples range from Mill’s theory of benevolent colonial despotism which exercised a shaping influence over the imperial Constitution in the late nineteenth century,9 to John Calhoun’s anti-majoritarian theory of the US Constitution which, despite being designed to justify the continuance of slavery, was built on a moral rhetoric so alluring that it influenced thinkers as liberal as Lord Acton.10 Overcoming these problems requires a fuller understanding of the strengths and limitations of the processes and techniques constitutional theorists use to derive theories and the manner in which they justify their belief in the reliability of those processes. In section II, we draw on the literature on social epistemology to show that the process of constructing any theory of a specific constitution involves two distinct forms of engagement with facts: first, identifying facts that have constitutional salience and, second, assigning significance to those selected facts. There are strong parallels between these processes and those implicated in the construction and transmission of traditions, an analysis of which sheds new and useful light on how constitutional theories construct the constitutional world they describe. In section III we build on the argument of section II to show that there are four broad families of approaches to selecting and assigning normative significance to facts into which most constitutional theories can be classified. Each of these families is united by its propensity for highlighting certain types of features, and its predilection to read certain types of normative significance into constitutional facts. As we show, a closer focus on these predilections not only helps us achieve a better understanding of how constitutional theories are made, but also of the limitations those methodologies impose on constitutional theory. In section IV, we argue that while these families have value, the failure to understand their aetiology and limitations has led to positions in constitutional debates becoming entrenched, polarised, and unyielding. Our purpose in making this point is methodological rather than merely critical. The methodology of constitutional theory suffers from the failure to recognise the extent to which constitutional theories are shaped by the predilections and predispositions that underlie them. We propose a new approach to constitutional theorisation and constitutional method which takes a heuristic approach, relating constitutional facts to the shifting needs of the polity, and opening up a more pragmatic space for theoretical debate and doctrine-making in constitutional scholarship.

9 On Mill’s theory and its impact, see J Pitts, A Turn to Empire: The Rise of Imperial Liberalism in Britain and France (Princeton, Princeton University Press, 2005) 133–162. 10 On Calhoun’s theories and slavery, see LK Ford Jr, ‘Inventing the Concurrent Majority: Madison, Calhoun, and the Problem of Majoritarianism in American Political Thought’ (1994) 60 Journal of Southern History 19. On the influence it exercised over Lord Acton, see, eg, Lord Acton, Essays on Freedom and Power (Glencoe, Free Press, 1948) 225–231.

106  TT Arvind and Lindsay Stirton II.  FACTS, THEORIES AND TRADITIONS: MAKING THE CONSTITUTIONAL WORLD

A.  The Social Epistemology of Constitutional Theory Constitutional theorists who state that a particular theory is superior to another are making a claim qualitatively different from – say – the claim that Beethoven is a better composer than Mozart. The latter statement deals with a matter of subjective opinion. However strongly one may hold a belief as to the relative superiority of one, few will argue that there is a theorist-external standard by which that belief can be judged. Constitutional theorists, in contrast, do make such claims. A theorist who claims that political constitutionalism provides a better account of the UK’s Constitution than legal constitutionalism,11 or that current US constitutional jurisprudence remains flawed by the inherent contradiction between the Constitution’s promise of liberty and its compromises and contradictions on the position of African Americans,12 is claiming that there is something about the institutional features of the relevant constitutional order that means that their theory describes the constitution with greater accuracy than others. It is, of course, possible to take a critical standpoint and treat these claims as mere rhetoric. Examination of debates in constitutional theory, however, suggests that the claims are seriously meant – that theorists believe, and hold themselves justified in believing, that the facts their theories assert about a constitutional order are true in a way that others are not. On what basis, then, do theorists form this belief? What gives them the confidence that their approach to theory-formation is epistemically effective and not just dialectically effective? Our case in favour of heuristics is founded on the understanding that there are inherent flaws in the answers which constitutional theories implicitly give to this question. Our focus in this section is, accordingly, on setting out our analysis of what those answers are. Our starting point lies in two observations about how constitutional theories work. First, all constitutional theories draw on a shared pool of facts about a polity. Disagreements between constitutional theorists do not relate to whether King John really signed a document in 1215 which included a statement that freemen would not be imprisoned save by the law of the land, or whether the Constitution of India really includes an Article that states that no person shall be denied equality before the law or the equal protection of the laws. They relate, rather to the institutional status of facts and to whether they have deontic power.13 Simply put, disagreements in constitutional theory relate to whether a specific fact in the shared pool carries any significance at all for our understanding of how the broader constitutional order operates today, and if so what. Second, in forming and justifying positions on these points, constitutional­ theorists are concerned not just with codified or uncodified constitutional texts as 11 See, eg, A Tomkins, ‘In Defence of the Political Constitution’ (2002) 22 OJLS 157. 12 See, eg, D Hall, ‘The constitution and race: A critical perspective’ (1988) 5 New York Law School Journal of Human Rights 229. 13 The language of institutional facts and deontic power is taken from the work of John Searle. See especially, JR Searle, The construction of social reality (New York, Free Press, 1995).

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  107 they stand in the present, but also with the accreted sum of practices, actions, and ideas inherited from the past that they hold to be constitutionally significant; and their purpose in so engaging is to derive a body of precepts that can be entrenched and transmitted to the future as determinative of the proper conduct of constitutional affairs. The remainder of this section discusses these processes, and how they purport to give theorists confidence in their epistemic effectiveness. B.  Taking Facts Seriously Let us begin with the first of these, namely, the processes underpinning a theorist’s treatment of a fact as having institutional significance and deontic power. This entails, first, selecting a subset of facts by attaching constitutional salience to specific facts taken from the common stock of facts bequeathed by the past. Second, it entails deriving lessons for the present from those facts in terms of actions or institutional configurations that they recommend, rule out, or warn against – in other words, deriving significance from those facts. The debate in the UK between political and legal constitutionalists illustrates the role that the attribution of salience plays in constitutional theory. In his influential treatise Law, Liberty, and Justice, one of the foundational works of modern legal constitutionalism, TRS Allan begins by linking Dicey’s relevance to the role factual considerations played in Dicey’s theory. ‘Contrary to orthodox opinion’, Allan declares, Dicey ‘was wise to seek an interpretation of the rule of law which reflected the traditions and peculiarities of English common law’. (Emphasis added)14 Allan then proceeds to build his own theory of the rule of law. In doing so, he is informed and influenced by his reading of contemporary legal and political philosophers, in particular Raz and Dworkin.15 But his account of that philosophy is interwoven with a detailed analysis of how judges in the twentieth century UK, working with the material of earlier eras, developed, applied, and extended the rule of law.16 Like Dicey, in other words, Allan’s theory is built on a deep engagement with the ‘traditions and peculiarities’ of the UK’s constitutional order and the process by which it emerged (an analysis that, in later work, is extended to also encompass the US and Australia).17 It is built on and underpinned by facts. This is also true of political constitutionalists. Crucially, however, they differ in which facts they emphasise. Adam Tomkins, for example, in Our Republican Constitution, engages not just with the political philosophy of civic republicanism, but with English constitutional history.18 But where Allan’s focus is on the

14 TRS Allan, Law, Liberty, and Justice: The Legal Foundations of British Constitutionalism (Oxford, Oxford University Press, 1994) 21. 15 ibid 23–29, 45–48. 16 ibid 32–44. 17 TRS Allan, Constitutional Justice: A Liberal Theory of the Rule of Law (Oxford, Oxford University Press, 2003). 18 A Tomkins, Our Republican Constitution (Oxford, Hart Publishing, 2005).

108  TT Arvind and Lindsay Stirton contribution made by courts to constitutional government and constitutional development, Tomkins’ focus is on episodes in which the courts failed to protect private persons against oppressive institutions.19 In the episodes he discusses, it was Parliamentary institutions that took the lead.20 As in Allan’s work, this engagement with the traditions and peculiarities of the UK’s constitutional order exercises a shaping influence. Constitutional theories, in other words, are underpinned by disagreements about which facts are of significance for understanding the constitutional arrangements of a polity. But, in addition to disagreements about which facts are significant, theorists may also disagree about what that significance is. They may agree on the facts that shed light on the present constitutional order, but disagree on the light they shed. The debate about the right to keep and bear arms in the US Constitution provides a good example. For much of the constitutional history of the US, African-Americans were denied the right to keep and bear arms notwithstanding the Second Amendment and its counterparts in State constitutions, and militias soldiered by arms-bearing white men were regularly deployed to suppress dissent by African Americans. In a recent treatise, Carol Anderson draws on the interweaving of the past history of arms-bearing and the systematic suppression of African Americans to argue that the Second Amendment is at its heart lethal and ‘steeped in anti-Blackness’, a ‘loaded weapon … just waiting for the hand of some authority to put it to use’.21 A strikingly different reading of the same history is presented in the concurring opinion of Justice Clarence Thomas in McDonald v City of Chicago.22 In contrast to Anderson’s view, Thomas argued for an extension of the Amendment by holding that the Fourteenth Amendment made the right it created valid against State governments. Thomas does not differ from Anderson’s reading of the past. Like her, he discusses the racialised history of arms-bearing and arms laws in the US. But Thomas attaches a very different significance to these aspects of the past. Where Anderson sees evidence of a right fatally tainted by racism, Thomas sees a warning against permitting State power to be deployed to disarm individuals. The significance of its history and the lesson it holds for the present is, for Thomas, that individual rights conferred by the Second Amendment need to be better secured against governmental interference. C.  Continuity and the Ascription of Deontic Power The differences discussed in the previous section are not random. As we show in this section, they are the result of processes which strongly resemble the processes by which traditions are developed, reworked, and adapted. A tradition is at its heart an attempt to create a pattern to guide future action. Traditions differ from evanescent action in that they outlive the specific circumstances



19 ibid

69–87. 87–109. 21 C Anderson, The Second: Race and Guns in a Fatally Unequal America (London, Bloomsbury, 2021). 22 (2010) 561 US 742. 20 ibid

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  109 against whose background they were originally enacted.23 Constitutional theory likewise differs from an individual application of legal doctrine in that a theory is intended to identify general patterns which can guide future action in different circumstances. Equally, both constitutional theory and traditions are principally concerned with the normative prescriptions, precedents, and conditions for action established by the material that constitutes the theory or tradition. Like constitutional theory, a tradition seeks to use its material to identify and establish exemplars of action. Finally constitutional theory, like a tradition, seeks to assemble not a disparate set of records of past action but a received image of patterns, practices, beliefs, and institutions which require, regulate, permit, or prohibit the enactment of particular patterns of future action.24 The idea that the law resembles a tradition is, of course, not novel.25 Nevertheless, a closer exploration of the analogy sheds useful light on the methodology of constitutional theory, as well as why that methodology has a propensity to produce disagreements that are sharp, long-lasting, and seemingly intractable. As theorists of traditions have pointed out, the material that constitutes a tradition can be classified into three elements: the aspects of the tradition that constitute its continuity, a canon of written texts that cumulatively embody the distinctive features of that tradition, and a core of teachings or precepts that are foundational to the tradition and cannot be rejected without rejecting the tradition.26 A tradition’s core is placed beyond criticism, whereas a canon is subject to refinement, evolution, and scrutiny.27 Continuities, canons, and cores are fundamental to constitutional theory. Theorists seek to uncover what they believe to be the best understanding of a constitution by examining textual evidence (treatises, cases, statutes, etc), as well as broader evidence of constitutional practices, beliefs, and outcomes over time. They decide not just how much weight to attach to a particular item of evidence and what that evidence tells us, but also what constitutes evidence of constitutional practice in the first place.28 And the starting point for that process is fundamentally determined by their views as to where constitutional continuity lies. Consider, for example, the respective positions occupied by the pre-Union Parliaments of England, Scotland, and Ireland in the UK’s constitutional imagination. The post-Union Parliament is, in principle, a successor to all three Parliaments; yet in practice constitutional theorists tend to draw exclusively on the precedents established by the English Parliament, even where those of other Parliaments speak squarely to issues of theoretical relevance. Take, for example, the question of whether Parliament can be prorogued against its will – a question that was of considerable interest to constitutional theorists in the UK in 2019. In December 1639 the Scottish Parliament, faced with an attempt by the King to prorogue it, declared that it could

23 E Shills, Tradition (Chicago, University of Chicago Press, 1981) 31–32. 24 ibid 12–13. 25 See, eg, G Postema, Bentham and the Common Law Tradition (Oxford, Clarendon Press, 1986); N Duxbury, Frederick Pollock and the English Juristic Tradition (Oxford, Oxford University Press, 2004). 26 J Alexander, ‘A systematic theory of tradition’ (2016) 10 Journal of the Philosophy of History 1. 27 ibid 23–24. 28 See, eg, Loughlin, Public Law and Political Theory (n 8) 59.

110  TT Arvind and Lindsay Stirton not be prorogued without its consent, and in June 1640 it abolished one of the main instruments of executive control over it.29 That this episode hardly figured in discussions of the purported prorogation of August 2019 is not because the Scottish Parliament is ignored or not studied. Rather, it is because modern constitutional theorists largely share Tomkins’s perception that the modern British Constitution ‘for good or ill … was made in England’.30 To inform a constitutional theory, and action, must be seen as forming part of an intellectual continuity with the present, and the Scottish Parliament’s pasts are not so regarded notwithstanding the legal continuity. The process involved in identifying continuity between the practices and beliefs of the past and those of the present day might be termed ‘memorialisation’. The facts implicated in memorialisation are public or social facts: the images of public actions, practices, and beliefs from the past that are seen by a theorist as having enduring relevance when it comes to determining how a polity’s constitutional affairs should be conducted.31 Disagreement in constitutional theory is frequently a consequence of different views about where continuity lies. The disagreement between Anderson and Thomas over the Second Amendment in the US, for example, reflects how each of them sees constitutional continuity. To Anderson, the continuity implicated in the right to keep and bear arms is the phenomenon of the consistent and systematic oppression of African Americans, and the Second Amendment’s full enforcement therefore necessarily entails the continuance of that oppression. To Thomas, in contrast, the continuity in question is the assurance to every individual of the right to acquire and possess the means to defend oneself against oppression, and the Second Amendment’s full enforcement therefore has the potential to operate as a powerful tool for African American rights. Similar differences exist elsewhere. The debate between political and legal constitutionalism in the UK is driven by whether one sees greater constitutional continuity with the present day in the actions of the mid seventeenth-century legislature or the early seventeenth-century and late eighteenth-century courts. Likewise, theoretical debates around post-colonial constitutions such as India’s have involved whether their independent constitutions marked the creation of an order radically different from any they had previously had, or whether there are continuities between the modern constitutional order and the constitutional ideas on which the independence movement was built, or even with the older ideas and theories underpinning the precolonial constitutional order. Differences as to where continuity lies, in other words, underpin disagreements about which of the shared store of facts deserve to be given deontic power. As the next section shows, these differences are in turn shaped by a theorist’s understanding of the other two components of tradition: cores and canons.

29 For the political context, see IM Smart, ‘The political ideas of the Scottish Covenanters, 1638–1688’ (1980) 1 History of Political Thought 167, 173. 30 Tomkins, Our Republican Constitution (n 18) 114. It is pertinent to note that while Tomkins discusses the history of the Covenanting Parliament, he does not mention the actions discussed here. 31 Alexander, ‘A systematic theory of tradition’ (n 26) 12.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  111 D.  Core and Canon in Constitutional Theory Two other processes operate alongside memorialisation: ‘elevation’, by which certain texts acquire canonical status, and ‘sacralisation’ or, less provocatively, ‘axiomatisation’ by which certain precepts acquire the status of an unchallengeable core. The three processes are closely intertwined in the making of constitutional theories, which are defined both by the processes themselves and by how they interact, as Figure 1 illustrates. Figure 1  The interaction between core, canon, and continuity

on an

Reimaginin g con tinu ity or rev isi ng

Re j

Memorialisation

Textual facts n

io

at

ev El

Formative facts

continuity

Public facts

ing agin reim or re co he gt tin ec

th ec

Continuity

Ax (sa iom cr at al isa isa t tio ion n)

Canon

Core

Re

v is i

ng th

e canon or rejecting the

cor

e

A canon includes not just the written text of a constitution (in jurisdictions that have one), but also other texts. It may include cases seen as being of high constitutional significance: for example, Chief Justice Marshall’s decision establishing the principle

112  TT Arvind and Lindsay Stirton of judicial review in the US in Marbury v Madison,32 or the decision of the Supreme Court of India in Kesavananda Bharati v State of Kerala33 establishing the doctrine of basic structure under which a constitution may not be amended in a way that cuts against its fundamental elements or values.34 It may also include texts that do not have the force of law but nevertheless are seen by a theorist as carrying weight. These could be historical documents, such as the Federalist Papers in the US, or they could be particularly influential commentaries, such as the commentaries of Ito Hirobumi on the Meiji Constitution in pre-War Japan,35 or even works of legal or political philosophy: for example, the status occupied by the works of Joseph Raz or John Rawls in many present-day constitutional theories in the common law world. As with continuity, theorists can and do differ on which texts belong in the canon: for example, the debate among constitutional theorists in the UK about whether the decision in Anisminic36 should be seen as an important part of the canon or as a case that borders on the heretical. In contrast to continuities and canons, theories of constitutional law do not need to affirm a core, although in practice most do. A core consists of central precepts whose importance to a constitution is seen as being so foundational that they are sacralised and placed beyond question or beyond criticism. The idea of parliamentary sovereignty constitutes precisely such a core in many modern theories of the UK’s Constitution, and jurisdictions such as India37 and Kenya38 which have a ‘basic structure’ doctrine holding certain aspects of their constitutions to be unamendable have arguably turned those aspects into a non-challengeable core. A core may be the subject of broad consensus. The idea of the ‘balanced constitution’ that dominated constitutional thought in the UK from the sixteenth century to the 1830s is an example: it was accepted as underpinning the UK’s Constitution by thinkers as diverse as Blackstone, Henry St. John Bolingbroke, Edmund Burke, Walter Moyle, and even Charles I.39 But what constitutes the core, too, is capable of being contested, as the strident debates over the place of doctrines such as human rights, or even devolution, within the UK’s constitutional order indicate. Equally, it is wholly possible for a constitutional theory to lack a core if no doctrines are seen as unquestionable. Lord Steyn’s controversial statement in Jackson40 that the supremacy of Parliament was a judicial creation which, like all judicial creations, 32 (1803) 5 US 137. 33 AIR 1973 SC 1461. 34 The doctrine in turn was influenced by the work of the German jurist Dietrich Conrad. For a discussion, see M Polzin, ‘The basic-structure doctrine and its German and French origins: A tale of migration, integration, invention and forgetting’ (2021) 5 Indian Law Review 45. 35 For an accessible overview, see T Masuda, ‘The Meiji Constitution: Theory and Practice, 1890–1913’ (1991) 1 East Asian History 125. 36 Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147 (HL). 37 Kesavananda Bharati v State of Kerala AIR 1973 SC 1461. 38 David Ndii v Attorney General, High Court of Kenya, decision of 13 May 2021, available at www.afronomicslaw.org/sites/default/files/pdf/BBI%20Consolidated%20Judgment%20-%20Final%20Version% 20-%20As%20Delivered.pdf. 39 On the Royalist contribution to the received understanding of the doctrine, see CC Weston, ‘English Constitutional Doctrines from the Fifteenth Century to the Seventeenth: II. The Theory of Mixed Monarchy under Charles I and after’ (1960) 75 English Historical Review 426. 40 R (Jackson) v Attorney General [2005] UKHL 56, [2006] 1 AC 262.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  113 could be judicially qualified or limited41 is an example of precisely such a position. To Lord Steyn, whilst the UK’s Constitution has a canon, none of its precepts in that canon are beyond questioning, with the result that there is nothing in the Constitution that can be said to constitute a core. As Figure 1 illustrates, the process of theorisation places core, canon, and continuity in constant dialogue – and, occasionally, conflict – with each other. Not every case or other text that might have a claim to canonical status will fit with a given theorist’s understanding of continuity, nor will it necessarily fit with that theorist’s view of the core. The same is true of the facts that might have a claim to form part of a polity’s constitutional continuity, and the precepts that might have a claim to form part of its constitutional core. In such a situation, the theorist faces the task of either resolving the seeming conflict, or altering their picture of the core, canon, or continuity by rejecting or reworking portions of them. The history of constitutional theory and constitutional conflict provides several examples of this process of rejection, harmonisation, and resolution. The eighteenthcentury Dutch debate between the Orangists and Patriots, for example, saw the Orangists reject the idea that political liberty formed part of the core of the Dutch Constitution, because they saw it conflicting with the prosperous commercial society underpinned by civil (rather than political) liberty, which they saw as being at the heart of constitutional continuity in the Netherlands.42 Similarly, the difference between the constitutional theories of Hozumi Yatsuka and Minobe Tatsukichi in pre-War Japan revolved at a fundamental level around the clash between core and continuity – and, in particular, whether the adoption of a new, western-influenced Constitution had created a new, and more liberal, core which limited the Emperor’s powers (as Minobe argued), or whether (as Hozumi argued) even the most liberal provisions in the Constitution must be read subject to the deeper constitutional continuity represented by the traditional idea of the state as a family with the Emperor at its head.43 An even starker example is provided by the support which many leading constitutional scholars in Turkey gave the military coup of 27 May 1960. The ousted government, they argued, had repeatedly violated the secular republican principles on which Ataturk’s Turkey was founded. Given the Constitution’s inadequacy when it came to defending that core, the only solution was to sweep away both the government and the Constitution, and put in their place a new set of arrangements. The ‘canon’ of the constitutional text could be rejected in defence of the deeper core represented by Ataturk’s secular republican vision.44 These examples are historical, but the pattern also holds in relation to present day constitutional theory. Consider, for example, Lord Steyn’s controversial statements on Parliamentary sovereignty in Jackson. Lord Steyn was, arguably, rejecting the idea that Parliamentary sovereignty constituted the core of the Constitution 41 ibid, [102]. 42 Kluit, Academische Redevoering (n 1) 104–108. 43 Masuda, ‘The Meiji Constitution’ (n 35) 135–138. For a somewhat unsympathetic but factually accurate discussion of Hozumi’s position, see R Minear, Japanese Tradition and Western Law: Emperor, State, and Law in the thought of Hozumi Yatsuka (Cambridge, Harvard University Press, 1970). 44 ‘The Principles identified by the Constitutional Commission’, Resmi Gazete, 1 July 1960, 1636 (Part I); 2 July 1960, 1644 (Part II); translation on file with the authors.

114  TT Arvind and Lindsay Stirton precisely because he was unable to reconcile it with what he believed to be an important element that emerged from the canon, namely, the historical commitment of the judiciary to protecting the polity from arbitrary and despotic government. In effect, when faced with a clash between canon and core, his position led him to reject the orthodox account of the core rather than revise the canon. To sum up: our purpose in this section has been to shed light on how constitutional theorists reach, and come to have epistemic confidence in, the position that their constitutional theories contain a more accurate description of constitutional facts than other competing theories. As we have shown, the process involves ascribing salience and significance to specific facts, and is shaped by the theorist’s views on the constitution’s continuities, canons, and core precepts. In the next section, we show that this process in turn is shaped by underlying predispositions stemming from how a theorist sees the polity in question, which cumulatively have a largely negative effect on constitutional theory. III.  A METHODOLOGY FOR CONSTITUTIONAL THEORY

A.  A Taxonomy of Constitutional World-Views The argument in section II has demonstrated that the claim that one constitutional theory is superior to another cannot be defended with reference to which theory is the better fit with salient and significant facts: questions of salience and significance of constitutional facts are themselves a product of processes of constitutional world-making that are highly contingent. In a companion piece to the present ­chapter, co-written with Simon Halliday,45 we proposed a taxonomy, derived from Mary Douglas’s grid-group cultural theory, for exploring and classifying the different constitutional understandings of the task of judicial review in relation to administrative justice. In order to understand the divergent epistemologies underpinning constitutional world-views, we propose a similar scheme, based on the same twodimensional classification. Douglas terms her dimensions ‘grid’ and ‘group’. The grid dimensions represent the spectrum between transactions mediated by rigid institutional classification and those where structure is minimal.46 This presents an obvious analogy with the formal/ substantive distinction familiar to constitutional theory. Formalist understandings of the constitution emphasise the authority and legal competence of constitutional actors to exercise particular powers on particular grounds, even where that comes ‘at the expense of justice or wise policy or efficiency in the individual case’.47

45 TT Arvind, S Halliday, and L Stirton, ‘Judicial Review and Administrative Justice’ in J Tomlinson and others (eds), The Oxford Handbook of Administrative Justice (Oxford, Oxford University Press, 2021). 46 M Douglas, Risk and Blame: Essays in Cultural Theory (London, Routledge, 1992) 106. 47 F Schauer, Thinking Like a Lawyer: A New Introduction to Legal Reasoning (Cambridge, Harvard University Press, 2009) 35.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  115 A substantivist approach in contrast accords less priority to structure, and more to the achievement of fundamental social values, or desirable outcomes.48 The group dimension, which represents the extent to which an individual is incorporated into a group, can similarly be interpreted in terms of the extent to which the constitution defines and embraces – and indeed imposes – broadly accepted social goals. This dimension is captured in Banakar and Travers’ distinction between ‘conflict’ and ‘consensus’ perspectives on society49 – though we prefer to use the terms ‘discord’ and concord’, to capture the distinction between a polity characterised by ‘conflict, bargaining and agreement’50 and one united by a conception of the common good sustained and realised through the constitutional order. Combining the two dimensions, we can identify four distinct ways of seeing constitutions, illustrated in Figure 2. These are discussed in subsections (B)–(E) below. Figure 2  Four ways of seeing constitutions

Formal

Truce

(formal, discord-oriented)

Rulebook

(formal, concord-oriented)

Discord

Concord

Shield

(substantive, discord-oriented)

Cornerstone

(substantive, concord-oriented)

Substantive

48 Cf the contrast between Allan’s account of the principle of the rule of law at TRS Allan, The ­sovereignty of law: freedom, constitution, and common law (Oxford, Oxford University Press, 2013) 12 and Dicey’s emphasis on the limits on the powers of government. 49 R Banakar and M Travers, ‘Introduction’ in Law and social theory (Oxford, Hart Publishing, 2014). 50 R Dahl, ‘Decision-making in a democracy: the Supreme Court as a national policy-maker’ (1957) 6 Journal of Public Law 279, 294.

116  TT Arvind and Lindsay Stirton B.  The Constitution as a Shield The lower left taxon sees the constitution as a shield to protect individuals against aggression by others and against intrusions by the state. This image of the constitution is exemplified by the seventeenth century English constitutional theorist, John Locke, who understood government as being brought into being by the Constitution, ‘… for the Preservation of every Mans Right and Property, by preserving him from the Violence or Injury of others …’.51 Locke’s account of government as emerging from the consent of the people to form a commonwealth for the better protection of their rights has influenced libertarian readings of the US Constitution. Randy Barnett, for example, has offered a modern, Lockean reading of the US Constitution.52 The conception of the constitution as a shield influences both the manner in which continuity with the past is read, and the manner in which the implications of the canon are selected. Janet Ajzenstat, for example, has argued that the fathers of Canada’s Confederation set about to create a civic government based on the actual consent of the people – people who unsurprisingly held different views on such issues as the extent of collective provision of welfare, civil liberties versus national security, and many other issues.53 She reads the debates on Confederation in the colonial assemblies as demonstrating that a real concern at the time was how such consent could meaningfully given – whether the assent of the assemblies was sufficient, or whether it required the assent of the people directly, through a plebiscite for example. TRS Allan, for his part, draws on the image of the constitution as a shield to frame the central problem of public law as that of ‘… devising means for the protection and enhancement of individual human rights in a manner consistent with the democratic basis of our institutions’.54 Here, the constitution’s ability to provide an assurance that natural rights are not violated takes the place of consent as providing the constitutional order’s core, thus combining a substantive conception of the law with a propensity to assign significance to facts depicting society as a field of discord. The link between rights-based approaches and a discord perspective on society is well captured by the claim of Sir John Laws that ‘the language of rights is not the language of morality but of conflict’.55 Because this tradition sees conflicts inevitably arise between the respective goals, plans, visions and values held by different persons, as well as between these and the collective interests of the polity, its measure of the worth of the constitution is its ability to secure an appropriately defined autonomous sphere in which individuals can pursue their goals, plans, visions and values. The proper role of the constitution

51 J Locke, ‘First Treatise’ in P Laslett (ed), Two Treatises on Government (first published 1689, Cambridge, Cambridge University Press 1960), §92. 52 RE Barnett, Restoring the Lost Constitution: The Presumption of Liberty – Updated Edition (Princeton, Princeton University Press, 2013). 53 J Ajzenstat, Canadian Founding: John Locke and Parliament (Montreal, McGill-Queen’s University Press, 2007). 54 TRS Allan, ‘Legislative supremacy and the rule of law: democracy and constitutionalism’ (1986) 44 CLJ 111. 55 J Laws, ‘The Constitution: Morals and Rights’ [1996] PL 622, 626.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  117 is preventing governments from pursuing an expansive conception of the common good where doing so would intrude into this protected sphere. Questions of institutional form are not ignored, but they take second place to questions of substance. The constitutional rulebook’s significance is only as a means to the end of upholding the principles that constitute the tradition’s core. Locke’s argument that the legislative power should be vested at least in part in an elected assembly, for example, was not based on a belief that the endorsement of the peoples’ representatives was sufficient to imbue any positive enactment with the requisite authority. Rather, Locke believed that if the approval of such a body were required for passing legislation or levying taxes, these would be no more oppressive nor restrictive of liberty than that which was necessary to the proper, limited purpose of government. The process of deliberation and assent, in other words, mattered because it made it more likely that the resulting laws would respect the rights of persons. C.  The Constitution as the Cornerstone of Social and Political Life Remaining at the substantive end of the vertical axis, but moving towards concord brings us to the image of the constitution as a cornerstone of social and political life. This image differs from the understanding of the constitution as a shield in that it allows for a fuller conception of the public good, and hence of a sense of purpose for the polity – a sense of purpose which it is the function of the constitution to nourish and sustain. The image of the constitution as a cornerstone comes from the subtitle of Granville Austin’s account of the founding of India’s Constitution: ‘the cornerstone of a nation’. The combination in this perspective of a substantive interpretation of constitutional law with a concordant, even purposive view of the polity, is seen in Austin’s account of the purposes of the constitution as understood by members of the constituent assembly: The Constitution was to foster the achievement of many goals. Transcendent among them was that of social revolution. Through this revolution would be fulfilled the basic needs of the common man, and, it was hoped, this revolution would bring about fundamental changes in the structure of Indian society – a society with a long and glorious cultural tradition, but greatly in need, Assembly members believed, of a powerful infusion of energy and rationalism.56

This illustrates a number of paradigmatic features of the lower right taxon. First, what is being constituted (or more accurately, re-constituted) is not just India’s government, but India’s society and polity, which was to be infused with a sense of purpose. The founders were effecting a social revolution with a number of objectives: national unity and domestic stability, which Austin tells us, were ‘considered necessary prerequisites for a social renascence’,57 the protection of minority interests, an efficient government and administration, the protection of national security, together with institutions to achieve them.58

56 G

Austin, The Indian constitution: Cornerstone of a Nation (Oxford, Clarendon Press, 1966) xi. xi. xi.

57 ibid 58 ibid

118  TT Arvind and Lindsay Stirton If Locke is the exemplar of a seventeenth century theorist of the constitution as a shield, then Thomas Paine writing a century later illustrates the image of the constitution as a cornerstone. For Paine, the purposes of government extended beyond the preservation of order. In Agrarian Justice Paine emerges as an advocate of a rudimentary basic income provided by the commonwealth. In his introduction to Paine’s Political Writings, Kuklick attributes the ease with which Paine combined the defence of liberty with the good of the polity to his Quaker roots: ‘in Quaker doctrine there was a concern not only for individual conscience also for the community of Friends’.59 A number of consequences follow from the expanded space for the public interest that this bottom right understanding allows, compared with the lower-left taxon. Instead of thinking wholly in terms of protection against rights violation, attention turns towards ensuring that the proper identification of the common interest. Much of the Rights of Man, Part I is taken up with a favourable contrast between the French Revolutionary Constitution of 1791 and the English Constitution – which Paine thought was no constitution at all, ‘a thing in name only’. Crucially the points of comparison are overwhelmingly concerned with political rather than civil rights, that is, with creating the means by which the governed community established its common purposes: the franchise, the allocation of seats to the Assembly, frequency of elections, and measures to limit corruption of elected representatives. The exceptions are prohibitions on game laws and monopolies, but even these can be seen as establishing a claim by members of the community on the common resources of the community. The emphasis on corruption and the degeneracy of putting private gain ahead of commonwealth is significant. Just as despotism – the subjugation of individual rights to the will of the ruler – is the primary pathology of government from the point of view of the ‘shield’ tradition, from the perspective of the cornerstone tradition it is the tendency of government and members of the society to degenerate into the advancement of narrow self-interest rather than the pursuit of the interests of the polity. D.  The Constitution as a Rulebook The last two conceptions of the constitution move from the substantive to the formal. In the context of constitutional theory, this leads their accounts of the constitution to emphasise traditional, formal conceptions of institutions, and place in the foreground institutionally oriented ideas such as sovereignty rather than the more free-standing principles that are the primary focus of the two substantively-oriented conceptions. If theories at the substantive end of the scale outlined in Figure 2 can be described as engaged in a search for principle, then theories at the formal end of that scale can be described as engaged in a search for the loci of sovereign authority.

59 B Kuklick, ‘Introduction’ in Thomas Paine, Political Writings (Cambridge, Cambridge University Press, 2000) xviii.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  119 In the upper right quadrant, the focus on institutions combines with a concordoriented view of the polity to produce a conception of the constitution as a rulebook for the institutions of state. Although such a conception does not deny the existence of the principles that are the focus of the ‘shield’ and ‘cornerstone’ conceptions, it tends to subordinate principle to institutions and to read core, canon, and continuity in a manner that emphasises the institutions themselves rather than the principles they embodied at any point of time. Constitutional theory, accordingly, does not simply ask what principles might best resolve a given issue, but which institution is constitutionally tasked with identifying, formulating, and applying the principles that should be used to resolve that issue. The practical consequences of viewing a constitution in these terms are illustrated by the recent jurisprudence of the US Supreme Court on voting rights. In Rucho v Common Cause,60 which concerned partisan gerrymandering, the court did not deny that the redistricting plan at issue was ‘highly partisan’,61 or that it was ‘incompatible with democratic principles’.62 The focus of its analysis, however, was on whether responsibility for dealing with the constitutional challenge posed by partisan gerrymandering lay with the courts, and it held that it did not. In institutional terms, it was to the state legislatures, checked and balanced by the federal legislature, that the issue was assigned.63 Notwithstanding the importance of the principle at stake, in terms of actual constitutionalism the question was who was tasked with upholding it. The idea of the constitution as a rulebook underpins a significant portion of modern constitutional theory, including virtually all of what is termed ‘political constitutionalism’. As the example of Rucho shows, its focus on the loci of institutional authority gives it a propensity to leave important social issues unaddressed if the relevant institution is unwilling or unable to act, and it is this propensity that has formed the basis of the most strident criticisms of it by ‘shield’ and ‘cornerstone’ theorists.64 From a conception-internal perspective, however, this subordination reflects the fact that ‘concord’ is not an organic phenomenon in an institutionally oriented conception. The inherent state of a polity is one of disagreement, not agreement.65 It is only through institutions – and particularly more open-ended political institutions – that we can carve out islands of concord in the polity, making the ‘painful compromises’ that are essential a stable and successful polity.66 ‘Politics is the medium of concilliation’,67 and institutions are necessary to make an attained concord workable. As the Jamaican leader Norman Manley put it, the best-written constitution in the world would be of little avail without institutions animated by a 60 Rucho v Common Cause (2019) 139 SCt 2484 (US Supreme Court). 61 ibid, 2491. 62 ibid, 2512. 63 ibid, 2496. 64 See, eg, the dismissive comments on ‘the slightly absurd game of “hunt ultimate sovereign”’ in N Barry, ‘Sovereignty, the Rule of Recognition, and Constitutional Stability in Britain’ (1994) 2 Hume Papers on Public Policy 10, 15. 65 J Waldron, Law and Disagreement (Oxford, Oxford University Press, 1999) 1. 66 C Forsyth, ‘Showing the fly the way out of the flybottle: The value of formalism and conceptual reasoning in Administrative Law’ (2007) 66 CLJ 325, 336. 67 Tomkins, ‘In Defence of the Political Constitution’ (n 11) 174.

120  TT Arvind and Lindsay Stirton democratic spirit, of the type which he held Jamaica to have ‘built up and preserved over the years’.68 The political philosophy underpinning this constitutional worldview is instantiated by Hobbes’ famous image of Leviathan as the body composite of all members of the polity, and similar ideas are seen in constitutional theories around the world. The Indian Constitution, for example, is typically seen as a cornerstone, as section III.C has discussed. But at the time of its framing, leading jurists such as B.R. Ambedkar, who chaired the committee that drafted it, saw it far more as an institutional rulebook. To Ambedkar, the Constitution’s core lay in its creation of strong centralised governmental institutions that could develop a consensus around the post-independence social revolution’s intervention in long-established social structures and practices.69 The Constitution’s social provisions, seen by many modern theorists as clear evidence of its place as a cornerstone,70 was to Ambedkar simply part of the rulebook, akin to the Instruments of Instructions that were an important part of colonial administrative practice.71 This does not mean these theories have no role for courts. The courts, too, have the capacity to provide a field on which individuals can articulate ‘the democratic demand of the citizen to be heard’.72 Constitutional rights can also have a place in a rulebook theory, as the debate on drafting Jamaica’s Constitution indicates. In his speech to Jamaica’s Legislative Council on the proposed Constitution, Manley endorsed the incorporation of a Charter of Fundamental Rights and Liberties, but not for the reasons that would be offered in a shield-based theory. The Charter, Manley said, would guide a well-meaning government as a framework within which it could work and outside which it must not stray. Equally, however, it was essential that such rules should not limit government discretion inappropriately. ‘If you are not careful,’ Manley added, ‘you may write yourself out of the right to do many things that are good and useful to the community.’73 In consequence, the role of courts and rights is also circumscribed by the constitutional rulebook, and a court that exceeds those bounds acts unconstitutionally, no matter how pressing the issue. The nature of the positions to which this leads becomes clear if we compare the positions on Miller (No. 2) taken by Paul Craig and John Finnis. Craig was writing from a perspective emphasising principles over institutions and discord over concord. Discord can affect institutions just as it can affect individuals; and if the constitution is a shield then sometimes that shield will need to also protect Parliament against an intransigent executive in times of discord. Indeed, to fail to so extend the shield calls into question the principles that are the core of

68 NW Manley, ‘The Independence Constitution’ in R Nettleford (ed), Norman Washington Manley and the new Jamaica: Selected speeches and writings, 1938–68 (New York, Africana Publishing Corp, 1971) 310. 69 BR Ambedkar, Federation versus Freedom (Pune, Gokhale Institute, 1939). 70 See, eg, T Khaitan, ‘Directive principles and the expressive accommodation of ideological dissenters’ (2018) 16 ICON 389. 71 See his speech to the Constituent Assembly introducing the draft Constitution, Constituent Assembly Debates, Vol VII, 4 November 1948, 7.48. 241–244. 72 DAO Edward, ‘The Role of Law in the Rule of Law’ (1994) 2 Hume Papers on Public Policy 10. 73 Manley, ‘The Independence Constitution’ (n 68) 304–5.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  121 the constitutional order.74 To Finnis, in contrast, the constitution is a rulebook, and to read down a distinction between ‘law’ and ‘convention’ that the rulebook formerly accepted is to severely disrupt the institutional configuration that is the core of the constitutional order.75 E.  The Constitution as a Truce The last category sees the constitution as a social and political truce between factions with significantly different levels of economic and/or political power. It shares with rulebook theories an emphasis on institutions but, unlike that view, is pessimistic about the ability of institutions to produce concord; and unlike theories that treat constitutions as akin to shields, it is also sceptical about the existence of principles capable of commanding sufficient assent to protect zones of personal autonomy. It is the state’s task to stop tensions building up ‘potential towards explosion’,76 and there are few effective constraints on it when it acts. Carl Schmitt’s theory instantiates this understanding. Schmitt used the repeated constitutional crises of the Weimar state to argue that it was the power of the ­sovereign – made manifest in the ability to invoke a state of exception and quell internal enemies – that constitutes the bulwark against discord threatening the political community.77 Other constitutional theories seeking to justify autocratic government similarly build on this conception. Tsuzuki Keiroku’s controversial writings on the Meiji Constitution, for example, took the view that the national interest must take precedence over public opinion. It was the right and the duty of the Imperial bureaucracy, in discharge of their responsibility to the Emperor, to reject the legislature’s views if they conflicted with the national interest. If the legislature persisted in its view, the appropriate course of action was ‘as many successive dissolutions of the Diet as necessary’ until it accepted the bureaucracy’s position.78 Likewise, in his critical study of West African constitutional systems after decolonisation, W. Arthur Lewis argued that political leaders such as Sekou Touré and Nkrumah, who believed their societies to be afflicted by tribal and ethnic antagonisms which menaced the unity of the state, had embraced totalitarian single-party rule as the only route to containing and overcoming these antagonisms and preventing the polity’s disintegration.79 Quite apart from the insight it gives us into the constitutional underpinnings of autocratic states – an issue rarely discussed in mainstream constitutional 74 PP Craig, ‘The Supreme Court, Prorogation and Constitutional Principle’ [2020] PL 248. 75 J Finnis, ‘The Unconstitutionality of the Supreme Court’s Prorogation Judgment’ (University of Oxford Research Paper 6/2020, 2020) http://ssrn.com/abstract=3548657 accessed 1 July 2021. 76 K Llewellyn, ‘The normative, the legal, and the law-jobs: The problem of juristic method’ (1940) 49 Yale Law Journal 1355, 1375–1376. 77 AJ Jacobson and B Schlink (eds), Weimar: A Jurisprudence of Crisis (Berkeley, University of California Press, 2000) provides an excellent overview of the broader context of Weimar constitutional theory and Schmitt’s place within it. 78 J Banno, The establishment of the Japanese constitutional system (JAA Stockwin tr, London, Routledge, 1992) 31–38. 79 WA Lewis, Politics in West Africa (Oxford, Oxford University Press, 1965) 50–2.

122  TT Arvind and Lindsay Stirton theory – this conception also sheds light on a neglected dimension of constitutional theory. The idea that the stability of a polity hinges on ‘painful compromises’ is never far from the surface even in mainstream constitutional theories; and while constitutional theorists rarely endorse this conception’s institutional prescriptions, they do nevertheless draw on the diagnostic resources it provides while looking elsewhere for solutions.80 Lewis himself instantiates this. Although he accepted that West African societies were characterised by mutual antagonism, strongman rule only created a ‘fascist state’81 without resolving the antagonism – unlike well-designed pluralist institutions. As he put it, … in the inflamed atmosphere of the plural society, the unspectacular Prime Minister, who shuffles along, but somehow manages to keep things going on an even keel, may build a permanent community, where the strong man may leave only fragments behind.82

The imagery of confrontation and barely stable compromise also underpins the recent critical literature on the South African Constitution, arguing that the Constitution represents an unsatisfactory attempt to paper over significant and growing differences between the different sides who negotiated the post-Apartheid settlement,83 but cannot meet the aspirations of the black majority without adopting an expressly anti-colonial perspective.84 Perhaps more strikingly, it was precisely this imagery that underpinned the justification of the Turkish military coup of 27 May 1960 by constitutional scholars: the ousted government had turned the state from ‘a social power dependant on the law’ to ‘an autocratic power that represents the interests of a group’. The discord it had sowed placed an obligation on the army to ‘reestablish the organizations of the state … and legitimate government’.85 In a truce conception, a group that holds itself unjustly marginalised by a constitutional order will also hold itself justified in rejecting that order ‘in the name of an alternative and deeper legal and moral foundation’.86 That the real world bears this out suggests there is reason to take the truce conception’s claims seriously. IV.  HEURISTICS AND THE LIMITS OF RHETORIC

The discussion thus far has sought to outline the implicit methodology of constitutional theory and critically examine its strengths and weaknesses. We have shown that constitutional theory has an ingrained tendency to lapse into what Hirschman 80 Cf Murray’s argument that the social rights in the Irish, German, and Mexican Constitutions were motivated by elites’ fear of social revolution: T Murray, ‘Socio-Economic Rights Versus Social Revolution? Constitution Making in Germany, Mexico and Ireland, 1917–1923’ (2015) 24 Social & Legal Studies 487. 81 Lewis to Keenleyside, quoted in RL Tignor, W. Arthur Lewis and the Birth of Development Economics (Princeton, Princeton University Press, 2006) 172. 82 Lewis, Politics in West Africa (n 79) 80. 83 P du Toit, C Swart, and S Teuteberg, South Africa and the Case for Renegotiating the Peace (Stellenbosch, SUN Press, 2016). 84 T Madlingozi, ‘Social justice in a time of neo-apartheid constitutionalism: Critiquing the anti-black economy of recognition, incorporation and distribution’ [2017] Stellenbosch Law Review 123. 85 ‘Principles identified by the Constitutional Commission’ (n 44). 86 S Halliday and B Morgan, ‘I Fought the Law and the Law Won? Legal Consciousness and the Critical Imagination’ (2013) 66 Current Legal Problems 1, 18–19.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  123 evocatively termed ‘rhetorics of intransigence’, whose orientation towards formulating and honing arguments exacerbates rather than ameliorates the challenges and divides that characterise modern constitutional law.87 In this final section, we offer a deeper diagnosis of why constitutional theory has a propensity to so lapse and suggest how the analysis set out in this chapter points to a possible path out of this intransigence. A.  Understanding the Impasse As section II discussed, constitutional theory differs from many other forms of theory in that it seeks not just to understand the past or present, but also to shape the future. In most areas of theory, it is possible to draw a clear distinction between descriptive or analytical theory, on the one hand, and normative theory on the other. An astrophysicist describing the nature of black holes, for example, is not thereby seeking to influence black holes to behave in the manner the theory describes. Similarly, an anthropologist describing the nature of a society is not doing so to persuade that society to reshape itself so it better fits with the description. Constitutional theory is different.88 Constitutional theorists seek not just to describe facts, but also to create facts. In formulating a theory of a constitution, a theorist is not just describing the existing constitutional order, but also seeking to provide a framework that can integrate new developments into the constitutional order and chart a path for that constitutional order in a still-unfolding future. This is not, and cannot be, a simple matter of describing and analysing facts. It is inherently concerned with making new facts, and it inherently straddles the boundary between analytical and normative theory. A constitutional theorist outlining a theory of the relationship between different governing institutions is – in sharp contrast to an astrophysicist or anthropologist – seeking precisely to persuade the institutions to behave in the manner the theory suggests. This means that in law, unlike in most other disciplines, the continued existence and future states of an institutional fact cannot be separated from belief in their continued existence and expectations of their future state.89 Legal theories deal not just with present facts but also with future facts, and they do so in a context in which whether or not a particular future fact eventuates depends in significant part on a ­theory’s success in influencing institutional behaviour. It is this that gives the predilections discussed in section III so much bite. The epistemic confidence that constitutional theories place in the validity of the factual underpinnings of their theories is not justifiable because that confidence is inextricably linked to the future facts 87 See AO Hirschman, The Rhetoric of Reaction: Perversity, Futility, Jeopardy (Cambridge, Belknap Press, 1991) 168. 88 Although we focus on constitutional theory in this section, the points we make here also apply to legal theory more generally. 89 Cf the discussion in A Ross, On Law and Justice (U Bindreiter tr, Oxford, Oxford University Press, 2019) 58–63. Note that the references to ‘legal politics’ in the English edition are the result of a mistranslation. Danish does not distinguish between ‘policy’ and ‘politics’ (both are ‘politik’), and in the overwhelming majority of cases Ross meant the former rather than the latter.

124  TT Arvind and Lindsay Stirton the theory seeks to produce, and the types of future facts to which constitutional theory relates are necessarily of a type whose eventuation is so tenuous that one cannot achieve a justifiable belief that they will come to pass.90 Constitutional theorists have not, of course, been inattentive to this point, but the solutions they have suggested have only served to exacerbate the problem. TRS Allan, for example, has sought to avoid the epistemic trap we have outlined above by suggesting that constitutional theorists are doing no more than reporting our own opinion, based on a view of constitutional practice that we find ­defensible – an account of our practice that shows how legal doctrine furthers the moral and political values that a good constitution ought to serve.91

There is an obvious allure to this way of framing the difficulty, but it is equally precisely the acceptance of this approach that has led to debates within constitutional theory assuming the form of mere rhetorical clashes between different styles of public law thought.92 If the starting point is that this account furthers the moral and political values that any good constitution would, then it follows that other accounts do not and, for that reason, must be rejected. Constitutional debate, thus, becomes an effort to slay the ‘misshapen monsters’, as Kluit termed them, created by competing theories. There is a strong parallel between the current state of constitutional theory and the problem of rhetorics of intransigence which was the focus of Hirschman’s work. Hirschman was motivated, as we are, by a concern that public discourse had come to be ever more characterised by an impasse in which opposing schools repeatedly deployed the same rhetorical tools against each other.93 Positions were advanced on the basis that they were necessary to avoid adverse consequences,94 and they were opposed on the basis that they would bring adverse consequences95 or jeopardise well-functioning institutions that had stood the test of time.96 Alternatively, they were advanced on the basis that they ‘fit’ with the existing political order97 and with the general trend of historical development,98 but were opposed on the basis that they sought to change characteristics that were so deeply entrenched in the social order that any attempt to change them was bound to fail.99 Hirschman argued that unless public discourse moved beyond these positions, it would descend into a ‘dialogue of the deaf’ that was far from friendly to the type of open-minded opinion-forming processes on which liberal democracy depends.100 90 This is why, as we see it, one cannot judge between constitutional theories on the basis of which provides the best fit, or for that matter the best moral justification for a constitutional practice. Judgements about the relevant practices themselves are baked into the selection of facts and practices that demand explanation, and which don’t. For a sophisticated argument to the contrary see S Lakin, ‘Why Common Law Constitutionalism is Correct (If It Is)’, ch 5 in this volume. 91 Allan, Sovereignty of law (n 48) 19. 92 Loughlin, Public Law and Political Theory (n 8) 62. 93 Hirschman, The Rhetoric of Reaction (n 87) 167–170. 94 ibid 152–153. 95 ibid 11–42. 96 ibid 81–132. 97 ibid 149–151. 98 ibid 154–159. 99 ibid 43–80. 100 ibid 168–170.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  125 B.  From Rhetoric to Heuristics: Towards a Resolution How, then, can we overcome this impasse? The answer lies in acknowledging that the epistemic confidence we can place in any single account of the constitution is necessarily limited. Rather than approaching constitutional theory from a starting point which holds that there is always one account of constitutional practices that is superior to its competitors, one should approach the area with the appreciation that no theoretical position in constitutional law does justice to the variety and diversity of the body of social facts from which constitutional theories are drawn.101 The construction of a theory is a process not just of expanding our constitutional world by reading deeper significance into constitutional practices, but also of shrinking and reducing our constitutional world by winnowing out social facts to which one refuses to attach constitutional significance – a process that as discussed in section III is fundamentally shaped by the theorist’s pre-existing worldview in relation to where the balance of significance lies as between the institutional and the principled, and the civil and the solidary. A constitutional theorist, in consequence, is operating not nearly so much like a theorist of the natural or social world as like an editor preparing a recension of a text attested in multiple fragments or in competing versions. Recensions are versions of a work produced by a process of critical revision,102 and represent an attempt to reconstruct an archetype, or best recoverable form, of a text from the available evidence of the contents of that text.103 An editor preparing a recension of a text works on the basis of a wide body of evidence as to its content: manuscripts and early printed evidence, but also translations and quotes in other texts. Crucially, and precisely like constitutional theorists, an editor preparing a recension must not just use sources providing evidence of the best recoverable form, but also reject the evidence of sources that contradict the reconstructed reading; and will typically do so based on a pre-existing view as to the nature of the text and its contents. This process is strikingly similar to the manner in which Allan, in the work quoted above, describes constitutional theorists exercising judgement and forming opinions to construct and advance the reading of the constitution they hold to be the most defensible.104 The difficulty with relying on a recension-like methodology in constitutional theory, however, lies in the wider implications of constitutional theory. As we have argued above, constitutional theories – unlike recensions of texts – are produced not simply for their own sake, but to exercise influence over how a given polity conducts public affairs. Shrinking a constitutional world by rejecting a particular 101 In this regard, we are in partial agreement and partial disagreement with Sarah Nason’s contribution, S Nason, ‘Methodological Pluralism and Modern Administrative Law’, ch 6 in this volume. As should be apparent from Arvind, Halliday, and Stirton, ‘Judicial Review and Administrative Justice’ (n 45), we agree that administrative law must be understood in pluralist and empirically grounded terms if it is to successfully solve the problems to which it is addressed, but as outlined in this chapter, we believe that a heuristic constitutional theory will show precisely such a problem-solving orientation. 102 C Baldick, The Oxford Dictionary of Literary Terms 3rd edn (Oxford, Oxford University Press, 2015) 304. 103 P Beal, A dictionary of English manuscript terminology, 1450–2000 (Oxford, Oxford University Press, 2008) 333. 104 Allan, Sovereignty of law (n 48) 19.

126  TT Arvind and Lindsay Stirton social fact from it is, therefore, not simply a matter of judgement. It seeks to eliminate or close off the possibility of particular forms of institutional action, or of priority being accorded to particular types of principles. Theories typically justify this with reference to the risks and downsides which those types of institutional action or principles carry, but a shrinking of the constitutional world also inevitably entails trade-offs carrying their own risks and dangers, the likelihood of whose eventuation will never be known with certitude.105 The solution to the difficulty this poses, we suggest, lies in moving towards a more heuristic conception of constitutional theory. The terminology of ‘heuristics’ requires a word of explanation. There is a significant strand of literature in legal theory that views heuristics with suspicion, seeing it as having a propensity to lead to systematic error in factual and normative judgements.106 Heuristics, however, has two distinct aspects.107 The focus of legal scholarship has largely been on only one of these aspects, namely, the use of mental shortcuts as a rapid way of answering challenging questions. Heuristics in this sense is closely associated with biases, and it is unsurprising that legal theory has been critical of it. The other aspect of heuristics is different. Here, heuristics are used because it reflects that the problem itself is intractable or poorly defined, and not merely to find a simple shortcut to make decisions that could also have been made in more optimal ways. A classic example is an environment characterised by multiple needs, drives, and goals that must be balanced, the relationship between which, and the content of each of which, is far from certain.108 In a situation like this, characterised not just by incommensurability of goals but also by uncertainty as to the extent and loci of the incommensurability, there is no nonheuristic way of resolving the problem.109 In his work on styles of public administration, Christopher Hood provides an elaboration of the difference between heuristic and non-heuristic conceptions of theoretical work, which also applies to constitutional theory.110 Non-heuristic ­ approaches to constitutional theory take their starting point in the social precommitments they believe to be embedded in a particular constitutional order. The theorist’s task is to elaborate on those pre-commitments, analyse their implications for the type of outcomes that should be promoted in the conduct of public affairs, identify situations where those outcomes are not in fact being promoted, and put forward suggestions for how the conduct of public affairs might be altered to promote those outcomes. It is this pattern that constitutional theory as conventionally done follows. A heuristic approach, in contrast, takes a very different starting point. Even if there is agreement that the constitution does in fact embed certain pre-commitments, there 105 Cf the very similar points made in Hirschman, The Rhetoric of Reaction (n 87) 168–170. 106 See, eg, CR Sunstein, ‘Hazardous heuristics’ (2003) 70 University of Chicago Law Review 751; CR Sunstein, ‘Moral heuristics’ (2005) 28 Behavioral and Brain Sciences 531. 107 For a fuller discussion, see C Engel and G Gigerenzer, ‘Law and heuristics: An interdisciplinary venture’ in C Engel and G Gigerenzer (eds), Heuristics and the law (Cambridge, MIT Press, 2006). 108 HA Simon, ‘Rational choice and the structure of the environment’ (1956) 63 Psychological Review 129. 109 There are obvious parallels between our discussion of heuristics and the concept of improvised order in D Howarth, ‘The British Constitution as Improvised Order’, ch 11 in this volume. 110 C Hood, ‘When extremes meet: “Sprat” versus “Shark” in Public Risk Management’ in C Hood and DKC Jones (eds), Accident and Design: Contemporary Debates on Risk Management (London, Routledge, 1996).

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  127 is uncertainty about what those commitments are, what their practical implications are in a given situation, and how one resolves a situation in which multiple competing commitments pull the conduct of public affairs in different directions. This, it should be stressed, is not because the heuristic conception sees polities as necessary located on the discord end of the spectrum. Rather, it is because of the epistemic limitations of non-heuristic approaches to constitutional theory discussed in sections II and III: the reductionism that is necessarily and always embedded in them, the impossibility of having epistemic confidence in the judgements on which that reductionism is based, and the shaping role played by theoretical predispositions in producing them. Constitutional heuristics, in consequence, is willing to modify initially held opinions in light of new evidence, and tends away from the all-or-nothing outcomes to which the current structure of constitutional debate leads.111 A heuristic approach to constitutional theory will differ methodologically from current constitutional theory in two ways. First, heuristics require theorists to be conscious of the predispositions that underpin their engagement with social facts – of where, in other words, they situate themselves on the spectrum between focusing on principles and focusing on institutions, and assuming civil existence or assuming solidary existence. They need, equally, to be conscious of the resulting predispositions in relation to what they expect to see in the material, and the risks and trade-offs inherent in the form of constitutional reductionism they favour. A heuristic approach is, accordingly, willing to countenance messiness, and to explore and accept forms of constitutional action that sit uneasily with its theoretical predispositions but nevertheless provide workable solutions to constitutional problems. Its focus is on the type of facts the theorist seeks to create, and the present and future social needs to which those facts seek to respond, rather than on cohering strictly to a tightly defined set of philosophical positions. In engaging with doctrine and practice its focus is not simply on how we might justify those doctrines or practices but, rather, on the work they do or can be made to do to meet present and future social needs. Second, and flowing from the first, a more heuristic constitutional theory also approaches sources differently. Where a non-heuristic approach uses and relies on Kuhnian exemplars – ideal-typical scenarios that showcase the normative power of the theory – a heuristic approach focuses on examples – empirically-grounded scenarios that instantiate and illustrate challenging questions facing constitutional theory, and which frequently sit uneasily with the more simplistic scenarios that exemplars represent. In consequence, its primary task is neither a search for a sovereign nor a search for principle, but a search for experience. As Arun Thiruvengadam has persuasively shown in his study of South Asian constitutionalism, constitutional courts regularly draw on and constructively use sources from other jurisdictions in a manner that sustains and furthers their constitutional tradition, and better equips it to meet emerging social needs.112 111 Hirschman, The Rhetoric of Reaction (n 87) 168–170. 112 AK Thiruvengadam, ‘In Pursuit of “the Common Illumination of Our House”: Trans-Judicial Influence and the Origins of PIL Jurisprudence in South Asia’ (2008) 2 Indian Journal of Constitutional Law 67. There is a strong parallel between the phenomena Thiruvengadam discusses and those that underpinned Alan Watson’s theory of legal transplants. See A Watson, Legal Transplants: An Approach to Comparative Law 2nd edn (Athens, University of Georgia Press, 1993).

128  TT Arvind and Lindsay Stirton This type of socially responsive and epistemically open engagement with, and willing to draw on, a diverse range of sources is characteristic of a heuristic approach. Legal and political philosophy still has a role in this process, but as a source of solutions we can borrow, or of problem-definitions we can use to gain greater insight into the competing policy considerations in an area, rather than the normative touchstones and sources of legitimacy they tend to be in non-heuristic approaches. It is this that perhaps constitutes the bigger break with the manner in which constitutional theory is done today but, as we have sought to show in this chapter, it as a break that is necessary if constitutional theory is to be able to move beyond the impasse to which it has been led. V. CONCLUSION

In this chapter, we have argued for a shift in the way we think about and do constitutional theory. Theories, as we have sought to show, are built on an implicit methodology of selecting and ascribing constitutional significance to facts, and memorialising, sacralising, and elevating them to a continuity, core, and canon. These are in turn shaped by predispositions which depend, first, on whether a theorist sees the essence of a polity’s constitutional arrangements as lying in a set of principles or in a configuration of institutions; and second, on whether they see the primary thrust of those arrangements as lying in working to shape concord or general agreement within the polity, or to manage the discord that will necessarily characterise any polity where interests and expectations conflict. We have also proposed an analytical scheme, drawing on the institutional theories of Mary Douglas, to better study and model the way in which these predispositions influence constitutional theory. While other analytical schemes exist, such as Loughlin’s contrast of normativism and functionalism, we believe that the scheme we propose has a number of advantages over existing approaches. Douglas describes her approach to culture as one of ‘social-accounting’, which, ‘selects out of the total cultural field those beliefs and values which are derivable as justifications for action’ and which constitute an ‘implicit cosmology’.113 Analysing these ‘thought styles’, as she terms them, enables us not just to classify theories within a scheme, but also to identify the relationship between justificatory accounts and their implied world-views. Such a focus is one that is particularly necessary in the context of constitutional theory, where little if any consideration has been given to the complex relationship between account-giving, reasons, and beliefs that underpins disagreements about how to read a particular constitution. These points do not, of course, negate the value of constitutional theory as currently done. There is much insight that constitutional theorisation has shed, and will continue to shed on the functioning of constitutional orders. Nevertheless, as the discussion in sections II and III has shown, there has also been a failure to recognise the ways in which the predilections built into constitutional theories act to shape



113 M

Douglas, ‘Cultural Bias’ in In The Active Voice (London, Routledge, 2011) 190.

Slaying the Misshapen Monster: The Case for Constitutional Heuristics  129 and direct the positions they take. We have sought to argue that a greater focus on these processes, and the limitations they impose on constitutional theory, will help move constitutional theory beyond its tendency to lapse into unyielding positions whose allure obscures their limitations. Rather than merely seeking to overcome other theories through persuasion, the greater use of heuristics will open the door to a less polarised constitutional world in which entrenched positions can be the subject of open-minded debate. But even for those who are not persuaded by the case we have sought to make in favour of heuristics, doing constitutional theory with a greater awareness of the biases built into it, and considering more deeply why the form of reductionism that theory takes is justifiable and why belief in the theory is justifiable notwithstanding the predilections that form of reductionism inevitably generates, will lead to a richer and more productive engagement than we currently see.

130

5 Why Common Law Constitutionalism is Correct (If It Is) STUART LAKIN* I. INTRODUCTION

H

ow can we establish that one account of a constitution rather than some other is the correct one – in the sense that it provides an accurate statement of fundamental propositions of legal and constitutional rights, duties, powers, etc? That is the methodological question I shall attempt to answer in this chapter. Call it the ‘motivating question’. I shall use two paradigmatic accounts of the British Constitution as vehicles for my arguments: the positivist, ‘orthodox’ one put forward by Jeffrey Goldsworthy, and the interpretivist, ‘common law constitutionalist’ one propounded by Trevor Allan.1 I shall tentatively conclude that Allan’s account is a better one than Goldsworthy’s – and is therefore arguably the correct one – on the interpretative basis that Allan offers the more morally attractive model of British constitutional practice. From that narrow conclusion about one constitution, I shall extrapolate to a broader conclusion about all constitutions: that it is only by use of the interpretative method that we can establish that one account of a constitution rather than some other is the correct one. My narrow conclusion, if correct, has all sorts of far-reaching implications for the nature and functioning of the British Constitution. It suggests that popular beliefs about (for instance) the relative powers of Parliament and courts, the nature of statutory interpretation, and the correctness or incorrectness of particular judicial decisions, may be mistaken. But that it not my primary interest in this chapter. Instead, I want to explain and defend the interpretative method itself, and to think about its implications for constitutional theory and practice. I shall do this by way of a critique of positivist methods of the sort exemplified in Goldsworthy’s work, * Thank you to Ana Cannilla and Jeffrey Goldsworthy for very thoughtful comments and discussions on an earlier draft. 1 See J Goldsworthy, The Sovereignty of Parliament, History and Philosophy (Oxford, Clarendon Press, 1999); J Goldsworthy Parliamentary Sovereignty, Contemporary Debates (Cambridge, Cambridge University Press, 2010); TRS Allan, The Sovereignty of Law; Freedom, Constitution and Common Law (Oxford, Oxford University Press, 2013).

132  Stuart Lakin but widely adopted or assumed both by British constitutional lawyers and in constitutional law writing around the world. According to these methods, we can only understand a constitutional practice by reference to empirical facts about selected constitutional actors’ beliefs, intentions, utterances and other observable practices. This is to say that constitutional analysis is a descriptive exercise. To bring controversial questions of morality into this domain, say positivists, is to blur the descriptive question of how the constitution does work, with the evaluative question of how it should work. In section II, I begin by setting out the two accounts of the Constitution described above. I then lay out the sense in which the two accounts represent rival models of British constitutional practice: models of which types of facts, configured in which way, are constitutive of the content of the law and the Constitution. The important and inescapable challenge presented by the motivating question, I explain, is to demonstrate which of these (or some other) model gives the correct understanding of the practice. In section III, I consider a range of positivist approaches to that question. These approaches, I contend, share a common flaw: they each, in different ways, assume the correctness of the orthodox model rather than provide some independent argument in its defence. In section IV, I defend the interpretative method as the right type of argument by which to resolve disagreements between rival models. This method correctly approaches the motivating question as speaking to the political legitimacy of a constitutional practice. It is in virtue of this method, I argue, that common law constitutionalism is (probably) correct. II.  TWO ACCOUNTS OF THE BRITISH CONSTITUTION

We begin with a summary of Goldsworthy’s orthodox account of the Constitution (GO) and Allan’s common law constitutionalist account (CLC). I have divided this summary into four distinct, non–exhaustive, propositions, each mirroring the other in their coverage. These propositions will serve as a reference point throughout the chapter. A.  Goldsworthy’s Orthodoxy (GO) GO(1) Every constitution possesses a Rule of Recognition. This rule sets out the criteria of legal validity in the constitution, and it identifies political institutions and their respective legal powers. The content of the rule of recognition is the ‘public, common standard of correct judicial’ that most officials, in all three branches of government, in fact accept.2 Evidence of what most officials accept is taken from ‘… [their] actual practice: [] the way in which courts identify what is to count as law, and [] the general acceptance or acquiescence in these directions’.3 Some aspects of this rule (and rules



2 HLA 3 ibid

Hart, The Concept of Law 2nd edn (Oxford, Clarendon Press, 1994) 116. 108.

Why Common Law Constitutionalism is Correct (If It Is)  133 of recognition in general) may be indeterminate.4 The rule of recognition may also change if, and only if, most officials accept such change. GO(2) Most officials in Britain presently accept, and have historically accepted, that Parliament has the power to make or unmake any law, and that courts do not have the power to strike down primary legislation. The precise rule of recognition in Britain is ‘What the Queen in Parliament enacts is law.’5 GO(3) The legal content of a statute is the legislative intention communicated through the text of that statute. Ascertaining these intentions may depend on clear rules or canons of statutory interpretation. Applying such intentions and rules is the primary adjudicative role of courts. If no clear parliamentary intention or rule is available, judges must fill the ‘gaps’ in the law using extra-legal discretion. Judges must use their discretion to enable statutes ‘to achieve their purposes without damaging the background principles that Parliament is committed to’.6 GO(4) The rule of recognition ‘What the Queen in Parliament enacts is law’ may be morally defensible, but a rule of recognition may exist even if most officials think that it is an unjust rule, or that there should be some better rule. If a particular law is unjust, but legally valid under the prevailing rule of recognition, then judges and citizens may have a moral duty to disobey it.7 B.  Allan’s Common Law Constitutionalism (CLC) CLC(1): [T]he content of the law … is a product of normative judgment in which we attempt to make good moral sense of an array of such familiar legal ‘sources’ as Acts of Parliament, judicial precedent and influential dicta. An account of English law on any specific subject is always a theory of how best to read the relevant legal materials, guided by notions of justice and coherence: we assume that law, correctly interpreted, should as far as possible serve the interests of justice, rather than injustice, and be broadly coherent rather than confused and contradictory. And this is true even when we disagree about what justice requires, or about what would make the law more coherent overall.8

CLC(2) Parliament does not possess absolute, sovereign legislative power. Legislative supremacy ‘may [only] operate within the constitutional framework of the rule of law’.9 ‘Parliament’s authority is confined by the limits of our ability (in any concrete context) to interpret its enactments as contributions to the public good.’10 CLC(3) Statutes do not mean what Parliament intended, in the sense of communicating a ‘speaker’s meaning’; the interpretation of a statute instead requires us to

4 ibid. See, generally, chs VI and VII. 5 ibid 102. 6 Goldsworthy, Contemporary Debates (n 1) 8 and, generally, ch 9. See further Goldsworthy’s discussion of the ‘principle of legality’ in this chapter, section III. 7 Goldsworthy, History and Philosophy (n 1) 18–19. 8 Allan, The Sovereignty of Law (n 1) 5. 9 ibid 133. 10 ibid 12.

134  Stuart Lakin construct the intent of the ‘ideal or representative legislator’ who seeks to reconcile ‘current policy and overarching legal principle’.11 There is then no conflict between parliamentary supremacy and the rule of law. These ideas are interdependent, embodying the twin imperatives of democracy and respect for individual dignity and autonomy.12 CLC(4) Given that the content of the law depends on notions of justice and coherence, the laws that obtain in a constitution are, perhaps with rare exceptions, necessarily legitimate.13 A statute is only recognisable as such if it can be read in a way that is compatible with the principle of equal citizenship.14 C.  Are GO and CLC Commensurable? Having laid out Goldsworthy’s and Allan’s account of the Constitution, let us restate our motivating question: how can we establish which of these accounts (if either) is correct? I want to begin by confronting a pre-emptive challenge to that question. In so doing, I hope to lay the groundwork for the rest of the chapter. The objection is that GO and CLC are incommensurable in that they pursue different aims and sit within different intellectual disciplines. As such, the objection runs, it makes no sense to ask which account is correct: correctness means entirely different things from one account to the other. This objection takes many different forms and runs in both directions. Supporters of GO sometimes contend that they are describing the Constitution ‘as it is in practice’, ‘in reality’, ‘on the ground’, and other such expressions and sentiments.15 Those who support CLC, they say, are theorising abstractly about how law and a constitution ought, ideally to function.16 What proponents of CLC present as an account of the Constitution, say GOists, is, in truth, ‘revisionist’, ‘dogmatic liberalism’ reflecting the moral convictions of the (purported) interpreter.17 From the other direction, CLCists sometimes argue that the interpretative method reflects how judges and lawyers truly reason about law and the Constitution.18 Those who support GO, they argue, adopt

11 ibid 193–4. 12 ibid 168 and, generally, ch 5. 13 Whether law or a law can be unjust or morally sub-optimal is a matter of intense debate among interpretivists. Compare, for instance, M Greenberg, ‘The Moral Impact Theory of Law’ (2014) 123 Yale Law Journal 1288–1342, and TRS Allan ‘Law as a Branch of Morality: The Unity of Practice and Principle’ (2020) 65(1) The American Journal of Jurisprudence 1–17. 14 Allan, The Sovereignty of Law (n 1) 33 and, generally, ch 4. 15 The seminal argument that there is a gap between CLC and constitutional reality is found in J Griffith, ‘The Political Constitution’ (1979) 42 Modern Law Review 1. See further, M Gordon, Parliamentary Sovereignty in the UK Constitution: Process, Politics and Democracy (Oxford, Hart Publishing, 2015), 30; D Oliver, ‘Parliament and the Courts: A Pragmatic (or Principled) Defence of the Sovereignty of Parliament’ in A Horne, G Drewry and D Oliver (eds), Parliament and the Law (Oxford, Hart Publishing, 2013), 309. 16 See, for instance, J Griffith, ‘The Brave New World of Sir John Laws’, (2000) 63 Modern Law Review 159. 17 T Poole, ‘Dogmatic Liberalism? T.R.S Allan and the Common Law Constitution’ (2000) 65 Modern Law Review 463; J Raz, ‘Dworkin: A New Link in the Chain’ (1986) 74 California Law Review 1103. 18 Allan, The Sovereignty of Law (n 1) 9, 22.

Why Common Law Constitutionalism is Correct (If It Is)  135 the detached perspective of a sociologist surveying the beliefs of officials and other constitutional actors, rather than ‘internal perspective’ of a c­onstitutional lawyer working out the content of the law and constitution.19 What GOists present as an account of the Constitution, say CLCists, is in truth no more than the m ­ isapplication of abstract, positivist legal theories.20 The conclusion of objectors on both sides is that their propositions and no others are apt for explicating the content of the contemporary British Constitution. In my view, this objection from incommensurability is flawed and unconstructive. It looks past common assumptions and concerns that both accounts very plausibly share, and which open the way for genuine disagreement and competition.21 The propositions in GO(1)–(4) and CLC(1)–(4), I suggest, plausibly address the same type of questions in relation to the same objects of explanation. As if to respond to Barber’s challenge, these questions give us ‘some sense of what counts as a successful exercise in constitutional theory, a set criteria against which accounts can be tested [so that we can] judge the merits of any particular piece of work’.22 First, they each address the constitutive question: how do the decisions and practices of institutions and officials in Britain impact on the legal and constitutional rights, duties, powers etc?23 How, for instance, are the enactments of Parliament, the judgments of courts, and – what are usually referred to as – constitutional conventions relevant to the content of the law and constitution? Locked within propositions GO(1)–(4) and CLC(1)–(4) are two different sets of answers to those questions. These propositions ascribe law-making or constitution-making significance to different aspects of these same practices and decisions. As I shall explain in detail below, for GOists the content of the law and the Constitution depends on empirically determinable aspects of those practices and decisions. For CLCists, this content depends on a combination of the empirical and moral aspects of the practice. Second, each account responds to – or at least lays the ground for – the doctrinal question: which rights, duties, powers, etc do obtain within the British Constitution? For instance, GO(2) and CLC(2) advance contrasting propositions about the powers of Parliament vis-a-vis courts in the British Constitution. A fuller statement of each account might include any number of doctrinal claims about the powers and duties of institutions and the rights of individuals in Britain. Importantly, the answer to any doctrinal question about the content of the law or the Constitution depends on one’s answer to the constitutive question. To put this same point more fully, the correctness of any doctrinal proposition of law or constitutional practice must depend on some constitutive

19 ibid 32. 20 ibid 38, 156–157. 21 For the possibility of rapprochement between positivists and anti-positivists, see, for instance, D Kyritsis, Where Our Protection Lies: Separation of Powers and Constitutional Review (Oxford, Oxford University Press, 2017) 18–20; D Kyritsis, Shared Authority: Courts and Legislatures in Legal Theory (Oxford, Hart Publishing, 2014) 19–20; G Letsas, ‘The DNA of Conventions’ (2014) 33 Law and Philosophy 535. 22 N Barber, The Constitutional State (Oxford, Oxford University Press, 2010) 1. 23 What I describe in this chapter as ‘constitutive’ and ‘doctrinal’ questions correspond to Ronald Dworkin’s ‘doctrinal’ concept of law. See, R Dworkin, Justice in Robes (Cambridge, Mass, Harvard University Press,) 2006, chs 1 and 8.

136  Stuart Lakin account of what makes any doctrinal claim true or correct.24 It follows that incorrect constitutive claims cannot be the basis of true doctrinal claims. As I shall explain below, establishing what makes a constitutive claim (and the doctrinal claims that flow from it) correct or incorrect is the central aim of this chapter and the thrust of our motivating question. Let us now consider the contrasting relations between the constitutive and doctrinal claims in GO and CLC. This will help to build a picture of the precise points of disagreement between the two accounts. For Goldsworthy, the doctrinal proposition GO(2) is true, and the parallel doctrinal proposition CLC(2) is false, in virtue of the empirical and historical constitutive proposition (or propositions) GO(1). This is to say that the powers of Parliament and courts depend on an accurate description of what most officials accept and have historically accepted. The corollary of the GO(1) is GO(4) and the rejection of CLC(4).25 Since Parliament’s powers depend on the empirical fact of what most officials accept and have historically accepted, those powers do not depend on a justificatory moral theory. Thus, Parliament (or, for that matter, any other law-making body) could have absolute legal power, even if there were no moral justification for such power.26 For Allan, by contrast, the doctrinal proposition CLC(2) is true, and the parallel doctrinal proposition GO(2) is false in virtue of the interpretative constitutive proposition CLC(1). The powers of Parliament and the courts depend on the moral theory that best justifies their practices and decisions. For Allan, those powers depend, for instance, on substantive conceptions of democracy and the rule of law, and on deeper values of equal citizenship and freedom.27 As he sees it, Parliament does not possess the legal power to legislate contrary to these types of principles and values.28 Importantly, this limitation on parliamentary power arguably holds for CLCists even if it can be shown that most officials accept absolute, unlimited parliamentary power.29 Here as elsewhere, the different constitutive claims in GO and CLC generate divergent – and potentially radically divergent – doctrinal claims. The corollary of CLC(1) is CLC(4). Since the powers of institutions depend for CLCists on the moral justification for those powers, it follows that those powers – if not every exercise of those powers – are necessarily morally legitimate.30 To the extent that absolute parliamentary power is morally indefensible, Parliament does not legally possess such power. Next, we have in both GO and CLC a constitutive-doctrinal pairing in relation to the meaning of statutes. It will be clearest if we work with a concrete proposition of

24 See R Dworkin, Law’s Empire (London, Fontana, 1986), 109–110; N Stavropoulos, ‘Legal Interpretivism’, Stanford Encyclopedia of Philosophy, https://plato.stanford.edu/entries/law-interpretivist/. 25 This is the relationship that legal positivists describe as the Separation Thesis. See Hart, Concept of Law, (n 2) 293; HLA Hart, ‘Positivism and the Separation of Law and Morals’ (1958) 71 Harvard Law Review 593. 26 See, eg, Gordon, Parliamentary Sovereignty (n 15) 21. 27 Allan, The Sovereignty of Law (n 1) 89 and, generally, ch 3. 28 ibid 37. 29 See the discussion of history in section III below. 30 For intra-interpretivist debate on this point, see above n 13.

Why Common Law Constitutionalism is Correct (If It Is)  137 English law. Take the following proposition from the Privacy International decision: PI It is the law that courts may not review the exercise of power by the Investigatory Powers Tribunal.31 For Goldsworthy, whether PI is true or correct depends on the empirical constitutive proposition GO(3). PI is true, either: a) if it was Parliament’s clearly communicated intention in section 67(8) of the Investigatory Powers Act to oust the jurisdiction of the High Court; or b) if, in the absence of any clearly communicated parliamentary intention, judges created a new legal rule PI. Again, the corollary of GO(3) is the truth of GO(4) and the falsity of CLC(1) and CLC(3). The fact that the legal content of a valid statute depends, for GOists, on what the legislature has communicated entails that the meaning of that statute cannot depend on a justificatory moral theory.32 Thus, a legally valid statute may nonetheless be unjust or morally sub-optimal. In that case, judges and citizens may have a moral duty, or may otherwise decide, to disobey the statute, or to lie about its content and effect.33 For Allan, by contrast, whether PI is true depends on the interpretative constitutive proposition CLC(3) (which is a more concrete application of CLC(1)). PI is true if it reflects the intent of the ‘ideal or representative legislator’ who seeks to reconcile ‘current policy and overarching legal principle’; and it is false if it does not. For CLCists, the text of a statute has no legal meaning independent of the correct reconciliation of all relevant principles and policies.34 Thus, even where the ordinary language meaning of a text clearly supports one doctrinal claim, that may not be the correct legal meaning.35 The corollary of CLC(3) is again CLC(4). For interpretivists, law is a moral concept.36 It only makes sense to say that a community is governed by law if its legal and constitutional practice can be understood in a morally defensible way.37 Hence, if it were not possible to read PI in a way that honours the fundamental moral principles of the Constitution, then PI would not be law at all. Having set out the constitutive-doctrinal relations in GO and CLC, the incommensurability objector might feel even more confident in his views. He may comment that Goldsworthy and Allan advance radically different constitutive claims, which will likely generate opposing doctrinal claims. Goldsworthy’s empirical, constitutive claims GO(1) and GO(3) sit squarely within the legal positivist tradition. Allan’s interpretative, constitutive claim CLC(1) sits squarely within the anti-positivist, interpretivist tradition. Allan rejects GO(1) and GO(3); Goldsworthy rejects CLC(1). Allan’s doctrinal claim CLC(2) fails by the lights of GO(1). Goldsworthy’s doctrinal claim GO(2) fails by the lights of the CLC(1). From this untidy tangle, he might conclude that there is simply no basis on which to assess the two accounts against each other. Empirical and interpretative arguments, he might say, are apples and oranges.



31 R.

(Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22, [2019] 2 WLR 1219. J Raz, Ethics in the Public Domain (Oxford, Oxford University Press, 1994) ch 10. J Goldsworthy, ch 2 in this volume, section III. 34 Allan, The Sovereignty of Law (n 1) 173–174. 35 ibid 214–223. 36 See R Dworkin, Justice for Hedgehogs (Cambridge, MA, Harvard University Press, 2011) ch 19. 37 Dworkin, Law’s Empire (n 24) ch 6. 32 See 33 See

138  Stuart Lakin This conclusion again unhelpfully listens for dissonance rather than consonance between the two accounts. It misses the important sense in which the two accounts speak both to the moral and empirical parts of the Constitution. Both accounts assume in common, we may suppose, that British constitutional practice is characterised by a range of fundamental constitutional moral principles: parliamentary sovereignty, parliamentary accountability, democracy, the rule of law, the separation of powers, individual liberty and so forth. Similarly, both accounts must recognise the plethora of local, moral principles relevant to discrete areas of law and constitutional practice, for instance, the principles of judicial review or the law of negligence. At the same time, both accounts assume in common that British constitutional practice is characterised by numerous empirical facts: the text of statutes and judgments, the intentions of legislators, the beliefs of officials and citizens about what statutes and judgments mean, the beliefs of officials about their constitutional duties, and so on. What we see in our analysis of the contrasting constitutive claims in GO and CLC above, I suggest, is two rival models or configurations of the moral and empirical parts of British constitutional practice taken together.38 GOists tell us that the moral principles and values of the Constitution, correctly understood, do not determine the powers of institutions and the legal content of statutes, but provide a (contingent) justification for such content as obtains as a matter of empirical fact. CLCists put things the other way round: the moral principles and values of the Constitution, correctly understood, are among the determinants of the power of institutions and the meaning of statutes. The moral parts of the Constitution determine how and why particular empirical facts make law. The philosophical disagreement I have just described, I suggest, is busy at work beneath the surface of many familiar debates about the UK Constitution. It explains the (now quite mainstream) view that there is a tension between different visions of the Constitution.39 Take for example the role and meaning of the rule of law in Britain. For GOists, the rule of law is a formal concept comprising a set of formal and procedural standards.40 Its (limited) role within the Constitution is to enable lawmakers better to communicate their legislative intentions or judge-made rules in legislation and judgments.41 But rule of law standards, on this view, do not determine or condition the meaning of parliamentary intentions or common law rules;42 and

38 As Greenberg puts it a model is, ‘a counterpart at the metaphysical level of a method of interpretation at the epistemic level. (A model’s being correct in a given legal system is what makes the corresponding theory of interpretation true’. See M Greenberg, ‘How Facts Make Law’ in S Hershovitz (ed), Exploring Law’s Empire: The Jurisprudence of Ronald Dworkin (Oxford, Oxford University Press, 2006) 245. For – what I think is – a similar use of the device, see G Gee and GCN Webber ‘What is a Political Constitution?’ (2010) 20 OJLS 273, 290–291. Compare A Young’s classification of models in ch 3, this volume. 39 See, for instance, RB Taylor, ‘The Contested Constitution: An Analysis of the Competing Models of British Constitutionalism’ [2018] Public Law 500; A Latham-Gambi, ‘Political Constitutionalism and Legal Constitutionalism – an Imaginary Opposition?’ (2020) 40 Oxford Journal of Legal Studies 737. 40 See J Raz, The Authority of Law (Oxford, Clarendon Press, 1979) ch 11; R Bellamy, Political Constitutionalism: A Republican Defence of the Constitutionality of Democracy (Cambridge, Cambridge University Press, 2007) ch 2. 41 Raz, The Authority of Law (n 40) 224–226. 42 For a clear explanation of the positivist distinction between law and the rule of law, see J Gardner, ‘Legal Positivism: 5 ½ Myths’ (2001) 46 American Journal of Jurisprudence 199, 207–211.

Why Common Law Constitutionalism is Correct (If It Is)  139 there are occasions where law works better without heeding these standards (as for instance, when officials require discretionary powers). For CLCists, by contrast, the rule of law is the central value in the Constitution around which all other constitutional principles and values orbit.43 It insists that the state may only deploy coercive force in accordance with the scheme of moral principles underlying the past enactments of Parliament and the past decisions of courts.44 These principles determine and condition the content of the law; and they explain the general political obligation to obey the law.45 We might give similar contrasting stories about any of the constitutional principles mentioned above: whether, for instance, parliamentary sovereignty is an extra-legal convention at the apex of the Constitution as in GO(2), or shorthand for a set of legal principles that determine the distribution of powers between Parliament and courts CLC(2).46 Whether democracy is a majoritarian concept that justifies the fact of absolute legislative power,47 or a constitutionalist, rights-based concept that imposes a legal limit on legislative power.48 Whichever view one holds on the role and importance of these types of big constitutional principles, I suggest, belongs to a broader model of the relationship between the moral and empirical parts of constitutional practice. Perhaps the most visible manifestation of these competing models – if not necessarily the most constitutionally important – of the moral and empirical parts of British constitutional practice is case law. Whether, or the sense in which, one deems a judicial decision to be correct or incorrect will depend on which model one espouses. One could choose almost any case to illustrate this point. The Anisminic and Privacy International decisions are an obvious place to start.49 For GOists, the decision by the House of Lords in Anisminic can be explained in two possible ways. Either it was an ordinary case of statutory interpretation: the court gave effect to the empirically determinable background assumption that Parliament did not intend to remove all legal control from the Commission.50 Or it was an instance of extralegal judicial pragmatism and dissembling: judges pretended to be giving effect to Parliament’s intentions, when in truth, they were (say) making furtive moves to reform the Constitution in line with their moral preferences or other goals.51 Similarly, the

43 Dworkin, Justice in Robes (n 23) 168–186; Allan, The Sovereignty of Law (n 1) ch 3; S Lakin, ‘Debunking the Idea of Parliamentary Sovereignty: The Controlling Factor of Legality in the British Constitution’ (2008) 28 Oxford Journal of Legal Studies, 709, 728–734. 44 Dworkin, Law’s Empire (n 24) 227. 45 ibid 110. 46 For the latter view, see J McGarry, ‘The Principle of Parliamentary Sovereignty’ (2012) 32 Legal Studies 577. 47 Bellamy, Political Constitutionalism (n 40) ch 3; Gordon, Parliamentary Sovereignty (n 15) chs 7 and 8. 48 R Dworkin, Freedom’s Law: The Moral Reading of the American Constitution (Oxford, Oxford University Press, 1996); J Laws, ‘Law and Democracy’ [1995] Public Law 72. 49 Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147; R (Privacy International) (n 31). 50 Goldsworthy, Contemporary Debates (n 1) 286. 51 Goldsworthy, History and Philosophy (n 1) 252; HWR Wade and CF Forsyth, Administrative Law 7th edn (Oxford, Oxford University Press. 1994) 737. See further Goldsworthy’s discussion of judicial pragmatism, ch 2 in this volume, section III.

140  Stuart Lakin reasoning by the Supreme Court in PI about the paramountcy of the rule of law, says Goldsworthy, is ‘more consistent with legal pragmatism than common law constitutionalism’.52 Any suggestion in the reasoning of judges that the rule of law places a limit on parliamentary sovereignty was necessarily incorrect or obiter dicta. For CLCists, by contrast, both Anisminic and PI epitomise the operation of CLC(3). The different effects of the ouster clauses in the two cases resulted from the proper and consistent interpretation of relevant constitutional principles, policies and facts – for instance, democracy, parliamentary sovereignty and the rule of law.53 In their disagreements about the interplay between these different standards, judges were engaging in the moralised constitutional theory characteristic of CLC. D.  The Motivating Question I have attempted to show above that GOists and CLCists are not talking past each other from the perspectives of different disciplines. To the contrary, they are engaged in a common set of inquiries. They are each attempting to answer the constitutive question: how do the decisions and practices of institutions in Britain determine legal and constitutional rights, duties and powers? On the back of their answer to that question, they are trying to answer the doctrinal question: which rights, duties and powers do obtain. Their answers to these questions, I have suggested, represent two competing models of the moral and empirical parts of British constitutional practice. To view GO and CLC in this way focuses the mainstream idea that there is a tension between these accounts of the Constitution. It helps us to understand that, behind the familiar disagreements about the Constitution found in law journals, judgments and debates, is a baseline of agreement about the object and aims of that disagreement.54 This is the key, I suggest, to a genuine contest between the two models. That realisation only gets us so far, however. It is one thing freely to elaborate one’s favoured model of the Constitution, happily ensconced in one’s own thought world. It is another to defend that model against other models that reject its premises and details. This is the challenge that I have set up in our motivating question viz. how can we establish which account of a constitution is correct? We can now reformulate that question as follows: what makes it the case that either an empirical

52 Goldsworthy, ch 2 in this volume text accompanying n 143. I gratefully adopt Goldsworthy’s GOist analysis of the judgments in both PI and R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5. For different models of the Supreme Court decision in (R (Miller) v Prime Minister; Cherry and Advocate General for Scotland [2019] UKSC 41, [2020] AC 373 approximating to CLC and GO, see in this volume Arvind and Stirton, ch 4, section III.D, Young, ch III, section IV. 53 On Anisminic, see TRS Allan, ‘Legislative Supremacy and the Rule of Law: Democracy and Constitutionalism’ (1995) 111 Cambridge Law Journal 127; Allan, Sovereignty (n 1) 214–15. On Privacy International, see HJ Hooper, ‘Balancing Access to Justice and the Public Interest: Privacy International and Ouster Clauses in the Broader Constitutional Context’, U.K. Const. Law (12 Feb. 2018) (available at https://ukconstitutionallaw.org/) (accessed 17 July 2021). Hooper’s blog relates to the Court of Appeal reasoning, but it captures perfectly the CLC interplay between moral principles and statutory text. 54 Dworkin, Justice in Robes (n 23) especially 169.

Why Common Law Constitutionalism is Correct (If It Is)  141 model of British constitutional practice (and the doctrinal claims that flow from that model) or an interpretative model (and the doctrinal claims that flow from it) is correct? This is a challenge that anyone seeking to give the correct account of the Constitution – or indeed, the correct answer to any given question of law or the Constitution – must confront. To adapt Dworkin, the challenge is ‘… silent prologue to any decision at law’.55 Before I consider the details of possible methodological arguments in favour of GO or CLC, I want to lay down some aspirational parameters for this inquiry. This is in a determined bid to rescue the promise of genuine disagreement between GO and CLC from an impasse. First, and most importantly, the correct method must enable an objective and unprejudiced comparative appraisal of different models, such that either GO or CLC (or something else) could emerge as the correct model. It must not give a priori weight to one or another model. For instance, one cannot build into one’s method the claim that law and morality are separate and that, in consequence, the adjudicative role of courts is necessarily to give effect to the intentions of Parliament. This is a method in name only. In substance, it is something like the GO model of the Constitution masquerading as a method.56 Far from enabling a comparative appraisal between GO and CLC, it crowns GO the victor before the contest begins. At least one such objective measure of comparison, GOists and CLCists can agree, is the extent to which a model fits the practice.57 A model must be able to account for, for instance, most of the judicial decisions widely held to be correct in the Constitution, most of the statutes widely held to be valid, most of the constitutional conventions widely held to be in force, and most of the constitutional principles widely thought to characterise the Constitution. Importantly, it must also account for the fact that people disagree about how to understand these decisions and practices. We have seen above that disagreement is a fundamental feature of British constitutional practice, both on the part of academics analysing the practice, and on the part of practitioners (judges, lawyers, citizens, officials) participating in it. A model that either fails to account for this and other key features of the Constitution, or which explains features in a way that is entirely alien to the participants in the practice, will likely fail for these reasons. Indeed, it would hardly be a model of the practice at all. But satisfying the requirement of fit as a mere threshold requirement of plausibility will not suffice to decide between GO and CLC.58 As I have demonstrated above, both models plausibly fit the decisions and practices of the British Constitution (as, arguably, do other models). GOists and CLCists must, second, give some additional argument(s) as to why their model gives a superior understanding of British 55 Dworkin, Law’s Empire (n 24) 90. 56 For criticism to this effect of M Gordon’s ‘positivist and political’ constitutional method, see S Lakin, ‘The Manner and Form Theory of Parliamentary Sovereignty: A Nelson’s Eye View of the UK Constitution?’ (2018) 38 Oxford Journal of Legal Studies 168 (a review of Gordon, Parliamentary Sovereignty (n 15)). For Gordon’s reply, see his conclusion in ch 9, this volume. 57 See Dworkin, Law’s Empire (n 24) 255. For the meeting of minds between CLCist and GOists on this point, see Goldsworthy’s ch 2 in this volume notes and accompanying nn 124–128. 58 See Kyritsis, Shared Authority, (n 21) ch 3.

142  Stuart Lakin constitutional practice than other models. Again, I think we can safely rule out some familiar techniques. It will not do simply to keep hammering the details and implications of one’s favoured model, explaining how that model explains this or that case or feature of the Constitution, and how other explanations fail by the light of that model. Proponents of GO or CLC can do this endlessly. But this style of argumentby-bombardment cannot settle the deep disagreement between GOists and CLCists as to which types of facts, configured in which way, are constitutive of which doctrinal content. The resolution of this disagreement requires a metric by which to establish that empirical facts, moral facts, or some combination of the two determines content. That metric, I shall argue below, must be independent of the very facts of British constitutional practice of which GO and CLC are rival models. Finally, GOist and CLCists must adopt a suitably conciliatory attitude to the inquiry. If, as I have suggested, British constitutional practice can plausibly support both the GO and CLC models, then proponents of each model must be open to the correctness of the other model (and other models besides). I have described GO and CLC at various points in this chapter as rivals and competitors, but I have also tried to emphasise that both models are joined in a common endeavour: to understand the British Constitution correctly. Truth rather than victory is the aim. In that spirit, I suggest that GOists and CLCists must employ a principle of charitable interpretation, making each model as good as it can be. If there are apparent gaps or deficiencies in one or other model, then we should think about how, consistently with that model, they might be rectified.59 By extension, name-calling, caricaturing or rhetorical bigging-up of models cannot advance this inquiry. For instance, I have followed the trend in this chapter of referring to GO as the ‘orthodoxy’; but it is for GOists to establish whether, and in what sense, its orthodox status counts in its favour. Objective, methodological argument alone can resolve the complex tension between GO and CLC. III.  WHAT MAKES GO OR CLC CORRECT?

I now propose to examine what are arguably the dominant approaches to the motivating question: descriptive (positivist) and interpretative (anti-positivist). These are the more abstract methodological counterparts of the constitutive propositions GO(1) and (3) and CLC(1) and (3). To begin with, I shall explain and critique two descriptive methods found in the work of Goldsworthy. I shall argue that these methods display a philosophical flaw common to all descriptive methods. As such, they cannot support GO, and they cannot refute CLC. In section IV, I shall argue that the very flaws in descriptive methodology gesture towards the correctness of the interpretative method. In a bid to establish the correctness of GO(2) and (3), Goldsworthy advances two methodological arguments, a ‘philosophical’ one, and an historical one.60



59 That

charitable attitude, I shall argue in section IV, is built into the interpretative method. History and Philosophy (n 1) chs 1, 2 and 10.

60 Goldsworthy,

Why Common Law Constitutionalism is Correct (If It Is)  143 The philosophical argument targets the following three-step argument made by (some) CLCists: that there are only two sources of law in Britain, statute and common law; that the source of Parliament’s legislative powers cannot logically be statute; and that Britain must therefore have a common law Constitution.61 In response to these claims, Goldsworthy argues that the common law can no more be the source of Parliament’s powers than Parliament (or statute) can be the source of judges’ powers. In each case, he says, an explanation is needed, independent of common law and statute, as to how the supposed power-conferring institution came by its powers. As he puts it: ‘[t]he authority of either Parliament, or the judges, or both, must be based on laws that neither was solely responsible for creating …’.62 That explanation, he concludes, is GO(1): the powers of Parliament and judges depend on a ‘widespread consensus among senior officials in all branches of government’.63 Goldsworthy derives the premise of his historical argument from his ‘philosophical’ conclusion that GO(1) is correct. Given the philosophical truth of GO(1), he says, ‘a history of the development of parliamentary sovereignty must be a history of [the consensus among the most senior officials of all branches of government] …’.64 Applying that premise, he conducts a detailed historical survey of judgments and academic commentaries, encompassing the work of Locke, Blackstone, Coke, Dicey, and many others.65 His conclusion is that there is overwhelming evidence that for centuries, lawyers and officials have accepted the correctness (of GO(2) and GO(3)).66 Therefore GO(2) and GO(3) are correct. Notwithstanding the monumental research behind Goldsworthy’s philosophical and historical arguments, I am afraid that they share a defect with all other descriptive methods. To begin with the philosophical argument, Goldsworthy is right to say that common law cannot be the source of Parliament’s powers (and vice versa), but we need to account carefully for why this is so. The reason is this: judgments by courts and statutory texts enacted by Parliament, are among the facts and features of British constitutional practice that GOists and CLCists are seeking to explain. They are part of the object of explanation (explanandum) of which GO and CLC are rival constitutive models (explanans). It follows that neither common law nor statute can confer authority on the other. This would be to assume, question-beggingly, the law-making authority of the power-conferring source or institution.67 It would be to treat part of the explanandum as the explanans. The problem with GO(1) as an explanation for GO(2) and (3) is that it begs precisely the same question. It rests on the circular claim that one part of the explanandum (the beliefs of most officials) can 61 ibid 238–242. For this version of CLC see, for instance, Laws, ‘Law and Democracy’ (n 48). For Allan and other interpretivists, moral values rather than any positivist source determine legal and constitutional content: see Allan, The Sovereignty of Law (n 1) 133. Both versions, and every other, I contend, are subject to the explanatory burden I set out below. 62 Goldsworthy, History and Philosophy (n 1) 240. 63 ibid. 64 ibid 6. 65 ibid passim. 66 ibid 7. 67 I rely here on the arguments in Greenberg, ‘How Facts Make Law’ (n 38). See further, N Stavropoulos, ‘Why Principles?’ (unpublished) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1023758 (accessed 30 July 2021).

144  Stuart Lakin determine the relevance of other parts of the explanandum (the authority of statute or common law).68 But the fact that most officials hold certain beliefs about the powers of Parliament and courts GO(1) has no automatic relevance or irrelevance to the content of the law or the Constitution. Nor do any other Hartian dispositions – what people have said, thought, written, and so on. These are among the very facts whose putative or presumptive legal relevance GOists and CLCists are attempting to explain (or explain away). Precisely the same problem stings Goldsworthy’s historical argument. Indeed, the two arguments are barely distinguishable. This argument again seeks to use one putatively relevant part of British constitutional practice – the historical beliefs of officials – to explain the legal relevance of other putatively relevant parts of the practice – the powers of Parliament and the intentions of Parliament GO(2) ­ and (3). But this again fails to recognise that history, the beliefs of officials, the doctrine of parliamentary sovereignty, and the intentions of Parliament are facts or features of the practice to which GOists, CLCists and others assign different meanings and significance (or insignificance). Let me amplify this point. It is at least plausible that historical facts about British constitutional practice are relevant to the content of the Constitution today. In their common law reasoning and statutory interpretation, judges plausibly apply standards laid down in previous cases or practices.69 One might even argue that Parliament has some sort of duty to legislate in a way that coheres with past legislation.70 History is a phenomenon, we may say, which anyone seeking to give an account of the British Constitution must reasonably explain or explain away. But leading commentators and judges disagree about whether and how history is relevant. They give different models of history.71 Among the commentators who arguably denies the relevance of history to the content of the Constitution is John Griffith. For him, the content of the Constitution does not depend on any previously established ‘fact or principle’.72 It depends on the here and now of ‘what happens’.73 Judicial decisions, he contends, are forward-looking, pragmatic judgments of politics or policy, shaped by the particular social class, education and ideology of the judges.74 In a different vein, Herbert Hart, on whose work Goldsworthy relies, arguably makes the content of the rule of recognition depend on the present rather than historical beliefs of officials.75 For him, the fact that a particular rule of recognition may have endured for centuries does not bear upon what the rule is today. CLCists agree with Goldsworthy that history is relevant,

68 As Greenberg puts it, the mistake is to suppose that the facts that make up a constitutional practice can determine their own relevance to the law. Greenberg, ‘How Facts Make Law’ (n 38) IV and VI. 69 For contrasting models of history in legal reasoning, see S Hershovitz, ‘Integrity and Stare Decisis’ in S Hershovitz (ed) Exploring Law’s Empire (Oxford, Oxford University Press, 2006) ch 5. 70 See Dworkin, Law’s Empire (n 24) 217 on ‘legislative integrity’. For criticism, see Kyritsis, Shared Authority (n 21) ch 4. 71 Compare in this volume, for instance, M Gordon, ch 9 at section III.B.ii, Arvind and Stirton, ch 4 at section II.B, N Barber, ch 1, n 59. 72 Griffith, ‘The Political Constitution’ (n 15) 19. 73 ibid. 74 J Griffith, The Politics of the judiciary 5th edn (London, Fontana, 1997) chs 8–10. 75 For an excellent discussion, see A Tucker, ‘Uncertainty in the Rule of Recognition and in the Doctrine of Parliamentary Sovereignty’, (2011) 31 Oxford Journal of Legal Studies, 61, 68–70.

Why Common Law Constitutionalism is Correct (If It Is)  145 but they disagree about how it is relevant. These differences reflect the contrasting models of moral and empirical facts in GO and CLC. For Dworkin, history is relevant, not in terms of what most officials have accepted, and not ‘over all historical stages of a community’s law’ (as Goldsworthy’s own model seems to suggest),76 but in terms of the ‘range of standards the community now enforces’. Those standards are specifically, he argues, the scheme of moral principle underlying past political decisions.77 It follows that, for CLCists, the relative powers of Parliament and courts,78 or even the distinction between Parliament and courts,79 may not be today what they were in the past. Evidence that most officials and commentators have consistently said or believed that GO(2) and (3) are correct is therefore no ‘embarrassment’ to CLCists, as Goldsworthy claims;80 for this is not the relevant aspect of history within their model.81 As ever, supporters of each of these different models of whether and how history is relevant can point to aspects of British constitutional practice that support their view.82 Each view arguably fits the practice. The challenge for each participant is to provide a methodological argument as to why their model is correct and others incorrect. In so doing, we must emphasise, they cannot use one putatively relevant part of the practice to explain some other putatively relevant part. This injunction applies as much to CLCists and rights sceptics as it does to GOists. For example, CLCists cannot move from the fact that a particular scheme of moral principles underlies British constitutional practice, or the fact that judges and commentators say they are reasoning morally, to the conclusion that moral principles limit the legislative powers of Parliament.83 Sceptics cannot move from the facts that judges and lawyers disagree and reason morally about the law, to the conclusion that there are no genuine legal rights and duties, and that public law (or all law) is politics.84 In each case, we need an argument about the meaning and significance of the (putatively relevant) explanatory facts.

76 Dworkin, Law’s Empire (n 24) 225. 77 ibid 227. 78 Support for such a view can arguably be found in Lord Steyn’s statement in Jackson v Her Majesty’s Attorney General [2005] UKHL 56 that the courts might establish a new ‘hypothesis of constitutionalism’ in the event of attempts by Parliament to abolish judicial review. See further J Jowell, ‘Parliamentary Sovereignty Under the New Constitutional Hypothesis’ [2006] Public Law 562; Arvind and Stirton, ch 4 in this volume at section II.D. Naturally, GOists dismiss such dicta as ‘demonstrably [ie historically] false’. See Goldsworthy, ch 2 in this volume at n 141. 79 N Duxbury, Elements of Legislation (Cambridge, Cambridge University Press, 2012). 80 J Goldsworthy, ‘The Real Standard Picture, and How Facts Make it Law; a Response to Mark Greenberg’ (2019) 64 The American Journal of Jurisprudence 163, 7. 81 There is an obvious analogue here with disagreement between so-called originalist and evolutive interpretation. For an illuminating discussion, see G Letsas, A Theory of Interpretation of the European Convention on Human Rights (Oxford University Press, Oxford 2007) ch 3. 82 For a Dworkinian account of British constitutional history rivalling Goldsworthy’s, see P Craig, ‘Public Law, Political Theory and Legal Theory’ [2000] Public Law 211; P Craig, ‘Theory and Values in Public Law: A Response’ in P Craig and Richard Rawlings (eds), Law and Administration in Europe Essays in Honour of Carol Harlow (Oxford, Oxford University Press, 2003). 83 I explore whether Allan adopts this structure of argument (concluding that he does not) in S Lakin, ‘Defending and Contesting the Sovereignty of Law: The Public Lawyer as Interpretivist’ (2015) 78 Modern Law Review 549 (a review of Allan, The Sovereignty of Law (n 1)). 84 See J Waldron, ‘Did Dworkin Answer the Crits’ in Hershovitz (ed) (n 38) ch 7.

146  Stuart Lakin Having called into question Goldsworthy’s philosophical and historical arguments in favour of GO(2) and (3), we can now take a step towards a sounder methodological argument. In order to avoid the explanatory bootstrapping difficulty identified above, the explanation for the correct model of British constitutional practice must be independent of the facts and features of British constitutional practice of which GO and CLC are rival models.85 As Greenberg puts it, the method must provide a ‘rational relation’ between different models and different doctrinal claims: an explanation as to why this model rather than any one of the countless alternative models makes a particular doctrinal proposition true or false.86 This conclusion gives concrete form to some of the parameters that I laid down above: that a successful methodological argument must include some objective metric by which to assess rival models, and that it must not give a priori weight to any one model in advance of that assessment. Before I propose such an argument, I first want to consider briefly one further attempt to argue descriptively for the correctness of GO. In a recent article, Goldsworthy makes a long and intricate argument against Greenberg’s ‘moral impact’ theory of law – which, for our purposes, approximates to CLC.87 His aim is to prove that ‘Legislative Supremacy’ – GO(2) – and ‘Legislative Intentions’ – GO(3) – are ‘actual fundamental doctrines of constitutional law in Anglo–American systems’.88 His strategy is to respond to criticisms by CLCists that GO(2) and (3) cannot adequately explain (ie do not fit) three characteristic features of Anglo-American legal practice: the fact that judges and lawyers disagree about the meaning of statutes, the fact that judges characteristically rely on moral standards, and the fact that laws are binding.89 In response to this criticism, Goldsworthy offers detailed explanations for each of these features, consistently with GO(2) and (3). To give just one snapshot of this exchange, applying CLC(3), Greenberg contends that only moral values can adequately explain why judges apply a principle of narrow construction to criminal law statutes, but not civil statutes.90 Goldsworthy replies from the perspective of GO(3) that this principle of narrow construction was historically invented by judges acting pragmatically ‘to blunt the brutalism of 18th Century Criminal Law’, but today the principle is an instance of ‘supplementing interpretation, due to statutory interpretations being ambiguous or vague. In this situation, [judges] are free to rely on [extra–legal] normative considerations …’.91 Let us grant that GO(2) and (3), as elaborated by Goldsworthy, have the resources to explain the features of legal practice highlighted by Greenberg. Where does that take us in terms of assessing GO against CLC? Consider the following passage from early in Goldsworthy’s article: [CLCists] cannot argue that [GO(2) and (3)] are merely a contestable theory of what constitutes legal content in Anglo–American law, which should be replaced by [CLC].



85 Greenberg,

‘How Facts Make Law’ (n 38) 245–251. 223 and 249. ‘The Real Standard Picture’ (n 80). 88 ibid 1. 89 ibid passim. 90 ibid 25. 91 ibid (footnotes omitted). 86 ibid

87 Goldsworthy,

Why Common Law Constitutionalism is Correct (If It Is)  147 [They] must show that these doctrines … are not fundamental elements of that legal content, whose constitutive determination requires theoretical explanation.92

Goldsworthy here seems to lay down a presumption in favour of GO(2) and (3). Provided that GO(2) and (3) can explain the features highlighted by Greenberg, he implies, then GO(2) and (3) are indeed ‘fundamental elements of legal content’. CLCists will not have rebutted the presumption; and they will not have made the case for ‘replacing’ these fundamental doctrines. There is simply no basis for such a presumption.93 The back-and-forth between Goldsworthy and Greenberg instead illustrate in admirable detail a point that I have made repeatedly thorough this chapter: that both GO and CLC fit British legal and constitutional practice. In other words, GO and CLC are precisely what Goldsworthy says they are not: ‘contestable (and contested) theor[ies] [or models] of what determines the legal content of statutes’. It is therefore premature to speak of CLC replacing GO(2) and (3). We have yet to establish, by some objective, independent, metric which model is currently the correct one. To hold otherwise is to violate the parameter laid down above that a method must not give a priori weight to any model. IV.  GO AND CLC AS RIVAL INTERPRETATIONS OF BRITISH CONSTITUTIONAL PRACTICE

Moral values are our metric; and interpretivism, or what Kyritsis describes as ‘moralised constitutional theory’, is our method.94 Or so I shall now briefly argue. Space precludes any lengthy explanation or defence of this position.95 Rather, I want to make fourshort points in light of the earlier arguments of this chapter. First, I have said above that an argument in support of GO or CLC must be independent of the very facts and features of British constitutional practice of which GO and CLC are rival models. It must supply some objective standard that can explain why a doctrinal proposition depends for its correctness or incorrectness either on GO or CLC (or some other model). The interpretative method meets this challenge. Moral values can tell us, non-question-beggingly, why empirical facts, moral facts, or some combination of the two determines the content of the law.96 No further (non-moral) argument is needed to establish this relationship.97 On this view, GOist values such as certainty and stability make it the case that the powers of institutions should depend on a consensus among officials GO(1).98 Likewise, (a political

92 ibid 12. 93 See Lakin, ‘Debunking the Idea of Parliamentary Sovereignty’ (n 43); Lakin, ‘How to Make Sense of the HRA 1998: The Ises and Oughts of the British Constitution’ (2010) 399 (a review of A Young, Parliamentary Sovereignty and the Human Rights Act (Oxford, Hart Publishing, 2009). 94 I borrow this phrase from Kyritsis, Where Our Protection Lies (n 21) 7. 95 The leading account is Dworkin, Law’s Empire (n 24). 96 Greenberg, ‘How Facts Make Law’ (n 38) 254–264. 97 ibid. See further, R Dworkin, ‘Objectivity and Truth: You’d Better Believe It’ [1996] Philosophy and Public Affairs 87. 98 See Goldsworthy, ch 2 in this volume, section VI; Goldsworthy, ‘The Real Standard Picture’ (n 80) 47 (perhaps conceding, in line with the interpretative method, that GO must have an evaluative basis).

148  Stuart Lakin constitutionalist understanding of) values such as democracy and separation of powers,99 authority,100 and the ability to evaluate the law,101 make it the case that parliamentary intent makes a proposition of (statute) law true GO(3). For CLCists, values such as equal liberty,102 integrity103 or shared authority104 make it the case that the powers of institutions, and the correctness of propositions of law depend on the scheme of principle underlying the practice CLC(1) and (3). Moral values may not be the only candidate to remedy the bootstrapping problems of descriptive methods, but I think that they are the best candidate.105 This brings me to my second point. We have seen that GOists and CLCists share highly targeted and distinctive aims. Their specific object of explanation is the British Constitution – or at least Anglo-American constitutions – rather than constitutions or legal systems at all times and in all places. And the types of explanations they offer are constitutive and doctrinal: they each want to explain what makes propositions of law and the constitution true; and they each want to detail the true extant propositions. Among their shared parameters for success, we have seen, is whether a model fits the practice. More specifically, both Goldsworthy and Allan agree that the correct model must be the one to which a political community is committed. The interpretative method, I suggest, explains the importance of these common questions and concerns about a constitutional practice.106 They speak, say interpretivists, to two fundamental questions of political legitimacy. First, why do the decisions and practices of these institutions and officials impose binding legal obligations (in so far as they do)? Why, for instance, does a text created by a body of people sitting in a particular place change the normative position of individuals? Second, why, in circumstances of disagreement about justice, rights, and so on, do the members of a community owe their allegiance to such decisions (in so far as they do)?107 In other words, on what basis do

99 See Bellamy, Political Constitutionalism (n 40). 100 For a moralised view of Joseph Raz’s theory of law, see, for instance, Dworkin, Justice in Robes (n 23) 198–212 and 227–331; Kyritsis, Where Our Protection Lies (n 21) 18–20. 101 J Waldron, ‘Normative (or Ethical) Positivism’ in J Coleman (ed), Hart’s Postscript: Essays on the Postscript to the Concept of Law (Oxford, Oxford University Press, 2001), 430. For discussion, see M Gordon, ch 9 in this volume at section II.B. 102 Allan, The Sovereignty of Law (n 1) ch. 3. 103 Dworkin, Law’s Empire (n 24) chs 6 and 7. 104 Kyritsis, Shared Authority (n 21) passim. 105 The chief rival involves ‘indirectly evaluative’ analysis of which model or features are ‘central to our self-understandings in terms of law’. See J Dickson, ‘Methodology in Jurisprudence: A Critical Survey’ (2004) 10 Legal Theory 117–156, 124; Raz, Ethics in the Public Domain (n 32) 301; A Zanghellini, ‘A Conceptual Analysis of Conceptual Analysis’ (2017) 30(2) Canadian Journal of Law and Jurisprudence 467. This explanation arguably runs into the bootstrapping problem identified in section II: it takes empirical facts about the psychological views of participants to make other facts constitutive. See Greenberg, ‘How Facts Make Law’ (n 38) 255. For further criticism, see Barber, The Constitutional State (n 22) 8–11. 106 There are a bewildering number of theories to which people refer as interpretative (or in an alternate spelling, interpretive). The version I endorse is that elaborated by Dworkin, Law’s Empire (n 24), and Stavropoulos, ‘Legal Interpretivism’ (n 24). For a comparison between other variants, see M Moore ‘The Interpretive Turn in Modern Theory: A Turn for the Worse?’ (1989) 41(4) Stanford Law Review 871. Within this volume, M Gordon (ch 9) contrasts the versions of interpretivism espoused by Dworkin and Martin Loughlin at section X; and A Young (ch 3) discussed the version of interpretivism advanced by N Barber, (ch 1, n 22) at section X. 107 See Stavropoulos, ‘Legal Interpretivism’ (n 24); See the chapters in this volume by Kyritsis (ch 8) and Dyzenhaus (ch 7). In posing these questions I am not assuming any general accounts of bindingness or

Why Common Law Constitutionalism is Correct (If It Is)  149 the values and principles of the practice take priority over the personal moral convictions of individuals? It is my argument that we must assess a model of the British Constitution – or indeed any constitution – by how successfully it responds to these questions.108 The justification for each model will constrain the ability of that model to explain a constitutional practice.109 If so, then GOists and CLCists each face an arduous interpretative challenge. They must first show that their answers are defensible as a matter of political morality. They must then show, crucially, that their convictions in political morality are reflected in British constitutional practice. It is by joining these two dimensions of constitutional argument – justification and fit – that both GOists and CLCists address the practice as it is rather than as it ought to be. This brings me to my third point, and a belated attempt to pick a winner between GO and CLC. Goldsworthy has argued that even if GO cannot be descriptively correct, it is interpretively so. As he puts it, the ‘… interpretive methodology is construed … as an attempt to clarify and harmonise the principles actually accepted by British legal officials’.110 To ‘disconnect the behaviour from the reasons that actually motivate those who engage in it, and provide it with a justification that they would disown’ he says ‘is to propose a new practice rather than interpret theirs’.111 This is, I am afraid, an attempt to smuggle the descriptive positivist method critiqued above back into the arena. It again assumes that a practice can only be that which people think or believe it to be. But CLCists give a rival account of what it means for a practice to belong to a political community, one that has as much purchase within British constitutional practice as the GO account. For them, it is people’s normative reasons rather than their motivational reasons that make the practice their own.112 For interpretivists there can be no predetermined, factual, view on which of these accounts is correct – whether in favour of GO or CLC.113 Correctness depends on which of the complex stories in political morality supporting each account makes best (moral) sense of British constitutional practice. So which model of the practice is correct, CLC or GO? Let us return to some of the facts in dispute between Greenberg and Goldsworthy above: that judges characteristically disagree about the meaning of statutes; and that in their reasoning they characteristically draw upon moral principles. In order to vindicate their model of the practice, GOists must show that despite these features, judges nonetheless

political obligation. My point is that we should understand both GO and CLC as representing, or entailing, such an account. 108 Goldsworthy contends that (non-moral) standards of rationality could perform the same role, specifically the tacit or implied assumptions about a practice: Goldsworthy, ‘The Real Standard Picture’ (n 80) 38–41. Admittedly, rationality can help to rule out what Greenberg describes as ‘bent’ (ie bizarre) configurations of a practice (Greenberg, ‘How Facts Make Law’ (n 38) 248–251); but such standards are not equipped, in my view, settle deep disagreements of the sort between GOists and CLCists. 109 See, Dworkin, Law’s Empire (n 24) 234. For this ‘holistic’ (as opposed to ‘threshold’) account of fit and justification, see Kyritsis, Shared Authority (n 21) ch 3. 110 Goldsworthy, History and Philosophy, (n 1) 254. 111 ibid 253. 112 For discussion of the distinction between motivating and normative reasons, see Goldsworthy, ‘The Real Standard Picture’ (n 80) 37–39; Letsas, ‘The DNA of Conventions’ (n 21). 113 See Lakin, ‘Defending and Contesting the Sovereignty of Law’ (n 83); Stavropoulos, ‘Why Principles?’ (n 67).

150  Stuart Lakin generally apply common standards of legislative intent, such that the content of the law is certain, stable and non-dependent on the moral preferences of judges. In effect, GOists must explain away such features. This endeavour leads Goldsworthy and other GOists to the most ingenious broadening of the notion of intent. For instance, Goldsworthy tells us that legislative intent is under-determinate,114 that judges sometimes exercise unacknowledged interpretive creativity,115 that judges sometimes deliberately depart from the law for consequentialist reasons,116 and that the suggestions by judges that moral principles limit the powers of Parliament are all necessarily obiter dicta and wrong. To every suggestion by CLCists that moral principles partly determine the content of the law, GOists respond with some creative rendition of legislative intent.117 No doubt legal reasoning within some constitutional systems does, has, or could honour GOist values, but it is far from clear that the British Constitution does so today.118 GOist’s ‘bed of Procrustes’ treatment of legislative intent, and their accommodation of frequent extra-legal judicial excursions off-piste, are arguably ill-suited to securing the certainty, stability and judicial restraint that they prize. The further that the practice of legislation and adjudication moves from clear and ascertainable legislative intention, consistently applied by judges, the weaker that a model based on these types of values looks.119 Indeed, I think this gap gives us reason to look away from GO to CLC. For CLCists, moral disagreement about the role and powers of institutions and the meaning of statutes, is not an illusion, masking the truth of GO; it is a defining feature of contemporary legal and constitutional practice. This is not to ‘celebrate’ moral disagreement for disagreement’s sake,120 but to recognise that an interpretative justification for the practice must embrace this constitutional phenomenon rather than apologise for it.121 In its ‘protestant’ model of legal and constitutional reasoning, CLC does just that.122 Rather than dismiss as constitutionally unwarranted and legally irrelevant the moral disagreements of judges in cases such as Privacy International, CLCists view such disagreements as a clear manifestation of the values underlying constitutional practice. They are, says, Allan, part of an ongoing attempt to ‘integrat[e] shared basic values in a larger scheme, able to reconcile divergent conceptions of legality, equality, and justice’.123 It is these values that plausibly explain both the bindingness of law and the (presumptive) allegiance

114 Goldsworthy, ‘The Real Standard Picture’ (n 80), 15–20. 115 ibid 21–24. 116 ibid 20. 117 For an exhaustive study on legislative intent, see R Ekins, The Nature of Legislative Intent (Oxford, Oxford University Press, 2012). 118 On the relevance of Dicey’s theory in different periods in British constitutional history, see Craig, ‘Theory and Values in Public Law’ (n 82). 119 Dworkin, Law’s Empire (n 24) 139–150. 120 See Goldsworthy, ch 2 in this volume text to n 134. 121 For an interesting political constitutionalist explanation for disagreement, see Gee and Webber, ‘What is a Political Constitution?’ (n 38) 277. 122 For the protestant approach to legal argument, see Dworkin, Justice in Robes (n 23) ch, 6; Allan, The Sovereignty of Law (n 1) 342–346; Gerald J Postema, ‘“Protestant” Interpretation and Social Practices’ (1987) 6 Law & Philosophy 283–319. 123 Allan, The Sovereignty of Law (n 1) 345.

Why Common Law Constitutionalism is Correct (If It Is)  151 owed by members of the British constitutional community to the law. GOists cannot shoot down this conclusion on the basis that it does not accord with their model, or that it is not the practice that most people or officials accept. The practice is what the best interpretation makes it. The only response available to GOists is to give a better account of disagreement and other key features of the Constitution, one that is consistent with the values underlying their models. This brings me to my fourth and final point. I have focused in this chapter on two illustrative, broad-brush accounts of the Constitution, GO and CLC. While this focus has been sufficient, I hope, to make the case against descriptive constitutional methods and in favour of interpretivism, it leaves much room for debate about the correct interpretative model of the Constitution. Positivist and interpretivist theorists disagree as much among themselves as they do with each other about which combination of values and facts generate which legal and constitutional doctrines. More generally, constitutional commentators disagree about the relative importance of legal and constitutional rights, duties, powers etc as against ‘the experience, impulses, the practical, detailed ongoing of politics’.124 I leave open the possibility that the correct interpretation of the Constitution may emerge from either the positivist or interpretivist camp, or indeed from some other camp. I also freely acknowledge that a model of constitutional practice must deal with the ‘messiness’ and complexity of practical politics as much as more orderly questions of normative constitutional content.125 What I have attempted to argue in this chapter is that there is no descriptive method available to settle these types of disagreements. It is only by understanding which values underpin the constitution that we can do so. There is a broader closing point to make here about the virtue of the interpretative method both as a way of understanding a constitutional practice, and as a way of conducting scholarly disagreements. Properly applied, this method honours the choices made by a political community, through its institutions practices and decisions, about how to organise its constitution and legal system.126 Hence, it is entirely conceivable for interpretivists that any morally defensible model of a practice could provide the correct understanding of the Constitution today, whether GO, CLC, some combination of the two,127 or something else. In this respect, the method satisfies the

124 See G Gee and G N C Webber, ‘Rationalism in Public Law’, (2013) 76 Modern Law Review 708, 722. 125 Many authors in this volume have emphasised this point. The emphasis on law and courts in this chapter is for illustrative purposes only. 126 There is no cause, then, for Goldsworthy to be ‘baffled’ by the interpretivist’s approach to the Mongolian legal system (see Goldsworthy, ch 2 in this volume, section VI). The interpretivist, as much as the positivist, seeks to make sense of the practice as it is, rather than as he would wish it to be. See P Craig, Public Law and Democracy in the United Kingdom and the United States of America (Oxford University Press, Oxford, 1990) 3. For a fascinating exchange of views on precisely how values relate to a particular constitutional practice, see the chapters by P Cane and P Craig in Law and Administration in Europe Essays in Honour of Carol Harlow (Oxford, Oxford University Press 2003). For criticism of both views, see Allan, The Sovereignty of Law (n 1) 333–340. 127 For argument to this effect, see Lakin, ‘Defending and Contesting the Sovereignty of Law’ (n 83). For a model that combines insights from interpretivism and legal positivism, see D Kyritsis, ‘What is Good about Legal Conventionalism’ (2008) 14 Legal Theory 135.

152  Stuart Lakin first parameter I identified above: it enables an objective and unprejudiced comparative appraisal of different models. This has important implications, I think, for how constitutional scholars engage with each other as much how they engage with different constitutional practices. It does no injury to a scholar’s work to assess the political values to which they are committed, and to examine whether a constitutional practice instantiates those values. To the contrary, this constructive process makes every constitutional scholar a collaborator with every other. V. CONCLUSION

My overarching aim in this chapter has been to clear the way of a constitutional theory roadblock: the mistaken view that one or other account of a constitution can be descriptively correct. This widespread view is responsible, I think, for the polarised and uncompromising feel of much contemporary constitutional writing. It explains the popular view, criticised above, that accounts of the British Constitution such as GO and CLC are so different as to be incommensurable. Having removed that roadblock, I have attempted to show that we can understand GO and CLC as rival models of the salient facts and features (both empirical and moral) of British constitutional practice. The question of how to establish which model is the correct one (our motivating question), I have argued, is an interpretative rather than a descriptive one. We must ask which model provides the most attractive justification for the practice. With a briefness that, I fear, will disappoint many of the readers drawn to this chapter by its title, I have tentatively concluded that CLC gives the superior ­interpretation.128 Perhaps more frustratingly still for such readers, I have enthused about the methodological possibility of some model of the Constitution other than CLC being interpretatively correct, whether now or in the future.



128 For

a comprehensive philosophical and doctrinal defence, see Allan, The Sovereignty of Law (n 1).

6 Methodological Pluralism and Modern Administrative Law SARAH NASON1

I

n this chapter I argue that modern administrative law theory is characterised by its plurality of methods, and by moving beyond traditional subordination to top-down constitutional theory. New approaches to administrative law, and specifically to judicial review and administrative law adjudication, take a wider approach to values, and a more systematic look at the facts of practice, facts including case law, legislation, the characteristics of administrative procedures and institutions, and various data and statistics on the use and effectiveness of accountability mechanisms. This results in opportunities for scholars to better understand and respond to the diversity and complexity of public administration and administrative law adjudication, and to develop more accurate and multi-faced analyses. I argue that these new approaches, and the theoretical perspectives they generate, can transcend traditional methodological categories of descriptivism, normativism, historical and doctrinal analysis. They also draw on new techniques such as organic and anatomical analysis, content analysis, machine learning, and computer visualisation. Understandably this presents challenges, including the need for adequate methods training, clear research design and ensuring sufficient reliability and validity in findings; the time and space necessary for inter-disciplinarity; universities and funders willing to invest in more innovative and experimental methods; and the challenges of ‘speaking a different methodological language’ both in terms of conducting research, peer review processes and communication of results in an impactful way. I first explain that whilst new approaches have been developing, there is still some mainstream subordination of administrative law to top-down constitutional theory, and argue that this is reflected in the example of failures to properly analyse judicial review from a sufficiently nuanced administrative and empirical perspective.

1 I would like to thank the editors for their comments on previous drafts, and all participants at the initial workshop, The Methodology of Constitutional Theory, School of Law, University of Reading, April 2017.

154  Sarah Nason I argue for an approach of constructive interpretation, reaching back to foundational principles about how we come to knowledge on any subject and recognising administrative law as reflexively socially constructed. I explain why I consider that most scholars take an interpretive approach to administrative law and present some new directions in doctrinal methods. This is followed by discussion of a perennial problem for administrative law, namely, whether there are such things as general principles of administrative law (usually seen as deriving from constitutional principles) or rather a range of administrative laws deriving from function or context (such as in housing, planning, health and education). I conclude with a brief discussion of the potential of computer science methods, and a summation of the opportunities and challenges provided by the examples of plurality surveyed in this chapter. I.  SUBORDINATING ADMINISTRATIVE LAW TO CONSTITUTIONAL LAW

In a 2019 workshop on New Directions in Administrative Law Theory the convenors noted: ‘Administrative law theory traditionally takes a back seat to constitutional law theory.’2 This issue, and the challenges of deploying various methods, are highlighted by contemporary events, including an Independent Review of Administrative Law (IRAL) which reported in 2021,3 and the UK Government’s response to it.4 Since the modern judicial review procedure was developed, reforms have sought to improve its efficiency, including interventions to address the increasing numbers of immigration and asylum claims which, at the time of writing, make up about 90 per cent of the combined judicial review caseloads of the Administrative Court and Upper Tribunal. These procedural and case-management reforms stemmed from Law Commission Reports,5 and a Review of the Crown Office List6 (as the Administrative Court specialist caseload was previously named); none were as contentious as reforms proposed from 2012 onwards by Conservative and Conservative-Liberal Democrat Coalition governments.7 The proposals developed from 2012 to 2015, were ostensibly based on concern that the growing judicial review caseload was part of a bundle of unnecessary ‘red tape’ draining public finances, damaging prospects for economic recovery, and stifling innovation. A growing body of empirical research work has begun to bust these various ‘myths’ about judicial review. In particular, research by Varda Bondy, Lucinda Platt 2 T Adams, H Dindjer and A Perry, ‘Showcase: New Directions in Administrative Law Theory’ International Journal of Constitutional Law Blog, 8 Sept. 2019, at www.iconnectblog.com/2019/09/ showcase-new-directions-in-administrative-law-theory. 3 Independent Review of Administrative Law Report (CP 407, March 2021) (‘IRAL Report’). 4 Ministry of Justice, The Government’s Response to the Independent Review of Administrative Law (CP 408, March 2021) (‘JR Reform’). 5 See, eg, Law Commission, Report on Administrative Law: Judicial Review and Statutory Appeals (1994). 6 Sir Jeffrey Bowman, Review of the Crown Office List (Lord Chancellor’s Department 2000). 7 Ministry of Justice, Judicial Review: Proposals for Reform (CP25/2012); Judicial Review: Proposals for Further Reform (Cm 8703/2013); Reform of Judicial Review: Proposals for the Provision and Use of Financial Information (Cm 9303/2015).

Methodological Pluralism and Modern Administrative Law  155 and Maurice Sunkin, ‘busts’ three particular claims as ‘at best misleading and at worst false’: First, that the past growth in the use of JR has been largely driven by claimants abusing the system, either deliberately or otherwise. Second, that the effect of JR on public administration is largely negative because JR makes it more difficult for public bodies to deliver public services efficiently. Third, that JR litigation tends to be an expensive and time consuming detour concerned with technical matters of procedure that rarely alters decisions of public bodies.8

In my own work, Reconstructing Judicial Review (RJR), I argued that the reforms could be seen ‘at least in part, as a politically motivated response to the apparent reformation or constitutionalisation of administrative law and judicial review’.9 The ‘reformation’ account can include the following elements: that there has been an evolution of the UK Constitution from one marked by parliamentary sovereignty, to a more balanced system based on constitutional values, including values of liberalism and democracy, and the twin or bipolar sovereignties of Parliament and the courts. I also suggested that this constitutionalisation sat alongside perceptions of a growing culture of justification under which public power must be exercised in accordance with some coherently reasoned justification. The culture being one of open governance and the centrality of rule of law values including clarity, consistency, and transparency. This, I proposed, was also seen as aligned with the apparent growth of rights-based reasoning and associated concepts of proportionality and balancing across the public law of England and Wales, including in the practice of judicial review. I argued that in terms of perceptions of empirical change, the reformation, constitutionalisation and culture of justification included an image of judicial review expanding in terms of caseloads, substantive legal doctrine and practical impact; for most reformationists this was seen as a good thing too. This perception of positive expansion was countered by those who argued that since the total number of applications for judicial review in the Administrative Court had increased alongside anecdotes about tactical litigation and ‘left-wing pressure group’ activity, and since the majority of claims do not result in a substantive ‘win’ for the claimant at final hearing, there must have been a large-scale abuse of the system. Both these accounts may well be examples of what Paul Craig has referred to as ‘path dependency’.10 A concept from political science indicating that policy choice may be constrained by history, and that intellectual thought can be equally so constrained such that we think about issues ‘along a well-trodden intellectual track’.11 In this volume, Arvind and Stirton explore some of the consequences of constitutional theorising that relies on rhetoric, whilst presenting itself as fact, noting how this perpetuates division and disagreement and renders it difficult to evaluate whether particular theories, that have become styles of public law thought, ought to be accepted by a polity.12 8 V Bondy, L Platt, and M Sunkin, The Value and Effects of Judicial Review: The Nature of Claims, their Outcomes and Consequences (PLP, University of Essex and London School of Economics, 2015). 9 S Nason, Reconstructing Judicial Review (Oxford, Hart Publishing, 2016). 10 P Craig, ‘Judicial Review, Methodology and Reform’ forthcoming Public Law January 2022. 11 ibid. 12 Arvind and Stirton, ch 4 in this volume.

156  Sarah Nason As far as judicial review is concerned Craig concludes that this focus on rhetorical well-trodden paths leads to; ignoring contrary evidence; disregarding the normal incidence of judicial review; disregarding circumstances where judicial review can favour the legislature or the executive; and failing to consider that judicial review may actually be too limited in certain circumstances.13 These are largely the points I raised in RJR. I concluded that whilst there is some evidence of a slightly larger proportion of judicial review claims over the years being more ostensibly based on direct judicial engagement with constitutional values and rights, the empirical case for the ‘reformation’ had not been made out, and its influence had been much exaggerated. My refutation of the reformation was based in part on findings from a methodology seeking to draw closer reflexive connections between the products of empirical research, commonly occurring doctrinal arguments, and the range of values associated with the functions performed by judicial review. The epilogue to the 2012–2015 battle over judicial review was the enactment of various reforms including reduced time limits for bringing planning and procurement claims; new fees for oral renewals of permission and removal of the right to oral renewal for claims classed as ‘Totally Without Merit’ on paper permission; new rules for financial disclosure, interveners and costs; and the introduction of Costs Capping Orders. These reforms were designed to deter claimants, and they were implemented alongside restrictions on access to legal aid funding. At the time the reforms were proposed, I surmised that whilst they were ostensibly based on an assumption that judicial review was a financial drain and economic inhibitor, under the surface was adherence to a constitutional theory advocating restriction on the reach of judicial review. In effect, I saw the battle lines over judicial reform reflecting, not empirically established impacts, but the poles of constitutional theory; namely and respectively, legal constitutionalism and political constitutionalism. Political constitutionalism can be associated with a form of legal positivism that values majoritarian conceptions of democracy, including the view that arguments we generally express in terms of rights or principles are ‘the very questions which divide not unify opinion’.14 Political constitutionalists contend that disagreement must largely be resolved by political means, that rights are largely carved out from politics, and, on some accounts, that freedom exists in non-domination, such that no institution, most certainly not the courts and judiciary, should be above control by the body politic. Political constitutionalists have different views on the purpose and value of judicial review, but many, especially those aligned to right wing think tank Policy Exchange and its Judicial Power Project, consider that judicial review has expanded too far into the domain of politics and policy. This is illustrated by Professor Richard Ekin’s submission to IRAL, in which he argues that there has been an ‘inflation’ of judicial power, influenced in part by human rights law and ‘European integration’, but which he suggests also arises in the context of ‘ordinary’ judicial



13 Craig, 14 JAG

‘Methodology and Reform’ (n 10). Griffith, ‘The Political Constitution’ (1979) 42(1) Modern Law Review 16, 20.

Methodological Pluralism and Modern Administrative Law  157 review, though he seems to admit his only evidence is ‘a number of high-profile cases decided between 2015 and 2020’.15 Whatever the constitutional values-based case for the Judicial Power Project’s arguments, the methodological approach is weak, and tells us little about the actual practice of ‘ordinary’ judicial review. Cherry-picking a number of high-profile cases (usually those decided on appeal) is no evidence of any wide scale ‘inflation of judicial power’. Paul Craig has developed a sustained normative, and to an extent empirical, critique of the Judicial Power Project.16 More recently he has also highlighted the lack of empirical evidence behind claims of judicial overreach. As he states: ‘There is indeed an inverse relationship between the force of such allegations, and the empirical evidence that underpins them.’17 Craig proposes that these allegations are beset by historical methodological errors that he characterises as ‘mining and lumping’. These combined approaches seek to sustain a proposition by searching for supporting evidence, aggregating favourable sources, and ignoring contrary evidence. Though Craig himself adopts a range of more nuanced methodologies, those adopting legal constitutionalism, an interpretation he supports, are not immune to mining and lumping tendencies. Characteristically legal constitutionalism tends to lean closer to non or anti-positivist accounts of law, where legal rights are more than political claims; they are in some sense antecedent to the constitutional order, part of the fabric of constitutional principles that may be prior to the social facts of majoritarian politics and literal interpretations of legislation. For most legal constitutionalists, democracy is underscored by equality understood as a thicker conception than strict majoritarianism, and the rule of law as associated not only with fair procedures, but also with the specific value of freedom; legal constitutionalism is often then also associated with some form of liberalism. Legal constitutionalists have more confidence in the wisdom of the common law, and its claim to constitute and reflect community moral and political values over time. On the whole, this leads to support for a wider account of the legitimate reach of judicial review, up to and including the value, and doctrine, of legality, that Parliament must be taken to legislate in conformity with common law constitutional values and rights, unless it explicitly states otherwise. To write on the perceived battle between legal and political constitutionalism some five years after publishing RJR is to experience a certain sense of de ja vu. In 2019, the Conservative and Unionist Party proposed to further reform judicial review, this time as part of a manifesto commitment to constitutional re-balancing; recognising that judicial review is important to protect the rights of the individuals against an overbearing state, but also suggesting that it has been ‘abused to conduct politics by another means or to create needless delays’.18 This is a more overt

15 Policy Exchange, Judicial Power Project submission to IRAL, https://policyexchange.org.uk/ wp-content/uploads/The-Case-for-Reforming-Judicial-Review.pdf. 16 P Craig, Judicial Power, the Judicial Power Project and the UK, Oxford Legal Studies Research Paper No. 68/2017, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3080385. 17 Craig, ‘Methodology and Reform’ (n 10). 18 Conservative Manifesto December 2019, https://assets-global.website-files.com/5da42e2cae7ebd3f8bd e353c/5dda924905da587992a064ba_Conservative%202019%20Manifesto.pdf.

158  Sarah Nason expression of support for political constitutionalism than in previous government proposals. The Conservative Government them launched IRAL in July 2020. Its Terms of Reference (ToR) included that it should: focus on judicial review’s impacts on central government; bear in mind the balance between legitimate individual interests in challenging the lawfulness of executive action and effective executive governance under the law; and consider data and evidence on the development of judicial review and judicial decision-making.19 Dissection of IRAL’s report and the evidence submitted to it, is beyond the scope of this chapter. From a methodological perspective, the whole process was heavily criticised across the evidence submissions. Many commented on the breadth of the ToR and the consequent inadequacy of the timeframe for addressing them, others commented on the panel’s lack of sufficient expertise, and some questioned the independence of the Panel’s Chair, Lord Faulks.20 Katie Boyle and Diana Camps in particular criticised the Panel’s Call for Evidence, arguing that ‘the methodology lacks rigour and risks undermining the reliability and validity of data generated’.21 The Panel itself was aware of the limitations of its project, noting for example, ‘we would not claim or wish to be thought of as having undertaken the “comprehensive assessment” of judicial review or the “review of the machinery of judicial review generally” for which the terms of reference ask’.22 The general response to IRAL has been that its Report was measured and cautious. IRAL’s conclusion was that ‘the government and parliament can be confident that the courts will respect institutional boundaries in exercising their inherent powers to review the legality of government action’.23 In a response that left many, including Lord Faulks himself, scratching their heads,24 the government stated: ‘The panel found courts were increasingly considering the merits of government decisions themselves, instead of how those decisions were made – moving beyond the remit of judicial review.’25 The government subsequently proposed additional reforms to restrict judicial review, far beyond IRAL’s recommendations, giving stakeholders just six weeks to respond to further consultation. It is hard not to see this as an extreme swing to political constitutionalism, and such is reflected in the subsequent Judicial Review and Courts Bill introduced to Parliament on 21 July 2021. The government consultation, and proposed legislation, expressly supports the position

19 IRAL, ToR online at, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/ attachment_data/file/915624/independent-review-admin-law-terms-of-reference.pdf. 20 See, eg, M Zander, ‘Heads I win, tails you lose? Michael Zander on the Faulks Review: will it end as a government stitch-up?’ New Law Journal, 5 February 2021. 21 K Boyle and D Camps, ‘(Lacking in) Methodological Rigour, Human Rights and Devolution: IRAL’s challenge is one of process as well as substance’ (UKAJI Blog, 16 November 2020), https://ukaji. org/2020/11/16/lacking-in-methodological-rigour-human-rights-and-devolution-irals-challenge-is-one-ofprocess-as-well-as-substance/. 22 IRAL Report (n 3) 8. 23 ibid, 132. 24 Lord Faulks to Joshua Rozenberg during ‘Law in Action’ BBC Radio 4, 25 March 2021, www.bbc. co.uk/programmes/m000td1g. 25 Ministry of Justice, Judicial Review Reform: The Government Response to the Independent Review of Administrative Law (CP408/2021) para 2: https://consult.justice.gov.uk/judicial-review-reform/judicialreview-proposals-for-reform/supporting_documents/judicialreviewreformconsultationdocument.pdf-1.

Methodological Pluralism and Modern Administrative Law  159 of the Judicial Power Project, and of other scholars who consider that judicial review should not at base be about rights and constitutional principles.26 It does so without considering the nuanced evidence-base on administrative law and judicial review in England and Wales that is being developed by the use of a plurality of methods. These methods are generating increased light, and much less heat, than the polarised political versus legal constitutionalism debate which has characterised, indeed caricatured, the topic. II.  NEW METHODS OF ADMINISTRATIVE LAW THEORY

In 2019, Adams, Perry and Dinjder noted: In the past several years, administrative law theory has become a more active area, with new scholars taking the subject in new directions. The best of this recent work is rigorous, analytical, interdisciplinary, and deeply engaged with legal doctrine.27

Much of this new interdisciplinary work seeks to draw deeper evidence-based connections between theory, doctrine and the practice of law and administration. This includes innovative, systematic means to understand the development of doctrine, and the use of computer science methods to support information retrieval, analysis, and presentation (including through computer-assisted visualisation). In what follows I explain some of these methods and their impacts. A.  Administrative Law and Socio-Legal Methods Much of the new administrative law scholarship includes empirical socio-legal elements. This research examines the nature and role of law in society, including law’s social effects, legal processes and legal services, considered in context of factors ranging across the economic and political, as well as the social. Professor Simon Halliday suggests that empirical legal research methods have contributed to three particular themes in constitutional and administrative law: improving knowledge about the structure of constitutional arrangements; revealing the way public bodies actually operate (which also offers insights into the practical manifestation of constitutional values over time); and examining the use and effectiveness of accountability mechanisms.28 Whilst general observation is already commonplace in constitutional and administrative law, I argue that theorists and textbook writers should take closer account of the findings of empirical socio-legal research. For example, Halliday notes that public law textbooks are ‘inevitably premised on an empirical claim about the identity of

26 eg, JNE Varuhas, ‘The Principle of Legality’ [2020] Cambridge Law Journal 578. 27 New Directions in Administrative Law Theory (n 2). 28 S Halliday, ‘The Role of Empirical Legal Research in the Study of UK Public Law’ in C Hunter (ed), Integrating Socio-Legal Studies into the Law Curriculum (Basingstoke, Palgrave MacMillan 2012).

160  Sarah Nason major landmarks on the map’.29 I consider that more rigorous empirical research, and more attention to research already conducted, is needed to correct poor mapping of administrative law’s landmarks. Empirical research can include both quantitative and qualitative methods. For example, quantitative analysis of judicial review caseloads has been used to uncover trends in the incidence of judicial review, with researchers then seeking to explain the how and why of the trends identified by adding qualitative methods including surveys and interviews.30 What is important, as Boyle and Camps note is that: ‘One must be able to demonstrate that a selected research method would produce similar results if the study were replicated under similar conditions, and that the results measure what they are supposed to measure.’31 They conclude that the design of IRAL’s Call for Evidence was inherently biased towards a particular position, namely that judicial review should be reformed, codified and curtailed, and that the questions were not designed according to best methodological practice, subsequently entailing various types of bias. The IRAL questionnaire was only directed to government Departments, and not to other stakeholders. The near exclusive emphasis on the perspectives of the very departments that the judicial review process is there to hold accountable, with vanishingly little input from those private citizens to which accountability is owed, could well have undermined the independence of the consultation process.32 As it happened only 14 Departments responded to IRAL’s Call for Evidence. The government even resisted publication of these submissions, considering them to be part of the policy making process, that their disclosure would be contrary to fostering a free and frank exchange of views and could damage the integrity of collective cabinet responsibility.33 A raft of Freedom of Information Act requests from academics, practitioners and legal journalists were not able to uncover the contents of the submissions, with Departments citing prejudice to the effective conduct of public affairs, and HMRC admitting that it had not made a submission.34 So much for a transparent ‘culture of justification’. The government published a summary of Departmental evidence, but this was little more than a collection of opinions, not attributable either to particular Departments, or to official or ministerial level, and provided no quantitative empirical evidence beyond a few estimates of claims and their outcomes, at least some of which were disputed as to their accuracy on the very day the summary was published.35 The matter on which the Departments seemed

29 Halliday, ‘The Role of Empirical Legal Research’ (n 28). 30 See, eg, L Bridges, G Mezaros and M Sunkin, Judicial Review in Perspective, 2nd edn (London, Cavendish, 1995). 31 Boyle and Camps, ‘(Lacking in) Methodological Rigour’ (n 21). 32 ibid. 33 Summary of Government Submissions to the Independent Review of Administrative Law; https:// assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/976219/ summary-of-government-submissions-to-the-IRAL.pdf. 34 See the collection of FOI requests and responses on UKAJI’s website: https://ukaji.org/2020/11/04/ collection-of-responses-to-the-independent-review-of-administrative-law-iral/. 35 On twitter, barrister Tom Royston of Garden Court North Chambers questioned the accuracy of statistics provided to IRAL by the Department for Work and Pensions (DWP), and the Child Poverty Action Group has written to the DWP also questioning this data.

Methodological Pluralism and Modern Administrative Law  161 to offer the most commentary was the justiciability of matters of high policy, as the evidence summary puts it: Departments raised how there is a constant question as to how the court’s role is evolving, and whether it may prevent those democratically best placed to take decisions in the public interest from doing so.36

Whilst there may be a perception, even ‘a well-trodden intellectual track’ as Craig puts it, that the role of courts conducting judicial review has evolved, the summary provides no reliable evidence to support this. IRAL’s own use of empirical data was also questioned. One of its two substantive recommendations was that Parliament should intervene to reverse the impact of Cart v Upper Tribunal37 (which provides for judicial review of certain Upper Tribunal decisions on the basis of the second-tier appeals criteria). This recommendation, and the government’s acceptance of it, was in large part based on a figure, much quoted in the press, that claimants had only been successful in 0.22 per cent of Cart Judicial reviews (that is 12 claims in around eight years). The Public Law Project’s Joe Tomlinson and Alison Pickup quickly queried this conclusion.38 It had reportedly been based on a ‘trawl’ of reported judgments on BAILII and Westlaw legal databases. However, Tomlinson and Pickup, followed later by Joanna Bell,39 noted that ‘success’ in judicial review is more nuanced than winning at a final substantive hearing, but also that the figure reached was both under-inclusive and misleading. Misleading because a better figure for success at a hearing would look at success as a proportion of the claims actually reaching a final hearing (45 claims – putting ‘success’ at 26.7 per cent) and also that a more nuanced analysis would need to determine the actual ‘result’ in the 5,457 cases issued which did not proceed to a final hearing. Many of these are likely to have settled on terms at least partially favourable to the claimant. Tomlinson and Pickup also noted that the special Cart judicial review procedure itself makes hearings, and therefore reported judgments, ‘inevitable extremely rare’. The Office for Statistics Regulation later responded to the Public Law Project that the Cart data supplied by the Ministry of Justice was too limited and that the analysis applied was ‘overly simplistic’.40 When introducing legislation to reverse the effect of Cart, the Ministry of Justice conceded not only that it was incorrect on Cart judicial review success rates, but also implicitly that its own oft-repeated approach to determining judicial review ‘success’ rates had been misleading. Departing from its usual characterisation of success as the percentage of issued claims concluding with substantive judgment in favour of the claimant

36 Summary of Government Submissions to IRAL (n 32) para 8. 37 R (Cart) v The Upper Tribunal and R (MR (Pakistan)) (FC) v The Upper Tribunal (Immigration & Asylum Chamber) and Secretary of State for the Home Department [2011] UKSC 28; [2012] 1 AC 663. 38 J Tomlinson and A Pickup, ‘Putting the Cart before the horse? The Confused Empirical Basis for Reform of Cart Judicial Reviews’, U.K. Const. L. Blog (29 Mar. 2021) (available at https://ukconstitutionallaw.org/). 39 J Bell, ‘Digging for Information about Cart JRs’, U.K. Const. L. Blog (1 April 2021) (available at https://ukconstitutionallaw.org/). 40 UK Office for Statistics Regulation, Use of statistics in judicial review reform (10 June 2021): https:// publiclawproject.org.uk/content/uploads/2021/06/OSR-letter-to-PLP.pdf.

162  Sarah Nason at final hearing (around 1 per cent-3 per cent), it instead classed success as the proportion of final hearings resulting in substantive judgment for the claimant (by its own admission between 40 per cent-50 per cent across all subject areas).41 The problems highlighted here, in qualitative and quantitative methods respectively, demonstrate the importance of rigour in research design, implementation and analysis, when seeking to use empirical socio-legal empirical methods, and the broader debates demonstrate consequences of failure to be open-minded about law’s socially constructed nature. B.  The Social Construction of Law: Facts and Values Legal and political concepts, especially those relating to administrative law and judicial review, purport to explain and refine socially constructed practices. They cannot be conceived of independently from society and since society is organic and evolutionary, so too will be the concepts we construct to understand it, to make analytical sense of it.42 Whilst the intellectual furrows may be well-trodden, they may no longer reflect the empirically observable landscape. Whether consciously or unconsciously, separating the outcomes of empirical and normative methods can lead to the production of legal theories, including the theories about judicial review, that are increasingly disconnected from reality. The risks of disconnection, however, work both ways. If socio-legal research designed to uncover empirical evidence about judicial review, and other elements of administrative law, does not openly engage to at least some extent with constitutional and administrative law theory, it risks producing results that are underpinned by particular interpretations of legal, political and administrative values, that are not made clear, and that are therefore not capable of being properly analysed and challenged. Research designs and methods, including in empirical socio-legal research, should grapple with ‘foundational’ issues affecting the nature of human enquiry. The meaning of ‘empirical’ requires an appreciation of how humans might come to empirically correct knowledge about any social phenomena. These foundational issues have a bearing on our methodologies to better understand both administrative and constitutional law. As Arvind and Stirton explain in this volume, constitutional theories are built on facts and as such: are necessarily built upon some type of epistemological framework which determines how theorists justify and have confidence in the accuracy of the facts that sustain their theory and the conclusions they draw in relation to the implications of those facts for the constitutional order.43

41 Ministry of Justice Press release, 21 July 2021: www.gov.uk/government/news/new-bill-hands-additionaltools-to-judges. 42 F Schauer, ‘The Social Construction of the Concept of Law: A Reply to Julie Dickson’ (2005) 25(2) Oxford Journal of Legal Studies 493. 43 Arvind and Stirton, ch 4 in this volume.

Methodological Pluralism and Modern Administrative Law  163 As they go on to note, these theories are rarely explicitly stated, especially when the theorists’ aim is to persuade the reader of its correctness rather than articulate how the theory was constructed and how it meets particular criteria for justified belief. It is this construction, and these criteria, that Arvind, Stirton and I each counsel should be made more explicit in the methodology of constitutional and administrative law theory respectively. Developing my own methodology in RJR, I circled back to core philosophical concepts, including epistemology itself, which is the study of knowledge, especially with regards to the methods through which we come to knowledge about the world around us, and the distinction between knowledge and justified beliefs or opinions. Empirical legal researchers benefit from being aware of such concepts because, whether consciously or not, they underpin the designs of our projects. For example, one epistemology is deductive reasoning, where the researcher might develop some prior theoretical hypothesis about a particular issue in or involving constitutional or administrative law and seek to test this by collecting and analysing data. On the other hand, is an inductive approach, which begins with an admittedly broad social understanding of administrative law, followed by collecting data from which to divine a more precise account of themes or patterns that could ultimately cohere. Inductive reasoning is more often associated with qualitative research; a method focused usually on observing research subjects in context (for example, in their natural working environment as administrators, legislators or solicitors specialising in administrative law) and collecting information from them about their experiences which need not necessarily be reducible to any particular unit of measurement.44 These distinctions help us to better understand the foundations of empirical legal research, but they are generalisations, rather than hard and fast rules of research design. The overarching research design of IRAL was not made clear, but may have been deductive, based on seeking evidence to confirm or disprove the government’s hypothesis that judicial review needs reform, and specifically, curtailment. The line between access to justice and administrative efficiency which IRAL was tasked to map, looks like a specific dueling pair or demarcation problem. I think that many of these boundary debates in administrative and constitutional law theory can be traced back to so-called conflicting pairs, that have traditionally occupied philosophical debate.45 Examples of conflicting pairs in philosophy are: a priori (things that are true independent from our experience of them) and a posteriori (things that are true only after our experience of them); necessary truths (which hold true by virtue of logic or other forms of conceptual intuition in any possible world regardless of sense experience) and contingent truths (which are true by virtue only of some empirical fact of the matter). Theoretical debates over constitutional and administrative law have also played out in this dualistic way; does administrative law

44 DJ Galligan ‘Legal Theory and Empirical Research’, in P Cane and HM Kritzer (eds), The Oxford Handbook of Empirical Legal Research (Oxford, Oxford University Press, 2012) 976, 981, and more generally A Bryman, Social Research Methods 4th edn (Oxford, Oxford University Press, 2012) ch 1. 45 C Elgin, Between the Absolute and the Arbitrary (Ithaca, Cornell University Press, 1997).

164  Sarah Nason constrain executive action or promote good administration, is the UK Constitution a legal or political concept? How one answers these, and other questions about constitutional and administrative law, and indeed whether one thinks they are capable of a single answer, stems back to our views on how we think we know what is true about the world around us, particularly to our views on what is true about law, and how we know that to be true. In administrative law in particular, the existence of philosophical dualism tracks not only contestation between the individual and the administration, but also as a conflict between the importance of ‘context’ to administrative law decision-making and adjudication, on the one hand, and concern to prevent unconstrained judicial discretion (or contextualisation) leading to an apparent ‘wilderness of single instances’, on the other.46 It had been philosophical orthodoxy to propose a distinction between analytic truths, true by virtue of their meaning alone (conceptual analysis would likely fall within this category), and synthetic truths, true by virtue of sense experience (empirical proof in accordance with natural scientific methods). On this basis both doctrinal conceptual analysis and empirical legal studies are capable of providing truths about law. However, the distinction between the analytic and synthetic has been questioned, and famously rejected by some, notably Willard Van Orman Quine. For Quine, analytic truths are true by virtue of their meaning independently of matters of empirical fact. Synthetic truths are true by virtue of empirical facts that ground them. Quine argued that aside from very basic statements of logic, such as, ‘a man who is unmarried is an unmarried man’, further statements that depend on synonyms, such as, ‘a bachelor is an unmarried man’, cannot be analytically true. He considered, for example, that the true meaning of the second statement (using the synonym) would have to depend on the definition of the synonym which is likely to be a matter of fact, (that is, who posited it and where, and how regularly is it applied), or some rules of language (semantic rules) or by prior appeal to the truth of analyticity itself. To take an example relevant to administrative law, we can say that a decision that is ultra vires is an ultra vires decision. However, basic logic alone cannot tell us that a jurisdictional error is ultra vires; whilst jurisdictional error (or the jurisdictional principle) is often taken as synonymous with ultra vires, this equation depends on some prior definition of the synonym jurisdictional error. More recently following the government’s response to IRAL, ‘nullity’ is back under the spotlight, and may well be another example of where analytic logic alone is not enough to fully resolve a legal doctrinal conundrum in social context.47 Quine argued that the only truths are empirical (synthetic) truths. However, such truths cannot be arrived at by ‘reductionism’, reducing purported truths down to singular statements that can be empirically verified. For Quine total science is more like a field or force, the boundaries of which are constantly being readjusted

46 D Feldman, ‘Comparison, Realism and Theory in Public Law’ in J Bell, M Elliott, JNE Varuhas, and P Murray (eds), Public Law Adjudication in Common Law Systems: Process and Substance (Oxford, Hart Publishing 2016). 47 M Elliott, ‘Judicial review reform I: Nullity, remedies and constitutional gaslighting’ (6 April 2021), https://publiclawforeveryone.com/2021/04/06/judicial-review-reform-i-nullity-remedies-and-constitutional-gaslighting/.

Methodological Pluralism and Modern Administrative Law  165 by experience. In this field, re-evaluation of some statements of truth requires a knock-on re-evaluation of others, and ultimately any ‘statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system’.48 For Quine, everything is up for grabs in light of experience, even the rules of logic, and what is true can only be determined in light of the whole of science. Perhaps much the same can be said of what is true about both administrative and constitutional law; anything can be true if we make drastic enough adjustments elsewhere in the system. We might say this was Lord Chancellor Buckland’s standpoint in pursuing reforms that require a significant, and questionable, reinterpretation of empirical evidence across the whole field of administrative law. What remains to be seen is whether we are prepared to accept such drastic adjustments. In socio-legal research, many of those who reject dualisms between fact and value also challenge epistemological realism (that true knowledge is ‘out there’ and we can come to it independently of sense experience). Such theories express a ‘constructivist social epistemology’ under which ‘law as an autonomous epistemic subject … constructs a social reality of its own’.49 For example, Habermas argues that we should not seek to separate sociological and legal analyses, proposing that what is required is ‘an analysis equally tailored to the normative reconstruction and the empirical disenchantment of the legal system’.50 Roger Cotterrell similarly refers to ‘law as ideal and reality’,51 and in legal philosophy Robert Alexy recognises law’s inherent ‘dual nature’.52 My own methodology is tailored to normative interpretation of the values underpinning the social practice of judicial review and its place in administrative and constitutional law, as well as seeking to provide further empirical disenchantment (of the ‘myth busting’ kind noted above with respect to Bondy et al’s empirical work). A broader theme of the constructivist approach I take to administrative law theory, is to reject bright line demarcations between methods that are descriptive, evaluative, so-called indirectly evaluative, normative and prescriptive. Debates over what is or should be the best method for general legal theory led to ‘thirty years off the point’ for jurisprudence,53 and it would be disappointing were the same to happen with regards to constitutional and administrative law theory. In RJR, I characterised the ultra vires debate, explained by Lewans in this volume as ‘whether judicial review of administrative decisions is necessarily justified by Parliamentary intent or common law principles’54 as an interpretive debate over the values of administrative law and 48 WVO Quine, ‘Two Dogmas of Empiricism’ (1951) 60 The Philosophical Review 20, 39–40. 49 G Teubner, ‘How the Law Thinks: Toward a Constructivist Epistemology of Law’ (1989) 23 (5) Law and Society Review 727, 730. 50 J Habermas, Inclusion of the Other: Studies in Political Theory Ciaran Cronin and Pablo De Greif (eds) (Cambridge, Mass, MIT Press, 1998). 51 R Cotterrell, Law, Culture and Society: Legal Ideas in the Mirror of Social Theory (Oxford, Ashgate Publishing, 2006) 32. 52 R Alexy, ‘The Dual Nature of Law’ (2010) 23 (2) Ratio Juris 167. 53 See, eg, A Halpin, ‘The Methodology of Jurisprudence: Thirty Years off the Point’ (2006) 19 Canadian Journal of Law and Jurisprudence 67; J Dickson, Evaluation and Legal Theory (Oxford, Hart Publishing 2001) ch 3; D Priel, ‘The Boundaries of Law and the Purpose of Legal Philosophy’ (2007) 27 Law and Philosophy 643; B Bix, ‘Ideals, Practices, and Concepts in Legal Theory’, in J Beltran, J Moreso and D Papayannis (eds), Neutrality and Theory of Law (Dordrecht, Springer, 2012) 33. 54 See Lewans, ch 14 in this volume.

166  Sarah Nason judicial review. On the other hand, Paul Daly sees this same debate as stemming from methodological divergence: Forsyth’s account was conceptual analytical, Craig’s more descriptive and historical, and Allan and Elliott’s contributions were interpretive.55 I disagree, I think each theorist offered an interpretation of judicial review based on a collection of evidentiary sources (mostly case law judgments) interpreted in light of values they saw as going to the purposes of judicial review, underscored by their own perception of the nature of law generally. That one theorist might have focused more on the perceived internal consistency of doctrine, and another used a broader range of historical cases, and another brought in a wider set of values, does not to me denote that they were following entirely different methods, it was more that they were choosing to emphasise various sources to differing degrees. Following Arvind and Stirton in this volume we might also say each theorist was following a different approach in assigning normative significance to particular facts. C.  The Plurality of Administrative Law Values and Methods The most common method of constitutional and administrative law theoretical scholarship is interpretivism. For example, in highlighting historical common understandings of administrative law, Lewans notes that each represents an interpretive approach which conceives there to be different points or purposes for constitutional law.56 An interpretation may, or may not, include careful construction from empirical sources (including case law, legislation and guidance, institutions, changing internal practices within government and administration, scrutiny documents from legislative committees, relevant statistics and so on). Constructive interpretation uncovers themes in the development of constitutional and administrative law that might well constitute or add to the exposition of new theories. Examples of constructive interpretation uncovering ‘themes’ includes: the emergence or so-called ‘New Public Management’,57 the expansion of the regulatory state, the rise (and fall) of administrative justice,58 increased proceduralism in administrative law,59 a bureaucratic turn in redress (particularly where government designs, operates and is the primary defendant in redress),60 my own theory of a ‘new administrative of Wales’,61 and interpretations 55 P Daly, Plural Public law (University Research Chair in Administrative Law & Governance, University of Ottawa Inaugural Lecture, 4 March 2020) https://supremeadvocacy.ca/wp-content/uploads/2020/03/ Plural-Public-Law.pdf. 56 Lewans, ch 14 in this volume. 57 See, eg, T Christensen and P Lægreid, The Ashgate Research Companion to New Public Management (Oxford, Ashgate, 2012). 58 M Adler, ‘The Rise and Fall of Administrative Justice: A Cautionary Tale’ (2012) 8 Socio-Legal Review 28 and T Mullen, ‘Access to Justice in Administrative Law and Administrative Justice’ in E Palmer, T Cornford, A Guinchard and Y Marique (eds), Access to Justice: Beyond the Policies and Politics of Austerity (Oxford, Hart Publishing, 2016) 91. 59 C Harlow and R Rawlings, ‘Proceduralism and Automation: Challenges to the Values of Administrative Law’ in E Fisher and A Young (eds), The Foundations and Future of Public Law (in honour of Paul Craig) (Oxford, Oxford University Press, 2019). 60 R Thomas and J Tomlinson, ‘Mapping current issues in administrative justice: austerity and the ‘more bureaucratic rationality’ approach’ (2017) 39 Journal of Social Welfare and Family Law 380. 61 S Nason, ‘The “New Administrative Law” of Wales’ [2019] Public Law 703.

Methodological Pluralism and Modern Administrative Law  167 of the variable success (and more recent apparent failure) of ‘regionalising’ judicial review. Lee Marsons uses this constructive interpretivist method to collate and analyse a set of ‘often seemingly disparate, technical changes’62 to constitutional law, alongside the more attention-grabbing developments, to ‘join the dots’, demonstrating a string of legislative proposals enhancing executive power and reducing judicial scrutiny. This systematic constructive interpretation ‘allows us to make better sense of overall patterns of constitutional change (the good and the bad), but also to draw much needed broader attention to the constitutional dynamics of this period’.63 In these examples, the data observed have been reflected on as a means to understand changing conceptions of constitutional and administrative values. Whilst each interpretation is important, I consider those with a more rigorous empirical grounding to be more reliable and of more practical use. Those who have developed theories about administrative law and judicial review, generally reveal themselves to follow interpretivist methodology. For example, Trevor Allan identifies as an interpretivist, noting that ‘[m]atters of empirical fact or official opinion obtain their relevance from a theory informed by moral and political values – a theory whose construction is foundational to any useful account of the practice of judicial review’.64 Paul Craig also accepts that his method is interpretive, specifically subscribing to Ronald Dworkin’s approach.65 Common among newer interpretive theories of administrative law, however, is that they draw on a wider range of values than the triumvirate rule of law, separation of powers and parliamentary sovereignty, in part because their proponents also recognise a wider range of sources of administrative law. RJR was part of this development, alongside the work of others such as Joe Tomlinson, particularly his method of proposing and dissecting various administrative justice values, including transparency, efficiency, accountability and equal treatment.66 Tomlinson suggests that the core task for administrative lawyers ‘should be to focus on examining techniques that provide for the efficient implementation of public policy while ensuring justice for individuals’,67 which seems to echo my own interpretive theory of judicial review for the advancement of individual justice and good governance. Paul Daly has developed a values-based account of administrative law, and more recently examined the plurality of public law, including its sources, values, and methods.68 Daly’s values-based account stressed the centrality of the rule of law, good administration, 62 L Marsons, Constitutional change in the UK – joining the dots (Legal Action Group, Legal Action Magazine March 2021), www.lag.org.uk/article/210365/constitutional-change-in-the-uk---joining-the-dots. 63 ibid. 64 TRS Allan, The Sovereignty of Law: Freedom, Constitution, and Common Law (Oxford, Oxford University Press, 2013) 211 and Appendix. 65 ‘I have stated … on numerous occasions … that I subscribe to a Dworkinian interpretivist legal Theory’. P Craig, ‘Theory “Pure Theory” and Values in Public Law’ [2005] Public Law 440, 440–41. 66 J Tomlinson, ‘The Grammar of Administrative Justice Values’ (2017) 39(3) Journal of Social Welfare and Family Law 524. 67 J Tomlinson, ‘Embracing Complexity and Diversity in the Principles of Administrative Law’ (Admin Law Blog 2 December 2020), https://adminlawblog.org/2020/12/02/joe-tomlinson-embracingcomplexity-and-diversity-in-the-principles-of-administrative-law/. 68 P Daly, ‘Administrative Law: A Values-Based Approach’ in J Bell, M Elliott, J Varuhas and P Murray (eds), Public Law Adjudication in Common Law Systems (Oxford, Hart Publishing, 2016) 23 and ‘Plural Public Law’ (n 56).

168  Sarah Nason democracy and separation of powers. Plural public law goes further in acknowledging the importance of examining how this plurality of values ‘operate in the real world of legislation, executive action, administrative adjudication and judicial review’69 and that such examination in the real world ‘sheds light on public law and provides public lawyers with tools to manage complexity and influence change’.70 Daly rejects the notion that there is one ‘unified field theory’ of public administrative law. Daly’s methods are both doctrinal and interpretive, but he also counsels the importance of bringing to bear a variety of methods, including empirical socio-legal research, and methods from related disciplines such as political science. Another modern account is Joanna Bell’s, Anatomy of Administrative Law.71 Her insight is not to develop a specific list of administrative law values, and indeed she acknowledges that providing an exhaustive register might be impossible. Her contribution is to pose that the kind of values highlighted in much of the existing body of work, such as Daly’s, as well as that of theorists in the historical ultra vires debate, are perhaps best (though not exclusively) understood as what she terms ‘common law values’, and that these values interact in complex and diverse ways with legislative purposes. For Bell the diversity of these common law values stems in part from the range of beneficiaries of administrative law, that is those whose interests administrative law seeks to protect. What these modern methodological approaches appear to have in common is their rejection of an interpretivist lens that views administrative law solely or primarily from the top-down through a prism of overtly constitutional values and principles. Through their broader and deeper focus on empirical evidence, including systematic analysis of judgments, these approaches allow, and perhaps invite, the conclusion that administrative law constructs a social reality of its own. As Kristen Rundle’s chapter in this volume demonstrates, the top-down approach can also be associated with more orthodox constitutional accounts of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by judicial mechanisms to control administration rather than to facilitate good administrative practice.72 As Rundle then proposes, this pattern of constitutional thought, what Craig might perhaps call a form of ‘path dependency’, then sidelines acknowledgment of the broader range of actors in the administrative state and the variable roles they play in giving concrete expression to relationships between government and the governed. I propose that socially constructed bottom-up methods are better equipped to appreciate the significance of various actors, roles and relationships. D.  Doctrine and Theory Empirical and theoretical methods are often directly contrasted with doctrinalism. Whilst dependent on empirical sources such as statutes and case law, the accuracy of

69 Daly 70 ibid. 71 J

‘Plural Public Law’ (n 55).

Bell, The Anatomy of Administrative Law (Oxford, Hart Publishing, 2020). Rundle, ch 15 in this volume.

72 See

Methodological Pluralism and Modern Administrative Law  169 any particular account of the internal consistency of the principles abstracted from those sources, would not normally be seen as something testable to the accuracy of empirical methods. Here law is viewed as a self-sustaining set of principles where there is seen to be no express need to seek reference from the social world outside the documentary sources. My concern with some supposed general administrative law doctrines is that they prioritise seemingly descriptive conceptual analysis, failing to address the contested nature of the values that underpin those doctrines, and how these evolve over time. Concepts function as mediating devices between the real world and our understanding of it by constituting an attempt to render explicit what is already implicit in our common understanding of social practice. Descriptive concepts purport to achieve this without overt reference to the very values that the modern cohort of administrative law theorists insist explicitly characterises administrative law. Purportedly descriptive doctrinal classification may provide a useful tool in the judicial and administrative armoury, but it provides only a partial aid to judgement. Classification is useful, but only when it continues to reflect administrative and judicial practice, values, and the diverse relationships, which Bell for example, demonstrates as characteristic to the anatomy of administrative law. Doctrinal classification is a tool to aid judgement, it should not become a proxy for it. A more ‘organic’ account of rational judgement that could be drawn on here, is that developed by Cliff Hooker – described as organic naturalist interactivist-constructivist!73 Hooker notes that human problems (including moral and legal problems) are ‘multi-normed, multi-constrained, under-defined and context sensitive’.74 Administrative law include norms of different legal regimes, statutory and common law, topic specific, national, regional and international, as well as norms of different areas of administrative practice. Courts are constrained by judicial expertise, the forms and limits of adjudication, and institutional characteristics. Problems may be underdefined, not least because of the difficulty of estimating possible consequences. Hooker draws on an oscillating or homeostatic balancing process that can be observed in nature, whereby organic forms (including human beings) function at their best through a process of constant readjustment. There is a natural unity to human forms, but it is a unity that rejects divisions imposed by human fiat. Human beings can still strive for ideals, but the standards that guide our striving must be accessible to us; they must be learnable and improvable. The process by which humans come to reach these ideals is more important than formalism’s location of reason in the character of the end product, namely certain truth. The apparently absolute nature of formal rationality (both legal and moral) suppresses the importance of human judgement. The inter-activist, dynamic element of Hooker’s naturalism takes the functional biological condition of self-regulation, taking inputs from the environment and divesting outputs, as a model for reasoning in morality and law in the real world of social practice. The roots of reason lie in a structured response process

73 C Hooker, ‘Rationality as Effective Organisation of Interaction and its Naturalist Framework’ [2010] Axiomathes 99. 74 ibid.

170  Sarah Nason that incorporates feedback to increase effectiveness. Judgement is central and it can be systematically improved through learning and training and complemented by the presence of both expert and lay decision-makers. This is particularly true in the case of judicial review, which has sometimes been described as a ‘topsy turvy’ process of ‘institutionalised second guessing’.75 On this organic and dynamic theme, Jonanna Bell has developed an anatomical methodology, which, as Liz Fisher puts it, suggests administrative lawyers should be more like botanists.76 Bell recounts a monologue by the theoretical physicist Richard Feynman known as his ‘Ode to a Flower’. This is an exchange between Feynman and an artist friend, in which the friend bemoans the scientific preoccupation with taking things apart, and the way that this detracts from the complete beauty of the rose. Feynman responds that what is aesthetically pleasing can exist in many forms, and that there are all kinds of questions that science adds into the mix of what is exciting, mysterious, and ultimately beautiful about the flower. From this anecdote, Bell explains Feynman’s recognition that the flower is made up of many small components and complex processes, and as such it cannot be neatly summed up in terms of a single idea. Bell suggests that administrative law has been preoccupied with searching for a single ‘organising concept’ a ‘master theory’, as Bell puts it a ‘monistic principle’, or as Daly states a ‘unified field theory’. The crux of Bell’s argument is that pulling apart administrative law doctrine in an anatomical way demonstrates that the legal structures at play in administrative law adjudication are both complex and varied. As Rundle’s chapter shows, this is not just true of administrative adjudication but may also be true of the administrative state. Bell proposes three core senses of complexity: that legislative detail matters enormously in administrative law adjudication; that administrative law pursues different normative goals; this variety is due both to the diverse legal origins of administrative law values, but also to the third core sense of complexity, namely that the beneficiaries of administrative law are also varied and as such sit in various kinds of relationships with the administrative decision-makers they seek to challenge. A key contribution of Bell’s method is to spotlight the importance of statutory schemes to administrative law adjudication. Her approach is not to maintain, as others have, that a connection, perhaps any connection, to statute is crucial to the constitutional legitimacy of judicial review, but instead to show the various ways in which administrative law adjudication interacts with statutory schemes. As she states: the grounds of review do not function as a series of freestanding legal tests which apply directly to the facts of the case and mechanistically determine the proper legal outcome … administrative law doctrine often interacts closely with the particular legislative framework in the background of the challenge. The legislative frameworks with which administrative law doctrine interacts, furthermore, are not of one kind.77

75 T Endicott, ‘Questions of Law’ (1998) 114 Law Quarterly Review 292, 297. 76 L Fisher, ‘Should Administrative Law Scholars Be More Like Botanists?’ (Admin Law Blog November 2020), https://adminlawblog.org/2020/11/17/liz-fisher-should-administrative-law-scholars-be-more-like-botanists. 77 Bell, The Anatomy of Administrative Law (n 71) 66.

Methodological Pluralism and Modern Administrative Law  171 There are many different kinds of legislative framework: from those that are primarily designed to regulate individualised decision-making (such as social security and housing benefit decisions, or licencing decisions); to those that regulate decisions about how people ought to be treated; to those which regulate decisions about the protection and promotion of collective public interests. Administrative law doctrines, the so-called general principles or grounds of review, interact in different ways with the structure and detail of these different types of statutory schemes, with the statutory background itself often becoming more complex over time; one need only think of planning and environmental law to illustrate such complexity. Moreover, as Bell acknowledges, there is reason to think that the background administrative schemes to various administrative law challenges are increasingly rich in detail. This is likely, at least in part due to a combination of the increased quantity and density of secondary legislation, and a growing role for ‘soft law’ sources, including government guidance and administrative policy, both in framing procedures and in structuring administrative discretion. Again, this is an approach, this time specifically to doctrinal analysis, that recognises a plurality of values and sources of administrative law, and that these interact, organically and constructively. Those commenting on Bell’s work have highlighted various themes in the methodology of administrative law theory, including that administrative lawyers should not neglect the realities of the state; that they should appreciate the dynamic evolution of their subject; and that such ‘lines of inquiry will often entail using empirical methods and more systematic modes of analysis’.78 E.  Specialist and Generalist Administrative Law, Functionalism and Content Analysis Bell’s work recognises the importance of context to administrative law adjudication, and she joins other scholars, notably Harlow and Rawlings,79 Arvind and Stirton,80 and more recently, Tomlison and Daly, in emphasising the messy and pluralistic origins of administrative law and encouraging the restoration of connections with administrative practice, and variable anchorage in different types of administrative power. My own contribution on this theme is organised around two key tensions. The first is that while individual grievance claims make up the larger share of judicial review caseload, the application for judicial review has generally been ill-suited to the resolution of individual disputes (in effect the individual/good administration dichotomy).81 The second tension is between a generalist approach to administrative law – one that seeks to articulate general grounds of review applicable across areas of administrative activity – and a more specialist or context specific approach. 78 Tomlinson, ‘Embracing Complexity and Diversity’ (n 67). 79 C Harlow and R Rawlings, ‘Administrative Law in Context: Restoring a Lost Connection’ [2014] Public Law 28. 80 TT Arvind and L Stirton, ‘The Curious Origins of Judicial Review’ (2017) 133 LQR 91. 81 S Nason, ‘Plus ç a Change? An Empirical Analysis of Judicial Review in Modern Administrative Law’ in TT Arvind, R Kirkham, D Mac Síthigh and L Stirton (eds), Executive Decision Making and the Court: Revisiting the Origins of Modern Judicial Review (Oxford, Hart Publishing, 2021).

172  Sarah Nason This second tension is one Bell also highlights in her work. Methodologically this tension has played out in the apparent divide between normative and ‘functionalist’ approaches to administrative law. The so-called ‘functionalist’ school focused on law in context and included scholars whose positions on the nature of good administration led to recognition of more subject-matter-specific administrative laws (based on the characteristics of particular public bodies, the types of powers they possess, and the context of, and limits on, their decision-making). These approaches are quite different to those laid out in modern textbooks which tend to be organised around general principles and techniques designed to impose legal control and accountability on administrative power across a range of subject matters. Attempts to develop a systematic and comprehensive body of general administrative law risk glossing over context specificity in administrative and legal practice. As Bell has since proposed, this obscures the importance of the content and complexity of the legislative background and plays down the dynamism of specific areas of administration (housing, immigration, planning, the environment and so on) both in terms of sources of law, values, and relationships between actors. Modern administrative lawyers have increasingly turned to ‘content analysis’ to try and better understand the tension between specialist and generalist administrative law. In short, if we analyse administrative law judicial review like a big cake, the layers are characterised by topics like planning and education, and not by other matters such as grounds of review, or activist versus deferential judicial attitudes. Content analysis involves the systematic examination of texts and can straddle the divide been quantitative and qualitative methods as it tends to involve thematic categorisation, as well as counting the frequency with which particular themes or codes appear. Texts can be reduced to codes by categorising items in the text and then counting their occurrence, allowing inferences to be drawn from the document. Content analysis can also be used to examine the structure of discourse, and can be more reflexive, moving back and forth between the text and the interpreter’s reaction to it. Both RJR and Bell’s ‘Anatomy’ are in this vein, but a more specific example is Richard Kirkham and Elizabeth O’Loughlin’s examination of judicial review claims involving ombuds.82 Their analysis stressed the risks of traditional doctrinal analysis, where conclusions about the broader practice of the judiciary in administrative law (especially so-called judicial activism or overreach) are seemingly presented as ‘facts’ but are actually based on a selection of high-profile final appellate judgments.83 As Kirkham and O’Loughlin note, in the absence of more systematic analysis there is ‘potential for abstract arguments about the law to become informed and driven more by conscious or unconscious bias than by the real practice of the law’.84

82 R Kirkham and E O’Loughlin, ‘A Content Analysis of Judicial Decision-Making’ in N Creutzfeldt, M Mason and K McConnachie (eds), Routledge Handbook of Socio-Legal Theory and Methods (London, Routledge 2019). 83 ibid. 84 id.

Methodological Pluralism and Modern Administrative Law  173 The results of these types of content analysis give further support to the view that administrative law in social practice resists the imposition of strict theoretical or doctrinal demarcations. As Joanna Bell puts it of the supposed demarcation between specialist and generalist administrative law: There is a simple reason why there is no choice: it does not reflect the legal reality. Administrative law adjudication is, by its nature, legally intradisciplinary. This is not a ‘pure’ legal field, characterised by neat boundaries and a distinct logic. Legal overlaps are at its heart.85

Just like other methods, content analysis does not remove the need for judgement at the stage of moving from the abstract hypotheses to concrete conclusion, but it does have the potential to more clearly separate these methodological stages than is the case with traditional doctrinal analysis, therefore providing greater transparency and more opportunities for scrutiny. It is a fruitful means to expose administrative law’s reluctance to confirm to traditional doctrinal and theoretical boundaries, but nevertheless has the potential to reveal different senses of internal coherence that might otherwise not be appreciated. F.  Computer Science Methods In addition to manual content analysis, computer science is contributing new m ­ ethods for understanding constitutional and administrative law; in particular, through use of computer programs to retrieve and classify large amounts of legal and administrative data. Corpus linguistics for example, is the study of language expressed in corpora (bodies of real-world text) which can include collections of case law judgments and statutes, as well as other legal and related sources. The corpus method is an approach that derives a set of abstract rules that govern a natural language from texts in that language, exploring how that language relates to other languages. Corpora derived from machine readable texts can be analysed entirely by human researchers, but increasing use has been made of Natural Language Processing (NLP) techniques which rely on machine learning to derive meaning from human texts. One form of automated content analysis using NLP techniques used in constitutional and administrative law research, is ‘structural topic modelling’. For example, David S Law has used this approach to uncover three constitutional archetypes from the text of constitutional preambles.86 As he explains: topic modelling ‘utilizes patterns of word occurrence to break a corpus of text into its component topics, measure the proportion of the corpus that concerns each topic, and identify the vocabulary associated with each topic’.87 A topic is a cluster that groups together words that are more likely to appear with one another across the corpus. As with manual content analysis, and

85 J Bell, ‘The Anatomy of Administrative Law: Reflections on the Reflections’ (Admin Law Blog, 23 December 2020), https://adminlawblog.org/2020/12/23/joanna-bell-the-anatomy-of-administrativelaw-reflections-on-the-reflections/. 86 DS Law, ‘Constitutional Archetypes’ (2016) 95 Texas Law Review 153. 87 ibid 190.

174  Sarah Nason other forms of empirical research discussed in this chapter, automated methods can only assist analysis ultimately based on judgement, they cannot replace it. David S Law notes ‘automated content analysis is no substitute for human interpretation and judgment, but it does extend human capabilities dramatically’.88 Similar techniques have been used to predict the judgments of particular courts, most often top appellate courts, including the European Court of Human Rights (decisions of which have a significant impact on constitutional and administrative law across Member States). These approaches are ultimately based on NLP techniques including topic modelling. Initial attempts focused on the US Supreme Court,89 with subsequent application of more advanced methods to the ECtHR90 and the French Court of Cassation.91 An initial attempt has been made to predict the decisions of the UK Supreme Court, but the researchers note that unlike other jurisdictions, the UK lacks a structured publicly available data set of cases.92 At the time of writing it has been announced that important judgments will be available for free online via the national archives from April 2022, though the permitted uses of such judgments has not yet been established. Computational methods can also assist in uncovering and explaining constitutional architecture, including through network analysis and other visualisation methods. For example, I have been involved in the co-design of a system to visualise administrative justice pathways to redress.93 Our method was to classify, collate and organise the underpinning data, from expert workshops, heuristic evaluation and expert critical reflection. In this work we developed an application design study of an explanatory visualisation tool (prototype named Artemus). With explanatory visualisation, the goal is to provide both a way to present facts and allow users to explore them, so that they can develop a deeper understanding of the information. The aim is to educate and instruct, and to elaborate concepts and processes, explaining what they are, why they happen and how they relate to other concepts. Artemus includes two in-depth case studies in administrative law and redress relating to housing and education. The visualisation allowed administrative law experts to observe gaps and inconsistencies in the redress process. Breaking the redress routes down into computational visual constructs of ‘nodes’ and ‘edges’94 represented a different way to think about administrative law routes to accountability and allowed us to see complexity 88 ibid 192. 89 D Katz, M Bommarito and J Blackman, ‘A general approach for predicting the behavior of the Supreme Court of the United States’ (2017) 12(4) PLoS ONE. 90 N Aletras et al, ‘Predicting judicial decisions of the European Court of Human Rights: A natural language processing perspective’ (2016) PeerJ Computer Science. 91 O-M Șulea et al, ‘Predicting the Law Area and Decisions of French Supreme Court Cases’ (2017): www.researchgate.net/profile/MarcosZampieri/publication/321711075_Predicting_the_Law_Area_and_ Decisions_of_French_Supreme_Court_Cases/links/5aac0527aca2721710f89eca/Predicting-the-Law-Areaand-Decisions-of-French-Supreme-Court-Cases.pdf. 92 B Strickson and B De La Iglesia, ‘Legal Judgement Prediction for UK Courts’ [2020] ICISS 204. 93 J Roberts, P Butcher, A Sherlock and S Nason, ‘Explanatory journeys: visualising to understand and explain administrative justice paths of redress’ forthcoming IEEE Transactions on Visualization and Computer Graphics 2022. 94 A node represents an entity, in our case an administrative justice redress/accountability body, and an edge exists between two nodes if the two corresponding entities are related, edges include different relationships like internal administrative review, appeals, judicial review and complaint processes.

Methodological Pluralism and Modern Administrative Law  175 in the redress systems, and to observe structures that would not otherwise have been recognised or appreciated. This also allowed us to see the administrative state as an inter-connected network rather than a top-down hierarchy exercising purely delegated power, and was a good illustration of Rundle’s claim in this volume that seeing the administrative state in this more variable and ground-up way also challenges much orthodox constitutional theory. III.  CHALLENGES AND OPPORTUNITIES OF METHODOLOGICAL PLURALISM IN ADMINISTRATIVE LAW

Themes uniting the methods presented in this chapter are that they allow us to better understand complexity, plurality, and dynamism in administrative and constitutional law, and that they transcend traditional methodological boundaries. Benefits include creating a more accurate map of social practice, enriching theory, recognising coherence (and divergence) where such might not otherwise have been appreciated, and developing more evidenced-based reforms. Understandably there are also risks. Many of these relate to what Paul Daly refers to as the need to avoid becoming methodological ‘dilettantes’ who sashay between disciplines without coming to fully understand their respective rigors.95 Daly counsels ‘epistemic humility’. This is one reason why I make such a detailed case from first principles for an empirically enhanced constructive interpretative method. Another risk for the researcher is becoming a ‘jack of all trades, master of none’. To combat these risks, we must be intellectually honest and transparent in developing research designs and methods, and aware of the limitations of our own knowledge and experience. We must engage with those who have more specific methodological expertise and/or seek training to develop our own abilities. It is rarely as simple as engaging, for example, a social or computer scientistic, and divvying up research tasks. In crossing both disciplines one has to learn much of the language of the other discipline or disciplines, in order to develop research collaborations in the most fruitful and impactful ways. Joanna Bell suggests this requires administrative law scholars to be both ‘nimble’, in deftly crossing traditional methodological boundaries, and ‘sensitive’, being receptive to the quickly evolving subject matter. Sensitivity can also relate to the learning of various languages across methods and disciplines, which can itself lead to new discoveries. Co-design is especially valuable; for example, in our administrative justice project, computer science colleagues were ‘embedded’ from the outset, attending and participating in focus groups and workshops with experts from across the administrative justice sector, leading to production of a co-designed visualisation tool. Innovative and experimental methodological projects can require funders to sometimes take a risk on novel or exploratory methods where research techniques, and with them pathways to impact, may be less tried and tested, and yet may produce extremely valuable practical outputs, whether specifically intended or as



95 Daly,

‘Plural Public Law’ (n 55).

176  Sarah Nason by-products. Embedded collaboration, which may extend to formal training in each other’s research methods, also takes time and space to develop, and crucially requires understanding that mistakes will happen and that such can be important learning experiences. Cross-disciplinary expertise, or at least a clear appreciation of it, should also be reflected in journal editing and peer review processes, and there will also be implications around how to communicate results in a way that properly demonstrates both methodological rigour and impact. Recognising plurality and complexity in sources, values, and methods also has implications for teachers, and for writers of textbooks on administrative and constitutional law, and for law school curricula that experience a range of regulatory and market pressures. At least some of these challenges will need to be addressed if administrative law, and judicial review especially, are to be recognised as so much more than political footballs in largely theoretical debates between legal and political constitutionalists, and if a range of diverse voices are to continue to be heard. Perhaps most apt to state in final conclusion is how these challenges, and opportunities, reflect Justice Scalia’s pronouncement: ‘administrative law is not for sissies’.96



96 The

Hon Antonin Scalia (1989) 3 Duke Law Journal 511.

Morality

178

7 The Constitution of Legal Authority DAVID DYZENHAUS*

The concept of the basic norm or of the constitution in the legal logical sense corresponds to a certain degree with the concept so significant to natural law theory of the original contract or social contract that first and foremost constitutes the state. At bottom, we find the same theoretical need for a presupposition of unity on which the unity of the state order in the multiplicity of legal acts of state could be grounded, and which could grasp actual power relationships as legal relationships, and that leads here necessarily to the basic norm as it does in natural law theory to the foundational contract.1 Hans Kelsen

The epigraph to my chapter contains three important insights. In reverse order: first, the task of philosophy of law is to grasp what otherwise might seem to be mere power relationships as legal relationships; second, such a grasp requires understanding the legal order as a unity, which in turn requires understanding the Constitution in the sense of the basic norm that gives it unity; third, that task is roughly similar to the natural law project which seeks to give an account of the authority of the state in terms of a social contract between free and equal individuals. These lines were penned by Hans Kelsen in an essay of 1924 on the ‘The doctrine of the three powers or functions of the state.’ The first, a mark of Kelsen’s legal positivist theory, is that law has to be understood as a matter of authority, not sheer coercive power, and hence that legal philosophy must seek to explain how might is transformed into right through its mediation by law.2 The same is true of the second: an understanding of the authority of law presupposes the unity of the legal order and the idea of the fundamental norm that gives it this unity. * For comments on and discussion of a draft, I thank Steven Coyne, Lars Vinx and the participants in a seminar at the Wissenschaftszentrum, Berlin, especially Mattias Kumm. 1 H Kelsen, ‘Die Lehre von den drei Gewalten oder Funktionen des Staates’ [1924] in H Klecatsky, R Marcic, H Schambeck (eds), Die Wiener rechtstheoretische Schule: Hans Kelsen, Adolf Merkl, Alfred Verdross (Vienna, Europa Verlag, 1968), vol 2, 1625, 1652. 2 See, eg, H Kelsen, Die philosophischen Grundlagen der Naturrechtlehre und des Rechtspositivismus, where he says that the attempt by the positivist jurist to comprehend law as an objective order through the medium of the Grundnorm entails that ‘in a certain sense might is transformed into right’, in Klecatsky, et al. (eds), Die Wiener rechtstheoretische Schule (n 1) vol 1, 281, 339. I shall refer to the translation that appears as an appendix to H Kelsen, General Theory of Law and State (Cambridge MA, Harvard University Press, 1945, Anders Wedberg, trans) 389, under the title ‘Natural Law Doctrine and Legal Positivism’, translated by Wolfgang Herbert Kraus. For the relevant passage, see 437, where ‘die Transformation der Macht zu Recht’ (Kelsen’s emphasis) is translated as ‘the transformation of power into law’.

180  David Dyzenhaus It is then only the third insight that surprises: Kelsen’s acknowledgement that these two insights bring his legal positivist theory of law close to the natural law contractualist and constitutionalist project. Moreover, he went on to say that the idea of an original contract is more or less synonymous with ‘the concept of a foundational law or of fundamental law, the so called lex fundamentalis’ that ‘plays a most significant role in the theory of natural law’ and which he said is given ‘almost pure expression in the legal theoretical line of thought of the basic norm or the constitution in the legal logical sense’.3 Kelsen, however, was adamant that there is no more than an affinity between the third insight and the others. He asserted that the idea of the social contract is a ‘primitive’ version of the hypothesis of the basic norm which should be in any case rejected because it required supposing that there was an actual original act of consent on the part of those subject to the law, even if the political regime is autocratic.4 In this chapter, I argue that if one accepts the first two insights, the third cannot so easily be rejected and indeed is fully consistent with Kelsen’s qualifications to it. In this light, one can appreciate how legal positivist philosophies of law, including HLA Hart’s version, are implicitly constitutionalist theories. At least they are so as soon as legal positivists embark on explaining law as a matter of authority which requires taking into account the ‘internal point of view’, as Hart termed it,5 of the addressees of law’s claim to be qualitatively different from what he called a ‘gunman situation writ large’.6 If this is correct, we can then recast the debate about legal positivism and natural law as a more productive debate about constitutionalism.7 But, as I shall also indicate, that recasting has the potential to allow the concerns of legal philosophy a new role in taking forward debates in constitutional theory. As we will see in section I, Hart regarded this duty as primarily pertaining to legal officials; hence, legal subjects who are not officials might comply with the law only because they fear the sanctions that attend disobedience. In addition, Hart argued that official acceptance of the law does not require moral endorsement and that, he thought, supported one of his main theses that a legal order can be both authoritative and illegitimate. But I argue both that acceptance cannot be confined to officials and authority cannot be disentangled from legitimacy. More precisely, de facto legal authority is always de jure because of the role law plays in transforming might into right. Section II sets out the elements in Hart that support this argument. Section III turns to Kelsen in order to advance the argument. Section IV shows how this discussion helps to reframe the debate in legal theory about the relationship between law and morality as one about the relationship between legal order and legitimacy, and how that reframing in turn can take forward debate in constitutional theory.

3 Kelsen, ‘Die Lehre von den drei Gewalten oder Funktionen des Staates’ (n 1) 1652–53. 4 ibid 1652–54. 5 HLA Hart, The Concept of Law 2nd edn (Oxford, Clarendon Press, 1994) 102. 6 HLA Hart, ‘Positivism and the Separation of Law and Morals’, in Hart (ed), Essays in Jurisprudence and Philosophy (Oxford, Clarendon Press, 1983) 49, 59. 7 For a different but illuminating reading of Hart’s book as an exercise in public law theory, see P Cane, ‘Public Law in The Concept of Law’ (2013) 33 OJLS 649.

The Constitution of Legal Authority  181 I.  HART ON THE CONSTITUTION OF AUTHORITY

The command theory of law of Hart’s positivist predecessors Jeremy Bentham and John Austin explained legal order in terms of sanctions attached to particular commands, issued by a legally unlimited sovereign, who is identified by the fact that he is habitually obeyed by his subjects. Their account of legal authority is thus generally taken to be reductive in that it reduces legal order to a system of sheer coercive force. Hart’s move away from the command theory is largely driven by his sense that law cannot be understood as matter of force alone. It has to be understood as a matter of authority or legal right. But, as Hart explained in 1958, he also rejected the command theory’s claim that law is ‘the creation of the legally untrammeled will of the sovereign who is by definition outside of the law’8 because of what he took to be a fact about legal orders: ‘nothing which legislators do makes law unless they comply with fundamental rules specifying the essential law-making procedures’. ‘They lie’, Hart said, ‘at the root of a legal system’ and ‘what was most missing in … [Bentham and Austin’s] scheme is an analysis of what it is for a social group and its officials to accept such rules’. Hart thus suggested that this notion of accepted fundamental rules, ‘not that of a command as Austin claimed, is the “key to the science of jurisprudence”, or at least one of the keys’.9 In The Concept of Law, Hart elaborated his account of fundamental rules by describing a ‘primitive’ society in which there are only ‘primary’ rules, rules that impose duties on the individuals in the society, and in which problems arise in regard to: the ‘uncertainty’ about what social norms counts as such rules; the ‘static’ nature of these rules since there is no clear way of changing them; and ‘inefficiency’ because of the lack of recognised means of determining rule violations and of rule enforcement.10 In his view, The simplest form of remedy for the uncertainty of the regime of primary rules is the introduction of what we shall call a ‘rule of recognition’. This will specify some feature or features possession of which by a suggested rule is taken as a conclusive affirmative indication that it is a rule of the group to be supported by the local pressure it exerts.11

The static quality of primary rules is remedied by the introduction of ‘rules of change’ and the problem of inefficiency by the introduction of ‘rules of adjudication’.12 Now from the 1958 essay it might seem that the fundamental law of a legal order among these three types of ‘secondary rules’ is the rules of change and that their existence shows that even in a legal order in which the Parliament is supreme there are legal limits on what it may do, thus refuting the command theory’s claim that the sovereign is legally unlimited. But in The Concept of Law Hart made it clear that of these secondary rules, it is the rule of recognition that is fundamental since



8 Hart, 9 ibid.

10 Hart, 11 ibid 12 ibid

‘Positivism and the Separation of Law and Morals’ (n 6) 59.

The Concept of Law (n 5) 92–3. 94. 95–7.

182  David Dyzenhaus it specifies the law-making procedures. In addition, it becomes clear that the rules of change do not so much limit what the legislature may do as constitute it as a body with power to do certain things, just as the rules of contract law do not so much limit what the contracting parties may do, but make it possible for them ‘to create structures of rights and duties for the conduct of life within the coercive framework of the law’.13 According to Hart, the ultimate nature of the rule of recognition is indicated by the fact that its existence is not certified by any other rule. It exists as a matter of fact in the practice of the officials of the system and they apply it because they take the ‘internal point of view’ towards it – they accept that it provides ‘a public, common standard of correct judicial decision’.14 The rule of recognition thus has much in common with Kelsen’s idea of the Grundnorm or basic norm of legal order developed almost half a century before. According to Kelsen, at the base of every legal order is a Grundnorm that must be hypothesised in order to make sense of the idea that there is a unified normative order. The Grundnorm tells the officials and subjects of the order that the constitutional norms of that order – from which all other norms derive their validity – must be obeyed.15 Hart thus shares with Kelsen the claim that sovereignty – in the sense of the ultimate lawmaking power of a society – is a juridical idea, which must be explained as a matter of authority, not in terms of an exercise of sheer unmediated coercive power. They both explain how might is transformed into right through its mediation by law. But there the similarities between Kelsen and Hart seem to end. Hart claimed that the idea that the Grundnorm prescribed obedience to law is redundant. All one need observe is that there is as a matter of actual official practice a rule that exists that stipulates criteria for the recognition of valid laws and that officials manifest their acceptance of this rule by applying the criteria and criticising those who do not. As Hart put the point: If a constitution specifying the various sources of law is a living reality in the sense that that the courts and officials of the system actually identify the law in accordance with the criteria it provides, then the constitution is accepted and actually exists. It seems needless reduplication to suggest that there is a further rule to the effect that the constitution … is to be obeyed. This is particularly clear where, as in the United Kingdom, there is no written constitution: here there seems to be no place for the rule ‘that the constitution is to be obeyed’ in addition to the fact that certain criteria of validity (e.g. enactment by the Queen in Parliament) are to be used in identifying the law. This is the accepted rule and it is mystifying to speak of a rule that this rule be obeyed.16

Hart also rejected Kelsen’s view that it is ‘logically impossible to regard a particular rule of law as valid and at the same time to accept, as morally binding, a moral rule forbidding behaviour required by the legal rule’. And he noted that

13 ibid 27–29. 14 ibid 116. 15 H Kelsen, Introduction to the Problems of Legal Theory (Oxford, Clarendon Press, 1992, Bonnie Litschewski Paulson and Stanley L Paulson, trans) 54–5. 16 Hart, The Concept of Law (n 5) 293.

The Constitution of Legal Authority  183 no such consequences follow from the account of legal validity given in this book. One reason for using the expression ‘rule of recognition’ instead of a ‘basic norm’ is to avoid any commitment to Kelsen’s view of the conflict between law and morals.17

In other words, Hart thought it an advantage of his approach that it permitted the preservation of what he regarded as a morally sound dualism about the legal and the moral domains, which he expressed in the Separation Thesis.18 Such dualism permits the good liberal citizen to decide whether to obey the law without being distracted by what Hart regarded as a natural law mindset that supposes that an unjust law is not law. That mindset, he suggested, issues in contrary stances: the anarchist stance that a law is invalid if it is unjust in the eyes of the beholder and the reactionary stance that if X is the law, it must also be just.19 Hart also thought that this dualism is sound for theoretical reasons. It permits the legal theorist to point out that acceptance does not require moral endorsement. On the one hand, he is clear that legal authority is something that is actively created: It is true … that if a system of rules is to be imposed by force on any, there must be a sufficient number who accept it voluntarily. Without their voluntary co-operation, thus creating authority, the coercive power of law and government cannot be established.20

But, on the other hand, he emphasised that, once one understands that the realm of legal obligations and the realm of moral obligations are distinct in that there is no necessary connection between them, one can also appreciate that the obligations of the legal realm can be accepted on the basis of ‘many different considerations’: calculations of long term interest; disinterested interest in others; an unreflecting inherited or traditional attitude; or the mere wish to do as others do. There is no reason why those who accept the authority of the system should not examine their conscience and decide that, morally, they ought not to accept it, yet for a variety of reasons to continue to do so.21

This relationship of voluntary cooperation could be confined to legal officials, though Hart seemed unsure how to understand the situations in which this occurred, which he diagnosed variously as pathological, unhealthy, and abnormal. On one account, obedience – the first condition for the existence of legal order is the only one which private citizens need satisfy: they may obey each ‘for his part only’ and from any motive whatever; though in a healthy society they will in fact often accept these rules as common standards of behaviour and acknowledge an obligation to obey them, or even trace this obligation to a more general obligation to respect the constitution.

In contrast, ‘acceptance’ or the ‘second condition must also be satisfied by the officials of the system. They must regard these as common standards of official behaviour and appraise critically their own and each other’s deviations as lapses’.22



17 ibid

(Hart’s emphasis). 293. 19 Hart, ‘Positivism and the Separation of Law and Morals’ (n 6) 53–4. 20 Hart, The Concept of Law (n 5) 201 (Hart’s emphasis). 21 ibid 203. 22 ibid 116–17 (Hart’s emphasis). 18 ibid

184  David Dyzenhaus If, however, the official attitude was confined to officials, this would present ‘an extreme case’: The society in which this was so might be deplorably sheeplike; the sheep might end in the slaughter-house. But there is little reason for thinking that it could not exist or for denying it the title of a legal system.23

Here Hart’s concern seems to be with the displacement of acceptance from the primary rules of a primitive society to the secondary rules of a modern, complex legal order, and from individuals in general to possible confinement to officials.24 With such displacement a habit of obedience to primary rules whatever their content could arise even when the majority find the content objectionable. This habit amounts to a kind of acceptance, since law is not the gunman situation writ large, though it is not the same as the active, critical acceptance that the official sector must display. It is the kind of passive stance of obeying rules whatever their content merely because they are certified as valid that Hart decried throughout his work and which he argued legal positivism, correctly understood, could guard against.25 This situation thus seems to contrast with another where there is widespread disobedience, which he describes as follows: The normal, unproblematic case where we can confidently say that a legal system exists, is just one where it is clear that the two sectors are congruent in their respective typical concerns with the law. Crudely put, the facts are, that the rules recognised as valid at the official level, are obeyed. Sometimes, however, the official sector may be detached from the private sector, in the sense that there is no longer general obedience to the rules which are valid according to the criteria of validity in use in the courts. The variety of ways in which this may happen belongs to the pathology of legal systems; for they represent a breakdown in the complex congruent practice which is referred to when we make the external statement of fact that a legal system exists. There is here a partial failure of what is presupposed whenever, from within the particular system, we make internal statements of law.26

But notice that it does not seem likely that Hart would describe a stable, slave-owning society in which very efficient and cruel sanctions preempted any thought of disobedience amongst the slaves as ‘healthy’ or ‘normal, unproblematic’, even if it were stipulated that it was the legal health of the society that was being described, not its

23 ibid 117. 24 He emphasised, ibid 202, that such ‘pathologies’ are made possible by the move from the simple societies to the more complex legal world legal philosophy sought to explain. While there are immense gains in ‘adaptability to change, certainty, and efficiency’, the ‘cost is the risk that the centrally organised power may well be used for the oppression of numbers with whose support it can dispense, in a way that the simpler regime of primary rules could not’. Hart is clear that in a primitive society that the majority must take the internal point of view: [T]hough such a [primitive] society may exhibit the tension, already described, between those who accept the rules and those who reject the rules except where fear of social pressure induces them to conform, it is plain that the latter cannot be more than a minority, if so loosely organized a society of persons, approximately equal in physical strength, is to endure: for otherwise those who reject the rules would have too little social pressure to fear (ibid 91–2). 25 For an excellent treatment of these issues, see MA Wilkinson, ‘Is Law Morally Risky? Alienation, Acceptance and Hart’s Concept of Law’ (2010) 30 OJLS 441. 26 Hart, The Concept of Law (n 5) 117–18 (my emphasis).

The Constitution of Legal Authority  185 moral health. For it is surely the case that missing in such a society is anything like Hart’s idea of a ‘complex congruent practice’ that he takes to be at the core of a society that has made the step from a ‘primitive’ society into the legal world Hart wishes to explain. Hart, however, neither here nor elsewhere elaborates his understanding of this phenomenon, of, as he also put it, the ‘complex, but normally concordant, practice of the courts, officials, and private persons in identifying the law by reference to certain criteria’.27 And it seems that he thought he could avoid this issue because the ‘existence’ of the practice ‘is a matter of fact’.28 II.  APPROACHING NATURAL LAW?

Joseph Raz argues that because legal officials are in the business of telling subjects what their obligations are, they must give the appearance of allying themselves with the law’s claim to legitimate authority. In other words, the officials – those who ‘use’29 the law – must at least pretend to believe that the laws they enforce are just, even if they do not in fact so believe.30 They act as if they adopt the point of view of the figure Raz names the ‘legal man’. He claims to find this figure in Kelsen: the legal man adopts ‘the law as his personal morality, and as exhausting all the norms he accepts as just’.31 In other words, an understanding of legal orders as authoritative involves understanding their claim to have legitimate authority, which requires the figure of the man who accepts the legitimacy of the legal order because he equates justice with the content the laws happen to have. In turn, the legal theorist or jurist must explain legal order from this figure’s point of view. But Raz thinks that presupposing this figure is perfectly consistent with the claim that legal orders are rarely just, so that in general their claim to legitimate authority fails.32 He seems then to make no concession to the natural law claim that there is an intrinsic connection between law and morality. 27 ibid 110. 28 ibid. 29 See J Raz, ‘Kelsen’s Theory of the Basic Norm’ in Raz (ed), The Authority of Law: Essays on Law and Morality (Oxford, Clarendon Press, 1983) 122, 141, where Raz criticises Kelsen for not seeing the difference between ‘jurists talking about the law, and the activities of lawyers and judges using the law’ (Raz’s emphasis). 30 Or at least this is what I take Raz to be saying in ‘Legal Validity’, ibid 146, 154–5, including fn 13, 155. See HLA Hart, ‘Legal Duty and Obligation’, in Hart (ed), Essays on Bentham: Jurisprudence and Political Theory (Oxford, Oxford University Press, 1982) 127, 155–57, 155: ‘Raz’s final view, after hesitation, seems to be that one necessary condition [if a legal system … is to constitute the law of a particular society] is that there be on the part of judges either belief or at least the pretence of belief in the moral justifiability of the law’. 31 Raz, ‘Kelsen’s Theory of the Basic Norm’ (n 29) 142–3; see also J Raz, ‘The Purity of the Pure Theory’, in SL Paulson and BL Paulson (eds), Normativity and Norms: Critical Perspectives on Kelsenian Themes (Oxford, Oxford University Press, 2007) 237. Raz’s claim to find this figure in Kelsen is not uncontroversial. For a detailed account, see S Paulson, ‘The Weak Reading of Authority in Hans Kelsen’s Pure Theory of Law’ (2000) 19 Law and Philosophy 131. I will deal with this issue in the text below. 32 It may even be the case, though Raz is rather ambiguous on this point, that he thinks that in general legal orders lack authority, not only legitimate authority. See J Raz, ‘Authority, Law, and Morality’, in Raz (ed), Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford, Oxford University Press, 1994) 194.

186  David Dyzenhaus Yet, Raz also says that his argument shows that the ‘gap’ between legal positivists and natural lawyers is not ‘as unbridgeable as is sometimes imagined’ because legal positivism must adopt in somewhat modified form two ‘semantic theses’ usually associated with natural law. First, there is the thesis that ‘normative terms like “a right”, “a duty”, “ought” are used in the same sense in legal, moral and other normative contexts’. The second thesis is that ‘legal statements are moral statements’; to say that someone has a legal duty to repay a debt is to assert a moral duty to repay the debt ‘arising out of the law’.33 Positivists, Raz suggests, are inclined to reject the second thesis, arguing that what the law is is a matter of social fact and its value is always a further question. While if the law is in force there will be some who make what Raz calls ‘fully committed statements’ – statements that endorse the law’s claim to legitimacy – it is possible for others to make the same statements from their point of view, but without endorsing it, just as a knowledgeable Catholic can inform an ignorant Jew about what his religious duties are without endorsing the Jewish faith.34 The jurist, the lawyer, and so on can make statements from the point of view of the legal man while leaving it as a further question whether those statements should be endorsed. That shows why positivist legal theory must accept the first semantic thesis. But it also shows that the second thesis must be accepted, albeit in modified form, because the ‘not committed detached’ statements of the lawyer or jurist are ‘parasitic on the full-blooded normative statements’ of the legal man.35 Recall though that for Raz, officials, including judges, have at least to give the appearance of making the same ‘full-blooded normative statements’, which can only mean that they must give the appearance of morally endorsing the legal order and its particular laws. This claim led Hart to worry that Raz’s analysis of what Hart called ‘authoritative legal reasons’36 threatened to undermine the Separation Thesis.37 In addition, he thought the analysis should be rejected, first, because it is unrealistic – judges need form no views about the moral merits of the law in accepting that they are under a standing duty to apply it, whatever the merits of its content; something he took Raz to recognise in the thought that the judges could be insincere in associating themselves with the legal man. Second, he rejected the claim in so far as Raz’s account presupposes a ‘cognitive account of moral judgement in terms of objective reasons for action’.38 ‘Far better adapted to the legal case’, Hart asserted, is a non-cognitive theory, according to which ‘to say that an individual has a legal obligation to act in a certain way is to say that such action may be properly demanded from him according to legal rules or principles regulating such demands for action’.39 Hart did note a difficulty for his argument. The virtue of Raz’s account, he suggested, is that it responds to the idea that when a subject is stated to have a duty



33 Raz,

‘Legal Validity’ (n 29) 159. 156–7. 159. 36 HLA Hart, ‘Commands and Authoritative Legal Reasons’, in Hart (ed), Essays on Bentham (n 30) 243. 37 Hart, ‘Legal Duty and Obligation’ (n 29) 153. 38 ibid 158–59. 39 ibid 159–60. 34 Ibid 35 ibid

The Constitution of Legal Authority  187 to act contrary to his own sense of his interests, it is assumed that there exist reasons that are objective in that they ‘exist independently of his subjective motivation’. On this assumption, ‘it would be difficult to deny that legal duty is a form of moral duty’ because that denial would seem to entail that there were ‘two independent “worlds” or sets of objective reasons, one legal and the other moral’. And Hart conceded that his non-cognitive account of legal reasons seems committed to the ‘paradoxical’ even confused conclusion that ‘judicial statements of a subject’s legal duties need have nothing directly to do with the subject’s reasons for action’. He also conceded that his account of judicial acceptance had become ‘whittled down’ to the point where it seemed quite different from the idea as he had ‘first introduced it in the model of a simple society whose members accepted a commander’s words as … [authoritative legal] reasons for doing what he commands them to do’.40 This latter kind of acceptance he described as ‘full blooded’ to contrast it with ‘institutionalised “whittled down”’ forms of acceptance. But in neither case ‘need there be, though there may often be, belief in the moral legitimacy of the legislature or the pretence of such belief’.41 Here Hart detected a puzzle for philosophy of law. If one of the features of law that requires explanation is law’s authority, sanction-based accounts of law will not do. Something is needed that attends to the way in which law provides reasons of a special kind to subjects – authoritative legal reasons. Hart’s term ‘full blooded’ does not help address this puzzle. It seems to mean something like ‘vigorous’ or ‘robust’ and is very close to ‘fully committed’ or ‘fully normative’, the terms Raz used in the passages to which Hart objected, and in which he made the argument that a judge must give the impression that the rule he accepts he is under a duty to apply to a legal subject is one that the legal subject is under a duty to follow.42 If authoritative legal reasons were always moral reasons of some kind, their normative force would be more easily explained. But since Hart is not willing to countenance that such reasons always have a moral quality to them, he confessed that he was at something of a loss. Indeed, in a philosophical obituary of Hart, Raz concludes that because Hart’s rule of recognition is to be explained as a matter of the facts of about its existence that includes the fact that criticisms are made of those who deviate from its duties, he put forward a non-cognitivist, ‘sanction theory of rights and duties’.43 In other words, Hart put forward a position of the sort he thought he had rejected in rejecting the command theory of law and what he took to be Kelsen’s version of it.44 The exchange between Hart and Raz does, however, help to illuminate this complexity. Raz’s point can be put as follows. When Hart talks of ‘official acceptance’, he tends to think of the relationship as one between the officials, on the one hand, and the fundamental rules of the legal order, on the other. Acceptance refers to the voluntary cooperation of each official with every other that involves using the secondary rules to produce, determine the content of, apply, and enforce the primary



40 Hart,

‘Commands and Authoritative Legal Reasons’ (n 36) 266–68. 268. 42 Raz, ‘Legal Validity’ (n 30) 146, 153, including fn 13. 43 J Raz, ‘H. L. A. Hart (1907–1992)’ (1993) 5 Utilitas 145, 149. 44 Hart, The Concept of Law (n 5) ch 3, ‘The Variety of Laws’. 41 ibid

188  David Dyzenhaus rules. But these officials do not have to have any regard for the individuals who are subject to their decisions. Once they have determined what the law is, they simply declare it, apply it or enforce it. But, as Hart acknowledged in his response to Raz, his own account then seems committed to the ‘paradoxical’ even confused conclusion that ‘judicial statements of a subject’s legal duties need have nothing directly to do with the subject’s reasons for action’.45 Even more telling is that Hart could not himself consistently maintain this paradoxical view. In The Concept of Law, chapter 9, ‘Laws and Morals’, Hart says that ‘a minimum of justice is realised whenever human behaviour is controlled by general rules publicly announced and judicially applied’. If law is to function as a system of social control, its rules ‘must be intelligible and within the capacity of most to obey, and in general they must not be retrospective …’.46 His Hobbesian argument is that both law and morals have to have a minimum content if they are to ‘forward the minimum purpose of survival which men have in associating with each other’ and that ‘in the absence of this content’ men would have ‘no reason for obeying voluntarily any rules’.47 The point about reason is important because we have to understand law as operating with reasons, not causes.48 These reasons must be addressed at least to the natural facts of human vulnerability, approximate equality, limited resources, and limited understanding and will.49 This ‘natural necessity’, he thus says, has to qualify the ‘positivist thesis that “law may have any content”’.50 Later in that chapter, Hart says that judges ‘may often display’ what he calls ‘characteristic judicial virtues’: ‘impartiality and neutrality in surveying the alternatives; consideration for the interest of all who will be affected; and a concern to deploy some acceptable principle as a reasoned basis for decision’.51 Hart is, I think, deliberately coy about the status of these virtues within his general argument about the necessary minimum content of natural law.52 But I assume that he sees them as the characteristics of the role of a judge who understands himself as operating in a system of authority rather than a gunman situation writ large, and who understands his legal order as catering to the minimum content of natural law he sketched. Now imagine such a judge confronted by a law that requires that a whole group of people be treated inequitably, for example, one of the segregationist statutes of the apartheid regime.53 That judge must – as a matter of legal duty – struggle to find an interpretation that displays both consideration for the interest of all those who will be affected and a concern to deploy some acceptable general principle, derived from

45 Hart, ‘Commands and Authoritative Legal Reasons’ (n 36) 266–8. 46 Hart, The Concept of Law (n 5) 206–7. 47 ibid 193. 48 ibid 194. 49 ibid 194–200. 50 ibid 199. 51 ibid 205. 52 In my view, his coyness is symptomatic of the kinds of issues about the rule of law and legality that arise in his challenges to Lon L Fuller, as depicted by J Waldron, ‘Positivism and Legality: Hart’s Equivocal Response to Fuller’ (2008) 83 New York University Law Review 1135. 53 For a study of this issue, see D Dyzenhaus, Hard Cases in Wicked Legal Systems: Pathologies of Legality 2nd edn (Oxford, Oxford University Press, 2010).

The Constitution of Legal Authority  189 the relevant legal materials, as a reasoned basis for decision.54 It is a struggle because a law that institutionalises inequality on its face does not display such consideration and is hard to interpret in a way that displays some acceptable general principle. That creates a tension internal to the legal perspective of the judge, since the judicial virtues are prominent among the norms that he or she accepts as defining the judicial role. Put most concretely, the judge will have a hard time answering the legal subject who asks: ‘But, how can that be law for me?’ Indeed, in remarks in 1958 that presage this chapter 9, Hart mentions as ‘facts’ that no system which utterly failed to ‘measure up some moral or other standard’ ‘has ever existed or could endure’ and that ‘the normally fulfilled assumption that a legal system aims at some form of justice colours the whole way in which we interpret specific rules in particular cases, and if this normally fulfilled assumption were not fulfilled no one would have any reason to obey except fear (and probably not that) and still less, of course, any moral obligation to obey’.55 He went on to say that if there were not some group that received the benefit of protection from the law, the system would ‘sink to the status of a set of meaningless taboos’ and ‘no one denied those benefits would have any reason to obey except fear and would have every moral reason to revolt’.56 One may paraphrase these passages by saying that if the commands of the powerful are incapable of sustaining a claim to be exercised with right on those within their power, the commands lack authority, and therefore start to lose their grip on their claim to legal status. For such commands undermine the mutual relationship between protection and obedience Hobbes outlined at the end of Leviathan – protego ergo obligo – to which we have already seen Hart is attracted, and which he states as follows, in giving a snapshot in chapter 8 of his argument to come in chapter 9: It seems clear that the sacrifice of personal interest which such rules demand is the price that must be paid in a world such as ours for living with others, and the protection they afford is the minimum which, for beings such as ourselves, makes living with others worth while. These simple facts constitute … a core of indisputable truth in the doctrines of Natural Law.57

Hart’s own treatment of this kind of issue relied on a clear distinction between an in- group that gets protection from the law and an out-group that does not. As I have 54 In ‘Positivism and the Separation of Law and Morals’ (n 6) 81, Hart says that there is ‘in the very notion of law consisting of general rules, something which prevents us from treating it as if morally it is utterly neutral, …’. Generality, Hart says, requires ‘[n]atural procedural justice’ which consists of ‘principles of objectivity and impartiality in the administration of the law’ and ‘which are designed to ensure that rules are applied to only to what are genuine cases of the rule or at least to minimize the risks of inequalities in this sense’. But such concessions do not, Hart claims, undermine the Separation Thesis. The protections do not have to be afforded to everyone, and laws ‘that are hideously oppressive’ can be applied with the ‘most pedantic impartiality’. Fuller’s and Dworkin’s theories of adjudication can, I think, fruitfully be seen as putting pressure on Hart’s suggestion that impartiality has to be pedantic. 55 ibid 79. It has been suggested to me that Hart at this point is setting out the foil for his argument, not his own position. But Hart clearly thinks that these are facts, though their existence does not trouble the substance of the Separation Thesis. Indeed, he thinks he can ‘go further’ than conceding the existence of these facts and still preserve the Thesis, and at 79–82 he sets out the basics of the argument he elaborated in ch 9 of The Concept of Law. 56 Hart, ‘Positivism and the Separation of Law and Morals’ (n 6) 82. 57 Hart, The Concept of Law (n 5) 181 and see text to n 48 above (Hart’s emphasis). I explore the relationship between Hobbes and Hart in D Dyzenhaus, ‘The Public Conscience of the Law: from Hobbes to Hart’ (2015) 45 Ragion Pratica 565.

190  David Dyzenhaus indicated, he thought that the in-group could be small, confined to the officials of the system, and the system would still be a legal system, albeit in a deplorable state.58 Only if there were no one in the in- group at all would the system become one in which there was the appearance of law but what passed for law had sunk to ‘the status of a set of meaningless taboos’. Hart thus did not attend to the important issues that arise for legal philosophy when one is in the in-group for some purposes but not for others; when, that is, an individual has the status of second-class citizenship.59 Yet, it seems fair to say that he must hold the view that the ‘normal’ or ‘healthy’ case of a legal order is one in which power is exercised authoritatively. To be a first-class citizen, or full legal subject, in such an order is to be able to accept that the legal order as a whole and its particular laws offer one the kind of reasons that are understandable as addressing one’s interests. Even when such subjects strongly disagree with the content of the law, they must be in a position to recognise that the legal order as a whole serves the interests that are in the character of a legal order to serve and that its particular laws are interpretable in light of those interests. When they ask the question, ‘But, how can that be law for me?’, a judge mindful of the judicial virtues can give an answer. This line of argument may seem undermined by Hart’s insistence that in any case in which the judge had to display these virtues in order to respond to a legal question raised by the parties, the judge has to exercise an extra-legal discretion as such a question will arise precisely because the positive law determines no answer. Such issues were the preoccupation in the debates in legal theory initiated by Ronald Dworkin in the last 40 years of the last century abut ‘hard cases’ – cases in which there is reasonable disagreement about the answer to the legal question posed by the case. Dworkin sought to show that ultimately at stake in such cases is the question of the political ideal that law serves and that legal positivism, despite the denials of its proponents, gives an answer that informs the way judges who endorse a positivistic ideal decide these cases.60 Because legal positivists responded to Dworkin by asserting that there is a categorical distinction between a theory of law and a theory of adjudication, legal theory has long appeared to be at an impasse.61 But my argument so far helps

58 Hart, The Concept of Law (n 5) 117. Compare 90-1. 59 Indeed, he did not seem to think there was an issue – see his treatment of a segregationist law, ibid 161. This passage at 117–18 is the closest he comes: Hence a society with law contains those who look upon its rules from the internal point of view as accepted standards of behavior, and not merely as reliable predictions of what will befall them, at the hands of officials, if they disobey. But it also comprises those upon whom, either because they are malefactors or mere helpless victims of the system, these legal standards have to be imposed by force or threat of force; they are concerned with the rules merely as a source of possible punishment. The balance between these two components will be determined by many different factors. If the system is fair and caters genuinely for the vital interests of all those from whom it demands obedience, it may gain and retain the allegiance of most for most of the time, and will accordingly be stable. On the other hand, it may be a narrow and exclusive system run in the interests of the dominant group, and it may be made continually more repressive and unstable with the latent threat of upheaval. Between these two extremes various combinations of these attitudes to law are to be found, often in the same individual. 60 R Dworkin, Law’s Empire (London, Fontana, 1986). 61 See L Murphy, What Makes Law: An Introduction to Philosophy of Law (Cambridge, Cambridge University Press, 2014).

The Constitution of Legal Authority  191 to soften that distinction in that it shows that Hart’s legal theory is committed to an outline of a theory of adjudication, albeit not to a substantive theory about how judges should decide such cases. On Hart’s theory, judges should come up with an answer to the legal question that meets two criteria. First, it shows how the law responds to legal subjects in a way that respects their subjectivity – their ability to respond to reasons. Second, the answer rests on a justice-like general principle that is supported by the relevant legal materials, a principle that shows that the legal subjects are such because they are not outside the class whose interests the law protects, even if the protection is rather minimal. My argument does not, however, fully support Dworkin’s position in part because it does not start on the terrain of adjudication, in the thick of exploring rival accounts of how judges do and should decide hard cases. Rather, it starts on Hart’s own terrain of an account of the authority of law. It shows how once one chooses to account for law in terms of the distinction between might and right and from the internal point of view, one inevitably approaches a position more accommodating of a natural law view of the relationship between de facto and de jure authority. Moreover, that position recognises that judges and other legal officials play an important role in mediating the legal relationship between lawmaker and legal subject, or, in political terms, between ruler and ruled. I shall now argue that it matters that the position is understood in political terms, more accurately, in legal/political terms. Such terms have to do with an idea of legitimacy that cannot be reduced to a moral position of the kind that Hart seemed to have in mind when he articulated the Separation Thesis that there is no necessary connection between law and morality. Rather, it has to do with the actual acceptance by legal subjects of the de jure authority of a legal order in a way that constitutes that authority. One might say that there is more to the analogy we saw Hart draw between the rules of contract law and his secondary rules than he recognised. As I shall show, it is important to hold onto Raz’s insight as to how the perspective of the legal subject is implicated in the claim to legitimate authority of the legal order with which judges have to appear to associate themselves. But we need to see that the theses he accepts from natural law about that perspective are not so much semantic as political. Moreover, we should reject his description of that perspective as that of the legal man who adopts ‘the law as his personal morality, and as exhausting all the norms he accepts as just’. Here Kelsen shows the way.62 III.  THE LEGAL MAN VS. THE LEGAL SUBJECT

In the last section, I put forward an unorthodox interpretation of Hart’s theory of law’s authority in which authority is constituted by the acceptance of subjects as well as of officials, and the authority is de jure because of the role of legal officials, 62 I am much influenced in this section by L Vinx’s discussion of the figure he calls the ‘law-abiding citizen’ in Hans Kelsen’s Pure Theory of Law (Oxford, Oxford University Press, 2007) 76, 154–6, 209, and of the need to recast the relationship as one between law and legitimacy rather than between law and a personal perspective on justice; see, ibid fn 50, 41.

192  David Dyzenhaus particularly judges, within the institutional framework of the law in giving answers to the question of the legal subject: ‘But, how can that be law for me?’ In this section, I put forward another (perhaps slightly less) unorthodox interpretation, this time of Kelsen. It is unorthodox because for the most part Kelsen supplies the basis of what a leading Kelsen scholar calls a ‘weak reading’ of authority according to which, as in Hart, laws’ authority seems addressed to legal officials and not legal subjects, though unlike Hart, this kind of authority is a matter of giving the officials permission to apply sanctions to subjects when they fail to comply with the law.63 It is the element of sanctions in Kelsen’s theory that led to Hart’s charge that he is an adherent of a reductive, command theory of law, one which reduces law ultimately to coercion. But as we have seen, Hart also cannot escape the charge, this time put by Raz, that his own theory of authority is ultimately a sanction-based one, though its basis also seems different in that it relies on the informal sanction of criticism that officials face when they deviate from the accepted practice of applying secondary rules, and not the sanctions that apply to subjects for non-compliance with primary rules. And even this difference may not be large. Recall that Hart said of the rules of change that they constitute the legislature as a body with power to do certain things, just as the rules of contract law make it possible for contracting ‘to create structures of rights and duties for the conduct of life within the coercive framework of the law’.64 In the course of his criticism, Raz emphasises that part of the reason that Hart puts himself in this predicament is that, as Kelsen argued, ‘judicial social practices do not confer binding force on the ultimate legal rules …’.65 Put differently, Hart’s claim that it ‘seems needless reduplication to suggest that there is a further rule to the effect that the constitution … is to be obeyed’ is wrong; and even if it seems ‘mystifying to speak of a rule that this rule be obeyed’,66 this is a mystery that legal philosophy has to solve, and which Kelsen at least attempted to do. This attempt requires the ‘strong reading’ of Kelsen’s theory of authority in which the legal subject is front and centre in the authority relationship. It is to be found in Kelsen’s treatments of norm conflicts, and is given expression in the kinds of statements that Raz adduces in evidence for his claim that the legal subject for Kelsen is the ‘legal man’ who adopts ‘the law as his personal morality, and as exhausting all the norms he accepts as just’.67 Hart, unsurprisingly, criticises Kelsen at just these points, in particular, Kelsen’s argument that it is a mistake to adopt dualism when it comes to the relationship between law and morality.68 In response to Kelsen’s claim that it is not possible to

63 For detailed discussion, see Paulson, ‘The Weak Reading of Authority in Hans Kelsen’s Pure Theory of Law’ (n 31). 64 Hart, The Concept of Law (n 5) 27–9. 65 Raz, ‘The Purity of the Pure Theory’ (n 31) 252. 66 Hart, The Concept of Law (n 5) 293. 67 Paulson, ‘The Weak Reading of Authority in Hans Kelsen’s Pure Theory of Law’ (n 31) 156–68. See Raz, ‘Kelsen’s Theory of the Basic Norm’ (n 29) especially 134–5, in which he quotes and cites an abundance of text to support his argument that Kelsen’s conception of normativity is one of ‘justified normativity’ and that this conception is essential to a distinction between might and right. 68 HLA Hart, ‘Kelsen Visited’, in Hart (ed), Essays in Jurisprudence and Philosophy (n 6) 286.

The Constitution of Legal Authority  193 serve two masters when it comes to law and morality,69 Hart says that while this point addresses the situation of conflict in which an individual is subject to duties by two norms he accepts as valid, it does not address the situation of an individual who wishes to make a moral criticism of the law for, say, requiring military service, even though he himself is not liable for such service.70 Hart has other criticisms. But I shall focus on the thought he found ‘alarming’ that Kelsen’s argument ‘excludes the possibility of a moral criticism of law’.71 Hart’s discussion omits two significant details.72 First, Kelsen’s claim about the law/morality relationship is invoked in support of his argument against dualism in regard to the national/international law relationship; and, second, he ends his discussion of the latter by saying that it ‘is by juristic interpretation that the legal material is transformed into a legal system’.73 Kelsen’s emphasis on the role of juristic interpretation in making an order legal is significant. Like Hart, he did not offer a theory of adjudication since such a theory was, he thought, a political theory beyond the scope of legal science, and he also emphasised that at each of the levels of legal activity below the level of legislation there is an irreducible moment of judicial or official creativity. But unlike Hart and his followers, Kelsen denied that there are gaps in the legal order.74 In this respect, Kelsen’s legal theory is closer to Dworkin than it is to the legal positivism of Hart and his students as, on his view, a legal order provides seamlessly for an authorised official to solve by a legal procedure any problem raised within the legal order.75 Indeed, his legal theory goes further than Dworkin’s since it rejects a

69 Hart refers to Kelsen, General Theory of Law and State (n 1) 373–5, 408–10. The discussion of national and international law is at 373. 70 Hart, ‘Kelsen Visited’ (n 68) 302–3. 71 ibid 302. 72 In my opinion, Hart’s other main criticisms misfire. One is that Kelsen’s account of a collision of duties is of a psychological collision, where a person thinks he is faced with conflicting duties when in fact he is not. Hart, ibid 303–5. But Kelsen’s point is that, given that the person is not faced with an actual conflict, the conflict in pulls has to be explained psychologically, not as a matter of normative collision – Kelsen, General Theory of Law and State (n 1) 375. Hart also says that there is no problem in asserting that there is a conflict between two ought statements of the sort ‘there are good reasons for doing A’, for example, ‘there are good legal reasons for doing A and good moral reasons for not doing A’. Hart, ‘Kelsen Visited’ (n 68) 307–8. But the issue for Kelsen is what must be presupposed by the person who has to decide between doing A or not, as Dworkin pointed out in the collection in which ‘Kelsen Visited’ was first published. (Hart’s essay was first published in HE Kiefer and MK Munitz (eds), Ethics and Social Justice (Albany, State University of New York Press, 1968) 171. For Dworkin’s response, see ‘Comments on the Unity of Law Doctrine (A Response)’, ibid 200. In my view, the same point undermines Hart’s other objections. In general, Hart seems to assume that because we can, say, assert without contradiction that the law of country X requires A and the law of country Y requires not A, we can thereby identify a contradiction between two valid systems of norms. But we need to ask that question not from the perspective of an external observer, but from the perspective of a judge faced with a dispute in private international law, who has to decide which system’s norms are valid for the purpose of determining the result. In ‘Kelsen’s Doctrine of the Unity of Law’ 340–2, Hart had something to say about this issue, but his answer exposes even further the problems in his account, as I argue in ‘The Janus-Faced Constitution’ in J Bomhoff, D Dyzenhaus and T Poole (eds), The Double-Facing Constitution (Cambridge, Cambridge University Press, 2020) 17. 73 Kelsen, General Theory of Law and State (n 1) 375. 74 H Kelsen, Pure Theory of Law [1960] 2nd edn (Berkeley: University of California Press, 1967, Max Knight, trans) 245–50. 75 For an illuminating discussion of the similarities between Kelsen and Dworkin, see T Honoré, ‘The Basic Norm of a Society’ in Paulson and Paulson (eds), Normativity and Norms (Oxford, Oxford

194  David Dyzenhaus distinction that Dworkin relies on between legislation – the instrument of policy and the province of the legislature – and adjudication – the realm of principle and the province of the judiciary.76 For Kelsen, the moment of creativity in legislation is much more heightened than in adjudication, but it is quantitatively rather than qualitatively different. Legislation is a legally authorised act that must respect the constitutional norms that govern its production.77 In setting out this position, Kelsen, was primarily concerned with explaining legal order from the internal point of view of one who accepts that the norms of that order are binding, not because of any threat of force, but because they are part of an authoritative order. For if one adopts the internal point of view, one assumes that the answer to any legal question will be produced by legal procedures and will make sense of that individual’s subjection to the material and formal norms of the legal order by displaying them as a unity. One assumes, in other words, that a principle of legality governs, and this assumption shows why Hart was wrong to claim that Kelsen’s theory ‘has a juristic Midas touch, which transmutes all questions about laws and their relationship into questions of the content of law or questions concerning what laws say; but the touch is perverse, for not all questions are of that kind’.78 Rather, Kelsen is mostly concerned to answer questions about what makes an order legal: a polity governed in accordance with the rule of law in the sense that law regulates its own production and all legal problems are solved in a dynamic process of norm application by legal procedures. Moreover, Kelsen at times accepted that the assumption of the unity of legal order introduced a natural law element to his legal theory, thus transcending the limits of a ‘strict positivism’.79 He recognised that the same point had to apply to the assumption when it was made about the international legal order. ‘In this sense’, he said ‘there is absolutely no contradiction between natural law and positive law’, although he also suggested that his theory’s positivistic character was preserved because the content of the legal order would be filled by legislation. Hence, he thought his thesis that law can have any content could remain intact.80 But that thesis must be qualified for reasons quite similar to the qualification we saw Hart inject into his own version of legal positivism in the last part; and this qualification supports a claim that the affinity Kelsen perceived between the idea of a social contract and the basic norm goes deeper than he was willing to recognise, as he University Press, 1999) 89, 109–12. Honoré argues that Kelsen’s basic norm is best understood not as a norm ultimately about the Constitution but as a social norm of cooperation and as such should, where appropriate, influence judges in their decision of hard cases; ibid 31. This argument is, in my view, correct, but (as I explain in the text below) it is much closer than Honoré supposes to Kelsen’s actual argument. 76 R Dworkin, Taking Rights Seriously 2nd edn (Cambridge MA, Harvard University Press, 1978). 77 See in L Vinx, The Guardian of the Constitution: Hans Kelsen and Carl Schmitt on the Limits of Constitutional Law (Cambridge, Cambridge University Press, 2015) 48–9. 78 Hart, ‘Kelsen Visited’ (n 68) 321. 79 Kelsen, ‘Natural Law Doctrine and Legal Positivism’ (n 2) 437. See J von Bernstoff, The Public International Law Theory of Hans Kelsen: Believing in Universal Law (Thomas Dunlap tr, Cambridge, Cambridge University Press, 2010), 116 and M Notermans, ‘Social Peace as Conditio Tacita for the Validity of the Positive Legal Order’ (2015) 34 Law and Philosophy 201, 213–14. 80 H Kelsen, Das Problem der Souveränität und die Theorie des Völkerrechts: Beitrag zu Einer Reinen Rechtslehre [1981] (Aalen, Scientia Verlag, 1981) 252. For an apparent change of mind in this respect, see Kelsen, Pure Theory of Law (n 31) 219.

The Constitution of Legal Authority  195 himself seemed to recognise through indicating at several points that the postulate of unity is not merely formal, since it is also a postulate of peace.81 His argument for the postulate of peace is that every legal order must be effective to some degree before it can be said to be valid. If there were total compliance with its norms, it would be superfluous; too little compliance and it would not exist as a legal order. For it to be effective somewhere in between these two points, Kelsen asserts, it must represent a ‘compromise’ between ‘conflicting interest groups in their struggle … to determine the content of the social order’. In turn, for it to be that compromise, it must be the case that none of these groups is ‘wholly satisfied or dissatisfied’.82 And that is what makes a legal order an order of peace.83 Now on one view of an order as an order of peace, peace is just the absence of conflict. On this view, and in line with the weak reading of authority, Kelsen’s claim that a legal order is necessarily an order of peace could be said to be the result of the fact that a stable or effective legal order is one in which the state’s claim to a monopoly on legitimate coercion holds. But Kelsen has, in my view, something else in mind in his reflection on a legal order as an order of peace, which has to do with the fact that the monopoly on legitimate force is necessary but not sufficient for effective or stable legal order. It is necessary because in every legal order there will be individuals who will conform to law only when they fear being sanctioned if they do not. But it is not sufficient because long term stability depends on the majority of those subject to the law being able to understand their subjection as serving their interests even when they disapprove of the content of some or many laws. Put differently, the compromise between conflicting interest groups is unlikely to be achieved if the content of the law as determined by the relevant officials relegates the individuals on the losing side of the struggle to second-class status. Such individuals will not be able to accept the compromise as a sufficient basis to make sense of their legal order as an order of legal right or authority, rather than one of unmediated coercive power. The kind of compromise that sustains a legal order is, then, not a compromise on any terms whatsoever –one that permits the law to have any content. It is a compromise that maintains the individuals who are affected by the law in their status as legal subjects.84 81 For the most striking formulation see H Kelsen, Law and Peace in International Relations – The Oliver Wendell Holmes Lectures, 1940–41 (Cambridge MA, Harvard University Press, 1948) 1. In my view, this thought makes Kelsen along with the American legal theorist Lon L Fuller the contemporary end point of what I call ‘the long arc of legality’, an arc which has its starting point in the legal theory of Thomas Hobbes. 82 Kelsen, ‘Natural Law Doctrine and Legal Positivism’ (n 2) 438–39. 83 See further Notermans, ‘Social Peace as Conditio Tacita for the Validity of the Positive Legal Order’ (n 79) 213–14. 84 Kelsen does say that ‘[o]nly one group may be interested in “peace,” namely, the one whose interests are better preserved by this order than those of other groups. These other groups may also fail to violate the order. They may maintain the state of peace not because they consider it just, but because, in view of their own weakness, they must be satisfied with the minimum of protection which this order affords to their interests’; ‘Natural Law Doctrine and Legal Positivism’ (n 2) 441. But while his point is that one should make too much of the fact that a legal order is an order of peace, if it is a legal order it will be maintained in part by the weaker group on the basis that it does protect their interests, albeit minimally. See further H Kelsen, ‘The Law as a Specific Technique’, in H Kelsen (ed), What is Justice? Justice, Law and Politics in the Mirror of Science (Berkeley, University of California Press, 1957) 231, 238: ‘The peace of the law is not a condition of absolute absence of force, a state of anarchy; it is a condition of a force monopoly of the community. … A community, in the long run, is possible only if each individual respects certain

196  David Dyzenhaus This point can be put as follows, adapting Hart’s formulation above: the voluntary relationship that constitutes authority is not constituted exclusively by the officials who make, interpret, and implement the law. It is also constituted by the acceptance of those to whom the law is applied – the legal subjects – as well as by the officials who accept the norms that regulate law-making, law interpretation and law-application.85 And this factor introduces into the legal order what we can think of as the perspective of the individual legal subject, one which Hart almost always portrays as a perspective entirely external to law and deployed as a resource to criticise the content of the positive law. The very fact that law made at the higher levels of the legal hierarchy must be progressively concretised as it journeys down to the moment of actual application to a particular legal subject entails that the officials must take the subject’s perspective into account if they are to be faithful to their role. This perspective will become altogether explicit when the legal order institutionalises mechanisms of review of and appeal against the interpretations of the law adopted as the basis for official implementation of the law. But even in the absence of such mechanisms, officials must take into account this perspective because their own internal point of view is incomplete without it. As Raz argues, officials cannot make the distinction between might and right without assuming that it holds for legal subjects. More precisely, their internal point of view is dependent on whether, in applying the law to legal subjects, they apply it by right from the perspective of those subjects. Of course, if no mechanisms of review or appeal are available to subjects, there is no way within the legal order for testing the claim that the law is so applied, other than by observing that subjects have what Austin called the habit of obedience.86 And just as the internal point of view is incomplete without this perspective, so a legal order is incomplete without such mechanisms, as it is incomplete without mechanisms of enforcement for use against those who wish to treat the legal order not from the internal point of view, but as a system of unmediated coercion – as a gunman situation writ large. It is incomplete in the first respect because it does not permit the vindication of the internal point of view. It is incomplete in the second respect because it fails to guarantee to those who take the internal point of view that their practice will not be undermined by those who regard the law as nothing more than commands backed by threats. I want then to suggest that the inclusion of this perspective in even its most austere form is the beginning of constitutionalism. That inclusion is part of an attempt to understand legal order from the position of a legal subject, one who wishes to make interests – life, health, freedom, and property of everyone else – that is to say, if each refrains from forcibly interfering in the sphere of interest of the other.’ See also Notermans, ‘Social Peace as Conditio Tacita for the Validity of the Positive Legal Order’ (n 79) 203 fn 9: ‘for Kelsen the peace of the law is not an “anarchical” order of absolute absence of force, nor a “utopian” state completely free from violence or conflict between individuals or groups; the legal order is rather a socially approved (be it “democratically” or “autocratically”) state of compromise, in which the use of force is monopolized just to prevent, reduce and counteract individual violence and social conflict’. 85 See Wilkinson, ‘Is Law Morally Risky? Alienation, Acceptance and Hart’s Concept of Law’ (n 25). 86 See Lecture VI of the first volume of J Austin, Lectures on Jurisprudence or The Philosophy of Positive Law 2nd edn (London, John Murray, 1861) 168. In general, Austin’s analysis of this idea is far more sophisticated than Hart’s cursory treatments of his theory suggest.

The Constitution of Legal Authority  197 sense of the norms of her order as valid or binding on her – as norms she obeys because she accepts their authority, not because she fears the sanctions that would attend non-compliance. Her obedience together with that of other like-minded subjects is an essential component of what we saw Hart call the ‘complex congruent practice’ of acceptance that constitutes the authority of a legal order. However, if the legal subject is the ‘legal man’ who adopts ‘the law as his personal morality, and as exhausting all the norms he accepts as just’, as Raz describes him, the introduction of this perspective into an account of legal order will seem to presuppose exactly Kelsen’s monism about law and morality that Hart found so ‘alarming’, and which led him to express disquiet about Raz’s argument.87 But, on closer inspection, Raz’s legal man should not perhaps be so much a cause for alarm as for puzzlement. As JW Harris pointed out, the analogy Raz draws between the legal man and the situation of a non-believer in a religion who adopts the point of view of a believer in order to explain to him his religious duties that in the case of the non-believer is unconvincing. With the non-believer, there is a real point of view which he can adopt, whereas the ‘legal man’ does not exist. … [He] has, by assumption, no extra-legal moral system on which he bases the normativity of the law; for if he did there could be no guarantee that uniform moral force would be transmitted to every conceivable legal enactment. … So far as substantive moral grounding is concerned, the hypothetical ‘legal man’ stands in the air.88

Harris’s analysis of what he calls Kelsen’s ‘pallid normativity’ can though point us in a different direction which may help to solve the puzzle. Kelsen was, of course, fully aware that, as Hart put it in his criticism of Kelsen’s rejection of dualism, ‘No human being is just a lawyer or just a moralist. Some at least think about both moral and legal norms and consider their meaning as norms and find that they conflict’.89 But, as Kelsen could have retorted, this is not his but Hart’s problem: Hart cannot explain why this individual should experience any normative conflict in the first place, given his general view that legal norms do not of themselves have any morally binding force.90 In contrast, the question for Kelsen is precisely how to make sense of legal order from the perspective of an individual faced with a conflict between her ‘personal morality’ – the moral norms she herself holds dear – and the norms of her legal order, which she also accepts as valid. Moreover, in the perspective of a legal theory that aims to understand law’s authority, this individual is the legal subject, and so does not stand on air, faced with a naked conflict between the norms of two valid orders. Rather, she stands on the ground of someone who is already subject to the rule of law – to the public order of law in which she happens to find herself. She must 87 Hart, ‘Legal Duty and Obligation’ (n 29) 153–61. 88 JW Harris, ‘Kelsen’s Pallid Normativity’ (1996) 9 Ratio Juris 94, 112–13 (my emphasis). 89 Hart, ‘Kelsen Visited’ (n 68) 306 (Hart’s emphasis). 90 As Lon L Fuller pointed out many years before in a different context – his debate with Hart about the Grudge Informer Case: ‘I hope I am not being unjust to Professor Hart when I say that I can find no way of describing the dilemma as he sees it but to use some such words as the following: On the one hand, we have an amoral datum called law, which has the peculiar quality of creating a moral duty to obey it. On the other hand, we have a moral duty to do what we think is right and decent. When we are confronted by a statute we believe to be thoroughly evil, we have to choose between those two duties’; L Fuller, ‘Positivism and Fidelity to Law: A Reply to Professor Hart’ (1958) 71 Harvard Law Review 630, 656.

198  David Dyzenhaus then attempt to make sense of the presumption that she should subject herself to the norms of that legal order even when these conflict with her personal morality. This is a kind of relativism.91 The legal subject accepts that the modern legal order is legitimate even when some or many of its norms do not correspond with her sense of justice, and so she should regard particular laws as binding on her whatever their content. She thus in a sense relativises her convictions in regard to both the convictions of her fellow subjects and the public judgments that are established as mandatory by the law. The idea here has something in common with Kelsen’s claim that law can have any content as well as with Hart’s account of the two features of the reasons law provides to its subjects – ‘authoritative legal reasons’.92 The first feature is that such reasons are ‘peremptory’ – that a commander ‘characteristically intends his hearer to take the commander’s will instead of his own as a guide to action and so to take it in place of any deliberation or reasoning of his own’.93 The second feature is that such reasons are ‘content-independent’ – they are ‘intended to function as a reason independently of the nature or character of the actions to be done’.94 It differs, however, from Hart’s account in respect of a feature that for Hart, and Raz following him, is implicit in the second – that in order to function as a content-independent reason the content of the law has to be determinable as a matter of fact, that is, ‘without using moral arguments’.95 As Lon L Fuller pointed out against legal positivism, a ‘statute [should be] … seen, not as a message addressed into a void, but as a message whose meaning is dependent upon the interpretation its addressee, as a reasonable and sensible man, would put on it’.96 Put differently, and somewhat paradoxically, in order for an authoritative legal reason to function as a content-independent reason its content must be both determinable and determined in such a way that the legal subject can make sense of her subjection to it. The clue that provides the start to the solution to the puzzle lies in Kelsen’s political theory. Kelsen’s first short but rich work on the nature of democracy reveals that the ground of the legal subject is very particular. It is the ground of the individual who, however sure she is of her moral convictions, knows that in order to live on peaceful terms with others equally sure but of very different convictions, she has to find with those others some common standards for public life despite their disagreements. That gives the individual reason to accept that it is worth living under the political order of democracy in the knowledge that there is no guarantee that the content of individual enacted laws will correspond with those convictions.97

91 See J Raz, ‘Legitimate Authority’, in Raz (ed), The Authority of Law (n 29) 3, 10–11, where he sketches in order to reject the idea of ‘relativized authority’, a way of referring to ‘what those people or societies accept or propose as legitimate authority without endorsing those views’. 92 Hart, ‘Commands and Authoritative Legal Reasons’ (n 36) 244, 252–3. 93 ibid 252–3. 94 ibid 254. 95 J Raz, ‘Legal Positivism and the Sources of Law’, in Raz (ed), The Authority of Law (n 29) 37, 47–8. 96 LL Fuller, ‘The Justification of Legal Decisions’, in Die juristische Argumentation, ARSP NS 7 (Wiesbaden, 1972) 78. 97 H Kelsen, The Essence and Value of Democracy [revised edition of 1929] (Nadia Urbinati and Carlo Invernizzi Accetti, eds, Brian Graf, trans, Plymouth, Rowman & Littlefield, 2013) 103–5.

The Constitution of Legal Authority  199 Kelsen argues further that a properly functioning democracy will govern in accordance with what he calls a ‘principle of legality’ – that all political decisions that will be backed by the state’s monopoly on legitimate coercion will be put into legal form, and so will enter as legal norms into the structured hierarchy of the legal order.98 As democratically produced norms, they carry with them whatever legitimacy accrues to norms merely because the majority has decided in their favour as a result of a political contest in which it prevailed over the opposition of what turned out to be the minority. That legitimacy can only be preserved, Kelsen argues, as long as the decision respects the conditions whereby the minority is itself preserved, so that not only is it capable of opposition in some future contest, but also has the prospect of prevailing.99 But democratic or political legitimacy is not the only legitimacy that these norms have. They also accrue legal legitimacy, gained through the transformation of might into legal right that accompanies their conversion from political into legal norms. Successfully converted norms are norms capable of being concretised by legal officials in ways that display the norms’ identity with the legal order understood as a meaningful whole. The basic norm has to be presupposed in order to make sense of this idea. This is not a mere logical requirement, necessary to make conceptual sense of the fact that there is a valid or binding set of norms. It is also deeply pragmatic in nature. It is an achievement of constructive juristic activity; and such achievement will depend on whether the material content of particular norms can be interpreted so as to display this identity. It is best described as a regulative assumption: an assumption that we need to make in order to make sense of our practice, but also one which we have to maintain by ensuring that our practice conforms with it. The more the legal order is not an order of peace in which subjects can recognise that the law protects their interests, the harder it will be to achieve this goal, and the more precarious the order will become. This factor sets both a sociological and a normative outer limit on legal order. The limit does not dictate what content the order should have. Rather, it marks not very precisely a border that is crossed when a law or set of laws puts a class of legal subjects beyond the pale. The limit is thus more negative than positive in nature, suggesting what kinds of laws will subvert the basis of legal order. It is not, however, wholly negative. Officials who exhibit the kinds of virtues on Hart’s list of judicial virtues, who adopt what Fuller called an ‘ideal of fidelity to law’,100 have a positive duty to ensure that a law that might on a literal interpretation breach this limit is interpreted and applied in a way that preserves to the greatest extent the subject status of those affected by the law. If they are unable to come up with such an interpretation, the law will be suspect as law no matter how well it complies with the formal criteria of validity in that legal order. The limit also does not dictate the institutional make-up of the legal order. But it does provide a kind of internal imperative to reform legal order in such a way as to



98 ibid 99 ibid

ch 7, ‘Administration’. 103. ‘Positivism and Fidelity to Law: A Reply to Professor Hart’ (n 90).

100 Fuller,

200  David Dyzenhaus make it better able to live up to the aspiration of legality; to deal, for example, with the category of laws that are legally suspect despite their formal validity. Put differently, to the extent that a legal order lacks the institutions necessary to implement, interpret, and enforce the law, it will be defective legally speaking. From the perspective of the legal subject, there is a dynamic inherent to legal order that requires that she be able to raise and get an answer to questions of the form, ‘But, how can that be law for me?’ But just as there is an institutional transformation in the transition from what Hart calls a primitive order to a complex legal order, so there is a transformation in the internal point of view. Raz’s description of the legal man applies only to the point of view of the individual of a primitive order. As Hart and Kelsen both pointed out, in such an order the individual regards the public norms as directly binding in that there is a complete coincidence between the individual’s sense of justice and these norms. More precisely, the individual has no sense of justice apart from the content of these public norms and so for him the issue of the legitimacy of the law does not arise. Indeed, for Hart it seems that there is no law present – only generally accepted public norms.101 The issue of the legitimacy of law can arise, then, only in a context in which at any time at least some individuals will find themselves subject to public legal norms whose content they find unjust. Such subjection involves what Kelsen in his work on democracy calls the ‘torment of heteronomy’: the pain that accompanies conceiving of oneself as an autonomous individual who can and should decide for herself how best morally speaking to live, but who accepts that she has reason to abide by mandatory public norms that she would not herself choose if she had the power to do so and which may even strike her as unjust.102 Democracy can alleviate the torment and help to explain why she should have this stance of acceptance because it preserves for her the equal chance of participating in making such norms. There is both a formal and a substantive aspect to such preservation. The formal aspect is that the political process is as open to her participation as it is to everyone else in the political community. The substantive aspect is that the maintenance of this openness recognises her status as a free and equal political subject. 101 See, eg, Kelsen, ‘Natural Law Doctrine and Legal Positivism’ (n 2) 422–3; Hart, The Concept of Law (n 5) 91–2. Note that Hart’s and Kelsen’s views about the transition from primitive to full-fledged legal orders are very different. Kelsen thinks that primitive legal systems are systems and have basic norms, but lack centralised organs of application and enforcement. Hart seems to argue that what is lacking in the regime of primary norms is not just his secondary rules but law itself; hence, it would be a misnomer to call someone a ‘legal man’ in a primitive order. (See further, Raz, ‘The Institutional Nature of Law’ in The Authority of Law (n 29) 103, 105.) Hart’s view is forced on him because there cannot be legal duty in his theory in the absence of a rule of recognition, whereas for Kelsen the kind of institution Hart has in mind is important to vindicate the basic norm, but not necessary for it. This difference lies behind Hart’s difficulties in accounting for the legal status of international law in ch 10 of The Concept of Law. Kelsen’s view is to be preferred for reasons I set out in ‘Hobbes on the international rule of law’ (2014) 23 Ethics & International Affairs 53. See further T Nardin, Law, Morality, and the Relations of States (Princeton NJ, Princeton University Press, 1983) 139–48. 102 Kelsen, The Essence and Value of Democracy (n 97) 27, where the translator gives ‘agony of heteronomy’ as the translation. For a very careful account of how Kelsen’s legal subject differs from Hart’s and Raz’s conception, see Vinx, Hans Kelsen’s Pure Theory of Law (n 62) especially 47–77.

The Constitution of Legal Authority  201 But, as I have already indicated, that the norms decided on in the political process must enter into the legal order before they can be concretised, applied and enforced also plays an alleviatory role with equivalent formal and substantive aspects. Here the formal aspect has to do in part with the fact that an answer to the question ‘But, how can that be law for me?’ is one that will be produced in accordance with legal procedures. But it also has to do with the formal criteria that have to be met for a public policy to be converted into law to take, that is, legal form, for example, generality.103 Substantively, the content of the norm must be interpretable as consistent with the content of other relevant substantive norms of the legal order. And since that content will be an order of peace – one that reflects a compromise of interests between the different social groups subject to law – if any particular norm is to support rather than undermine the postulate of unity, it must be interpretable as maintaining the subject status of those subject to it. Just as the democratic political order must maintain the political subject status of those who end up forming the opposition to a particular law, so the particular laws of a legal order must maintain the legal subject status of those subject to them. The distinction between these two statuses should not, however, be exaggerated. The legal subject status is not apolitical.104 The very conception of the individual as a legal subject is political in that it is formulated in order to explain why it might be rational for an individual to subject herself to the public norms established in a legal order. And it is intrinsically connected to the status of the political subject of democracy. This is the case even when one notes that some proponents of social contract theory, notably Hobbes, thought that democracy is only one of the forms of government consistent with this claim; a point which we saw Kelsen emphasised in his eventual dismissal of contract theory. For while one can try to imagine an autocratic political order in which the leader governs in accordance with the rule of law, it is unclear why the leader would accept this kind of constraint, even whether it is really accurate to call this political order autocratic.105 The leader’s submission to this kind of constraint creates a tension within the autocratic system since he thereby accepts that all his political judgments have to be put into proper legal form and so be interpretable as answering the subject’s question ‘But, how can that be law for me?’ He is no longer, then, a tyrant or autocrat, a natural person with more power than anyone else who uses the form of law when and only when it is convenient to exercise his political judgment that way. Rather, he is an artificial person, constituted by law,

103 These criteria get their most elaborate formulation in the work of LL Fuller, The Morality of Law revised edn (New Haven: Yale University Press, 1969) ch 2. I discuss these different aspects of formalism in public law in D Dyzenhaus, ‘Process and Substance as Aspects of the Public Law Form’ (2015) 74 CLJ 284. 104 This point was very clearly made by the Nazi lawyer E Forsthoff, ‘Der Formalismus in Öffentlichen Recht’ (1934) 4 Deutsches Recht 347, 347–49. 105 See J Finnis, Natural Law and Natural Rights (Oxford, Clarendon Press, 1980) 273: a ‘tyranny devoted to pernicious ends has no self-sufficient reason to submit itself to the discipline of operating consistently through the demanding processes of the rule of law, granted that the rational point of such self-discipline is the very value of reciprocity, fairness, and respect for persons which the tyrant, ex hypothesi, holds in contempt’ (Finnis’s emphasis).

202  David Dyzenhaus and relates to his subjects through law and thus in a way that makes his judgments answerable to them. One may say that constitutionalism is set in motion. If the motion continues, the direction is democracy because of the combination of two ideas: that the acceptance by legal subjects is constitutive of the authority of the legal order and that the judgments made by public officials must thus be answerable to the subjects. Any form of political rule that is undemocratic is in tension with this combination. And here we have, in my view, the reason why early modern writers, notably Hobbes, who were concerned about the dangers they saw in democratic rule could find only pragmatic arguments tied to their contexts to support the claim that political power is best concentrated in the hands of a monarch. The same reason explains why they at the same time insisted that the monarch enjoyed something like democratic legitimacy because he got his authority from a social contract between the individuals who were transformed by it into his subjects.106 But the other direction is always a practical possibility and even when a democratic order is in place, there is no guarantee that either the political or the legal subject statuses will be maintained. One way to try to secure such maintenance is to put in place an entrenched Bill of Rights – both political and individual – and to give to a formally independent judiciary the authority to invalidate laws that it regards as violating the rights. In other words, one puts in place the kind of constitutional order that is nearly always presupposed in discussions of constitutionalism. But it is clear that no guarantee comes with even this kind of constitutional dispensation. In addition, it is not at all obvious that legal orders in which there is parliamentary supremacy and an unwritten constitution do a worse constitutional job of maintaining these statuses. For this reason, legal theory should be agnostic on the question whether this kind of constitutional judicial review is necessary for maintaining the authority of legal order. But there is widespread agreement in constitutional theory that it is a requirement of legal authority that in every national legal order there should be a body of officials, a judiciary, independent of those who implement and enforce the law, who have the role of ensuring that the latter act within the limits of the law, with such limits understood as both formal and substantive.107 Such agreement strongly supports a claim that this kind of review authority is a necessary condition for the maintenance of legal authority.108 It is also usually taken

106 See K Hoekstra, ‘Early Modern Absolutism and Constitutionalism’ (2013) 34 Cardozo Law Review 1079; L Foisneau, ‘Sovereignty and Reason of State: Bodin, Botero, Richelieu and Hobbes’ in HA Lloyd (ed), The Reception of Bodin (Leiden, Brill, 2013) 323; D Lee, Popular Sovereignty in Early Modern Constitutional Thought (Oxford, Oxford University Press, 2016); B Straumann, Crisis and Constitutionalism: Roman Political Thought from the Fall of the Republic to the Age of Revolution (Oxford, Oxford University Press, 2016). 107 Indeed, in Kelsen’s view, the first step from a primitive order to a more complex, indeed properly functioning legal order, takes place with the establishment of courts and so he emphasises the ‘surprising fact that the centralization of the law-applying function – that is, the establishment of courts – precedes the centralization of the law-creating function – the establishment of legislative organs. Long before parliaments as legislative bodies came into existence, courts were established to apply the law to concrete cases’; H Kelsen, Peace Through Law (Chapel Hill, University of California Press, 1944) 21. 108 Hart’s and Kelsen’s disagreement, noted above, about whether a primitive order is a legal order will affect their views on whether this body is necessary for there to be legal order at all. Hart’s answer would be ‘yes’ and Kelsen’s ‘no’, which is why Kelsen has no trouble for accounting for international law’s status

The Constitution of Legal Authority  203 as a necessary condition for the maintenance of the rule of law. For when a legal official makes a decision affecting the rights and interest of a subject, the subject has no way of ensuring that the decision is according to law – that it is law rather than personal whim that lies behind the decision – if she is unable to get an answer to her question, ‘But, how can that be law for me?’ from an official independent of the first. However, from a legal theory perspective, more important than the content of the actual debates about appropriate constitutional design is that the subject matter of these debates is about how to realise the project of constitutionalism set in motion by the very attempt to understand legal order as a matter of authority, which is to say from the point of view of the legal subject. As I have argued, this person is not Raz’s airborne legal man, but rather the legal subject – someone with her feet planted firmly on the ground. I have also suggested that this ground is very particular. One might say that it is political rather than moral ground so that the reasons that count in its space are more political than moral and that it is both sociological – based in acceptance in fact – and normative. In the next and final section, I shall try to provide the basis for an elaboration of these claims. IV.  ACCEPTANCE, LEGITIMACY, AND THE SOCIAL CONTRACT

The claim above that the ground is political more than moral so that the reasons that count in its space are more political than moral is intended to indicate an important but not hard and fast distinction. I have suggested that the Anglo-American debate in legal theory over the last 50 years or so has been shaped by a dualism about law and morality in which it is assumed that the key issue for philosophy of law is whether a necessary connection can be demonstrated between the content of the positive law of any legal order and the personal morality of individuals.109 Left largely unaddressed in this debate have been the questions – ‘Which individuals’ personal morality is at stake and what amounts to such a morality?’ At times, legal positivists have suggested that the answers to such questions do not matter as the Separation Thesis can be shown to hold for any set of values one chooses. But when Hart put forward his position he usually assumed that the values are the values of the concerned liberal citizen, anxious to maintain a critical distance from the law so that he can bring to bear his moral judgment on the question of as law while Hart has considerable trouble. But the point in the text is not about legal order as such but about the maintenance of legal order. 109 That Hart made at times many concessions that seem to undermine the Separation Thesis is evidence of the tension in this regard in his overall theory, and not, as some of his followers claim, against his avowals, that he never proposed this thesis. Raz often moves in this direction as well, for example, his essays ‘The Inner Logic of the Law’ and ‘The Politics of the Rule of Law’, both reprinted in Ethics in the Public Domain (n 32) 222 and 354, or ‘The Rule of Law and its Virtue’ in The Authority of Law (n 29) 210. (For illuminating analysis of the last, see K Rundle, ‘Form and Agency in Raz’s Legal Positivism’ (2013) 32 Law and Philosophy 767.) In addition, see Raz’s discussion of legitimacy in his essay ‘On the Authority and Interpretation of Constitutions: Some Preliminaries’ in L Alexander (ed), Constitutionalism: Philosophical Foundations (Cambridge, Cambridge University Press, 1998) 152 at times moves his theory in the direction my chapter suggests.

204  David Dyzenhaus how to react to the law, unclouded by natural law illusions of a necessary connection between law and morality.110 Raz introduced a different dimension with his theory of authority, according to which a directive is authoritative only if it reflects the objective reasons that already apply to the subject. It can do so by solving coordination problems – which side of the road to drive on – or transmitting expert judgments in the form, for example, of environmental or public health regulations, or by reflecting the moral reasons that apply to the subject in any case, but which the authority is better at getting right than the subjects would if left to their own devices. On this view, law is a transmission belt for objective reasons and its authority depends not on its being law, but on (at least generally) getting the reasons right. But, as we saw, in Hart’s legal theory, law has authority whether or not it is liberal in content, whereas for Raz the law merely claims authority, and so when it pronounces on moral matters it will have authority only when its demands happen to coincide with the demands of liberal morality. It is difficult then to understand how Raz’s position differs from the anarchist one that Hart saw as one of the two pitfalls inherent in natural law theories that assert a necessary connection between law and morality.111 But the alternative Hart offers, as we saw him acknowledge, is implausible. While law, in his view, has authority, and does not merely claim it (as in Raz), the idea of right inherent in law is normatively empty. It is so because when Hart discusses the Separation Thesis and its advantages, he equates morality with the stock of liberal moral values on which an individual will draw in criticising the law. That same equation is to be found in Raz’s ‘normal justification thesis’ – that in moral matters a directive has authority when and only when it reflects reasons that already apply to the subject of the directive.112 On the one hand, Hart’s and Raz’s legal theory of legal authority is that it is de facto authority, based on the fact of acceptance of a certain set of institutional arrangements for making, applying, interpreting, and enforcing mandatory public norms. As such it is normatively inert.113 That is why, on the other hand, Hart and Raz are committed, willy-nilly, to an anarchist theory according to which law becomes normatively active only when it has the content it should have according to liberal morality, in which case it is morality not law that is the activating agent. Both fear that if they were to concede any normative force to law as such, they would be forced to embrace the reactionary horn of the dilemma that Hart took to be the problem

110 See GJ Postema, Legal Philosophy in the Twentieth Century: The Common Law World (Heidelberg, Springer Dordrecht, 2011) Vol 11 of A Treatise of Legal Philosophy and General Jurisprudence, 339–40. 111 In retrospect, it is difficult to understand why Raz thought he had managed to answer comprehensively the challenge posed by RP Wolff, In Defense of Anarchism (New York, Harper & Row, 1970). See, eg, Raz, ‘Legitimate Authority’ (n 91) 11–10, 25–27. For it may turn out that Raz’s cases of unproblematic legitimate authority, eg solving coordination problems and making judgments in certain complex technical areas, are not the issues that Wolff had in mind, and those where Raz thinks that there is unlikely to be legitimate authority are the issues that troubled Wolff. 112 Raz, ‘Authority, Law, and Morality’ (n 32) 198. 113 This is J Gardner’s useful and telling phrase in ‘Legal Positivism: 51/2 Myths’, in J Gardner, Law as a Leap of Faith (Oxford, Oxford University Press, 2012) 19, 23.

The Constitution of Legal Authority  205 of natural law theories; and their fear then drives them in their different ways to the other anarchist horn.114 Raz’s position is, of course, more nuanced in that the subject who rightly accepts law’s authority does so not only because the law reflects reasons that in any case apply to him, but also because he knows that those who make the law are generally better at making such judgments than he himself. The latter is the ‘second order’ reason for not deciding for himself on the content of certain ‘first order’ reasons, those on which the law of his legal order has pronounced.115 The idea of a reason to submit oneself to reasons not of one’s own determination brings Raz’s subject of authority closer to Kelsen’s legal subject. But Raz’s position remains highly individualist in that it is an individual question for me whether I should defer a decision to be taken in light of the reasons that apply to me individually (which may not be the same as those that apply to you). So there is a sense in which the decision of whether to defer or not seems abstracted from politics or membership in a political community and from the fact that many of the factors that play into the decision stem from such political embeddedness. It is, one might say, a moral rather than a political conception of the subject, which is why when Raz describes what he takes to Kelsen’s legal man, the description is entirely in terms of the man’s personal morality. Kelsen, in contrast, has (at least on my argument in the last section), a political rather than moral conception of the legal subject. It includes the subject’s personal morality, while explaining why the subject might find reason to subject herself to laws that not only fail to reflect that morality, but also should be rejected from the standpoint of liberal moral theory.116 The main reason is the constitutionally-mandated, dynamic or constructive juristic activity that applies general norms to the concrete situation of the legal subject in a way that preserves that individual’s subject status. As we have seen, Hartian legal positivism consigns such activity to an extra-legal, political realm.117 In so consigning it, this kind of legal positivism not only manages to evade the responsibility of offering a theory of the role of interpretation in constructing and maintaining legal order, but also that of developing a theory of the role of constitutionalism in performing the same tasks. But, as I have tried to show, the evasion could not be total as long as Hartians took it as incumbent on philosophy of law in general to offer an account of law’s authority, one that does not reduce the explanation of law’s bindingness to coercion. Hart’s idea of the internal point

114 For a similar criticism, see J Weinrib, Dimensions of Dignity: The Theory and Practice of Modern Constitutional Law (Cambridge, Cambridge University Press, 2016) 41–7. I differ from Weinrib in that, following Kelsen, I propose a political rather than a moral solution to the dilemma. 115 See J Raz, Practical Reason and Norms (Princeton NJ, Princeton University Press, 1990) 39–40. 116 Here it is instructive to compare a recent and ongoing debate in political theory initiated by self-styled Realists who argue that the fact of acceptance of a political order, whatever the content of its laws, is the basis of legitimacy. See especially B Williams, ‘Realism and Moralism in Political Theory’ in B Williams, In the Beginning was the Deed: Realism and Moralism in Political Argument ((Princeton, Princeton University Press, 2005, Geoffrey Hawthorn ed) 1. 117 ‘Inclusive’ legal positivism sought to remedy this problem. But it seemed to suggest – rather improbably – that judges have no discretion when and only when the law happens to incorporate liberal norms and in the case at hand such norms in fact provide through correct moral argument an answer.

206  David Dyzenhaus of view and his short excursion into natural law sow the seeds of constitutionalism and undermine the dualism of the Separation Thesis. If these seeds were permitted to sprout, Hartian legal positivism would have to move even further away from the command theory of law than Hart ever contemplated. However, I do not want to suggest that the terminus of the move would be either a full blown liberal theory of how judges should decide hard cases or the conventionalist or the fact-driven theory to which Dworkin claimed Hartian positivism is committed.118 Indeed, it seems to me that the way in which Dworkin’s challenge to Hartian legal positivism established the terrain for legal philosophical debate in English had several unfortunate consequences for philosophy of law. First, Kelsen, perhaps the most important figure in philosophy of law of the last century, was and remains largely marginalised. Second, the challenge led to an impasse. Some have tried to break this impasse by resurrecting versions of an Austinian, sanction-based theory of law.119 But they do so at the cost that Hart had vividly pointed out in 1958: law is seen as the gunman situation writ large. Others are putting forward a theory in which law matters only when, according to some full-blown independent moral theory, it makes a positive moral impact on its subjects.120 Here the cost is that legal philosophy ceases to be interested in what is distinctive about legal order. I have suggested that a return to Kelsen could locate the debate on more productive terrain than that of theory of adjudication vs. theory of law – the terrain of constitutionalism. On that terrain, one can properly address the topics of how law’s de jure authority gets constituted by the acceptance in fact of those subject to it and how that acceptance in turn conditions the shape of the legal institutions that make up the legal order as well as the content of its laws. That acceptance does not quite amount to a social contract, in either the sense of the actual contract which Hume is generally thought to have successfully debunked, or a hypothetical social contract, which it is thought to have met the same fate as the actual one, this time under fire from Dworkin.121 Rather, it partakes of elements of both the ideas of actual acceptance and the acceptance of the hypothetical reasonable person.122 118 Dworkin, Law’s Empire (n 60). 119 Eg, Fred Schauer, The Force of Law (Cambridge MA, Harvard University Press, 2014). 120 See M Greenberg, ‘The Moral Impact Theory of Law’ (2014) 123 Yale Law Journal 1289 and S Hershovitz, ‘The End of Jurisprudence’ (2015) 124 Yale Law Journal 1160. 121 D Hume, ‘Of the Original Contract’, in F Watkins (ed), Hume: Theory of Politics (Austin, University of Texas Press, 1953) 193; R Dworkin, ‘Justice and Rights’ in Dworkin (ed), Taking Rights Seriously (n 76) 185. 122 Hume’s rejection of the social contract and consent theory is not as categorical as is often thought to be the case. See P Winch, ‘Certainty and Authority’, in AP Griffiths (ed), Wittgenstein: Centenary Essays (Cambridge, Cambridge University Press, 1991) 223, 227–28, commenting on Hume’s case of an hereditary prince who returns after having been deposed by a usurper who remained unpopular and is greeted with enthusiasm by the people as their legitimate ruler. In this case, Hume says, ‘they consent because they perceive him to be already by birth their lawful sovereign’. (See Hume, ‘Of the Original Contract’ (n 121) 206.) Winch says (his emphasis): consent does indeed play a role in the relations between citizens and ruler in this case., but not the role described by social contract theorists. It is not the source of their sense of the ruler’s legitimacy; rather, their recognition of his legitimacy is expressed in the role played by the thought of his birth in the way they consent to his rule, and the importance they attach to this is of course rooted in the hereditary institutions which belong to their form of political life. It goes without saying that this form of life involves certain ingrained habits of obedience towards people occupying certain positions within it.

The Constitution of Legal Authority  207 It partakes of actual acceptance in just the way that Austin’s much maligned idea of a habit of obedience suggests, but understood in terms of the actual beliefs of those in the political community about what it takes to vindicate claims of right, not fear of sanctions.123 It partakes of the idea of the hypothetical reasonable person in that it postulates or better bets that the tacit habit of acceptance would become explicit if put to the test of scrutiny by the reasonable subject. This test is not altogether hypothetical, however. It depends on the fact of widespread acceptance, though on condition that this fact cannot plausibly be claimed to be the product of coercion.124 And the more legal subjects have available to them the institutional means of contestation of the mandatory or legal public norms the content of which seems unreasonable to them, the less easy it will be to make the claim that acceptance is the product of coercion. Philosophy of law cannot settle what these means have to be. But as I have reconstructed it here, as long as it seeks to understand law as a matter of authority, particular positions must take a stand on the terrain of constitutionalism. In this regard, we should note that the fact-driven doctrine of political responsibility that Dworkin attributed to legal positivism is no fairy tale despite the fact that in The Concept of Law Hart has little to say about constitutions let alone about anything that one might describe as constitutionalism. One could explain Hart’s lack of attention to these topics as due to the fact that he was writing in the context of a legal order in which Parliament is taken to be able to make any law it pleases, and well before the surge in constitutionalism of the late twentieth century. But not only is there the persistent theme in Hartian legal positivism that the content of the law is what is determinable as a matter of fact, but also the command theory of law seems designed to supply the theoretical basis for a legal order in which Parliament is supreme with the result that it can make any law it pleases. Both Bentham and Austin thus argued against entrenched Bills of Rights that might limit the scope of legislation by giving authority to judges to determine the limits, with Bentham providing a political argument based on the maximisation of utility, while Austin suggested that such limits are conceptually unavailable since constitutional law is not law properly so called, but merely positive or conventional morality. In addition, Bentham’s efforts to find ways to diminish the opportunities for judges to impose their views on the content of legislation support the kind of

On this interpretation of Hume, his views are not that different from the idea of ‘attributed consent’ developed by Kinch Hoekstra in a bid to understand the role of consent in Hobbes’s de facto theory of authority: ‘The de facto turn in Hobbes’s political philosophy’ in T Sorell and L Foisneau (eds), Leviathan After 350 Years (Oxford, Oxford University Press, 2004) 33, 67–8. 123 On habit and legitimacy, see C Misak, ‘A Pragmatist Account of Legitimacy and Authority’ in D Rondel and S Dieleman (eds), Pragmatism and Justice (Oxford, Oxford University Press, forthcoming). 124 Note that this condition is accepted by the realist mentioned above – see Williams, ‘Realism and Moralism in Political Theory’ (n 116) 8, where he confronts the example of a group that accepts the dominant ideology that deems it to be inferior and so justifiably relegated to a subordinate position in the society. As a result, he resorts to the ‘critical theory principle’ – ‘the acceptance of a justification does not count if the acceptance itself is produced by the coercive power which is supposedly being justified’.

208  David Dyzenhaus theory of adjudication that requires that the content of the law be determinable by factual tests.125 Similarly, despite Hart’s rejection of the command theory of law, the image of a supreme Parliament that is legally constituted but may make any law that it pleases – a law with any content in positivist parlance – dominates his imagination not only in The Concept of Law but also other writings. Like Bentham and Austin, he seems haunted by the potential in natural law theories to provide the resources for judges to impose substantive limits on the ability of such a body to make a law with any content. For example, he specifically warned against the temptation to ‘exploit the moral implications latent in the vocabulary of the law and especially in words like ius, recht, diritto, droit which are laden with the theory of Natural Law’, lest it lead to saying that ‘enactments which enjoined or permitted iniquity should not be recognised as valid, or have the quality of law, even if the system in which they were enacted acknowledged no restriction upon the legislative competence of the legislature’.126 These moments in his thought make highly plausible the suggestion that it is the rules of change that are fundamental for him, not the rule of recognition. But if that’s right, a major gap opens up in his thought that can be plugged only by a version of what he took to be Kelsen’s superfluous idea of basic constitutional norm, and which argues for the superiority of the kind of legal order in which parliament is supreme.127 Kelsen, in contrast, did have a lot to say about the proper role of constitutional adjudication and the place and kind of constitutions in the construction of legal order. Again, context can be invoked as part of the explanation, since Kelsen was one of the designers of the Austrian Constitutional Court – the first of its kind – as well as a judge on its first bench. But Kelsen argues that the commitment to legality inherent in the idea of the Rechtsstaat is best fulfilled when there is an entrenched written constitution, with constitutional guarantees as narrowly framed as possible, and with an independent court which has full power to review both statutes and executive acts for their compliance with the constitution.128 In addition, he argued for monism about

125 They did disagree on some matters pertinent to my discussion. Austin did not share Bentham’s faith in the wisdom of the judgments of ‘the people’, and so advocated a large role for judicial elites in legislating. And Bentham not only coined the term ‘international law’, but wished to promote its development, while Austin denied that it was law properly so called; like constitutional law, it is merely positive morality. 126 Hart, The Concept of Law (n 5) 208. 127 See my ‘The Idea of a Constitution: A Plea for Staatsrechtslehre’ in D Dyzenhaus and M Thorburn (eds), Philosophical Foundations of Constitutional Law (Oxford, Oxford University Press, 2016) 9, 19–21, discussing J Waldron, ‘Who Needs Rules of Recognition?’ in M Adler and K Himma (eds), The Rule of Recognition and the US Constitution (New York, Oxford University Press, 2009) 327, 342. See further S Perry, ‘Political Authority and Political Obligation’ in L Green and B Leiter (eds), Oxford Studies in Philosophy of Law (Oxford, Oxford University Press, 2013) 1. At 73 Perry cites Waldron’s paper and says that ‘the priority Hart gives to the rule of recognition over power-conferring secondary rules seems to me to be clearly in tension with his fundamental insight that the “introduction of into society of rules enabling legislators to change and add to the rules of duty, and judges to determine when the rules of duty have been broken, is a step forward as important to society as the invention of the wheel”. [Hart, The Concept of Law (n 5) 41–2.] In other words, at the foundational level of law, it is legislative and judicial powers … that are most important, not judicial duties.’ 128 H Kelsen, ‘The Nature and Development of Constitutional Adjudication’ in Vinx, The Guardian of the Constitution (n 77) 22.

The Constitution of Legal Authority  209 international law, for the position that a national legal order is subordinated to the norms of the international legal order. He thus favoured the establishment of international courts that could decide conflicts between the norms of international law and those made by sovereign states.129 He did, in short, provide a pretty full theory of constitutionalism, though he portrayed the choice for a constitutional court as well as for monism of the sort he favoured as one that had to be decided by political not legal theory. But it is highly plausible to reconstruct Keslen’s thought so as not to understand him as a philosopher with two hats, a legal theory hat, on the one hand, and a constitutional/political hat, on the other. Rather one should see his constitutional/political theory as the quite seamless elaboration of his bid to understand the modern legal order in a pure or non-ideological way.130 One should explain his legal theory, that is, as an account of a peaceful and orderly testing ground for competing political ideologies, at least for those prepared to confine their contest to the terms set by one willing to take the perspective of the legal subject. These claims about both Hart and Kelsen cannot in this chapter go beyond the suggestion of plausible reconstruction. But given the current impasse in philosophy of law, and the possibility of a productive exchange between legal philosophy, on the one hand, and theories of constitutionalism and political legitimacy, on the other, in which a range of common concerns are recognised, that suggestion has, in my view, much to recommend it.

129 See, eg, Kelsen, Introduction to the Problems of Legal Theory (n 15) 116–17. 130 See Vinx, Hans Kelsen’s Pure Theory of Law (n 62) and The Guardian of the Constitution (n 77), also Honoré, ‘The Basic Norm of a Society’ (n 75) and Notermans, ‘Social Peace as Conditio Tacita for the Validity of the Positive Legal Order’ (n 79).

210

8 Constitutional Law as Legitimacy-Enhancer DIMITRIOS KYRITSIS*

I. INTRODUCTION

C

onstitutional lawyers are no strangers to moral principles. Concepts like the rule of law, democracy, and human rights figure prominently in constitutional texts and judicial decisions. And yet, the place of political morality in constitutional doctrine is contested and precarious. For many it is a key resource in resolving constitutional disputes, while others warn that it is unduly volatile and fraught with risks, though, perhaps, still relevant to the evaluation of extant constitutional law. In this chapter I side with the former camp. I argue for what I label moralised constitutional theory (MCT). MCT, as I understand it, does not merely say that political morality is sometimes a determinant of our constitutional law rights, duties, powers etc. Rather, it makes the bolder claim that in the by and large well-ordered political societies with which we are familiar, constitutional law rights, duties, powers etc are of necessity partly grounded in moral principles.1 MCT is a broad church. For instance, according to one of its variants, constitutional law is underpinned by the moral principle of equality.2 According to another, constitutional law’s touchstone is the rule of law.3 It is important to stress that these

* I would like to thank participants in the workshop on the methodology of constitutional theory at the University of Reading in 2017, at the Jurisprudence section of the 2020 SLS Conference at the University of Durham, and at a meeting of the Public Law Cluster of the University of Essex for their helpful comments. I am especially grateful to Ana Cannilla, David Dyzenhaus, Theodore Konstadinides, Frank Michelman, and Maurice Sunkin for invaluable feedback on previous drafts. 1 I should add that the fact that they are so grounded does not entail that moral principles will necessarily figure expressly in the reasoning of constitutional judges or other officials, although it is highly likely that they will sometimes. Elsewhere I argue that such moral reasoning may be tempered or even on occasion completely pre-empted by doctrines of judicial deference, which are themselves based on higher order moral principles. See D Kyritsis, Where Our Protection Lies: Separation of Powers and Constitutional Review (Oxford, Oxford University Press, 2017) ch 7. 2 TRS Allan, The Sovereignty of Law: Freedom, Constitution, and Common Law (Oxford, Oxford University Press, 2013). 3 D Dyzenhaus, The Constitution of Law: Legality in a Time of Emergency (Cambridge, Cambridge University Press, 2006).

212  Dimitrios Kyritsis theories do not simply say that equality and the rule of law are ideals to which constitutional law should properly aspire. The thought is, rather, that for a certain norm to be part of the constitution, it must flow from or conform to equality or the rule of law, correctly construed. I propose a route that differs in important respects from the aforementioned options. I maintain that in its essence constitutional law enhances the legitimacy of a political regime in a distinctive way. Now, I am hardly the first to note a connection between law and the theory of legitimacy. Ronald Dworkin and Joseph Raz, to name two examples, have in their very different ways elaborated such a connection in their theories. Dworkin maintains that, for a political society to be legitimate, it must act with integrity.4 That is, it must treat all citizens according to the same conception of justice that it had committed to in the past; by doing so it realises an ideal of equal concern and respect. Judges who interpret legal practice by identifying a set of principles of political morality that best explain and justify past political decisions, as Dworkin’s theory of law proposes, are thereby upholding integrity. Conversely, for Raz the law claims legitimate authority, so we can understand a lot about how it purports to operate in our practical reasoning by reflecting on the conditions of legitimate authority.5 In particular, we can grasp the exclusionary force that it claims for its edicts. What is distinctive about the approach I advocate here (apart from the fact that it makes a claim specifically about constitutional law rather than law in general) is that it singles out a dimension of legitimacy that is less pronounced in those other theories, if it figures at all. I label it the assurance dimension of legitimacy (or assurance legitimacy, for short). Assurance legitimacy calls for guarantees that government will reliably and systematically act in a morally justifiable way. In section II I spell out this dimension of legitimacy and argue that constitutional law is especially suited to furnish the guarantees it recommends. Constitutional law may have other roles as well, but this one, I suggest, is rather central and crucial. In section III I go on to draw out the methodological implications from accepting this understanding of the role (or point, as I will also sometimes say) of constitutional law. I contend that, in order to work out the right answer to a question of constitutional law, we ought to ask which answer contributes to the fulfilment of this role. In other words, we ought to determine which arrangement of constitutional rights and duties enhances the legitimacy of a political regime in the special fashion that the constitution is supposed to. This is a thoroughly moral inquiry, but one that directs us to pay attention and give effect to political decisions that have been made in the past, including the momentous decisions that defined the institutional fundamentals of the regime in question. MCT thereby bridges the moral force of the constitution and its contingent character as a political settlement. In stressing the assurance-providing character of the constitution, the approach defended here seems to clip the moral aspiration that some would like to pin on constitutional law. It maintains that constitutional law should not be equated with



4 R 5 J

Dworkin, Law’s Empire (Oxford, Hart Publishing, 1998) ch 6. Raz, The Morality of Freedom (Oxford, Clarendon Press, 1988) chs 2–4.

Constitutional Law as Legitimacy-Enhancer  213 the pursuit of perfect justice: rather, it is primarily there to ensure that this pursuit is subject to certain conditions regarding the proper use of the state’s coercive force. In an important sense, then, it fixes more on the threats that come from the existence of the state and less on the full set of goods that it makes possible. I do not think that this ‘modesty’ is a cause for regret. For one thing, it seems to better resonate with political reality. Even in the by and large well-ordered political societies that we are familiar with, the state’s orientation towards justice cannot be taken for granted. This makes taming political power by constitutional means all the more urgent. As I shall argue, when it is constrained by constitutional – and other – guardrails,6 the state deserves our continuing – though often grudging – allegiance despite its missteps. In addition, constitutional law’s ‘modesty’ is not only a recipe for broadly well-ordered but far from perfect political societies. Even the steadfast pursuit of justice – if we can ever be sure of such a thing – should not be untrammelled. For, state coercion triggers a set of moral concerns which are largely absent from the theory of justice. These are concerns that legitimacy characteristically addresses. II.  MORAL FORCE AND SETTLEMENT

As already mentioned, this chapter champions a moralised methodology for identifying the correct answers to questions of constitutional law. But how should we go about defending this (or any other) methodology? This, we might say, is a metamethodological question. It asks what are the criteria for selecting a sound (or the most sound) methodology for constitutional law. One thing we can do is show that other tools that are on offer are inadequate. We evaluate competing legal methodologies such as semantics or history and conclude that they cannot furnish determinate or convincing answers to pressing doctrinal questions. Morality is the last method standing, so to speak. This is a line of argument characteristically pursued by Stuart Lakin.7 Alternatively, like Trevor Allan, we can argue that MCT best fits the selfunderstanding and discourse of legal practitioners.8 We survey the arguments that practitioners offer and find compelling and observe that they are typically underpinned by or drawn from political morality. There is much to recommend either strategy, but in this chapter I follow a different – though to some extent complementary – route. I argue that MCT is sound because it helps us capture salient features of constitutional practice; in fact, it can capture features that critics of MCT claim it misses or distorts. These critics contend that MCT is caught in a paradox, because it must straddle two desiderata seemingly pulling in opposite directions. On the one hand, it must account for

6 S Levitsky and D Ziblatt, How Democracies Die (New York, Crown Publishing, 2018). 7 See for instance S Lakin, ‘How to Make Sense of the HRA1998: The Ises and Oughts of the UK Constitution’ (2010) 30 OJLS 399–417 and Lakin, ‘Why Common Law Constitutionalism is Correct (If It Is)’, ch 5 in this volume. 8 See TRS Allan, ‘Principle, Practice, and Precedent: Vindicating Justice, According to Law’ (2018) 77 Cambridge Law Journal 269.

214  Dimitrios Kyritsis constitutional law’s moral force, and on the other, for the fact that it embodies a political settlement.9 Needless to say, the first desideratum can more straightforwardly be satisfied by MCT. A big part of the reason, surely, why constitutional law is thought to have moral force – at least in the by and large well-ordered societies with which we are familiar – is that it enshrines or guarantees important principles of political morality such as representative democracy, the rule of law, and human rights. MCT operationalises and vindicates this connection. It contends that these principles are not idle words or rhetorical flourishes. Rather, they determine the content of the constitutional law, such that more specific constitutional norms must be shown to flow from or adhere to them. MCT can also explain why these principles pertain specifically to constitutional law: The thought is that, because of its fundamental character, constitutional law can effectively express a polity’s commitment to these principles. Where the constitution enjoys hierarchical superiority over ordinary law, constitutional law does so by turning these principles into conditions of validity for all downstream state action. But arguably this commitment also finds a number of institutional expressions in jurisdictions without a formal constitution, where the doctrine of parliamentary supremacy is thought to preclude the imposition of such conditions on the validity of primary legislation.10 Of course, there are many constitutions that flout these principles outright, and it would seem that MCT has nothing to say about them. However, MCT’s goal is not to be descriptively ecumenical but to determine the role and content of constitutional law in those jurisdictions where constitutionalism is not an empty promise, and, as Sartori puts it, that promise is for ‘a fundamental law, or a fundamental set of principles, and a correlative institutional arrangement, which would restrict arbitrary power and ensure a “limited government”’.11 At the same time, though, constitutions are creatures of historical contingency. They mark a special political achievement, the establishment for the first time or the more or less infrequent updating of the basic terms and commitments of political life in a particular society. This feature is captured by the second desideratum, and it is here that MCT is thought to run into problems. In fact, it may be argued that the more emphasis it places on fundamental moral principles, the more likely it is that it will lose sight of the settlement aspect of constitutional law. In so doing it presumably becomes vulnerable to the twin charges of over-rigidity and disloyalty. Let’s start with the charge of over-rigidity. Typically, constitutional settlements are in part procured by compromise, as a result of which our commitment to one

9 These two desiderata bear some similarity with the two functions, ‘regulatory’ and ‘justificatory’, that the constitutions of liberal constitutional democracies perform according to Frank Michelman. See F Michelman, ‘Political-Liberal Legitimacy and the Question of Restraint’ (2019) 1 Jus Cogens 59, 60–1. 10 See, eg, J Laws, ‘Law and Democracy’ [1995] Public Law 72. 11 G Sartori, ‘Constitutionalism: A Preliminary Discussion’ (1962) 56 The American Political Science Review 853, 855. For Sartori, we should of course still study constitutions that do not fulfil that promise, but without thinning down our definition of constitutionalism in order to include them within it; rather we should treat them as ‘a loosening up of the concept’ for the purpose of corrupting it or using it as a ‘trap word’. Accordingly, he distinguishes constitutions properly so called from what he dubs nominal and fake ones (ibid 861ff).

Constitutional Law as Legitimacy-Enhancer  215 or the other moral principle may need to be watered down. Perhaps, we have had to curtail the constitutional protection of socio-economic equality or accept limits on the justiciability of executive action at times of emergency. But the applicability of moral principles cannot be similarly qualified or cast aside. Hence, if there is a tension between the moral principle that MCT designates as the guiding star of constitutional law and a specific constitutional provision, MCT would seem to recommend that the provision be discarded or interpretively neutralised so as to bring the content of constitutional law in that jurisdiction more in line with the precepts of, say, equality or the rule of law. Arguably, though, such ‘rewriting’ of the law flies in the face of the constitutional settlement. Let us now turn to the second charge. The compromises that have gone into a constitutional settlement serve important purposes. For example, sometimes they are expressions of a political community’s self-determination and, as such, deserve a high degree of respect. Other times they are the price for getting a certain part of society on board in the constitutional project. So, we have strong reasons to remain loyal to the content of the constitutional settlement. In the first case we want constitutional law to give effect to collective self-determination, and in the second we want it not to upset a hard-won and necessary political alignment. By contrast, the charge goes, MCT favours adherence to some sweeping moral principle over loyalty to the Constitution, warts and all. Those who press similar objections need not deny that we can employ moral principles to specify what is morally worthy about this or that constitutional settlement, identify ‘the deeper justificatory rationale for the particular constitutional rules that a legal system has adopted’,12 critique a constitutional settlement’s failings or propose its reform. Nonetheless, these normative inquiries are in an important sense ‘meta-constitutional’.13 They kick in at an analytically later stage, only after the often morally defective object of our loyalty or disloyalty has been fixed in a non-moral way, say, by reference to its text or the history of its creation or amendment. I accept that the charges of over-rigidity and disloyalty point to genuine theoretical and practical concerns: How can moral principles be made to apply more flexibly so as to capture the contingent character of the constitution? And how can MCT be true to the political moment that led to the constitution’s creation? But I believe that MCT can accommodate these concerns and thus satisfy the desideratum of settlement. I am not going to delve into how the variants of MCT discussed above try to reconcile the dual character of constitutional law (though I shall have more to say about one of them in section IV). Rather, as already mentioned, I shall pursue a different path. I shall contend that the point of constitutional law is not ideal justice but political legitimacy. Legitimacy can explain why constitutional settlement matters morally such that, provided certain conditions are met, the law should be determined by that settlement, properly understood, and not ideal justice. I shall further claim that a legitimacy-oriented MCT is immune to the charges of over-rigidity and disloyalty.

12 P Craig, ‘Constitutions, Constitutionalism and the European Union’ (2001) 7 European Law Journal 125, 127. 13 R Kay, ‘Preconstitutional Rules’ (1981) 42 Ohio State Law Journal 187.

216  Dimitrios Kyritsis Before I turn to these claims, I should address a preliminary and more radical criticism. It could be objected that what I propose is not much of a defence of MCT. Surely, the critic goes on, methodology is supposed to be analytically prior to and constrain our substantive theories. But here the order of explanation runs in the opposite direction: I derive my methodology from a substantive claim about constitutional law, the very thing that the correct methodology is supposed to guide us towards. Does this not make the argument circular? Presumably, the circularity is avoided, if you accept that accounts of constitutional phenomena should track the common understanding of those phenomena.14 On this view, we assess the soundness of different substantive accounts by appeal to an independent standard, whether or not they match the beliefs about the phenomena held by the community. That should be ‘a matter of fact’, which in principle could be gleaned from ‘empirical study, through opinion polls and the like’.15 However, we do not have to accept the understanding of the relationship between methodology and substance that this criticism assumes and the controversial theory of concepts on which it is based.16 On the view I propose, method and substance are interconnected; constitutional method is designed in light of our higher order premises about the nature of constitutional practice. We then employ it to make more concrete judgments about the content of the practice. This view is not meant to deliver a knock-down argument in favour of a certain constitutional method. Rather, the fittingness of the method is judged holistically, by the ability of the overall theory within which it is housed to give satisfactory answers to the theoretical and practical puzzles we encounter as we explore constitutional practice. III.  LEGITIMACY VS. JUSTICE

A.  Legitimacy as Assurance Political legitimacy sets a moral standard for the proper exercise of state coercion, which I said a moment ago is more modest than justice; as common wisdom has it, a state does not need to be fully just to satisfy it.17 This, however, begs the question: Why should constitutional law not be oriented towards the more demanding moral

14 Nick Barber outlines this view in N Barber, ‘The Significance of the Common Understanding in Legal Theory’, ch 1 in this volume: ‘When we are considering the community’s understanding … the claim is false if as a matter of fact – and it is a matter of fact – the account does not square with our common understanding’. 15 ibid. 16 For that theory of concepts, as applied to law, see for instance J Raz, ‘Two Views on the Nature of the Theory of Law: A Partial Comparison’ (1998) 4 Legal Theory 249. 17 See among others J Rawls, ‘Legal Obligation and the Duty of Fair Play’ in S Freeman (ed), Rawls: Collected Papers (Cambridge (MA), Harvard University Press, 1999) 117, 119, and J Rawls, ‘Political Liberalism: A Reply to Habermas’ (1995) 92 Journal of Philosophy 132, 175–6. In the former article Rawls speaks of the obligation to obey the law on the part of citizens rather than the right of government to coerce, but the distinction is not important for my purposes. On the relationship between legitimacy and justice, with particular emphasis on Rawls see S Langvatn, ‘Legitimate but Unjust; Just but Illegitimate: Rawls on Political Legitimacy’ (2016) 42 Philosophy and Social Criticism 132.

Constitutional Law as Legitimacy-Enhancer  217 standard? Why is legitimacy good enough? This question becomes more pressing when we consider that for some theorists legitimacy itself has ‘an essential connection with justice’,18 inasmuch as we often define what is legitimate by reference to how far we can tolerate divergences from what is just. An essential connection between legitimacy and justice would seem to reinforce the thought that the former is merely a pale shadow of the latter, lacking independent explanatory value. I am unpersuaded by this line of argument. I side with Philip Pettit and others who claim that ‘the concern with legitimacy is distinct from the concern with justice’.19 Although it is true that legitimacy is intertwined with justice, it is also shaped by moral considerations that are special to it. As a general matter, these considerations have to do with the conditions under which an actual political regime has a moral warrant to bind those subject to it or exercise coercive force over them given its moral imperfection and the existence of pervasive and ineradicable disagreement about moral issues that it must contend with. I cannot explore these considerations in full. Rather, I will focus on a subset that are particularly pertinent to the moral role of constitutional law and explain in what way they differentiate legitimacy from justice. These considerations comprise what I call the assurance dimension of legitimacy, a dimension that standard theories of political legitimacy do not pay much attention to. Typically, standard theories of legitimacy contend that the justice or injustice of a given political society does not exhaust the question of its legitimacy in the sense that it cannot account for the contingent character of the moral bond between persons and that political society. They ask: What gives rise to a moral obligation to obey the law of this particular society? But, for the main part they take as their object the content of a legal order, as if at a snapshot, and do not especially heed the way it is being developed overtime. Rather, they focus on whether we have consented to being subjected to this legal order,20 or whether we receive benefits from this legal order in return for which we owe it our allegiance,21 or whether the legal order treats us with equal concern and respect.22 But what gives us confidence that, when we give our consent, we do not sign a suicide pact, that the legal order to which we are bound will continue to provide the benefits or extend the same treatment in the future? Standard accounts of legitimacy have little to say about this, and yet it seems absolutely crucial. Surely, without such confidence it would be perfectly rational to withhold or hedge one’s obedience, which, after all, is supposed to be for the long haul. Political societies are dynamic, capable of changing direction, and – unless they are suitably constrained – not necessarily for the better. In order, then, legitimately to exercise power over us, they must also offer credible and reasonable assurances that their moral profile will not drastically deteriorate.23 This is the crux of assurance legitimacy.

18 Rawls, ‘Reply to Habermas’ (n 17) 175. 19 P Pettit, ‘Legitimacy and Justice in Republican Perspective’ (2012) 65 Current Legal Problems 59. 20 AJ Simmons, ‘Justification and Legitimacy’ (1999) 109 Ethics 739. 21 Rawls, ‘Legal Obligation’ (n 17). 22 Dworkin, Law’s Empire (n 4) ch 6. 23 On the notion of a society’s moral profile see M Greenberg, ‘The Moral Impact Theory of Law’ (2014) 123 Yale Law Journal 1288.

218  Dimitrios Kyritsis The content and rationale of this dimension of legitimacy needs some unpacking. Someone might argue that it sounds overly roundabout. Why should we focus on the existence of assurances? Why can we not simply say that what matters for legitimacy is whether our political society has got it right? In each case we would then have to ask: Does this or that state policy improve or worsen the moral profile of a political society?24 However, that would be to take an overly narrow view of the issue. Legitimacy does not only operate at the retail level of specific political decisions. A system of government typically affords some persons or bodies general authority to decide on matters of state policy based on a judgment about the common good. The most characteristic example is the authority vested in the national legislature, but the same can be said about the other branches of government. This general authority is passed from one group of people to the next, each with its own distribution of ideologies, characters and motivations. Trust might attach to this or that officeholder, but it would be irrational for citizens to trust the institution as a whole across time, unless there are standing and effective constraints on the exercise of its power. Such constraints help to organise government such that it will reliably and systematically act in a justifiable way in the long run.25 Without them, citizens would be left at the mercy of unscrupulous and corrupt rulers who happen to wield political power.26 In turn, the possibility that this might happen would make the stakes of their membership in political society prohibitively high. It would thus encourage attitudes that undermine allegiance to a regime, such as free-riding and a desire to ensure victory at all costs in order to stave off the grave peril that might attend defeat. Given what I have said about assurance legitimacy, it should not come as a surprise that it has a special affinity with constitutional law. Constitutional law in the by and large well-ordered societies that we are familiar with is meant to structure the system of government, setting up its most important branches and their decision-making procedures so as to give effect to principles such as democracy and separation of powers, and building checks and balance mechanisms into these procedures. Moreover, as a general matter, constitutional norms enjoy a degree of fixity and insulation from the rough and tumble of ordinary politics. As a result, they can effectively establish ground rules and place outer limits on how far a political society may go.27 So both in its content and through its endurance constitutional law is well placed to furnish the assurances that matter for legitimacy.28 I am not saying, of course, that constitutional

24 Although I cannot spell out the notion of a moral profile here, I would venture the following qualification. Improvements should include cases where a policy is morally sub-optimal from the perspective of ideal justice, but, say, its coordination benefits (the fact that others are prepared to comply with it and widespread compliance is desirable) outweigh its moral defect. 25 Kyritsis, Where Our Protection Lies (n 1) ch 1. 26 Being at someone else’s mercy is the condition that republican freedom is famously supposed to guard against. It is an interesting question whether there are connections between republicanism and the account I advocate. See relatedly Pettit, ‘Legitimacy and Justice’, above n 19, and T Hickey, ‘The Republican Core of the Case for Judicial Review’ (2019) 17 ICON 288. 27 I believe that this is to some extent independent of whether the constitution is written or unwritten, but I cannot argue this point here. 28 For an extended argument to this effect see D Kyritsis, ‘Williams and Rawls in Philadelphia’ (2021) 27 Res Publica 203.

Constitutional Law as Legitimacy-Enhancer  219 guarantees are panacea. Although constitutional law can characteristically enhance assurance legitimacy, it is not the only thing that does. Nor can it always make up for other deficiencies in the political system. As recent history amply confirms, once the other enhancers of assurance legitimacy fail, say, because the party system has been eroded or fallen prey to extremism, the constitution will typically not be able to put up a fight for long. Let me finish with a couple of remarks on the ‘drastic deterioration’ clause in the definition of assurance legitimacy I gave earlier. First, it would be unreasonable for a citizen to demand that their political society be disabled from improving its moral profile, just because that improvement would not be in his narrowly construed self-interest. That would be the case of the slave-owner in the antebellum South conditioning his allegiance to the federal government on the continued existence of slavery and of the constitutional mechanisms that entrenched it. But assurance legitimacy is there only to block undue setbacks to our interests, and the abolition of slavery is strongly mandated by justice. Second, it would also be unreasonable for a citizen to insist on assurances that the moral profile not deteriorate at all. As long as the possibility of suffering a particularly grave or flagrant injustice or being placed at a systemic political disadvantage has been minimised, citizens must be prepared to comply with any sub-optimal decisions falling below this threshold and only try to reverse them using the avenues of political contestation made available by the political system.29 In fact, the availability of effective and sufficient avenues of contestation is itself an important element of assurance legitimacy. Jacob Weinrib has argued that legitimacy cannot demand that the political system be reliably and systematically oriented towards justice and at the same time tolerate discounts on justice; if a regime satisfies this demand, then it cannot but be just.30 But that is too quick. A regime’s orientation towards justice can be reliable and systematic, even if on occasion it misfires. When legitimacy requires that we live with these mistakes, it is because we purchase something morally important with our reticence, namely a standing assurance that the risk of suffering a grave injustice at the hands of our government is relatively low. Precisely thanks to the regime’s reliable and systematic pursuit of justice, the fact that it might fall short now and then need not undermine our overall trust in it. To be sure, we do not always have to tolerate the failures of an otherwise reliably justice-oriented system of government. Although legitimacy ‘imposes weaker constraints [than justice] on what can be done’,31 neither is it completely indifferent to the content of state policy. To identify which failures are beyond the pale as far as legitimacy is concerned, we must strike a balance between considerations mandating adherence to the outcomes of binding procedures (or considerations of institutional design, as we may call them) and considerations pertaining to the content of the decisions that have resulted from such procedures. Accordingly, constitutional law plays its role as legitimacy enhancer by incorporating this balance in the working of



29 On 30 J

this way of cashing out the upshot of legitimacy see Pettit, ‘Legitimacy and Justice’ (n 19). Weinrib, ‘Rights in Search of Protection’ (2020) 40 OJLS 403. ‘Reply to Habermas’ (n 17) 175.

31 Rawls,

220  Dimitrios Kyritsis a constitutional order. Systems of constitutional review are an example of how this can be done.32 We could say that they are morally justified insofar as they effectively and accurately police departures from the range of suboptimal or unjust but still legitimate outcomes (for want of a better term, let us call it the range of acceptable decisions). The striking down of primary legislation that falls outside that range could be seen as an institutional method for ensuring that the political system as a whole stays on the right side of legitimacy. B.  Assurance and Disagreement As the example of constitutional review shows, assurance legitimacy will sometimes require legal officials making moral judgments concerning, for instance, how seriously a statute infringes on a fundamental right, because the legitimacy of the primary legislation before them – and hence its having an effect on our rights and duties – partly depends on that very question. Judgments of this kind are notoriously contested, so conditioning the legitimacy of political action in this way introduces a degree of open-endedness in the constitutional order. It is much easier to ascertain whether a Bill was supported by the constitutionally mandated majority of parliamentarians for it to become law than it is to agree on what counts as a violation of the right to private life. It could be countered that this implication defeats a key purpose of having constitutional processes for deciding contested issues of state policy. We want these processes to overcome our intractable substantive disagreements about, for example, whether homosexual couples have a right to adopt, not necessarily by making some of us change our minds but by reaching a resolution that everyone can accept. In Frank Michelman’s words, what we aim to achieve is a deflection of divisive questions of substance (does this law or policy merit the respect or rather the contempt of a right-thinking person?) to a different question (is this law or policy constitutional?) for which the answer is to be publicly apparent, or at any rate is to be ascertainable by means that are an order of magnitude less open to divisive dispute than are the deflected substantive disagreements.33

Some theorists claim that only a purely procedural standard of legitimacy can achieve this kind of deflection. For Jeremy Waldron, for example, this standard is satisfied by a majoritarian procedure that gives all citizens an equal vote.34 Such a procedure need make no reference to the merits of the substantive issue being put to a vote. Rather than ask whether justice requires that homosexual couples have the right to adopt, we ask whether this issue was settled – for now and until we collectively revisit it – in a way that gives everyone’s opinion the same weight. By so settling contentious matters a political society arguably displays an attitude of respect towards those 32 It is not necessary for present purposes to take sides on whether constitutional review is universally or contingently required by legitimacy or merely permitted by it. 33 F Michelman, ‘“Constitution (Written or Unwritten)”: Legitimacy and Legality in the Thought of John Rawls’ (2018) 31 Ratio Juris 379, 384. 34 J Waldron, Law and Disagreement (Oxford, Oxford University Press, 1999) 113.

Constitutional Law as Legitimacy-Enhancer  221 who disagree with the enacted policy: It does not tell them that they were wrong but only that they lost in a fair contest. It thereby also deflects our substantive disagreement: Satisfaction of the procedural standard is presumably ‘an order of magnitude less open to divisive dispute’.35 This benefit, so the argument goes, is lost if we opt for the mixed standard of legitimacy proposed here, because that standard seems to refer us back to the substantive question dividing us, albeit as one consideration among many. Moreover, when a court strikes down a law as unconstitutional, it does so (partly) on the basis of what it takes to be the correct understanding of the right to private life. I find this argument unconvincing. Let me mention two reasons. First, since legitimacy dictates that we abide by decisions made by those who have moral authority to govern that fall within the acceptable range, the space of what we can all agree to abide by widens. You will abide by a decision, because you think it is the right one; I will abide by it, because I believe that, though mistaken, it is still acceptable. In this sense there is a significant degree of deflection after all: Prior to the decision, the issue dividing us was ‘which is the right decision’. After the decision, the issue becomes whether this here is an acceptable decision that merits our allegiance. The latter issue seems significantly less divisive, precisely because among people who otherwise hold conflicting moral views there is bound to be a greater amount of overlap regarding what is acceptable. Second, this critique rests on an over-demanding understanding of the deflection strategy. Undoubtedly the assurance function is bound up with a degree of legal certainty. Decision-making procedures subject to overly vague limitations are unlikely to gain citizens’ trust in a political regime because of their perceived susceptibility to manipulation. Nonetheless, certainty is not the only thing that matters for assurance, and neither does assurance assign it absolute priority. Our allegiance to decisionmaking processes depends not only on the clarity of the decisions they produce but also on the extent to which they are the right decisions: our trust in them is likely to be undercut if there is a significant risk that they might lead to gravely unjust outcomes. Hence, it is at least plausible to argue that we should be prepared to accept some open-ended limitations on otherwise binding political processes, when doing so mitigates such a risk.36 The disagreement that these limitations will inevitably give rise to among citizens and officials and the attendant lack of legal certainty might then be a price worth paying. This understanding of the deflection strategy goes hand in hand with a certain way of thinking about the place of disagreement in the theory of legitimacy. For proceduralists like Waldron, due to the fact of reasonable disagreement, legitimacy must be a matter of pure procedural justice.37 Legitimate is whichever decision has been arrived at by the correct procedure, because reasonable disagreement bars the 35 The truth of this proposition has been questioned by those who claim that we disagree about what counts as a fair procedure as much as we disagree about the substantive issues that the fair procedures are supposed to settle. See T Christiano, ‘Waldron on Law and Disagreement’ (2000) 19 Law and Philosophy 513, 521–522; A Kavanagh, ‘Participation and Judicial Review: A Reply to Jeremy Waldron’ (2003) 22 Law and Philosophy 451. 36 Whether this is true may well be a contingent matter, varying from one jurisdiction to another. 37 J Rawls, A Theory of Justice (Cambridge MA, Belknap Press, 1971) 85.

222  Dimitrios Kyritsis invocation of the substantive merits of that decision as its ground. By contrast, on the understanding put forward here, getting it right is an independent standard to which procedures are also subject, whether citizens accept it or not. This does not mean that disagreement is irrelevant to legitimacy. Quite the opposite, as Bernard Williams has suggested, we suffer a loss in a valuable liberty when we are coerced to obey laws that we believe, correctly or mistakenly, to be unjust.38 Had the law gone our way, we would be free to do something that we are now prohibited from doing. This is a burden that we often find ourselves having to shoulder given that we participate in a political society with others who think differently from us. Of course, we cannot demand that our political society not pass laws to which we object. Even so, there are things a legitimate polity must do to ease these ‘strains of commitment’39 on dissenters short of giving them veto power over individual laws. Consider democratic decision-making. It can enhance political legitimacy in no small part because, all else being equal, it gives those on the losing side an assurance that their political defeat is not entrenched; if trying to reverse it in the future is not pointless, it becomes more palatable to accept it now. It is thus no accident that majority rule is not a peripheral feature of the legitimate polities that we are most familiar with, but their defining procedure.40 It is one of the key features of constitutional orders marked by persistent conflict and disagreement that can make them hospitable to winners and losers alike. C.  Constitutionalism of Fear? So far I have sought to show that, for a state to be legitimate, it must assure those who disagree with state policy that they are purchasing a significant benefit with their continuing obedience, namely that their state’s system of government is systematically oriented towards justice, although it may on occasion fail to be fully just. I now wish to consider a serious objection to this claim, which takes a number of different guises. Its nub is that, if taken to furnish the moral point of constitutional law, assurance legitimacy vitiates the moral ambition embodied in the Constitution. Either it puts individual self-interest at the explanatory driver’s seat rather than the ‘better angels of our nature’, or it tailors the content of the constitution to what we fear rather than what we aspire to. Thomas Hobbes’s theory was likely guilty of both errors. For him, persons join and continue to adhere to the social contract out of rational self-interest, and their aim is to avoid the threats to their self-preservation and security present in the state of nature.41 The limits on state power imposed by the social contract are accordingly minimal.42 Even a state with deeply unjust law can

38 See B Williams, ‘From Freedom to Liberty: The Construction of a Political Value’ (2005) 30 Philosophy and Public Affairs 3. 39 Rawls, Theory of Justice (n 37) 153–160. 40 See H Kelsen, The Essence and Value of Democracy (Plymouth, Rowman & Littlefield Publishers, 2013) 69. 41 T Hobbes, Leviathan (Oxford, Clarendon Press, 2012) ch 17. 42 ibid ch 18.

Constitutional Law as Legitimacy-Enhancer  223 be effective in securing broad obedience and thus averting the debilitating insecurity associated with the state of nature. Admittedly, there are some interesting continuities between the Hobbesian account and the one I am defending. Both focus on the importance for legitimacy of the ability of the state to assure. In one case the assurance is that others will comply with the laws of nature – thus making it rational to do so oneself.43 In the other it is that political society will not descend into tyranny – thus lowering the stakes of one’s participation in it. The two accounts are also structurally similar. They describe certain salient interests or concerns that persons have and then specify what political society should look like in order adequately to promote those interests or address those concerns. That being said, the account proposed here does not suffer from the aforementioned defects that sting Hobbes’. First, it does not identify the good of a legitimate political society purely in terms of individual self-interest. As already mentioned, it assumes an independent standard of justice which institutional arrangements must reliably and systematically track and by which we can define whether something counts as an improvement or deterioration of a society’s moral profile. Moreover, it is a standard, which, when state policy excessively deviates from it, strips it of its legitimacy. So, contrary to Hobbes, the account is not morally minimalist. As we shall see below, this difference has important ramifications. Second, it is not committed to a Hobbesian understanding of human psychology. It is compatible with viewing persons who contemplate whether to pledge their allegiance to a polity as partially – and even overridingly – motivated by considerations of political justice. However, it stresses that there are certain risks that are inherent in or are compounded by one’s participation in a political community; legitimacy crucially depends on mitigating those, too. For instance, these are risks stemming from the vast power of the coercive and bureaucratic machinery of the modern state and its susceptibility to being captured and mobilised towards factionalist ends. It is true that the diagnosis I just offered foregrounds aspects of life in a political society that we have reason to be wary of. As a result some might be tempted to draw parallels between this brand of MCT and Judith Shklar’s liberalism of fear.44 Rather than ‘celebrate the blessings of liberty,’ Shklar invites us ‘to consider the dangers of tyranny and war that threaten it’.45 The latter inquiry is made especially pressing by the fact that ‘some agents of government will behave lawlessly and brutally in small or big ways most of the time unless they are prevented from doing so’.46 Against this background, Shklar argues, liberalism should be defined in terms of its opposition to the evil of ‘cruelty and the fear it inspires, and the very fear of fear itself’.47 ‘Systematic fear’, she explains, ‘is the condition that makes freedom impossible’.48

43 ibid ch 17. 44 GS Kavka, ‘Right Reason and Natural Law in Hobbes’s Ethics’ (1983) 66 The Monist 120. 45 J Shklar, ‘The Liberalism of Fear’ in N Rosenblum (ed), Liberalism and the Moral Life (Cambridge MA, Harvard University Press, 1989) 21, 27. 46 ibid 28. 47 ibid 29. 48 ibid.

224  Dimitrios Kyritsis Its prevention can be the foundation for normative prescriptions, ‘a first principle … on which liberalism can be built’.49 It is important to clarify, though, that my claim is not that fear is the touchstone of liberalism. Such a claim, without further qualification, would no doubt be vulnerable to the criticism that it takes an unduly truncated view of liberalism’s moral sources and transformative potential. I make the narrower claim that fear of a certain kind is key to understanding the point of constitutional law. In fact, Shklar herself is sympathetic to the narrower claim, as she emphasises the importance of traditional constitutional guarantees (eg ‘[l]imited government and the control of unequally divided political power,’50 ‘a constant division and subdivision of political power’51). Furthermore, as I have stressed time and again, this understanding of constitutional law’s point makes reference to an independent standard of justice, and although I have not said much about that standard here, nothing I did say excludes the possibility that it may be very capacious or at any rate more capacious than the liberalism of fear. Neither is my account necessarily underpinned by cynicism about politics or pessimism about the realisation of justice. Even those who look to the state to solve serious social problems and be the engine of reform and enlightenment should also see in it a source of threats. Constitutional law provides a means for curbing those. Surely, when it achieves its purpose, it also facilitates the quest for justice. Despite these clarifications, it might be argued that my proposal is still morally myopic in its preoccupation with fear, in the sense that it is partial to the limited government strand of constitutionalism. It thus loses from sight the other major strand, whereby the point of constitutions is to constitute a community and enable collective self-government. Taken at face value, this objection ignores that the principle of collective self-government enters the account at several points. First, it is one of the considerations that can bestow assurance value and moral authority more broadly to a constitution-making process. A constitution whose creation involves popular participation commands respect, precisely because it instantiates a weighty political principle. Furthermore, more people can identify with it, in great part due to the fact that it is more likely to reflect their interests; this increases its stability and efficacy. Second, when collective self-government informs ordinary political processes under the Constitution, this can bolster the overall legitimacy of the political order. I have already intimated one reason for this in my discussion of disagreement in the previous section. It is important for the assurance legitimacy of a political society to provide avenues for meaningful and fair contestation of state policy, and democratic processes are well-suited for this purpose, as they give every citizen an opportunity to influence decisions, at least in principle. You may think that what I have said so far misses the point. Beneficial though these facets of self-government may be, they fail to capture the character of the constitution as an expression of constituent power by a sovereign people. Two responses can be made to this claim. To begin with, there are legitimate constitutions that cannot be said to have been popularly authored or authorised. The account proposed here

49 ibid 50 ibid 51 ibid

30. 28 30.

Constitutional Law as Legitimacy-Enhancer  225 tries to cast a wider net so as to identify the various ways in which assurance can be procured. More importantly, it is helpful to distinguish to some extent the question of authorship from the question what exercises of constituent power have as their proper aim. Those who create a constitution undertake a special political responsibility; they do not engage in revolutionary gymnastics. Here I try to explain how this responsibility should be discharged (though, as noted in the previous paragraph, authorship also bears on that responsibility). IV.  TWO MORALISED METHODOLOGIES FOR CONSTITUTIONAL THEORY

In the previous section I highlighted the strong link between constitutional law and the assurance dimension of legitimacy. But I have yet to explain how this link affects the methodology of constitutional theory. That is the aim of this section. I shall begin by elaborating how an MCT oriented towards the assurance dimension of legitimacy accounts for the desiderata of moral force and settlement. I shall call this type of MCT the assurance model. This is the model I favour. I shall then go on to contrast the assurance model to a type of MCT that is guided by a very different understanding of the moral point of constitutional law. I shall call it the epistemic model. The epistemic model is exemplified in the work of TRS Allan and others and offers an alternative vision of the relationship between the content of constitutional law and institutional history. I hope that the contrast will help clarify the assurance model’s distinctive methodology as well as the way in which it seeks to reconcile the two desiderata. A.  Constitutions as the Moral Footprints of Settlement Recall that the desideratum of settlement seeks to capture the fact that the content of constitutional law is shaped by political acts and decisions such as those made by a constitutional assembly. Earlier we said that this feature of the constitution is commonly thought to be at odds with MCT: MCT allows the content of the constitution to be determined by normative standards that are not validated or invalidated by what happens; by contrast, the settlement aspect of constitutions means that they are sensitive precisely to what happens, that is, a political society’s institutional history. In the assurance model, I shall now argue, this tension is resolved. According to it, some aspects of institutional history in more or less legitimate political societies enjoy a special moral authority. What bestows them this authority is in large part the fact that they furnish the kind of assurance that bolsters legitimacy. This, as we have seen, they achieve under two main conditions, both of which have an essential moral component. First, they themselves embody political moments or processes whose outcomes are morally salient. Second, they structure the system of government such that it reliably and systematically acts in a morally justifiable way. On this model MCT has a complex role to play in the determination of constitutional law rights and duties. For ease of exposition, I shall distinguish two levels at which it will operate.

226  Dimitrios Kyritsis i. At the most abstract level, MCT will spell out the general conditions under which institutional history may acquire constitutional significance. What credentials must political moments, processes, or practices (henceforth I shall refer to them collectively as events) possess in order to have the moral warrant to create or update a constitution?52 What must a system of government look like in order to be reliably justice-oriented? MCT will provide answers to these questions. It will also develop a theory of how these two types of consideration combine to vest constitutional significance in an event. It will be the aim of such a theory to specify possible trade-offs between them, for example, what discounts on good government can be made up for by an event that carries high moral authority, say, a constitutional assembly. But the most fundamental question that MCT will have to answer at this first level is why the decisions of a constitutional assembly or any other event should have any constitutional significance at all. Assurance legitimacy is key here. There are many possible systems of government – including the most just – that could be suitable for a political society, but assurance legitimacy can only be satisfied by one that is by and large adhered to. A system of government that is not efficacious in this sense does not provide the requisite assurance, because it cannot as a matter of fact organise and constrain public power. (This is definitely something Hobbes has taught us.) Still, efficacy is not sufficient for assurance legitimacy. Some political regimes, though efficacious, are not rightful. What is missing is that they are not supported by the two types of consideration sketched above. This is how the assurance model of MCT seeks to capture the moral essence of constitutional settlement. It is not the mere fact that there has been a settlement that creates constitutional law (conquests can settle, too), but the fact that a particular settlement justifiably speaks for and governs a political society.53 Hence, it is no accident that constitutional assemblies typically have constitutional significance. Given their democratic pedigree they can rightfully claim to represent a political society’s self-commitment and are capable of commanding broad public support. ii. MCT is not simply an exercise in political philosophy, however. Its ambition, ultimately, is to determine the content of constitutional law.54 So, at a second level MCT seeks to interpret the constitutionally significant events in a particular 52 It might seem more difficult to speak of credentials in the case of a political society with an unwritten constitution. But an unwritten constitution also has sources, so the nature of the inquiry remains unchanged. For each source, say, ordinary statute or conventional practice among institutional actors, we must ask to what extent they enhance assurance legitimacy and can thus rightfully contribute to the content of the constitution. 53 I am simplifying here. I do not wish to exclude the possibility that a settlement might start off with tainted credentials but overtime acquire the requisite salience to provide the framework for a legitimate polity. 54 In earlier work I have used the term ‘doctrinal relevance’ to signal this ambition. See Kyritsis, Where Our Protection Lies (n 1) 7

Constitutional Law as Legitimacy-Enhancer  227 jurisdiction and thereby specify the bulk of existing constitutional norms at any given time in that jurisdiction. Constitutionally significant events come in various shapes. Characteristically they culminate in a canonical text, but sometimes they do not. In the latter case a more fine-grained analysis is called for to read determinate constitutional norms off the event.55 It is important to stress, however, that in either case we do not primarily aim at discovering the meaning of a text or the content of a practice or agreement but determining (some aspects of) its moral footprint. That is, we do not merely describe what individuals and groups believe or intend or accept or can effectively impose on others. Rather, we explicate what follows from all these things in terms of what the state may legitimately do. To do this, MCT will consult the moral credentials of political actors, the principles of good government and the trade-offs between these two considerations, as these have been elaborated at the previous level. It will ask whether those considerations, suitably combined, count in favour of such and such a decision or agreement shaping the constitutional law of our political society. It will also ask what moral footprint the decision or agreement is warranted to have in light of those considerations. Importantly, that footprint need not coincide fully with what those who authored the decision or agreement had in mind. Recall, what we are looking for is how such an event matters for legitimacy. The moral considerations that underpin judgments of legitimacy will select what is, is not and cannot be relevant for this purpose. Thus understood, MCT has a vindicatory upshot. It gives us strong (though not necessarily overriding) moral reason to be bound by our political society’s decisions and agreements. To be sure, not all of them will come out as binding. And those who do will not necessarily have the sheen of perfect justice. But they command respect as elements of a successful though imperfect effort to set up a legitimate polity. I suggest that this way of thinking about MCT reconciles the desiderata of moral force and settlement, but at the same time it tempers our moral aspirations from constitutional law. It tells against reading into constitutional law the complete set of rights and duties that we would have under the best theory of justice. For, it views a legitimate polity as a political achievement that we share with others who might be unconvinced by such a theory. Constitutional law should track the decisions and agreements that record and sustain this achievement, even though on occasion they sacrificed perfect justice (within limits) for the sake of securing broad support of a constitutional order that works reasonably well for everyone. You might think that there is an inconsistency between, on the one hand, honouring a political settlement, as I said the assurance model aims to do, and, on the other, allowing for the possibility that constitutional norms might diverge from, say, the intentions and beliefs of a constitution’s authors. Surely, the argument goes, constitutional law must hew closely to them, because they are the ones that embody the settlement. This line of reasoning lies at the heart of originalist theories of

55 This is especially true of legal orders without a formal constitution, some of whose constitutional norms are grounded in conventional practices and are not codified. But it may also be true of orders that have a formal constitution, especially in cases where the formal amendment procedure laid down in the constitutional text is not the only avenue for changing constitutional norms. On this possibility see B Ackerman, We The People Volume 1: Foundations (Cambridge MA, Belknap Press, 1993) ch 10.

228  Dimitrios Kyritsis constitutional interpretation. Their stated aim is to retrieve the meaning that constitutional provisions had for those who drafted them, because ‘original meaning’, as they call it, is thought to encapsulate the political agreement that the enactment of the constitution represents. Originalists contend that the original meaning may not fully live up to the dictates of political morality, but this does not make it any less binding. If, as the assurance model insists, we have good moral reason to uphold that momentous political agreement – at least until such time as we collectively revisit it – we should not sanction modes of constitutional interpretation allowing an institutional actor – typically, in the originalist story, the judiciary – to rewrite it unilaterally. To do otherwise would arguably be to expose citizens to an unanticipated change in the system of government, precisely the type of change that assurance legitimacy condemns. A similar thought also drives some textualist theories of constitutional interpretation. Here, the text of the constitution takes the place of the authors’ beliefs and intentions as the focal point of political settlement. It goes beyond the scope of this chapter fully to evaluate this kind of claim. It seems to me that the truth of originalism or textualism is a substantive matter, unlikely to be settled at the level of methodology. All I want to do now is show why the opponents of originalism and textualism cannot be dismissed at that level either. To begin with, we cannot take it as a given that only considerations of text and intention can truly capture the content of a settlement. Ronald Dworkin, for one, has argued that fundamental rights guarantees of the sort contained in the US Constitution stand for a choice on the part of the constitution’s authors to enshrine into law moral principles such as equality and freedom of speech.56 If he is right, then appealing to those principles in constitutional reasoning is not an affront to the constitutional settlement but an affirmation of it. One may well disagree with Dworkin, but his position is not beyond the pale, conceptually speaking. It would equally be too quick to say that incorporating morality into constitutional reasoning fails as an assurance. As was suggested above, a system that allows some institutional actors to resort to moral principles may in fact contribute to the system’s overall assurance legitimacy despite the loss in legal certainty. It may be for instance that giving them this power introduces an element of checks and balances or enables the system to adapt to changing circumstances and improved moral insight. Again, this proposal will have to be judged on the merits, but it seems to me at least plausible. Many defences of constitutional review, for example, rest on it. B.  Assurance vs Moral Construction I said above that on the assurance model, MCT’s task is to navigate our constitutional practices and figure out what difference, if any, they make for legitimacy. I also said that they can make a relevant moral difference by structuring government in a way that reasonably addresses citizens’ concerns regarding their submission to the coercive power of the state. By contrast, on the epistemic model of theorists like 56 R Dworkin, Freedom’s Law: The Moral Reading of the US Constitution (Oxford, Oxford University Press, 1999) 7.

Constitutional Law as Legitimacy-Enhancer  229 TRS Allan the aim of constitutional theory is the ‘moral construction of legal practice’ with a view to spelling out principles of egalitarian justice that will govern the resolution of cases before us as well as future cases.57 Allan’s model, too, recommends that we pay attention to what our constitutional order has decided in the past. He contends that moral construction ‘[seeks] harmony between moral principle, on the one hand, and practical manifestation of principle on the other’.58 In order to achieve this kind of harmony, constitutional theory gives ‘presumptive if provisional force to the assumptions embedded, on careful analysis’59 within the practice. But here, unlike the assurance model, the reason for this backward-looking exercise is epistemic. For Allan ‘the law is normally our best guide to decision and action’,60 ‘because such practice is itself the principal forum for working out what justice requires in the context of current social conditions moral construction’.61 Hence, legal practice becomes a key repository of a political community’s ‘collaborative quest for justice’.62 Our previous efforts to identify and apply principles of egalitarian justice constitute ‘a source of enlightenment from which we can seek moral guidance’.63 We can learn from the moral distinctions officials drew in the past and reflect on the way they articulated the meaning of key moral concepts. Furthermore, by invoking our previous efforts at identifying the demands of justice we appeal to a tradition that we share with our fellow-citizens and that can therefore constitute a basis for our common political life. However, the force of the past on this model is only presumptive. Given that the aim of our engagement with the past is epistemic, it does not make sense to assign any weight to legal practice, when we are confident that it contains a moral mistake. Doing so would take us further away from justice, not closer to it.64 Accordingly, we are not disrespectful to the past when we disregard elements within it which we know are moral mistakes. Quite the opposite, only in this way do we truly vindicate the past, because we genuinely carry on the collective endeavour for justice that it embodies. In Allan’s words: Our conscientious resistance to official demands we reject is only adherence to law as, in our best judgment, we discern its requirements. Purportedly wicked measures, adopted in breach of fundamental principle, are aberrations, calling for interpretative ingenuity and, if necessary, repudiation. But that repudiation is only an appeal to the requirements of practice, consistently followed. We are rejecting misconstructions of the law – departures from practice, correctly conceived – rather than genuine assertions of official authority.65

57 Or, as Alan Brudner has put it, with whose theory Allan’s moral construction shares important similarities: ‘the aim of constitutional theory is to gain the public standpoint from which to sift from the constitutional jurisprudence of liberal-democratic courts principles of political justice for a liberal democracy’. See A Brudner, Constitutional Goods (Oxford, Oxford University Press, 2007) 12. 58 Allan, ‘Practice, Principle, Precedent’ (n 8) 270. 59 ibid 271. 60 ibid 279 61 ibid. 62 ibid. 63 ibid. 64 For this reason Allan criticises Dworkin’s willingness to allow that morally sub-optimal decisions can sometimes shape the best interpretation of the law. See ibid 290ff. 65 ibid 274.

230  Dimitrios Kyritsis For Allan, then, constitutional theory seeks to arrive at true principles of political justice by purifying the contingent, understood as a political community’s institutional and deliberative practice where these principles are latent, ‘known, forgotten, half recalled’. Its ambition is to attain a ‘perfection of the real’,66 and then apply the principles distilled in this process of perfection to determine the correct resolution of future cases. For this reason, it seems particularly vulnerable to the charge that it sacrifices the settlement aspect of constitutions for the sake of elevating their moral force. The assurance model has the opposite aim: according to it, constitutional theory seeks to identify the moral force in the contingent, understood now as the enterprise of governing a diverse and populous society by means of law such that it can gain people’s trust and justifiably command their allegiance, never veering far from the demands of justice but having the moral warrant to enforce even some of its mistakes. Thus, the assurance model furnishes the resources to temper the alleged over-rigidity of principles of justice. Sometimes, it says, we should not give one principle its full effect, but only in the name of other moral principles, those pertaining specifically to legitimacy, as I have described it.67 A second contrast has to do with the way the two models conceive of the bond between participants in constitutional practice. For theorists like Allan the bond is constituted by ‘mutual adherence to a shared tradition’.68 Moreover, in part by virtue of this bond participants in constitutional practice have the moral obligation that they do to comply with the norms of the practice. Insofar as we interpret in good faith from the same stock of constitutional precedents we can rightly demand the cooperation of our fellow-participants despite the existence of ‘sustained and vigorous moral disagreement’.69 Presumably, that demand would be out of place unless we shared the same practice. On the assurance model participation in the same interpretive practice does not suffice to sustain the allegiance of those who disagree with official – and coercively enforced – interpretations. We are more than fellow-travellers in a theoretical exploration, even one with the practically urgent goal of determining the meaning of justice. We constitute a political community in whose name dissenters may be coerced. State coercion raises the stakes. To demand our fellow-citizens’ compliance with the law, we cannot simply appeal to their moral conscience. We must also offer them guarantees that, even when they are on the losing side, they will be adequately shielded from abuses of state power. These guarantees strengthen their bond to their political community. Conversely, without them the political community’s legitimacy is seriously undercut.

66 Brudner, Constitutional Goods (n 57) 1. 67 On my account the moral force of practice is not typically presumptive. Either an institutional decision produces the assurance that legitimacy demands, or it does not. It makes little sense – except, sometimes, on pragmatic grounds – to adopt a defeasible presumption that it has the requisite moral effect. 68 Allan, ‘Principle, Practice, Precedent’ (n 8) 272. 69 ibid 273.

Constitutional Law as Legitimacy-Enhancer  231 V. CONCLUSION

If it is true that morality is necessarily a determinant of the content of constitutional law, as MCT contends, this has immense methodological significance. It affects how we go about reasoning through doctrinal issues and justifying our conclusions. But just as consequential is the answer MCT gives to the question which part of morality determines the content of constitutional law. In this chapter I have argued that MCT should be oriented towards political legitimacy, because the purpose of constitutional law is to bolster legitimacy by furnishing assurances that government will reliably and systematically act in a morally justifiable way. I have also argued that, when constitutional law is understood in this way, it makes sense to say that it stands for a morally binding settlement. Constitutional settlements are morally binding insofar as they advance constitutional law’s moral purpose, the enhancement of legitimacy. But this does not mean that our constitutional past holds absolute sway over the present. Although it prizes settlement, this brand of MCT assigns judges and other officials of today an active role. According to it, the past’s authority is mediated by the various moral considerations that pertain to political legitimacy, and it will often be the responsibility of those officials to weigh up such considerations in resolving constitutional disputes here and now.

232

9 A Positivist and Political Approach to Public Law MICHAEL GORDON*

I. INTRODUCTION

T

his chapter explores the relationship between legal positivism and politics in the context of public law methodology. In general, it might be thought that there is a tension between positivist and political approaches to public law: if we are taking a ‘positivist’ approach we are severing law from a wider set of values, whereas if we are being ‘political’ this must be instrumental or normative. If so, legal positivism is focused on establishing the autonomy of the law, and its detachment from any ‘external’ norms or principles, which might, among other things, be political in nature. Perhaps as a result, there is a further detachment between positivist and political approaches: they tend to be analysed in different legal spheres. Debates about legal positivism primarily occur on the plane of legal philosophy, and are generally preoccupied by the law-morality relationship, whereas debates about the relationship between law and politics primarily occur in the field of public law. A positivist approach may therefore be positioned to provide insights into the nature of law at a general level, while a political approach may have greatest resonance in the more specific contexts of constitutional and administrative law and theory. The aim of this chapter, however, is to outline, analyse and defend an approach to public law which is both positivist and political in its orientation. To do this, the chapter draws on the work of Jeremy Waldron on legal positivism and Martin Loughlin on political jurisprudence. At first sight, while there are important thematic overlaps, their work appears to exhibit the tensions and detachment identified above: Loughlin is a strident anti-positivist whose work is focused on the idea and foundations of public law, while Waldron is a legal and political philosopher whose work sits at a more general level and spans a range of concepts which include, but also go beyond, the core concerns of public law. However, there is arguably an underlying compatibility to their work which in combination provides a template for an approach to public law which is both positivist and political. * I’m very grateful to Stuart Lakin for his comments on my draft text.

234  Michael Gordon This chapter will explore the idea of a positivist and political approach to public law. First it considers the work of Loughlin and Waldron, explaining the apparent tensions between Loughlin’s idea of political jurisprudence and legal positivism, before demonstrating that a shift to the normative conception of positivism defended by Waldron resolves this incompatibility. Second, having established this basis for positivist and political public law, the chapter moves on to examine its nature, considering potential challenges to its coherence. Third, the chapter outlines the potential value of this as a methodology which provides a framework for explanation, justification and critique, before attempting to illustrate the utility of this framework through application to a number of practical examples in the field of public law. The aim of this chapter is not to prove that this positivist and political approach is definitively correct to the exclusion of all other methodologies of public law – indeed, as the variety of methods on offer in this volume illustrates, that would be an essentially futile goal. Instead, my aim in this chapter is to argue that a positivist and political approach is a compelling and attractive one on the spectrum of potential methods. Constitutional methodology is inevitably a selective process, for theorists and legal or political actors alike. In developing this account of an approach to public law which is both positivist and political, my objective is to explain how these two important strands of constitutional studies interconnect, and how they can shape and enhance our understanding of the subject. This is, therefore, an exercise in theorising rather than grand-theorising – in Carol Smart’s terms, it is ‘an attempt to make sense of experience or social order’ derived from ‘concrete’ knowledge, and not to ‘proclaim … unique truth above all other … systems of thought’.1 In making explicit the basis, nature and utility of a positivist and political approach to public law, the consequences and the opportunities of adopting this methodology will hopefully become clear, as well as opened up for further evaluation. For in employing an approach based on law’s distinctive nature as a social practice, but which is crucially located in a wider system of political activity and principles, it is essential to recognise that, ultimately, public law is a discipline which is constructed and reconstructed to serve the purposes we adopt for it. II.  A BASIS FOR POSITIVIST AND POLITICAL PUBLIC LAW

This chapter builds on claims developed in an earlier article exploring the work of Martin Loughlin and Jeremy Waldron.2 The article sought to reconcile the positions of Loughlin and Waldron to establish a conceptual basis for a positivist and political approach to public law. Before developing the argument further in this chapter, it is first necessary to set out the basis of this reconciliation of Loughlin and Waldron. There are attractive elements to both of their theoretical accounts of the nature of law, yet there is also an apparently fundamental incompatibility between the different approaches they defend. 1 C Smart, Feminism and the Power of Law (London, Routledge, 1989) 71. 2 M Gordon, ‘A Basis for Positivist and Political Public Law: Reconciling Loughlin’s Public Law with (Normative) Legal Positivism’ [2016] Jurisprudence 449–477.

A Positivist and Political Approach to Public Law  235 A.  Loughlin’s Theory of Public Law as Political Jurisprudence Loughlin’s conception of public law as an exercise in ‘political jurisprudence’ is pioneering in explaining the inherently political underpinnings of constitutional order and legal authority. He has developed the idea of ‘public law as political jurisprudence’ across a number of books,3 but a recent restatement of the theory illuminates a key aspect of Loughlin’s work which is of primary relevance here: that his account of public law is explicitly anti-positivist. In the Introduction to Political Jurisprudence, Loughlin foregrounds the idea that his theory is diametrically opposed to legal positivism: ‘Political jurists explicitly reject legal positivism’.4 Political jurisprudence is distinguished from other theories of law because it maintains that ‘the key to understanding the nature of legal order’ is ‘the question of how political authority is constituted’.5 This, according to Loughlin, is the ‘central question’.6 But it is one which legal positivism (‘the dominant school’) has ‘shunted to the margins of the discipline’.7 For Loughlin, a legal positivist approach means ‘the authority of the legal order is presupposed’, with law treated as ‘an object to be explained empirically or as a logically self-authorizing set of norms’.8 Consequently, ‘the province of jurisprudence narrows’ and becomes ‘limited to the task of explaining the structural form of positive law’.9 Loughlin therefore rejects what he considers to be the focus of a legal positivist approach to the understanding of law, but his objection goes deeper. His approach to public law as a form of political jurisprudence also challenges the core thesis at the heart of legal positivism – the idea that law is separable from morality, and therefore that the validity of legal rules within a legal system does not inherently depend on their merit, whether judged against principles of morality, justice, or any other external normative values.10 Loughlin, however, rejects ‘the separation postulated by legal positivism’, arguing instead that ‘fact and value unite in an experienced political reality’.11 His alternative method is interpretive – it fuses ‘description and evaluation in a way that positivists are unwilling to recognise’.12 This is inevitable and unavoidable, because according to Loughlin, ‘[w]e are always interpreting’ when we are ‘concerned to generate meaning’.13 Yet Loughlin’s approach to interpretation is crucially distinct from the moralised approach to ‘constructive interpretation’ associated with

3 See especially M Loughlin, The Idea of Public Law (Oxford, Oxford University Press, 2003) and Foundations of Public Law (Oxford, Oxford University Press, 2010). 4 M Loughlin, Political Jurisprudence (Oxford, Oxford University Press, 2017) 3. 5 ibid 2. 6 ibid 4. 7 ibid 10. 8 ibid 2. 9 ibid. 10 See, eg, J Austin, The Province of Jurisprudence Determined (1832); HLA Hart, ‘Positivism and The Separation of Law and Morals’ (1958) 71(4) Harvard Law Review 593–629; HLA Hart, The Concept of Law (Oxford, Oxford University Press, 1961); J Coleman, ‘Negative and Positive Positivism’ (1982) 11(1) The Journal of Legal Studies 139–164. 11 Loughlin, Political Jurisprudence (n 4) 9. 12 M Loughlin, ‘Theory and Values in Public Law: An Interpretation’ [2005] Public Law 48–66, 62. 13 ibid 63.

236  Michael Gordon Dworkin, framed around the idea that there is a right answer to legal questions which must present the law in its most appealing light.14 Loughlin criticises this natural law infused mode of jurisprudence as flawed because it ‘assumes the moral authority of an idealized legality’, converting ‘legal interpretation into a type of moral reasoning’ based on ‘a set of universal moral axioms’.15 So Loughlin’s theory of public law as political jurisprudence is anti-positivist, but not in the usual way. Loughlin’s approach, in contrast, is rooted in ‘the political world’. From the rejection of positivism and moralised legalism, Loughlin concludes that ‘the relationship between law and authority can neither be presupposed nor assumed to rest on a set of universal values’.16 Instead, public law, understood as an exercise in political jurisprudence, views authority as ‘generated through a political process that draws people together in a common understanding’.17 This produces a ‘special type of political power’ based on ‘the capacity of a collective singular, “the people”’, which is ‘consolidated through institutionalisation’ in a system of public law.18 The ‘constitution of political authority’ therefore ‘determines the nature of law’, but as the constitution of authority is ‘continually contested’, there is no singular conception of law. Instead, to understand public law as an exercise in political jurisprudence means to acknowledge the different ways of conceptualising the political role of law. In his most recent formulation, Loughlin calls these conceptions ‘Law’, ‘Legality’ and ‘Superlegality’.19 For Loughlin, ‘Law’ is found in ‘a set of practices operating in a discrete sphere of government’ reflected in the experience, reason and precedents of legal actors.20 In contrast, ‘Legality’ is law as an ‘instrument of political power’ to be deployed by the state through legislation.21 And ‘Superlegality’ is ‘a set of abstract principles that limits the range of legitimate action of governing majorities’ rooted in fundamental norms derived from political ideas of liberty and equality.22 The relative priority between these different conceptions of law will vary between different legal systems, but is to be determined as a matter of ‘political judgment’.23 The function of public law is to frame and facilitate the competition between these different conceptions of law: its ‘task is to negotiate between the various conflicting accounts of political right that form part of its evolving discourse’.24 This is a distinctive and challenging account of the nature of public law, offering a crucial insight into the political foundations of law. Loughlin’s sophisticated account demonstrates that public law has an essential political character, which is primarily

14 See, eg, R Dworkin, Law’s Empire (Oxford, Hart Publishing, 1998). 15 Loughlin, Political Jurisprudence (n 4) 4, 3. 16 ibid 4. 17 ibid. 18 ibid. 19 There is overlap between this new scheme and Loughlin’s earlier account of ‘law as custom’, ‘law as command’, and ‘law as right’ in M Loughlin, Public Law and Political Theory (Oxford, Clarendon Press, 1992). 20 Loughlin, Political Jurisprudence (n 4) 6. 21 ibid 6. 22 ibid. 23 ibid 7. 24 M Loughlin, Foundations of Public Law (Oxford, Oxford University Press, 2010) 164.

A Positivist and Political Approach to Public Law  237 evident in its key role in generating constitutional authority. The political dimensions of law also translate into each of Loughlin’s three specific conceptions of law’s mode of operation (‘law’, ‘legality’, ‘superlegality’) revealing the complexity and variability of the relationship between the legal and the political. But while Loughlin’s work has been pioneering in explain the fundamentally political nature of public law, it is not clear that his approach to political jurisprudence is incompatible with legal ­positivism. There are two key reasons for this – the first concerns the positivist method, and the second concerns the function of law within Loughlin’s account of political jurisprudence. B.  Political Jurisprudence and Normative Legal Positivism The first reason to doubt that political jurisprudence is inconsistent with legal positivism is rooted in the idea that positivists must separate considerations of ‘fact and value’ (to use Loughlin’s words) at a methodological level. This may be the aspiration of the analytical positivism of HLA Hart and his successors,25 but this does not mean all legal positivist work must adopt the same stance. Jeremy Waldron’s explicitly normative account of legal positivism provides a compelling alternative. According to Waldron, the choice of a positivist theory of law (of some sort) ‘cannot credibly be presented as a matter of pure “analysis”’; instead, for the choice ‘to be intelligible, it must be motivated’.26 Waldron’s account of normative positivism is based on the idea that law might have a different relationship with moral values at the ‘retail level’ and at the ‘wholesale level’.27 This distinction is designed to capture the difference between decisions concerning the validity and meaning of specific legal rules at the ‘retail level’, and decisions concerning the purpose and operation of law as a social institution at the ‘wholesale level’. It is to draw a distinction between ‘the law’ and ‘law’ – the rules of law within a system, as compared with the nature of a legal system itself. And for the normative positivist, an insistence on the separation of law and morality at the ‘retail level’ – the level of determining the validity and meaning of specific legal rules – is pursued because at the ‘wholesale level’ there are normative reasons to favour this separation. There would be a number of different ways of justifying this separation, but in broad terms it might be explained on the basis that a legal system can best fulfil its coordinating functions if ‘it is set up in a way that enables people, by and large, to determine what the law is on a given subject without having to exercise moral judgment’.28 Equally, normative positivism is also the legal theory which provides

25 See HLA Hart, Concept of Law 2nd edn (Oxford, Oxford University Press, 1994), 239–240. There are also reasons to doubt whether Hart’s theory was successful in avoiding reliance on ‘evaluative’ claims; see, eg, SR Perry, ‘Hart’s Methodological Positivism’ in J Coleman (ed), Hart’s Postscript: Essays on the Postscript to the Concept of Law (Oxford, Oxford University Press, 2001), 323. 26 J Waldron, ‘Normative (or Ethical) Positivism’ in J Coleman (ed), Hart’s Postscript: Essays on the Postscript to the Concept of Law (Oxford, Oxford University Press, 2001), 433. 27 ibid 415. 28 ibid 430.

238  Michael Gordon the conceptual underpinning for Waldron’s famous argument about the role of law in circumstances of pervasive disagreement about justice.29 If there is inevitable disagreement about the requirements of justice among a substantial group of citizens, the unilateral judicial determination of these questions through constitutional review is exposed as undesirable. For it overlooks the virtues of democratic debate and of resolving disagreement through collective decision-making, and instead imposes one conception of justice over others on the basis of legalised moral argument among a select, unrepresentative group. This defence of legal positivism as a normative thesis is not a marginal or mutant strand of positivist thought. Indeed, the tradition of ‘political positivism’ within which Waldron’s work is located has deep roots.30 Due to its similarity with the approach of the positivist forebears Bentham and Hobbes,31 it may actually represent the ‘classical’ tradition of legal positivism, which ‘saw theorizing about law as part of theorizing about morals and politics’.32 This connection between positivism and political jurisprudence is even explored in depth in Loughlin’s own writing on Hobbes. In a recent chapter length study, Loughlin recognises that ‘[m]odern legal positivism – the conviction that positive law forms an autonomous system of law that includes its own criteria of right and wrong, just and unjust – has its origins in [Hobbes’] work’.33 While Loughlin reconceptualises Hobbes’ work ‘as an exercise of political jurisprudence – of addressing the issue of legitimacy and not simply accepting the authority of positive law as a postulate of thought’,34 this does not eliminate the positivist underpinnings of this theory of law. Instead, it reveals that the positivism of Hobbes was founded in an account of ‘political right’. For as Loughlin convincingly argues, Hobbes viewed natural law as only binding ‘when expressed in the form of positive law’, while he also ‘converted’ the laws of nature ‘into precepts of political right’.35 As Loughlin argues, this ‘transformation of natural law was of critical importance in refashioning the instruments of modern political rule’, based on the idea that ‘to realize liberty and equality … humans must first be subjected to government’.36 But crucially, this is a scheme of political jurisprudence which depends on the positivisation of law in order to establish the basis of the modern state. Ultimately, therefore, Loughlin’s account fails to reflect the methodological variation which exists within legal positivism, and as a result, underestimates the important political dimensions of a specifically normative positivist approach. As such, we can start to see that it is wrong to dismiss every approach to public law which is positivist as incompatible with the insights derived from Loughlin’s political jurisprudence. Indeed, on the normative approach it is crucial that the way we understand

29 J Waldron, Law and Disagreement (Oxford, Oxford University Press, 1999). 30 See, eg, R Stacey, ‘Democratic Jurisprudence and Judicial Review: Waldron’s Contribution to Political Positivism’ (2010) 30(4) Oxford Journal of Legal Studies 749–773. 31 See generally J Bentham, Of Laws in General (HLA Hart ed, London, Bloomsbury Publishing, 1970) 1; T Hobbes, Leviathan (London, Penguin, 1985), ch 17, [391]. 32 See D Priel, ‘Toward Classical Legal Positivism’ (2015) 101 Virginia Law Review 987–1022, 992. 33 Loughlin, Political Jurisprudence (n 4) 26. 34 ibid 27. 35 ibid 30, 33. 36 ibid 32–33.

A Positivist and Political Approach to Public Law  239 law should reflect its distinctive nature as a means of coordinating human behaviour, but one which operates within a wider system of political government. As Waldron argues, ‘we should expect what we say about what law is to have some sort of connection to why law is seen as an importantly distinct mode or aspect of governance’,37 and this approach seems largely consonant with that underpinning an account of public law as political jurisprudence. There is a second reason to doubt the idea that political jurisprudence is inherently incompatible with legal positivism, beyond the methodological similarities discussed above. While on Loughlin’s account there can be no singular conception of law, each of the three competing understandings of law which he sets out are compatible with legal positivism. ‘Law’, ‘legality’ and ‘superlegality’ are all non-moral conceptions of how the law may operate within a political system – ‘law’ representing ‘traditional practice’, ‘legality’ representing ‘legislative will’, and ‘superlegality’ representing ‘universal reason’.38 Law as practice is ‘artificial reason’ not moral reason, legality as legislative will represents a collective political decision, and superlegality is ‘an expression of a society’s fundamental political principles’ rather than universal moral norms.39 In substance, therefore, each of these different relationships between law and politics can be understood within a positivist framework – in each conception law reflects a different mode of politics, but law remains distinct from, rather than becomes a mere function of, political (or moral) values. Law takes a different form depending on the different understanding of politics in play, but whether legal precedent, legal statute or legal principle, each of these conceptualisations can be seen to recognise the independent force of law, with its status and authority separate from the other kinds of obligations and powers which exist within a society. Indeed, this is acknowledged by Loughlin, who observes that ‘[o]perating in these ordinary ways, law evolves as an autonomous practice’.40 Yet even if it is right that these three different substantive relationships between law and politics are compatible with the overall idea of law as a positivised construct, there is a further challenge. Loughlin also argues there is a continuous conflict between these different conceptions of law, which ‘cannot be resolved by legal method’ but instead requires ‘the exercise of political judgment’.41 As such, ‘rather than making any claim to universal truth, political jurisprudence explains how a particular way of conceiving the political is conceived in a particular regime’.42 Loughlin is surely correct that the ‘interminable interplay between the relative authority of text, precedent and principle’ has no absolute solution, and that a balance between them can be reached in different ways in different legal systems.43 But legal positivism is not rendered irrelevant by the existence of this potential conflict. For even if the balance between text, precedent and principle cannot be determined by a scientific legal method, a normative account of legal positivism would view this as a matter

37 Waldron,

38 Loughlin, 39 ibid

5, 6. 6. 41 ibid 6, 5. 42 ibid 9. 43 ibid 5. 40 ibid

‘Normative (or Ethical) Positivism’ (n 26) 420 (emphasis in original). Political Jurisprudence (n 4) 6.

240  Michael Gordon of making normative choices concerning the constitutional structure in a given legal system (at ‘the wholesale level’, to import Waldron’s terminology). Moreover, while normative positivism can account for the existence of the conflict and the need for (political) choices about the characterisation of law within a specific state, it can also establish the scope of the dispute. For it can suggest that the texts, precedents and principles in play in constitutional interpretation should be those texts, precedents and principles recognised as legal according to the public law of the state in question. Indeed, the fact that these are different conceptions of law which frame constitutional authority and decision-making within the state – rather than, for example, conceptions of morality, justice or religion – is inherent in Loughlin’s theory. All three conceptions – ‘law’, ‘legality’ and even ‘superlegality’ – have an essential and specific legal character. And so while the way in which these different concepts fit together may be a matter for normative political judgement, at the ‘retail level’ of constitutional decision-making the relevant sources and values are those recognised as internal to the system of law, while external values which are not grounded in precedent, legislation or legal principle lack authority and are excluded, at least from a determination about the current state of the law. This does not mean an end to controversy about the answers to legal questions, or the elimination of conflict between different forms of law. But a (normative) positivist approach is consistent with the existence of different ideas of law, all operating and competing for supremacy within a particular legal system. And it can narrow the scope of controversy at the ‘retail level’ of constitutional decision-making, by ensuring that those disputes are focused on norms, sources and values which are determined as legally relevant within the legal system, rather than exposing public law to a moral free-for-all. This approach may have some outer limits, and does not become an automatic or mechanical process of rule application, depending on which of Loughlin’s conceptions of law is dominant within a particular state. But in large part it can shape the choices of legal actors, establishing for them the norms which are material to reaching valid decisions about the law in operation in their society. Fundamentally, then, the idea that political jurisprudence generates no singular conception of law is consistent with – and, indeed, can enrich – a normative positivist legal framework. There are consequently good reasons to believe that Loughlin’s theory of political jurisprudence is compatible with a normative account of legal positivism, rather than standing in opposition to it. That may feel like a relatively thin position, but it is significant in establishing a basis for a positivist and political approach to public law. Indeed, it may even understate the substantive overlaps between Loughlin’s work and normative positivism. For example, if we think about Waldron’s normative approach to legal positivism specifically, which views the insulation of law from external values at the point of its application as reflecting a democratic political choice, we can identify a number of strong affinities between these two ostensibly competing theories. In particular, both approaches highlight the contingency of legal rules, and the extent to which law is simply a product of real political choices and circumstances, rather than derived from a rational repository of superior moral values.44 44 See, eg, Loughlin, Political Jurisprudence (n 4): ‘All phenomena … are products of history’ (8); ‘political jurisprudence is both anti-rationalist and relative’ (3). See, eg, J Waldron, ‘Constitutionalism: A Skeptical View’ in Political Theory (Cambridge, Mass, Harvard University Press, 2016).

A Positivist and Political Approach to Public Law  241 And both theories embrace and foreground the idea that disagreement is a central feature of legal and political experience, rather than an aberration to be pushed to the fringes of our understanding of constitutional concepts.45 In that sense, we can begin to see that while these theories start from quite different places, in combination they can provide a basis for a positivist and political approach to public law. III.  THE NATURE OF POSITIVIST AND POLITICAL PUBLIC LAW

To this point, the chapter has considered the possibility of a theory of public law which is simultaneously positivist and political, grounded in the work of Loughlin and Waldon. But if we go beyond thinking about this as a mere possibility to an actual method, what does a combination of Loughlin’s political jurisprudence and Waldron’s normative legal positivism offer? Loughlin may rely on a reductive view of legal positivism which underestimates its implications as a normative theory – yet a normative account of legal positivism can still be enhanced by drawing from the deeper political foundations of Loughlin’s rich work on public law as political jurisprudence. This demonstrates the fundamental need to examine public law in its political context, in light of its political origins and the constitution of its authority, and not to assume that positivism is only concerned with the positive law. Instead, we also need to think about the advantages and disadvantages of rule by (positive) law – or as Waldron puts it, ‘to grasp the desirability of being governed in certain ways (eg by law) rather than other ways (eg by decree or managerial direction)’.46 In my view, this can only be done in the context of an account of the authority and legitimacy of positive (public) law, which we can generate from Loughlin’s theory of political jurisprudence. In this section we therefore consider the nature of a positivist and political approach to public law: how it might be defined and some potential objections to the coherence of this approach. A.  Defining a Positivist and Political Approach A positivist and political approach to public law is based on a normative account of legal positivism, understood in a political context which is necessary to explain the authority, functions and limitations of law. On this approach, law can be seen to retain its role as a distinctive social construct for structuring and organising human activities, but its political dimension is essential to understand the context in which law operates, the manner and extent to which legal arrangements are justified, and to provide a basis for critique of the legal system and its operation. Positive law and the political framework in which that law exists 45 See, eg, Loughlin, Political Jurisprudence (n 4): ‘For political jurists these disputes are not just symptoms of crisis; they are intrinsic to modern political and legal practice’ (4). See, eg, Waldron, Law and Disagreement (n 29). 46 Waldron, ‘Normative (or Ethical) Positivism’ (n 26) 427.

242  Michael Gordon are inherently and unavoidably linked, but this connection is not at the point of assessing the legality validity of specific rules. Instead, the connection is broader and more fundamental. First, as Loughlin’s work demonstrates, there is a constitutive relationship between positive law and politics – public law constitutes and is constituted by political necessity within the state. And second, as Waldron’s work illustrates, this relationship requires normative judgements to be made at a systemic level – constitutional arrangements can be structured in different ways, and so choices about how positive law will take shape in a particular political context must be determined by reference to political principles. As a result, the constitutive relationship between positive law and politics does not displace normative questions about the structure of constitutional systems – instead it demands them, given the variation possible in the way in which public law is operationalised in constitutional reality. As Loughlin shows, there are political judgements to be made about the balance between different conceptions of law, legality, or superlegality in a specific legal order and this is inevitably in part a normative task. Equally, the positivist separation of legal validity from non-legal values does not mean that law exists in a vacuum – instead it necessitates that law must be conceptualised as part of a wider political system for this choice of separation to be meaningful. Whether the positivist instinct to insulate legal decision-making from abstract ideas of morality or justice at the point of determining the legal validity of legal rules derives from the pursuit of basic legal certainty, or to reflect the virtues of a democratic law-making process, it is this political rationale which makes a defence of legal positivism constitutionally intelligible. A positivist and political approach can therefore provide us with an account which deals with how the validity of specific legal rules is determined, and the wider authority and functions of law as a political construct. Crucially, the positivist and political approach demonstrates that specific questions about validity are essentially connected to broader questions about the political purposes and (potential) legitimacy of law. As Waldron’s account of normative positivism demonstrates, there is no need to assume that the relationship between law and politics must be the same at the ‘retail level’ of specific legal decision-making as at the ‘wholesale level’ of an overarching theory of law. But it does illustrate the importance of considering the relationship between the ‘retail’ and the ‘wholesale’ to have a full understanding of law. Or, as we might characterise this for application in a public law context specifically, what Waldron calls ‘retail’ and ‘wholesale’ may equally be thought of as representing the different considerations and questions in issue at the level of constitutional practice and the level of constitutional design. What I have set out here is inevitably a broad template for an approach to public law. It is also obviously contestable in a number of respects. However, this positivist and political approach provides us with the basic underpinning for a framework to engage with public law. As I will argue shortly, that framework has three elements: it provides us with the tools to explain, to justify and to critique the law. Before we turn to assess the value of this framework, and the positivist and political perspective which generates it, we must first consider some broader objections to the coherence of the approach I am defending in this chapter.

A Positivist and Political Approach to Public Law  243 B.  The Coherence of a Positivist and Political Approach – Some Potential Objections There are three potential objections to the coherence of a positivist and political approach I want to explore here. The first is whether normative legal positivism is itself a coherent theory of law. The second is whether the method considered here is better understood as a form of positivism or of interpretivism. The third is whether this positivist and political approach is simply a rebranded version of ‘political constitutionalism’. i.  Is Normative Positivism Coherent? While I have argued above that normative accounts of legal positivism have potentially deep roots in the ‘classical’ positivist theories of Bentham and Hobbes, the idea of normative positivism remains controversial. This is especially the case in the dominant modern analytical tradition of legal positivism. From the perspective of descriptive positivism, normative positivism might be viewed as an irrelevant or parasitic gloss: either there is a conceptual separation between law and morality as a matter of empirical fact, or there is not. As Coleman has argued, ‘normative positivism is a thesis about legal positivism, not a version of legal positivism’.47 As such, normative positivism might be thought to add little other than a series of reasons for why an inevitable state of affairs is normatively attractive. The difficulty with this position is that it views normative theories as hierarchically inferior to descriptive or analytical theories, rather than simply different from them. The very starting point of a normative account of positivism of the kind developed by Waldron is that a descriptive account of the world as it is cannot be completely severed from the reasons which motivate that position – instead the argument is ‘normative all the way down’, and ‘does not really presuppose any purely descriptive, conceptual, or non-normative phase at all’.48 The dispute about whether a normative account of positivism can be coherent is therefore a function of a wider dispute concerning the very nature of legal theory. In practice this is likely to be irresolvable, given the persistence of real world disagreement about the truth of competing legal theories, whether internal to legal positivism, or between positivists, natural lawyers and others.49 In such circumstances, it seems problematic to entirely discount one set of arguments about the nature of law as incoherent – at the very least, the question of the coherence or otherwise of normative theories must become a central part of the debate about law, rather than imposing descriptive truth as a precondition cutting off further discussion.

47 J. Coleman, ‘Negative and Positive Positivism’ (1982) 11 Journal of Legal Studies 139–164, 147. 48 Waldron, ‘Normative (or Ethical) Positivism’ (n 26) 421. 49 For a sample reflecting the endurance of these debates, see, eg, N Simmonds, Law as a Moral Idea (Oxford, Oxford University Press, 2007); M Kramer, Where Law and Morality Meet (Oxford, Oxford University Press, 2008); S Shapiro, Legality (Cambridge, Mass, Harvard University Press, 2011); K Rundle, Forms Liberate: Reclaiming the Jurisprudence of Lon L Fuller (Oxford, Hart Publishing, 2013).

244  Michael Gordon There is a second layer to this critique, however, which follows from the claim that the truth of legal positivism as a descriptive thesis is logically prior to any normative discussion of positivism. As Gardner argues, from this perspective normative legal positivists become simply ‘positivity-welcomers’ who reach the ‘rather self-congratulatory conclusion’ that ‘the positivity of law is not only something we have to live with, but also something we can be proud of’.50 Yet as Waldron notes, the normative positivist does not have ‘to view law as a good thing’.51 Moreover, to consider the normative implications of positivism is not to assume that legal positivism provides a perfect model for legal order in a society. Instead, it is to undertake a relative assessment of the advantages and disadvantages of a system based on positive law. And by identifying the limitations of law as a means of organising human activity, a normative approach may suggest that legal positivism is simply the least-worst conception of legal practice available to us. In this sense, engagement with normative argument, and by extension a normative evaluation of positivism, can be justificatory or critical in nature (or indeed, incorporate elements of both justification and critique). However, this raises a final challenge: is critical positivism even possible? As Carol Smart argues in a radical challenge to the idea of a specifically feminist jurisprudence, such an approach to critical legal theory fails to ‘de-centre law’ – ‘[i]t encourages a “turning to law” for solutions, it fetishizes law rather than deconstructing it’ – which ‘ends with a celebration of positivistic, scientific feminism which seeks to replace one hierarchy of truth with another’.52 Even a feminist theory of jurisprudence would therefore take on ‘the mantle of a positivism which assumes that there must be an ultimate standard of objectivity’.53 This is a powerful critique, yet if positivism is understood as existing in a deeper political context, there is no reason to assume it provides a complete hierarchy of objective truth. Rather, positivism may offer an initial basis for critical understanding of the effects of law, by exposing the state and character of legal order within a given society. An account of legal positivism which emphasises the contingent authority of law should therefore provide a basis to facilitate rather than suppress critical deconstruction. Indeed, it is difficult to see how law can be redeemed (if that is possible) from a feminist or any other critical perspective other than through political action, rather than through attempts to make legal theory internalise specific values. In this sense, the goal of ‘de-centring’ law can arguably best be achieved through a positivist and political approach, which offers relative clarity about the content of law so that its full effects can be politically evaluated and confronted, rather than legal outcomes fetishized. ii.  Positivism or Interpretivism? A second and related challenge concerns whether the approach defended in this chapter can actually be understood as a positivist theory at all, or whether normative



50 J

Gardner, ‘Legal Positivism: 5½ Myths’ (2001) 46 American Journal of Jurisprudence 199–228, 205. ‘Normative (or Ethical) Positivism’ (n 26) 428. 52 Smart, Feminism and the Power of Law (n 1) 88–89. 53 ibid 71. 51 Waldron,

A Positivist and Political Approach to Public Law  245 legal positivism is better viewed as a form of interpretivism. There are two different varieties of interpretivism potentially in play here. The first is the moral interpretivism of Dworkin, according to which we choose a theory of law on the basis of whether it fits with our understanding of the practice, and also whether it presents that practice in a morally attractive light.54 There may be some overlap between this approach and that of normative positivism, in so far as a normative assessment is an important part of defending a legal theory. But unlike Dworkin’s approach, normative positivism is not necessarily rooted in a constructive approach to interpretation – the aim is not to develop a theory which presents the practice of law in its best possible light. Instead, as discussed above, a normative account of legal positivism can be justificatory and/or critical in nature, distinguishing it from the interpretivism of Dworkin which has an overriding moral purpose. Furthermore, there is a substantive difference – as Waldron argues, unlike natural law theories, ‘normative positivism need not commit itself to any particular meta-ethics, nor to any particular account of how we arrive at an understanding of the goods or values that law serves’.55 Whether or not Dworkin’s interpretivism is classified as a natural law theory, this distinction applies equally to his theory of law as integrity, which operates to serve a specific conception of ‘equal concern and respect’.56 The second version of interpretivism relevant here is the ‘non-moral’ approach relied on by Loughlin.57 Loughlin’s version of interpretivism rejects freestanding moral arguments, instead viewing law as a social phenomenon like any other, which must be interpreted to be understood. Loughlin argues that the ‘assumption that there are indisputable facts in this field that can be acquired by objective empirical investigation is erroneous: we are constantly seeking to fit this evidence to our preexisting assumptions about the subject’.58 Instead, public law values such as ‘liberty, democracy and accountability’ can only be determined by ‘the way in which they are set to work within a defined field’ – these ideas ‘take on more precise meanings’ only when they are located in ‘specific social and historical contexts’.59 The values which animate public law do not therefore derive from abstract principles, but the principles obtain meaning from their deployment in practice. As noted above, there are important methodological parallels between normative positivism and Loughlin’s interpretive approach. But there is an important difference too, which flows from a positivist recognition of the distinctive nature of law. While Loughlin views law as interpretative in the same way as history and other social sciences – ‘history is made by historians and their task is thoroughly interpretative’60 – there are important differences between law and history. Unlike

54 Dworkin, Law’s Empire (n 14) 90. 55 Waldron, ‘Normative (or Ethical) Positivism’ (n 26) 418. 56 See, eg, Dworkin, Law’s Empire (n 14) 381; R Dworkin, Taking Rights Seriously (London, Duckworth, 1978). 57 See Goldsworthy, ch 2 in this volume. 58 M Loughlin, ‘Theory and Values in Public Law: An Interpretation’ [2005] Public Law 48–66, 63. 59 ibid 65. 60 ibid 63.

246  Michael Gordon history, law has direct and concrete consequences for the actions of humans, and authoritative institutional actors designated with the power to conclusively resolve disputes about the meaning or application of law for these purposes. Interpretations of history can be continuously debated and updated. However, law makes demands on our behaviour and while we can disagree about the demands it is making, the courts are systematically empowered to impose a final interpretation on everyone else. Normative positivism tries to recognise both of these aspects of law’s nature – that at a structural (‘wholesale’) level law is a concept which is subject to interpretation like any other social science, but at the (‘retail’) level of practice legal rules have a definitive existence, and not all interpretations of them are equally authoritative. Is this combination – positivism in terms of the internal operation of the legal system, interpretivism in terms of the choice of system itself – better understood as ‘positivism plus’ or ‘interpretivism minus’? While it could arguably be the latter, the core idea of normative positivism is that it is based on an account of the relationship between the retail level and wholesale level, or between the nature of specific legal rules and the nature of law as a social construct. It would therefore be wrong to suggest this is an interpretive theory by focusing on this dimension in isolation from the overall method. As such, an approach which is focused on understanding positivism in its political context for political reasons is not best characterised as interpretive, but as grounded in normative legal positivism. It may also be worth asking whether the label applied to this theory has great significance. Waldron questions whether it really counts against normative legal positivism if it can ‘no longer easily be distinguished from modern natural law theory’, concluding that the dilution of these distinctions ‘should not be counted as a loss’, as the ‘preservation’ of the positivism-natural law divide ‘is not to be regarded as a methodological imperative’.61 Yet while there may be little point in agonising over the retention of categories simply for the sake of having categories, there is still a signalling function to be fulfilled through the use of appropriate theoretical labels. And from this perspective, it seems more important to emphasise the part of the theory which is a normative choice (here, to defend a positivist approach within a legal system) over the part which is arguably unavoidable (the normativity of the idea of law as a social construct). While ‘legal positivism is a label associated with a very broad cluster of theories’,62 it is better to view normative positivism as part of that tradition, rather than absorbing it into the (arguably even broader) category of interpretivism, which would risk diminishing its specific positivist features. iii.  Is This Just Political Constitutionalism? A final question concerning the coherence of a positivist and political approach is whether this is just a rebranding of political constitutionalism. And if so, does that simply make it an elaborate way of defending something that already exists?



61 Waldron, 62 ibid

432.

‘Normative (or Ethical) Positivism’ (n 26) 418.

A Positivist and Political Approach to Public Law  247 The idea of political constitutionalism has featured prominently in public law scholarship in recent decades. However, it remains contested – from outside the tradition and within it, both conceptually and, increasingly, politically.63 Loughlin in particular has been critical of the idea that a ‘so-called’ theory of political constitutionalism could be engineered from the functionalist writings of John Griffith, most famously in his influential lecture on ‘The Political Constitution’,64 and then subsequently juxtaposed with a model of legal constitutionalism.65 Loughlin is clearly right that theoretical debates which create a superficial competition between law and politics miss the key question, which is ‘not whether we have a legal or political constitution: it is how the idea of law within the political constitution (i.e. the constitution of the polity) might best be conceptualized’.66 Yet this does not necessarily mean the idea of political constitutionalism must be rejected – instead, it can be seen as attempting to acknowledge and frame the relationship between law and politics. This is a relationship where law is not the exclusive or inherently superior mode of constitutionalism, but which emphasises ‘the political character of the constitution, and the potentially constitutional character of the political’.67 However, there remains considerable disagreement about the parameters and purposes of political constitutionalism. It is a broad school of thought, with few (if any) absolute unifying tenets. Crucially for present purposes, political constitutionalism is not inherently positivist, nor is it unquestionably a normative theory. There are nevertheless some strong connections between legal positivism and political constitutionalism. For example, in his famous Chorley Lecture, Griffith defined his approach as follows: I am arguing then for a highly positivist view of the constitution; of recognising that Ministers and others in high positions of authority are men and women who happen to exercise political power but without any such right to that power which could give them a superior moral position; that laws made by those in authority derive validity from no other fact or principle, and so impose no moral obligation of obedience on others.68

That Griffith’s approach was not just positivist, but ‘highly positivist’, suggests the choice of a version of positivism on normative grounds, rather than the simple acceptance of a fixed idea of positivism as a descriptive inevitability. Yet even so, political constitutionalism in its current form exists independently of any specific position adopted by Griffith,69 and there is little to be gained from attempting to divine a ‘true’ theory from his work.

63 The literature is considerable; for a recent sample, see, eg, the special edition of the King’s Law Journal edited by Gee and McCorkindale, ‘The Political Constitution at 40’ (2019) 30(1) KLJ. 64 JAG Griffith, ‘The Political Constitution’ (1979) 42 MLR 1–21. 65 M Loughlin, ‘The Political Constitution Revisited’ (2019) 30(1) KLJ 5–20, 18. 66 M Loughlin, ‘Towards a Republican Revival? (2006) 26 OJLS 425, 436. 67 M Gordon, ‘Parliamentary Sovereignty and the Political Constitution(s): From Griffith to Brexit’ (2019) 30(1) KLJ 125–147, 125. 68 Griffith, ‘The Political Constitution’ (n 64) 19. 69 For contrasting accounts, see, eg, A Tomkins, Our Republican Constitution (Oxford, Hart Publishing, 2005); R Bellamy, Political Constitutionalism: A Republican Defence of the Constitutionality of Democracy (Cambridge, Cambridge University Press, 2007); G Gee and G Webber, ‘What Is a Political Constitution?’ (2010) 30 OJLS 273.

248  Michael Gordon The positivist and political approach defended in this chapter could therefore be understood as a further instantiation of political constitutionalism, within that wider and increasingly fragmented school. It speaks to the broader aims of political constitutionalism in attempting to explain the nature of the relationship between law and politics in a way which recognises the relevance and limitations of both, and the essential connections between them. But it also attempts to expand on the claims of political constitutionalism, especially the legal theoretical claims which might underpin it, which have often been overlooked. Equally, however, this positivist and political approach is certainly not an attempt to provide a definitive statement of political constitutionalism, which is no more likely to be achievable than providing a definitive statement of the nature of law. In that sense, the positivist and political approach explored in this chapter stands (or falls) on its own independent merits, entirely apart from whether political constitutionalism is thought to be a meaningful strand of constitutional theory. IV.  THE VALUE OF A POSITIVIST AND POLITICAL APPROACH TO PUBLIC LAW

I have argued in section II that the basis for a positivist and political approach to public law can be found in a reconciliation of the work of Loughlin and Waldron, and in section III I have outlined the essential nature of this approach. But how can this method be used to enhance our understanding of public law? The final aim of this chapter is to respond to this question, by setting out the value of this approach for analysing public law. This section therefore offers an initial account of the value of a positivist and political method, in two parts. First, it establishes a framework for public law analysis which flows from the adoption of a positivist and political approach. Second, it explores the utility of that framework, by applying it to illustrative examples of public law issues. A.  A Positivist and Political Analytical Framework An approach which combines normative legal positivism and political jurisprudence provides a basis for the explanation, justification and critique of public law. It generates the potential to develop a distinct analytical framework for understanding the operation of public law. The three elements of this framework – its explanatory, justificatory and critical dimensions – are inevitably interconnected, given they emerge from a normative theory of law which combines considerations of ‘what is’ and ‘what ought to be’. But these elements draw on political ideas and concepts in different ways for different purposes. They can also be understood to operate at two analytical levels – the ‘retail level’ of constitutional practice, and the ‘wholesale level’ of constitutional design. We will first consider how this positivist and political framework operates at the ‘wholesale level’ of the design of constitutional structures, before moving on to discuss how it might work at the ‘retail level’ of constitutional practice. The purpose of setting out this framework is not to try to establish an absolute monolithic structure, but to demonstrate that a positivist and political approach

A Positivist and Political Approach to Public Law  249 has multi-layered possibilities as a public law method, while attempting to show the potential implications of working within this conceptual scheme. B.  Explanation, Justification and Critique of Constitutional Structures First, the explanatory value of a positivist and political approach is that it reveals that the purpose of public law is defined by reference to its political context. As Loughlin shows, the idea of public law is developed to constitute an autonomous political realm based on the sovereignty of the state, to which all other claims to authority become subject.70 This demonstrates both that public law is not simply concerned with the limitation of government, but has a crucial empowering character. In addition, this empowerment of the state through law creates a system of government based on representation, with the public power generated capable (at least in principle) of superseding other forms of power – crucially including religious ­authority and the private power flowing from control of capital. This provides key context for understanding the potential of positive law, which forms part of and operates within a distinct model of political order. It explains why we might understand law as a set of generally authoritative and enforced rules which exist and have effect regardless of their substantive compatibility with standards of justice: positive law is the product of a specific set of governmental arrangements, and the authority of law flows from the fact it is located in such a political system, rather than because legal rules exist to serve any preordained set of moral ideals or a particular scheme of justice. Second, there is an important justificatory strand to this approach to public law. Understanding law by reference to its political purpose raises key questions for constitutional design. It requires normative political choices to be made about the shape and nature of constitutional structures, depending on what values and functions we want law to serve. This is not therefore about the complete elimination of moral considerations, but about the appropriate place of non-legal values within our legal systems, and the powers or privileges we give to legal actors to determine their reach and practical consequences. In essence, this is about how legal arrangements can be justified – in what ways legal systems should be organised, and what powers legal institutions should have. When legal questions arise about the nature of the political system (as they inevitably do) solutions are required. But in circumstances of real-world disagreement about justice, if public law is understood as inherently connected with morality, then that disagreement will simply be reproduced in the legal arena.71 It may be impossible to avoid normative debate about the proper structure of political systems at a theoretical level, but we can try to insulate dispute resolution at the practical level from disagreement at the level of the principled normative debate. This separation of the validity of positive law from moral standards is justified by normative values,



70 Loughlin, 71 Waldron,

Idea of Public Law (n 3) ch 5. Law and Disagreement (n 29).

250  Michael Gordon and in particular, the dominant modern political principle of democracy.72 On this basis, a normative positivist approach supports a system giving priority to legislative decisions over judicial reasoning or Crown/executive prerogatives, when legislation is produced in a democratic framework based on universal suffrage, which respects the political equality of citizens, allowing them to contribute equally to decisions in circumstances of disagreement about justice.73 The positivist and political approach therefore shows us that the place of moral or political values is in the debate about whether the arrangements of a legal system can be justified at the structural level. And then at this structural level, democratic ­principles can be understood to justify the general separation of legal rules from moral or political values at the point of determining their validity or meaning in the context of specific public law disputes. Legal positivism is therefore justified for normative political reasons, because this way of structuring a legal system can afford citizens ‘equal voice and equal decisional authority’ within their political community.74 The third dimension to this framework is a critical one, which must temper any conclusions about the extent to which normative positivism establishes a ‘good’ system of government, as opposed to one which is merely the most defensible way of structuring legal arrangements. For this cannot be an unreflective or complacent acceptance of democratic political systems or the legal institutions of which they are comprised. There must also be normative political evaluation of law’s effects, in specific contexts and systematically. And this requires critical normative reflection on the political limitations of legal order.75 It is obvious (in theory, and as a matter of historical record) that a positivist system of law may exist for ‘bad’ reasons, whether that is the consolidation of elite power, colonial exploitation, suppression of the wider population or discrimination against particular groups. In this sense, any normative defence of legal positivism must be relative – law is better understood from a positivist perspective than a moralistic perspective because it allows coordination for good (democratic) political reasons as well as bad, and has countervailing advantages, such as legal clarity and predictability.76 In this sense, legal positivism emerges simply as the least worst conception of law, because it creates space for democratic politics to potentially (but not necessarily) legitimise legal authority. As such, a democratic system of legislative decision-making is justified in relative terms only – it may exhibit greater political legitimacy than other modes of legal reasoning, but it is far from perfect. This establishes the need for critical scrutiny of the political legitimacy of (legal) public power, exposing failings and (if possible) exploring ways to enhance the system itself.

72 See, eg, J Dunn, ‘Conclusion’ in J Dunn (ed), Democracy: The Unfinished Journey 508 BC to AD 1993 (Oxford, Oxford University Press, 1992) 239: democracy is ‘today the overwhelmingly dominant, and increasingly the well-nigh exclusive, claimant to set the standard for legitimate political authority’. 73 See Waldron, Law and Disagreement (n 29) esp chs 3 and 13. 74 J Waldron, ‘The Core of the Case Against Judicial Review’ (2006) 115 Yale Law Journal 1346, 1389. 75 This is all the more important due to the friction often evident between democratic theory and traditional analytical positivism; see, eg, M Chun, ‘The Anti-Democratic Origins of Analytical Jurisprudence’ [2021] Jurisprudence 1–30. 76 See, eg, Raz, ‘The Rule of Law and its Virtue’ (1977) 93 LQR 95.

A Positivist and Political Approach to Public Law  251 It could therefore never be enough to view law in purely positivist terms, as if legal standards are the only ones against which a judicial decision can legitimately be evaluated. But legal positivism does provide the foundation for deeper critique, because it retains the distinctiveness of law as a social practice. If a crude form of ‘legal realism’ becomes the basis for critical theory, then law simply becomes a cypher for political action, and critique becomes exclusively political because law is merely a means to an end.77 Such an approach has limitations, because it overlooks the complexity of law and narrows the scope of critical theory. So, while the political character and ends of law are absolutely crucial to a critical understanding of legal practice, critique of law should not be purely instrumental. Instead, the specific and distinct contribution made by law to the functioning of our societies must be acknowledged rather than sublimated, for this too can be the target of critical reflection. This must be in addition to, rather than a replacement for, reflection on the political ends to which law is put. Consequently, a positive and political approach provides a foundation which requires us to think critically about the effects of our overall structures of law – such as the process of law formation and validation, or the role and composition of our legal institutions – and not only the substantive justice or injustice of legal outcomes. C.  Explanation, Justification and Critique of Constitutional Practice At the level of constitutional practice, a positivist and political approach allows us to understand legal decisions in a political context. It first provides us with a way of explaining what the law on a particular matter is, by offering insights into the determination of the validity and meaning of legal rules. To understand this as operating in a political context does not mean that legal validity becomes merely a function of political desirability or political pragmatism. Indeed, a crucial explanatory virtue of a positivist approach is that it highlights the distinctiveness of legal reasoning (whether compared to moral, ethical, or political reasoning, among others). But understanding this in context is essential to show that a positivist approach is explicable for political reasons. And as discussed above, those reasons are primarily based on the relative political legitimacy of a democratic legislative process (where one exists) which is therefore a mode of legal decision-making which ought to attract judicial deference, with the judicial role accordingly more limited than on a moral account of law’s nature. But even the way in which a positivist approach establishes the process by which courts (and other actors) determine the validity of law must be explained in a political context. A positivist approach to public law functions as a constraint on judicial power, which is justified in light of the non-democratic credentials of courts. Yet this will not be a complete or absolute constraint. Judicial discretion could never be eliminated (if that were even desirable) for the conceptual reasons identified by Hart

77 See, eg, EA Purcell Jr, ‘American Jurisprudence Between the Wars: Legal Realism and the Crisis of Democratic Theory’ (1969) 75(2) The American Historical Review 424–446, esp 434–437.

252  Michael Gordon concerning the inevitable indeterminacy of law, which means there will always be potential cases where the positive law ‘runs out’ and the courts are expected to fill the legal gap.78 Judicial discretion is also unavoidable in practice because at the point that a legal decision is made, the extent of judicial power is effectively self-determined, and decisions which appear to disregard the clear terms of positive law can sometimes be reached.79 However, even in the face of such decisions, legal positivism remains a central part of the background framework which shapes and frames judicial power. It provides the conceptual tools to critique those decisions which overlook or disregard positive law – sometimes within the courts themselves, in the form of dissenting judgments.80 But it also discourages the courts from engaging in interpretation which prioritises the judges’ preferred outcome over the terms of positive law. There continue to be numerous examples of judicial decisions where ‘the law’ is applied even despite explicitly stated judicial dissatisfaction with the outcome it produces,81 which is evidence of the influence of a positivistic understanding of law.82 The impact of positivism in establishing a distinctive form of legal reasoning in relation to public law is therefore not just conceptual, but also political, in structuring the interactions between constitutional institutions in practice. Second, the justificatory element of a positivist and political approach prompts us to consider the legitimacy of constitutional practice in an expansive way. Rather than thinking exclusively about the substantive outcomes of legal decision-making, we must also recognise that the political legitimacy of decisions is affected by the manner in which they were reached. The legitimacy of the political processes according to which law is made, and within which different constitutional institutions perform different roles, is also an important part of the legitimacy of any system of public law. And, therefore, the justice or injustice of legal outcomes (while still significant, as will be discussed momentarily) must not displace consideration of the justice or injustice of legal processes. There are two consequences of taking this position. First, it creates space for concerns about the extent of judicial power even where the judges might be thought to have reached some ‘good’ decisions in constitutional practice. Of course, what counts as a good or just judicial decision will be a matter of intense political debate, which is a key part of the problem. But even in principle, there are freestanding political reasons to favour a more limited conception of the judicial role, which are not negated if, for a particular period of time, some moderately good outcomes are being achieved in the courts. Second – and more significantly – if political legitimacy derives from legal processes and not just outcomes, then this requires sustained attention to

78 Hart, The Concept of Law (n 10). 79 See, eg, R (Evans) v Attorney General [2015] UKSC 21. 80 See, eg, Lord Wilson in Evans, [168]: ‘in reaching its decision, the Court of Appeal did not in my view interpret section 53 of FOIA. It re-wrote it.’ 81 For a recent example, see, eg, R (Black) v Secretary of State for Justice [2017] UKSC 81. 82 See, eg, the discussion of the limits of judicial interpretation of legislation of the Divisional Court in Hertfordshire County Council v Secretary of State for Housing, Communities and Local Government [2021] EWHC 1093 (Admin).

A Positivist and Political Approach to Public Law  253 be given to developing politically strong processes for the creation of law. Limitations on non-democratic actors are not enough – instead a positivist and political approach requires us to explore how public law practice can be enhanced to produce ever more legitimate arrangements. Accepting a traditional separation of powers and functions between a legislature, executive and judiciary will not be sufficient to meet this requirement – instead the challenge is to attempt to re-examine the legitimacy of the very institutional foundations of our legal and political systems. The third dimension of a positivist and political approach is to facilitate a critical attitude towards constitutional practice. This is based on an assessment of which legal norms are problematic, and which legal arrangements should be altered. Positivism preserves a set of legal standards against which we can critique what courts do, but such internal legal critique must be accompanied by wider political critique of decisions reached within the system, and of the system itself. After all, the impact of law is not just a consequence of its formal legal validity or effects – the way in which the law is enforced is inevitably important, but it can also have wider social ramifications. The social or political impact of law is difficult to assess or measure (and any measurements can be contested), but it remains crucial to think critically about the broader consequences of legal decisions or enactments. While positivism may demarcate the boundaries between legal and political critique, it does not establish a hierarchy between the two; rather, the existence of the boundary highlights the need for both. A positivist and political approach therefore positions us to understand the limitations of law as a means of generating change. While change through law is possible, law is also the tool which exists to reinforce the power of the state, and often therefore the interests of the status quo.83 Engagement with law is unavoidable in practice, but a positivist and political approach offers a corrective to the increasing legalisation of constitutionalism and moralisation of legal interpretation which provides a false perception of law as marking the path to utopia. D.  The Utility of Positivist and Political Public Law The framework which flows from the positivist and political approach I have outlined above is inevitably contestable, not least given it is grounded in normative legal theory. Consequently, the results generated by the theory become key, because a normative theory must offer something for it to be worthwhile. The normative dimension therefore shifts the parameters for testing the approach – what is offered in terms of our understanding of law becomes more significant than whether it is conceptually true or false, which is likely to be the subject of endless, unresolvable argument. This does not mean that such arguments are pointless, but instead that disputes about true methods may ultimately miss out all of the interesting constitutional questions, and 83 See, eg, KD Ewing and CA Gearty, Freedom Under Thatcher: Civil Liberties in Modern Britain (Oxford, Clarendon Press, 1990); GN Rosenberg, The Hollow Hope: Can Courts Bring About Social Change? (Chicago, University of Chicago Press, 1991); R Hirschl, Towards Juristocracy: The Origins and Consequences of New Constitutionalism (Cambridge, Mass, Harvard University Press, 2004).

254  Michael Gordon perhaps rob us of the agency and responsibility for choices about how legal systems are structured and operate. If we therefore take the positivist and political approach as one theory on a wider spectrum of constitutional methods, its vindication depends in significant part on the application of the theory. Therefore, to try to illustrate the value of this approach, in this final section I explore three examples of UK public law issues. The aim is not to offer a comprehensive analysis, but to show the insights this framework offers, in explanatory, justificatory and critical terms, into public law. The first issue is how we analyse ideas of parliamentary sovereignty. While this is a legal doctrine, applied by the courts, which establishes the legally unlimited legislative authority of the UK Parliament, it is a doctrine which can only be explained and justified in wider political context. In particular, the absence of legal limits can be seen to make sense when we recognise that this is because Parliament operates in a wider network of political constraints and on the basis of democratic political inputs. The choice here is not to have an unconstrained legislature, but to rely on political rather than legal factors to shape the exercise of legislative power, with Parliament’s status as the constitutionally ultimate decision-maker justified by its democratic credentials (in contrast to the unrepresentative nature of the judiciary). Further, while the doctrine emerged following a key political conflict in the seventeenth century,84 this does not mean that if the modern political conditions underpinning sovereignty have changed, so too must the law be seen to have changed. Loughlin and Tierney, for example, argue that ‘the absolutist legal doctrine has been qualified by political developments’ and should be dispensed with in favour of a modern concept of sovereignty.85 A positivist and political approach, however, would reject the premise that when political context changes, the law necessarily also changes too. While we might argue the law needs to change as part of a critical assessment of political legitimacy, the positivist authority of the legal doctrine is independent from the underlying dynamics of the political situation. Drawing a sharper distinction between law and politics at this level retains a firm line which the courts cannot transgress, despite some recent obiter dicta gesturing at the idea of new ‘rule of law’ limits on parliamentary sovereignty.86 Indeed, it is this distinction which lays the ground for both legal and political critique of any attempts by the courts to subvert this fundamental constitutional norm. The second issue which illustrates the distinctive nature of a positivist and political approach is the relationship between the courts and the political institutions as characterised in the two famous Miller cases, which arose during the process of the UK exiting the European Union. Taking a positivist and political approach to these cases reveals the limitations of courts focusing on substantive outcomes over the wider structures of political power. In both cases the UK Supreme Court relied on

84 See J Goldsworthy, The Sovereignty of Parliament: History and Philosophy (Oxford, Oxford University Press, 1999). 85 M Loughlin and S Tierney, ‘The Shibboleth of Sovereignty’ (2018) 81(6) MLR 989–1016, 1016. 86 See, eg, R (Jackson) v Attorney General [2005] UKHL 56, [102], [104]-[107], [159]; Moohan v The Lord Advocate [2014] UKSC 67, [35]; R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22, [144].

A Positivist and Political Approach to Public Law  255 expansive (and pretty vague) ideas of constitutional principle to constrain executive powers in ways which are legally questionable, while both cases also reveal the importance of analysing the wider political consequences of judicial decision-making. In Miller (No.1),87 a majority held that the government could not begin negotiations under the foreign affairs prerogative because ‘[i]t would be inconsistent with long-standing and fundamental principle for such a far-reaching change to the UK constitutional arrangements to be brought about by ministerial decision or ministerial action alone’.88 Parliament was therefore required to legislate to authorise negotiations (despite the fact this was not an explicit requirement in the statute governing the UK’s relationship with the EU),89 in addition to the legislation which would always have been required at the end of the process to reflect the outcome in domestic law. Parliament promptly legislated to give the government the necessary formal powers,90 in effect providing a statutory endorsement of the referendum result which could then be used politically to show that the legislature had ‘voted for Brexit’ throughout the rest of the process. The majority’s focus on broad principles over the detail of statute law can therefore be criticised from a positivist perspective, and while the decision had little immediate impact (because negotiations were authorised and commenced on schedule) it also had important (no doubt unintended) political consequences for the Brexit process. In Miller (No.2)91 the Supreme Court drew on the principle of parliamentary accountability in a novel way, apparently giving legal force to what is better understood as a foundational idea of the political constitution. The Court unanimously held that the government’s attempt to prorogue Parliament for five weeks, to prevent the legislature from obstructing an apparent plan to leave the EU without agreeing a withdrawal deal, was unlawful because there was no justification for executive action ‘which had such an extreme effect upon the fundamentals of our democracy’.92 While the Divisional Court had held that the prerogative power to prorogue was to be regarded as non-justiciable because ‘it is impossible for the court to make a legal assessment of whether the duration of the prorogation was excessive by reference to any measure’,93 the Supreme Court had no such reservations. The Supreme Court tried to present this as a unique case – ‘a “one-off”’94 – and the intervention is only truly explicable when understood in its exceptional political context. But given the result has been to push the government to launch a process to reform (or narrow) judicial review powers,95 it is unclear whether the intervention was worthwhile. Especially when, prior to the decision that prorogation was unlawful, Parliament had already legislated to protect its role in the Brexit process, placing the government under a

87 [2017] UKSC 5. 88 ibid [81]. 89 See ibid, the dissent of Lord Reed. 90 EU (Notification of Withdrawal) Act 2017. 91 [2019] UKSC 41. 92 Ibid [58]. 93 [2019] EWHC 2381 (QB), [54]. 94 [2019] UKSC 41, [1]. 95 See the Independent Review of Administrative Law (July 2020–Jan 2021): www.gov.uk/government/ groups/independent-review-of-administrative-law.

256  Michael Gordon legal obligation to seek a further extension to negotiations with the EU,96 and thereby defeating the purpose of the five-week prorogation before it had even begun. Finally, the third issue which highlights the distinctive possibilities of a positivist and political approach is the status of human rights within constitutional law. This approach gives us the tools to ask the key theoretical questions about the law-morality relationship established by modern human rights law. For the analytical debate about whether there is an inherent conceptual separation between law and morality starts to seem less significant in the age of legalised fundamental human rights, which are premised on a fusion of moral values and positive law, and (at least in general) purport to have overarching authority to which most other categories of legal rules are subject. From this perspective, the need for normativity in public law theory is clear, to allow us to ask questions about the consequences and effects of using positive law to make moral ideals legally effective standards with (usually) overarching status. From this vantage point, the debate is about much more than simply whether something like the Human Rights Act 1998 should be repealed or amended (or indeed lauded). Instead, the pervasive potential of human rights law makes this question peripheral. But that is also not to be fatalistic about the existence and implications of capturing human rights standards in positive constitutional law. The critical dimension of a positivist and political approach creates space for scepticism about the ultimate wider consequences of the legalisation of moral rights to be enforced by the judiciary. And this critical reflection can draw in both the structural and practical elements of constitutionalism. In a structural sense, this can include analysis of the overall impact of the human rights revolution on ideas of justice in our society, or whether even in this form the law leaves underlying inequalities unaddressed.97 And at the level of practice, we might ask political questions about who human rights law empowers in our communities.98 Or how the abstraction and formality of law might work to separate morality from politics.99 Or how the apparent universality of fundamental rights standards may be in tension with the selective way in which these norms can be actually deployed.100 In highlighting the issues relating to parliamentary sovereignty, the courts and the government, and human rights, I am not suggesting that a positivist and political approach prescribes any particular overall substantive theory of public law. Instead,

96 EU (Withdrawal)(No.2) Act 2019. 97 See, eg, J Whyte, The Morals of the Market: Human Rights and the Rise of Neoliberalism (London, Verso, 2019); S Moyn, Not Enough: Human Rights in An Unequal World (Cambridge, Mass, Harvard University Press, 2019). 98 See, eg, JAG Griffith, The Politics of the Judiciary 5th edn (London, Fontana, 1997). 99 See, eg, Roberts [2015] UKSC 79, where the UK Supreme Court upheld discriminatory practices of ‘stop and search’ as compatible with the ECHR, while dubiously endorsing a highly problematic stereotype when noting that young black and ethnic minority people, who were disproportionately targeted by these powers, ‘will benefit most from the reduction in violence, serious injury and death that may result from the use of such powers. Put bluntly, it is mostly young black lives that will be saved if there is less gang violence in London and some other cities’ [41]. 100 For eg, in relation to the right of access to justice, compare the rhetoric of UNISON [2017] UKSC 51 in the context of employment tribunal fees with the complete absence of this concept from the judgment in Begum [2021] UKSC 7 in the context of removal of citizenship.

A Positivist and Political Approach to Public Law  257 these examples demonstrate the way in which public law can be understood, and the ways in which specific public law challenges can be assessed, when a positivist and political approach is adopted. V. CONCLUSION

This chapter has explained and defended a positivist and political approach to public law. This approach is rooted in the (somewhat unlikely) combination of Loughlin’s political jurisprudence and Waldron’s normative legal positivism. But this combination establishes an approach to public law which is distinctive in its own right. I have argued that this approach is conceptually coherent and has considerable value as a means of understanding public law. In particular, it provides a framework for the explanation, justification and critique of public law, both at a structural and a practical level. And I have tried to demonstrate the utility of this framework using three examples to show the kind of insights it can generate, and the kind of questions it prompts us to ask about public law. No constitutional method will be perfect, and inevitably the methods we choose shape the conclusions we reach. As Lakin has argued, the positivist and political approach may be seen as ‘loading the dice’ in favour of certain values or outcomes.101 But that is likely to be true of any method. In circumstances of methodological disagreement, therefore, perhaps the best we can do is to be clear about what we are trying to do, how we go about it, and why, so that the method can be contested as much as the different positions we reach in relation to substantive legal or constitutional matters. As I have tried to show, a positivist and political approach is based on the idea that we should accept that legal method is not abstracted or insulated from political context or political values, while also appreciating that this does not mean law (or legal analysis) simply becomes a function of political or moral interpretation. Perhaps this is an accommodation which will satisfy few people, but it is based on the democratic idea that while law is a tool of public power, legal actors can and must still be challenged for their actions and decisions by reference to a broad set of political values. Recognising the connections between law and politics is not a path to legal perfectionism, where all law is engineered to be compatible with our chosen political principles. Instead, it has a more modest and realistic aim: to allow us to explain law’s political role as a distinct social construct, while also giving us an expansive set of tools to evaluate and criticise its effects.

101 S Lakin, ‘The Manner and Form Theory of Parliamentary Sovereignty: A Nelson’s Eye View of the Constitution?’ (2018) 38(1) OJLS 168–189, 175–178.

258

Social Theory

260

10 The Material Study of the Constitutional Order MARCO GOLDONI

T

his chapter’s aim is to show what it means to study a constitutional order from a material point of view. In a nutshell, this method entails the observation of two distinct, but related, objects of study: (1) the identity of a constitutional order;1 and (2) the relations (in particular between the social and the legal dimensions) that makes it what it is. The material study of the constitutional order focuses primarily on the fundamental norms of the order and on the reasons of their formation and permanence, but for a proper understanding it has to take into account other relevant phenomena. The chapter has been organised in four sections. The first one delineates the methodology of studying the constitutional order by juxtaposing it to other ‘political’ approaches and it offers a reconstruction of the two legal theories from which it draws its inspiration. Then, the second section will broach the political theory undergirding the material study of constitutional orders with a view to show that core aspects of social organisation and their legal forms are the bases of the constitutional order. Once the theoretical foundations of the material study have been explained, in the third section the object of the study will be reconstructed by focusing on its constitutive elements (unity, subjects, government, political aims). Sections four will illustrate this methodology by taking up an example from contemporary constitutional debate: constitutional change and transformation. The conclusion will sum up the main benefits of the material study, but will also end with a humbling reminder. I.  THE LEGAL THEORY OF THE MATERIAL STUDY

It seems difficult to dispute that in the last decades the dominant ways of approaching constitutional theory have inclined toward a legal understanding of the constitutional 1 The identity of the constitutional order can also be tracked by observing the lawmaking procedure, as it is advocated by Hans Kelsen: cf L Vinx, ‘Hans Kelsen and the Material Constitution’ (2021) 12 Jurisprudence (forthcoming); a similar understanding is proposed by J Colón-Rìos, Constituent Power and Law (Oxford, Oxford University Press, 2020) ch 8.

262  Marco Goldoni order according to which the juridical moment would be its constitutive and distinctive element. In its different expressions, this hegemony has been far from being monochromatic: from the empire of law to the different versions of global and constitutional pluralism,2 a common understanding of the constitution as a legal creature has been shared by all these streams of constitutional theory. Even in the case of radical pluralist analyses, the language of law has been assumed as the universal jargon for the understanding of the interaction among orders.3 The reaction against the dominant mode of doing constitutional theory has been to resort to the study and valorisation of the political dimension of constitutions.4 While the legal study of the constitution has been predicated on the basis of law’s values, such a reaction has been inspired by the rediscovery of the ‘autonomy of the political’. Two different versions of this approach can be retrieved. The first one is normative, and it advocates how constitutions should address conflicts and disagreement. This type of theory supports a form of parliamentary sovereignty as the best constitutional form for protecting fundamental rights and upholding democratic government.5 The second version has linked the formation of the constitution to a particular achievement, the growth of an autonomous sphere of politics that carry with it its own meaning. The constitution is therefore a political creation obtained in virtue of the principle of sovereignty. This is realised either in the form of a political jurisprudence conceived as a relation between potentia and potestas,6 or as the expression of a sacrificial logic according to which political sovereignty generates the constitution.7 This version of the political approach, however, is not normative. It rather aims at explaining under which conditions modern constitutions emerged and can stabilise. A legal and a political approach can shed light on the development of modern constitutions. Yet, both seem to understand questions of asymmetry between the legal and the political dimension of the constitutional order in terms of primacy or precedence of one over the other. In both approaches, the question of how the constitution is tied to the social order is obscured by postulating that either the law is constitutive of social interaction or that the political dimension of constitutions enjoys autonomous ordering properties. Both points are valuable and at times correct, but neither of the approaches thematise the relation between constitutional ordering and societal formation (ie, the organisation and development of fundamental social relations). It is that relation that concerns primarily the material study of the constitution. If the relation between constitutional order and societal formation is the core concern, then the relevant constitutional theory ought to recognise the centrality of 2 The list of significant works would be too long. Prominent examples of these approaches are: R Dworkin, Law’s Empire (Cambridge MA, Harvard University Press, 1986); R Alexy, A Theory of Constitutional Rights (Oxford, Oxford University Press, 2012). 3 P Schiff Berman, Global Legal Pluralism (Cambridge, Cambridge University Press, 2012). 4 See, most recently, M Nogueira de Brito, L Pereira Coutinho (eds), The Political Dimension of Constitutional Law (Dordrecht, Springer, 2020). 5 See the classic references: J Waldron, Law and Disagreement (Oxford, Oxford University Press, 1999); A Tomkins, Our Republican Constitution (Oxford, Hart Publishing, 2005); R Bellamy, Political Constitutionalism, (Cambridge, Cambridge University Press, 2007). 6 See M Loughlin, Political Jurisprudence (Oxford, Oxford University Press, 2017). 7 P Kahn, Putting Liberalism in Its Place (Princeton, Princeton University Press, 2004).

The Material Study of the Constitutional Order  263 social organisation. The definition of the approach as ‘material’, although focussed mostly on the political economy of social organisations, is not intended as a way to obscure other important factors of societal ordering. The distinctive move of the material approach is to determine the relation as internal.8 The latter means that it is not possible to study the constitution as either a precondition for society or as an overdetermined outcome of societal organisation. The idea of an internal relation shows that the constitutional order is an internal formation of society and cannot be simply applied over or to society. Accordingly, a fully reductionist view of the constitution, as it is sometimes propounded by materialist approaches,9 whereby the constitutional structure of a jurisdiction is fully determined by the undergirding political economy is not available. But we shall also hasten to add that this does not deny the crucial importance of the political economy of modern constitutional orders. Two legal theories have been substantively concerned with the previously mentioned internal relation between societal formation and the constitutional order: legal institutionalism and societal constitutionalism. This is not to say that other approaches were or are completely indifferent to the question,10 but what is of particular interest in these two streams is their concern for the intrinsic qualities of the internal relation. For this reason, a material study of the constitution can learn from both approaches and we will focus on these in the following sections. In its first wave, some of the main representatives of legal institutionalism11 – Santi Romano, Maurice Hauriou, Herman Heller, Costantino Mortati – advanced a theory of the institution that was supposed to overcome certain limits of legal positivism. What is of interest and of direct relevance to the material study of the constitutional order are three ideas common to these authors. First, the core function of legal orders is to organise and stabilise social relations. In fact, organisation is the main theme of this stream of scholarship. Where there is social organisation, there is legal order (ubi societas, ibi ius). The concept of organisation can be split into two processes: differentiation and specification. The former entails a crucial distinction between those who govern and those who are governed.12 This is not only a political, but a social distinction as well. The other process, specification, indicates the distribution of roles and functions within differentiation. Although not fully thematised, this emphasis on organisation fits perfectly with the material engine of the first modern constitutional orders, the rise of the principle of division of labour and its political import: political representation.13 While legal institutionalists do not focus mainly on this aspect, an understanding of their theory that accommodates the key function of division

8 This should not be equated with the idea of internal metaphysical relations as explained by D Armstrong, A Theory of Universals (Cambridge, Cambridge University Press, 1978) 84–5. 9 A classic statement is C Beard’s influential An Economic Interpretation of the Constitution of the United States (New York, Simon & Schuster, 1986). 10 For example, one can see a similar concern in the adaptation of Hart’s positivism to a constitutive theory of constitutional law by M Berman, ‘Our Principled Constitution’ (2018) 166 U Pennsylvania Law Review 1325. 11 See, for an overview, Loughlin, Political Jurisprudence (n 6) ch 7. 12 See, on this point, M Loughlin, The Idea of Public Law (Oxford, Oxford University Press, 2003) 5. 13 Cf N Urbinati, Political Representation (Chicago, Chicago University Press, 2006).

264  Marco Goldoni of labour, and more generally the division of social functions, is legitimate and compatible with their theoretical framework.14 The second point concerns the role of institutions as an articulation of social relations and legal orderings. In other words, institutions are a way of assembling social relations to weave together the social order. Hence, as it has been noted, they can also be defined as a self-description by society of sectors of society itself.15 This means that institutions (eg, family, corporation, trade union, property) are at the same time legal and social: according to institutionalists, each complex social formation already contains within itself a principle of legal ordering, otherwise it could not emerge as an organisation.16 It is important to note that an institution is always legal in virtue of its organising norms. No institution can be normatively organised without being at the same time legally shaped. The third point that deserves to be mentioned is the recognition that the constitutional order is not only the total sum of constitutional norms but also contains a core identity which is shaped by fundamental directive ideas (Hauriou), objectives or fundamental norms (Romano, Mortati). It is important here to note that these ideas or aims cannot be fully detached from the relevant social organisation (ie: they cannot be external to it). They are, rather, included within the social organisation and for this reason cannot be fully detached from at least some of the means employed to achieve these objectives. In other words, ideas or objectives are not autonomously imposed on the social order as if they were decided from the outside. This is a reflection of how institutions function according to legal institutionalists: they are always organised around the pursuing of what are fundamental objectives, where the fundamentality is constitutive of the institution itself. Constitutional orders do have a finality (in this sense, they are telic legal orders) and their organisation is driven by it as well.17 This means that legal institutionalists give a different twist to the common positivist idea that the legal system precedes – logically and as a condition of existence – legal norms.18 Institutionalists also believe that the whole order sustains constitutional norms, but the unity of the order is granted by a directive or guiding idea or objective.19 Societal constitutionalism can be read as a sophisticated and innovative evolution of this first wave of legal institutionalism.20 The key insight of societal constitutionalism is the idea of the constitution as an evolutionary achievement which, according

14 Behind this issue one can see another classic theoretical challenge: what keeps together a society organised around the principle of division of labour? On this point, see the following section. 15 G Teubner, Constitutional Fragments (Oxford, Oxford University Press, 2012) 21. 16 See, eg, how Mortati understands constituent power in La teoria del potere costituente (Rome, Quodlibet, 2020) 23–5. 17 For an account of how the commitment to the objectives shape the constitutional imagination see R Cover, ‘Nomos and Narrative’ (1983) 94 Harvard Law Review 4, 43–5. 18 M Berman, ‘Law as an Artificial Normative System’ in D Plunkett, S Shapiro, K Toh (eds), Dimensions of Normativity: New Essays in Metaethics and Jurisprudence (Oxford, Oxford University Press, 2019) 66–86. 19 It is a further question whether the driving idea takes the form of a norm as well. 20 A Golia, G Teubner, ‘Societal Constitutionalism’ (2021) MPIL Working Papers 1.

The Material Study of the Constitutional Order  265 to Niklas Luhmann, is understood as the structural coupling between the political and the legal system.21 This achievement has a double quality: it provides reserves of public power by bringing different systems to share structural elements but, at the same time, it preserves the autonomy of the social sub-systems and their functional differentiation. The coupling also externalises the internal paradox of each system and in this way it favours the stabilisation of the constitutional order. The second important contribution is represented by Teubner’s further expansion of Luhmann’s intuition to the complexity of other social sub-systems. Teubner assumes that social sub-systems (not only the political system) can reflexively couple themselves with law and become constitutionalised. The state remains a fundamental institution, but ceases to be the exclusive site of constitutional systems. In this way, Teubner has opened the processes of the constitutionalisation of society to the autopoietic contribution of social fragments. Yet, as he notes, constitutions outside the state need to satisfy the requirements of a “material” concept of constitution, according to which a constitution establishes a distinct legal authority which for its part structures a societal process (and not merely a political process, as is the case with nation-state constitutions).22

Not all couplings successfully establish constitutional orders, but only those that achieve a certain level of structural stability and reflexivity that invest a societal sector. The third important contribution is the belittling of the role of political will in the formation of the constitutional order. In this way, any residue of political opportunism is removed from constitutional theory. Societal constitutionalism does not exclude the existence of political will, but the latter has to be always mediated by the medium and the code of each social sub-system. Under this aspect, societal constitutionalism is capable of providing a more nuanced account of the function of law in the organisation of society and one that is not subject to the ‘occasionalism’ of legal institutionalists. The latter at times seem to postulate that the whole effort of ordering is on the shoulders of political institutions. Legal institutionalists (perhaps with the exception of Santi Romano) have tended to idealise the state as ‘the institution of institutions’ and, accordingly, have often given too much space to the autonomy of politics while underestimating the importance of other actors.23 This has come at a high price: the risk of accepting a version of political occasionalism, that is, a view which states that a particular order has been organised, within society, by the victory or hegemony of one or more political forces. In this way, legal and social organisation end up being determined by political will, with no adequate analysis of the social embeddedness of political action. Societal constitutionalism has amended this view thanks to its attention to the multiple societal forces and sub-systems, and their impact on the development of the social order. Yet, there are two important differences with the previous institutionalist version that still apply and should be emphasised. First, systems in societal constitutionalism

21 N Luhmann, Recht der Gesellschaft (Frankfurt, Suhrkamp, 1993) 427. 22 G Teubner, Constitutional Fragments (n 15) 74. 23 For a nuanced reading of Hauriou’s views on the subject see A Salvatore, ‘Il diritto della vita’ in M Hauriou (ed), La teoria dell’istituzione e della fondazione (Macerata, Quodlibet, 2019) 125–50.

266  Marco Goldoni are not organised around aims or purposes, but functions. The point of coupling subsystems is indeed to allow them to perform their functions by employing law to supplement social reflexivity with secondary legal rules. As noted above, structural couplings consolidate or stabilise social processes and avoid the risk that they become paralysed by their own paradoxes.24 A crucial difference with previous versions of legal institutionalism is the recognition of growing social differentiation and the importance of preserving the operative autonomy of each sub-system. According to constitutional sociology, some of the social subsystems have uploaded to the transnational or even global level, while others have remained at the national or subnational level.25 Once a society is assumed to be organised according to these coordinates, the institutionalist idea of an order as a project or as containing a directive idea26 becomes secondary and basically irrelevant for constitutional analysis. The second main difference is to be identified on a point that is of apparent convergence: both streams of thought recognise that institutions express the legal dimension of social relations. Yet, while institutionalists like Santi Romano and Hauriou directly equate institutions with law, systems theorists hold that institutions are the product of the structural coupling between social sub-systems and the legal system. Under this latter description, law and politics maintain their full autonomy, while for legal institutionalists one is impossible without the other. Allegedly, the innovative contribution of societal constitutionalism implies the recognition of the reflexive moment as essential for the analysis of the constitutional order: social institutions in themselves do not contain any immanent principle or rule, and they become endowed with a material constitutional order only through a self-reflexivity engendered by structural couplings. This is where constitutionalisation does most of its work: it is supposed to stabilise the normative expectations of individuals and, at the same time, open up constitutional orders to the inputs of social actors (eg, social movements). Notwithstanding this crucial contribution, societal constitutionalism has not properly thought through all the relations between the legal systems and the other systems, and this is the case for two reasons. First, it has underestimated the capacity of the state (or other encompassing legal orders, like the Church) to mediate and steer the relations among systems. This is possibly the case because the capacity of the state to coordinate functional differentiation has been discarded by looking at totalitarian and authoritarian experiences of the last century. Although sociological, this approach still remains committed to a modern idea of constitutionalism as a mechanism not only for generating power, but of limiting it as well. Second, it has often portrayed the formation of new structural couplings as forms of juridification of one social sub-system: science, for example, has been conceived as an autonomous system whose reach has upgraded well beyond the state through a form of reflexivity

24 C Thornhill, A Sociology of Constitutions (Cambridge, Cambridge University Press, 2011) 18; Teubner, Constitutional Fragments (n 15) 107. 25 C Thornhill, A Sociology of Transnational Constitutions (Cambridge, Cambridge University Press 2016); P Kjaer, Constitutionalism in the Global Realm: A Sociological Analysis (Abingdon, Routledge, 2016). Though, according to Marcelo Neves, a form of transconstitutionalism has also emerged in the last few decades: M Neves, Transconstitutionalism (Oxford, Hart Publishing, 2013). 26 This is the classic thesis defended by S Romano, The Legal Order (Abingdon, Routledge, 2017).

The Material Study of the Constitutional Order  267 that requires the coupling with law. Societal constitutionalism has thus represented these new formations as constitutional fragments, but actually an analysis more attentive to their material differences, would have shown that these are cases of asymmetry. Overall, the functional drive of social systems makes it possible, for societal constitutionalists, to observe institutions only as a coping mechanism for the internal paradoxes that affect each system.27 Talks of objectives or aims is almost non-­sensical from this perspective. Institutions arise in order to enhance or consolidate the functionality of systems. While legal institutionalism and societal constitutionalism form the backbone of the material analysis of the constitutional order, methodologically the study of the constitutional order will have to bear in mind, apart from the insights previously mentioned, that in modern times it is impossible to ignore the role of the political economy, here understood as the organisation whose form is dictated, among other things, by the creation of economic value. The latter should not be seen as the only telos of the constitutional order, but as one of the most important ordering forces that are operative in shaping the relations among systems and, so to say, ‘command’ the organisation of social production. In the next section, the key importance of the principle of division of labour as the main constitutional organising factor will be elaborated upon further, but it is important to bear in mind that the material study of the constitution emphasises in particular the connection between the legal order and the social organisation of labour.28 II.  THE POLITICAL THEORY OF THE MATERIAL STUDY

As a methodology, the material study rests on a conception of the role of constitutional analysis and theory, but also on certain assumptions about the relation between politics, law and the economy as it is explained by political theory. To make the claim for the material approach more robust, a political view of societal formation must be sketched out. This is necessary for the identification of the constitutive social facts. A fundamental aspect of the formation of social order (and, accordingly, of constitutional orders) has to be grasped as it explains not only its rise, but also helps in understanding why the order took up a particular form. An explanation of the main political mechanism that lies at the heart of the ordering process allows us to study the constitutional dynamic more accurately. The following does not intend to be an exhaustive overview of the political theories behind modern constitutional orders, but a discussion of the main available options in the current landscape, with a view to show why the material dimension provides the most fundamental level of analysis. The first idea is the social contract, possibly the most influential modern political conception of the genesis of the social order. According to this theoretical device, the social and then constitutional order is brought about through a hypothetical or

27 N Luhmann, Law as a Social System (Oxford, Oxford University Press, 2002) ch 7. 28 This link is the red thread of the ground-breaking analysis in R Dukes, The Constitution of Labour (Oxford, Oxford University Press, 2014).

268  Marco Goldoni real agreement by all the subjects to the order itself. As known, in the most common versions the constitutional order emerges against the background of the state of nature. The latter is portrayed as a pre-socialised state inhabited in a way that makes it rational to intend to get out of it. Otherwise, when it is deemed to be only a fiction (through the device of the original position), it is presented as a way to assess our ethical theories and their underlying assumptions.29 The problem is that the fiction is built in a way that obscure the role of certain fundamental social facts. The social relations that pertain at the time of the so-called origin (whether state of nature or the original position) are then concealed, nullified by virtue of a thought experiment but one that acts as a distorting lens because it de-politicises existing material conditions and social relations. Once translated into constitutional law, the social contract also legitimises taking matters out of the ordinary political and social processes of contestation. In other words, the material aspect of the constitutional order is entirely (and deliberately) concealed from view. The political theory of the social contract is then one that rationalises the relation between the rise of social order and the formation of the constitution as an external relation. Accordingly, the reasons for having a political association can be left outside of the development of the constitutional order. The social contract theorises the constitutional irrelevance of cooperation and interaction for the formation of the constitutional order. Strangely enough, this is actually a device that, at least in certain versions, is supposed to give moral reasons to enter into society. But the logic of contract – if truly contractarian – is usually one of self-interest.30 As noted by a prominent critic: ‘Those who enter contracts typically are concerned with their own benefits and need not care about the benefits to their partners in trade.’31 The constitutionalisation of certain principles and rights is admitted as a device to keep the order together (say, for example, with the constitutionalisation of the maximin principle), by protecting some of the interests that allegedly exist already before the entry into society. But the point of observation provided by the social contract does not allow to question the modality of formation of the order. The emphasis is on the normative deduction of basic moral principles which should then take up constitutional form.32 All in all, the process of constitutionalisation is deprived of its materiality and is idealised as a deliberation (guided by reflective equilibrium) on the main tenets of the basic structure.33 This leaves a lot of room for the Humean criticism against contractarianism. In fact, the most important alternative to the social contract tradition is represented by

29 The classic reference is J Rawls, A Theory of Justice (Cambridge MA, Harvard University Press, 1971) § 3. 30 Hence, it is debatable, for example, whether Rousseau or Rawls are truly contractarian authors. 31 R Hardin, ‘Why a Constitution?’ in D Law and M Versteeg (eds), Social and Political Foundations of Constitutions (Cambridge, Cambridge University Press, 2013) ch 2. 32 See, for a famous application of the maximin principle in constitutional law, F Michelman, ‘On Protecting the Poor through the Fourteenth Amendment’ (1969) 83 Harvard Law Review 7. 33 The latter is the version expounded by the late Rawls in Political Liberalism (New York, Columbia University Press, 1993) ch VI. Note that it could be possible to understand the basic structure as the material dimension of the constitutional order (Rawls mentions, as elements of the basic structure: the political constitution; the organisation of the economy; recognition of property rights; the nature of the family).

The Material Study of the Constitutional Order  269 the idea of conventionalism. This is actually one of the most developed and sophisticated explanations and justifications of constitutional orders and particularly of liberal-democratic orders.34 Since Hume, the idea that the organisation of society and the government are the products of a social convention has acquired a lot of traction. In recent times, the study of conventionalism has been revived by David Lewis who has provided an extensive analysis of what a convention is.35 In legal and political theory, the idea has gained enormous influence for two reasons. First, conventions are an apt device for explaining the coordination necessary in complex and pluralist societies. As they do not exercise a mandatory force on the basis of substantial reasons, conventions provide a point of convergence for different perspectives and values. In fact, according to the classic understanding, a convention is followed mainly because other people follow it.36 In other words, a convention can solve problems of coordination, deep disagreement and game-theory challenges in a more effective way than the social contract device. It does not entail the heavy normative expectation of a general agreement on constitutional essentials. One can see why the idea gets a lot of attention in the analysis of liberal-democratic constitutional orders. It is possible to explain the formation of the order without making reference to substantive values or objectives (though they are not necessarily excluded). Close to this explanation of a liberal-democratic order is the political theory of the ‘empty place’, famously put forward by Claude Lefort.37 Although couched in a different language (that of the phenomenological tradition), such a view of democratic constitutional orders entails that political and legal institutions can never be fully occupied by substantive principles or aims, and this relative emptiness is what allows democratic politics to unfold.38 But the biggest influence of conventionalism in constitutional studies can be seen in the impact made by a conventionalist reading of the rule of recognition. HLA Hart himself came to endorse that interpretation of the rule of recognition in the postscript to The Concept of Law.39 One could indeed study the relation between politics, law and the economy as an equilibrium set by certain conventions among officials. This is not a demanding conception as it does not mobilise the substantive norms of a constitutional order (or, at least, not directly). But precisely for these reasons, it does not capture the process of forming a constitutional order.40 However, one can see that conventionalist explanation of the constitutional order might come across as compatible with the material study. For this reason, it is worth unpacking further the notion of a conventional rule of recognition. According to one of his most prominent theorists, Andrei Marmor, the conventional rule respects two conditions: a condition

34 For a classic defence see R Hardin, Liberalism, Constitutionalism. Democracy (Oxford, Oxford University Press, 1999). 35 D Lewis, Convention (Oxford, Blackwell, 1969); for an in-depth discussion of legal conventionalism see T Spaak, ‘Legal Positivism, Conventionalism, and the Normativity of Law’ (2018) 9 Jurisprudence 319. 36 G Postema, ‘Coordination and Convention’ (1982) 11 The Journal of Legal Studies 164, 167. 37 C Lefort, Democracy and Political Theory (London, Polity, 1988). 38 For a criticism of this political approach to constitutionalism see A Somek and M Wilkinson, ‘Unpopular Sovereignty’ (2020) 83 MLR 955, 965–7. 39 HLA Hart, The Concept of Law 2nd edn (Oxford, Clarendon Press, 1994). 40 For a different reading which tries to adapt the rule of recognition to the metaphysics of grounding and anchoring see B Epstein, The Ant Trap (Oxford, Oxford University Press, 2015) ch 7.

270  Marco Goldoni of dependency and a condition of arbitrariness.41 The former postulates that one of the reasons for following a convention is that others do the same. The latter assumes that there is an element of arbitrariness in the choice of the convention. The conventional rule can be arbitrary (that is, it could have been otherwise) even if it reflects moral or political convictions.42 Ultimately, it is the rule-following behaviour that makes it a conventional rule more than its substantive content. The reason for following conventional rules is the adherence to the convention of the relevant agents. In other words, the key organising factor of the constitutional order is conventional, while other substantive principles and aims enter the picture only if allowed and introjected via the fundamental social convention. The social convention on which the constitutional order rests is indifferent to the content of the order itself. Therefore, it cannot provide an entry-point for the material study. Only political theories that postulate that the formation of the social order is driven by a purpose or a telos can provide a solid basis for the material study. The recognition of this point leads to a third prominent strand of political theory: political theology. In constitutional studies, the notion has been introduced by Carl Schmitt during the 1920s, but it has been expanded by political and constitutional theorists in the last few decades.43 It has witnessed indeed a revival not only among political and social theorists, but in constitutional discourse as well. The gist of this political theory as currently used in constitutional studies is that social and constitutional orders are formed around the principle of sovereignty. Unlike contract or convention, sovereignty mobilises a different energy and cannot be simply explained in terms of game theory or coordination issues. Sovereignty bonds legal and political communities because it constitutes an autonomous sphere of meaningful action to which every actor can relate. Hence, the most important feature for the constitutional order is the creation of a space for autonomous political action. This is not parasitical on a previous pre-political arrangement because it makes the rise of the political the key moment of community formation. The condition for its emergence is given by what is at its core a sacrificial logic. Already in the work of René Girard it is possible to identify the unfolding of this logic. The unity of the social order is based on the condition of a sacrifice which is supposed to interrupt the cycle of mimetic desires and potential revenge. This is made possible by the figure of the scapegoat, which is produced by ‘a deliberate act of collective substitution performed at the expense of the victim and absorbing all the internal tensions, feuds, and rivalries pent up within the ­community’.44 This is a non-reciprocal logic whose aim is to contain and channel

41 A Marmor, Social Conventions (Princeton, Princeton University Press, 2008) 14–5. 42 A Marmor, Objectivity and Legal Positivism (Oxford, Oxford University Press, 2001) 21. 43 See P Kahn, Political Theology (New York, Columbia University Press, 2011); G Agamben, The Kingdom and the Glory (Stanford, Stanford University Press, 2011); E Böckenforde, Constitutional and Political Theory (Oxford, Oxford University Press, 2016) ch 2; F Atria, La forma del derecho (Madrid, Marcial Pons, 2016). 44 R Girard, Violence and the Sacred (Baltimore, John Hopkins University Press, 1972). Sacrifice becomes the political device to protect the constitutional order from its own violence: ‘The sacrifice serves to protect the entire community from its own violence; it prompts the entire community to choose victims outside itself.’ One of the problems of Girard’s account is that it implies that sacrifice is an anthropological feature of all social orders, as it is the origin of the social bond.

The Material Study of the Constitutional Order  271 the latent unleashing of social violence. But it is far from being only a ­limitation, as it also participates in the constitution of the social order. The intuition, however, has been given a twist by modern constitutional theorists. Instead of looking into the rational choice and preference-based types of explanation or justification of the political community, political theology assumes that a political community is possible only if subjects are willing to sacrifice something for the existence or preservation of the constitutional order. This means that, as brilliantly described by Moshe Halbertal, in modern constitutionalism this logic takes the form of ‘sacrificing for’ rather than the ancient practice of ‘sacrificing to’.45 The potential unevenness that lies at the core of the social order is due to the exposure of people (willingly or unwillingly) to sacrifice for the sake of the constitutional order.46 The latter is portrayed as a rationalisation of a certain political dynamic, but even more importantly, it is productive of order. The unevenness that one can detect in the relation between the subject and the sovereign is a consequence of the idea that the social order lies not on contractual exchange or mutual agreement over a convention, but rather on a surplus that is required to (or freely given by) the citizen. According to this perspective, the possibility of an emergent constitutional order is premised on the capacity of citizens to bear the exposure to a fundamental political challenge. Will rather than reason is the relevant constitutional faculty.47 This position offers a powerful critique to the idea that a political community is a creation of a contractual mechanism or a conventionalist procedure. But the pars construens is less convincing and leaves to an impalpable and flimsy concept like political will the function of holding a political community together and developing a constitutional order. At best, constitutional theology can provide a critique of the formalism affecting the social contract tradition and conventionalism by emphasising the necessary symbolic dimension of constitutional orders. But it severs the relation between the symbolic and the organisation of society. Perhaps, it is also true that the sacrificial logic behind it is capable of triggering constitutional imagination in virtue of its capacity to reach the symbolic sphere, but it is not clear why this would be the only available logic and why it would shape a constitutional order with a certain form.48 Nor it is clear why it is always the case that sacrifice represents the cornerstone of political commitment. As noted by Wendy Brown, ‘sacrificing for’ can also become a device for hampering imagination when it is associated to an economic-driven logic.49 As Brown remarks, the sacrificial logic can be turned on its head and become a shared sacrifice for the sake of only some parts of the community. For example, this is the logic undergirding the constitutionalisation of austerity.50 But this latter instantiation cannot be fully 45 M Halbertal, On Sacrifice (Princeton, Princeton University Press, 2011). 46 P Kahn, The Reign of Law (New Haven, Yale University Press, 1997). 47 See, on this point, the brilliant analysis in P Kahn, Putting Liberalism in Its Place (Princeton, Princeton University Press, 2004) ch 4. More recently, see the application of these faculties for a comparative analysis by P Linden-Retke, ‘History, System, Principle, Analogy: Four Paradigms of Legitimacy in European Law’ Columbia Journal of European Law (forthcoming). 48 This is the hypothesis advanced most famously by Girard, Violence and the Sacred (n 44). 49 Brown notes that the constitutionalisation of austerity has brought with it a discourse on how to share its burden which has soon become a ‘shared sacrifice’ for the health and growth of the economy: W Brown, Undoing the Demos (New York, Zone Books, 2015) 213–4. 50 E Christodoulidis, The Redress of Law (Cambridge, Cambridge University Press, 2021).

272  Marco Goldoni grasped if sacrifice is understood as generative of a symbolic dimension. Rather than remaining anchored to untenable notions of life and body, charisma and sovereignty, the logic can be grasped only when attached to processes of social differentiation and organisation. Although a material approach to constitutional theory recognises that sets of uneven relations might comprise the core of societal and constitutional formation, it cannot accept that the account given by political theology exhausts all the possibilities of constitutional ordering. Political theology might be able to explain the myth and narrative of a constitutional culture, but it does not provide a general method to apprehend the internal relation between societal formation and constitutional order. In brief, political theology cannot offer a universal theory of the origin of the constitutional order, although it claims it can deliver it. The social relations that are addressed by the material study of the constitutional order concern fundamental traits of social organisation. Politically, this means that one focuses on the relation between those who are governed and the governed as the foundational distinction.51 Socially and economically, the focus on social organisation entails that labour and the activities necessary for the reproduction of society are assumed to be the directive factors of organisation (at least within the political economy of modern constitutional orders). In a nutshell, political organisation and social differentiation constitute the key moments of articulation of the constitutional order.52 Legally, the outcome is the recognition that order is a process of assemblage which entails operations of different nature: legal relations are then embedded in the level of cooperation and solidarity, or of sheer conflict and antagonism, that strengthens their binding nature.53 Accordingly, a distinction between the legal system or order (Ius) and the aggregation of legislation, decrees and the other legal instruments of regulation (Lex) is usually inherent to the relation that sustains both societal and constitutional formation. A warning is necessary, though, at this stage: what counts as division of labour and organisation of fundamental reproductive activities cannot be essentialised. Such an organisation is contextual to the type of society the constitutional theorist is observing. For this reason, only modern societies with complex organisation are the object of the material study. And even for these cases, one ought to remain alert to the different compositions of productive factors and forces because these aspects of the social organisation of production and reproduction vary according to epoch and jurisdiction. It goes without saying that the assembling of the social and constitutional order also requires the conveying of a constitutional imagination. But there is no reason to postulate that the only force capable of supporting the social and constitutional order is an imagination driven by the sacrificial logic. Imagination can be fed by other driving forces. This entails a crucial change of perspective: the material study shifts the 51 It should be noted that Loughlin accepts the fundamental role of the distinction between those who rule and those who are ruled as well: Loughlin, The Idea of Public Law (n 11) 5. 52 The intellectual points of reference of this conception of societal formation are to be found in contemporary anthropology and sociology of constitutions: cf C Thornhill and P Blokker (eds), Sociological Constitutionalism (Cambridge, Cambridge University Press, 2017). 53 Cf A Supiot, Homo Juridicus (London, Verso, 2010) ch 2; Christodoulidis, The Redress of Law (n 50) ch 3.

The Material Study of the Constitutional Order  273 preference accorded by the social contract and political theology to the extraordinary in favour of the ordinary.54 It also shifts the focus on the study of how the ordinary is imagined and performed.55 The production of normality and its forms of organisation are the focus of the material study. Not only the constitutional disposition of the citizen to sacrifice or to calculate its own self-interest, but the creation of institutions as patterns for the social order. The norm and not the exception is the primary interest for the constitutional theorist.56 III.  THEMATISING THE CONSTITUTIONAL ORDER AS LEGAL ORGANISATION

As already noted, the material study of the constitutional order is not a normative approach, though it recognises that it has to do with normative concepts. This means that it does not provide normative reasons for the design of constitutional orders, but it aims at the understanding of the reality of existing constitutional orders. The material study’s main focus is indeed on the formation and development of constitutional orders. A pause for clarification is necessary at this point. First, it should be stressed that order is here stipulated to be in contrast with the concept of system. This is more a definitional point, but as mentioned before, the material study is concerned primarily but not exclusively with a system of norms. There are indeed other important ordering factors that qualify as constitutional. Accordingly, the material study aims at giving an account of those factors that have produced a constitutional order and of their organisational principles. Of course, this is not to undermine the role of norms as they remain essential for the definition of the legal aspect of the order. The organisation of an order is indeed obtained through norms. Third, the material study maintains that the form of the constitutional order is as important as its matter. But given the central role of organisation, as highlighted in the previous section, the constitutional order is coupled with a social basis, and its relation with constitutional form (codified written constitution, constitutional statutes, constitutional conventions) is not exhaustive of the reality of the constitutional order itself. In fact, as mentioned at the beginning of this chapter, an inquiry into the nature of the constitutional order is necessarily two-fold: on one hand, it aims at identifying the characterising elements of the constitutional order (what makes it what it is) and, on the other hand, the processes and actors that have shaped it (under the assumption that these processes cannot be fully severed from the order itself). Following these preliminary remarks, the main aspects of the constitutional order that should be studied by public lawyers concern the conditions of its unity, its objectives or fundamental norms and its ‘bearing subjects’. These features go necessarily together because without one of the three elements the constitutional order would

54 It should be noted that conventionalism is not open to the same type of criticism as it is also focussed on ordinary action. 55 Cf M de Carolis, The Anthropological Paradox (Abingdon, Routledge, 2018). 56 The exception and the emergency are still relevant for the constitutional theorist, but as devices for protecting normality. This is the interesting interpretation of Schmitt’s institutionalist turn offered by M Croce and A Salvatore, The Legal Theory of Carl Schmitt (Abingdon, Routledge, 2013).

274  Marco Goldoni be ineffective and seriously unstable. However, for analytical purposes they can be distinguished and should also be studied as related but distinct features of the constitutional order. This analytical distinction will pay its dividends when the object of the inquiry is the determination of constitutional identity and transformation. The starting point of the analysis is that the constitutional order can only emerge as unitary order and this is usually a political achievement. Although there are variations on different forms, the modern state remains the main form of political unity, though not exclusive (one could still think of empires and federations, for example).57 The reference to the state allows us to identify a crucial factor of ordering in the creation of political unity. The latter introduces spatial and temporal coordinates which enable the legal order to make sense of certain social interactions under its own terms. Of course, the construction of political unity within the constitutional order is a way to draw fault lines of inclusion and exclusion.58 But part and parcel of the drawing of these lines is a moment of selection which defines what can be included or excluded in that order. This is not surprising: there is a typically modern attitude in the definition of political unity given that it takes place against the background of an increasingly complex social organisation driven by the principle of division of labour. In fact, the formation of political unity (and its determination as a concrete political unity) is inextricable from the organisation of productive and reproductive processes of social relations. But the relation between the organisation of productive activities and differentiation is not of a causal nature: the political moment, precisely because of the selectivity involved in the process, still plays a key role in assembling and re-organising social relations.59 It is worth repeating that, at the constitutional level, unity is the outcome of processes of differentiation and specification. In modern circumstances, differentiation is the first step towards the unfolding of the division of labour. Specification implies the ascription of roles and positions to certain social groups or elites. It goes without saying that specification is not a neutral process, but one that brings about social and, accordingly, legal hierarchies. Specification is propaedeutic to the individuation of the governing class. It should be noted immediately that these two processes (differentiation and specification) do not undermine the creation of political unity. To the contrary: under conditions of modern capitalism, they are part and parcel of the construction of unity. The latter is the end-result of a particular mix of differentiation and specification. Ultimately, unity will be granted by fundamental norms; however, these norms cannot function as a priori conditions of knowledge. Fundamental norms will have to include or express (depending on whether they are formalised or not) the organising principles or values behind differentiation and specification. Moreover, given that differentiation and specification are imbricated in the political economy of a concrete society, contradictions within the

57 See for an enlightening constitutional reflection S Larsen, The Constitutional Theory of the Federation and the European Union (Oxford, Oxford University Press, 2021). 58 H Lindahl, Fault Lines of Globalisation (Oxford, Oxford University Press, 2013). 59 A materialist analysis of state formation is quite useful for understanding this point: see, for an overview, B Jessop, ‘Critical Theory of the State’ in E Christodoulidis, R Dukes and M Goldoni (eds), Research Handbook on Critical Legal Theory (Chestertown, Edward Elgar, 2019) ch 7.

The Material Study of the Constitutional Order  275 organisation of society might always play out and disrupt (or at least undermine) political unity in the longer term. Unlike other constitutional theories, the material study does not give a prominent place to the principle of sovereignty as a crucial factor of unity. While sovereignty is an important modern phenomenon, for the study of political unity it is more important to understand how the art of governing is exercised. Processes of differentiation and specification have to be steered, otherwise they can become disruptive of political unity.60 This is an activity that is in charge of managing organisation of powers and fundamental social relations. It is usually defined as governing and the analysis of it entails a move beyond a formalist understanding of the basic principle of constitutional organisation: separation of powers.61 Constitutional doctrine understands it as the separation of functions (typically: legislative, executive, judicial) and their distribution to different organs. While the idea of different functions sheds a light on the nature and the working of constitutional organs by describing their activities (eg, adjudicating or executing), it remains insufficient for explaining the formation and development of the constitutional order. Besides (or together with) functional differentiation, a governing activity that operates as an engine of unity for the different functions is necessary. There is political unity only as long as there is the capacity for unitary governing activities, at least for the pursuing of certain fundamental goals.62 Yet, the construction of unity is not for the sake of unity, but for the purpose of setting up and consolidating a concrete constitutional order with its social organisation and its aims. Governing can mean several things and historically its relationship with the constitution has mutated. In the twentieth century, a continental constitutional tradition has tried to understand the art of governing as a fourth autonomous function63 to be placed ‘above’ the classic three functions. In this way, the governing function would not be confused with executive power and would steer all the other functions in order to ensure that, despite their difference in kind, they do not pull toward radically diverse directions. This is a key point. Functions might be distributed to different organs (the legislative function, for example, can be shared by the legislature and the executive, as is the case in most constitutional orders) but they always assume typical forms: the legislative function, for example, is characterised by lawmaking processes. But is it possible to typify the governing activity? There is no typical form of the art of governing. For this reason, it is better to talk of a governing activity rather than

60 This is a theme that cuts through the history of modern constitutional thought. The classic starting point is Benjamin Constant’s reflections on neutral power (pouvoir neutre): J-P Feldman, ‘Le constitutionnalisme selon Benjamin Constant’ (2008) 75 Revue française du droit constitutionnelle 675. 61 For an updated and original analysis of the principle see C Möllers, The Three Branches (Oxford, Oxford University Press, 2013). 62 A critical point for political unity is represented by the other important constitutional principle of organisation of powers: federalism. Things become more complicated with the distinction between federation and confederation. While for the former term, it is relatively uncontroversial that the distribution of powers would still be compatible with a unitary form of government, it is less clear that the same can apply to confederations. A study of the federation which is compatible with the perspective advocated here is O Beaud, Théorie de la Federation (Paris, PUF, 2008). 63 See C Mortati, Il governo nell’ordinamento italiano (Milan, Giuffré, 2001).

276  Marco Goldoni function. Governing is an activity that can be undertaken in many ways and it is necessarily compatible with the forms of the standard three functions. For the same reason, there is no need to assume that the governing activity is entrusted with one supreme organ (though, this has often been the case) or it is an exclusive state competence.64 During the twentieth century, the art of governing has been increasingly the domain of the political system and, more specifically, of political parties. The awareness of the importance of political parties is far from being a novelty, even for constitutional studies,65 and it cuts across different jurisdictions and constitutional schools: from the US and France to the Commonwealth model.66 But in other realities as well, the political direction of the constitutional order is still determined by political parties.67 This brings us to the second important factor of ordering: the ‘bearer’ or subject of the constitutional order. The material study assumes that for the purpose of constitutional analysis, abstractions like ‘people’ or ‘nation’ are insufficient as ordering forces because they are hugely underdetermined. For the purpose of tracking constitutional reality, it is necessary to identify the concrete subjects, often having collective nature and an institutional form, that leave an imprint on the order itself. These subjects can be the monarch, an institutionalised church or clergy, and of course political parties. One could also add other institutions or subjects: heads of state or constitutional courts are other potential subjects capable of shaping in a structural way the constitutional order. The bearing subjects can also be a combination of these subjects and institutions in what could be defined a ‘constitutional bloc’.68 The crucial role of the bearing subjects is that they are the guarantor of the coupling between societal formation and constitutional order. In other words, for identifying the subjects that are concretely bearing a constitutional order one has to look at their capacity of working as relational force between the social and the constitutional level. The bearers of the constitution have to be capable of generating fundamental political aims and pursue them, but also make sure that the organisation of society is inspired by some of those aims. A vivid example of this function is still represented by the Chinese Communist Party and its role in shaping both Chinese society and its constitutional order.69 Nonetheless, formal institutions might be able to perform this organising 64 Another important distinction should be drawn with the principle of sovereignty. In a nutshell, the distinction is not a matter of category (as it is for example in Rousseau), but of different functions. For an opposite view see R Tuck, The Sleeping Sovereign (Cambridge, Cambridge University Press, 2015). Governing can be an art exercised without sovereignty (as it is the case for derivative legal orders). In the contemporary context, the European Union is the most important example of a governing polity without sovereignty: cf N Walker, ‘Late Sovereignty in the European Union’, in N Walker (ed), Sovereignty in Transition (Oxford, Hart Publishing, 2003) 3–32. 65 The classic reference is M Duverger, Political Parties (Ann Arbor, Michigan University Press, 1964). 66 D Levinson and R Pildes, ‘Separation of Parties, not Powers’ (2006) 119 Harvard Law Review 2236; Bellamy, Political Constitutionalism (n 5) ch 5. 67 For an overview of the role played by political parties in four European countries see P Ignazi, Party and Democracy (Oxford, Oxford University Press, 2017). 68 The current Iranian Constitution seems to provide a vivid example of an order based on two bearing forces: political party and religious leadership. See B Ackerman, Revolutionary Constitutions (Cambridge MA, Harvard University Press, 2019) ch 9. 69 J Shigong, ‘Written and Unwritten Constitutions: A New Approach to the Study of Constitutional Government in China (2010) 36 Modern China 12. Cf H He, ‘The Party’s Leadership as a Living Constitution in China’, in T Ginsburg, Simpser (eds), Constitutions in Authoritarian Regimes (Cambridge, Cambridge University Press, 2013) ch 11; B Ngoc Son, ‘China’s Material Constitution’, in M Goldoni and

The Material Study of the Constitutional Order  277 activity, but it cannot be predetermined whether they are only of one type (eg, only political parties) or a mix of different subjects and institutions.70 It is vital for the material study to be aware of the connecting role of the bearing subjects. They perform the vital organising function of coupling social and constitutional orders. Without this recognition, the risk is to fall prey of a vulgar version of political decisionism whereby the bearing subjects of the constitutional order are such just out of sheer political will. If the latter were assumed to be true, then there would not be any internal relation between societal formation and the constitutional order because the organisation of society would be obtained by a political decision emerging out of nothing. This brings us to the third and most defining feature of the constitutional order: its fundamental political objectives. This is the most distinctive element of the material study: the majority of theories of the legal or constitutional order either entails a neutral perspective or the recognition of a general but vague objective. In the first version, typical of legal constitutionalism, the constitutional order enables the possibility of pursuing goals or interests that are left to the free determination of individual subjects. In the second version, the objective is reduced to a vague type of achievement: to create a legal order or to maintain peace and stability. A crucial difference between the material study and other methodologies is that the fundamental political aims of a constitutional order have to be taken into account not only for their political value, but also for their legal salience. Fundamental political objectives do enjoy legal force and they radiate (as long as the relevant social facts obtain) across large sectors of society. This is the point where it is recognised that not only politics, but law as well cannot be reduced to a relation of means to ends. The legal world, and in particular the coupling of law and politics that is the constitutional order, contains and formalises its own objectives. At this stage, a crucial point should be emphasised: the objectives are not external aims for whose achievement a set of (legal) means is generated. The objectives are internal to the formation of the social and constitutional order and they are implicit or explicit in the social facts that ground the constitutional order. Hence, they cannot be diminished to the functions of the constitutional order. Rather, objectives are norms, and in this case fundamental norms.71 They might take up the form of principles or values72 as long as these do not take up only a deontological content. It is also possible that there are contradictions or just tensions between fundamental norms and the more this is the case the less stable the constitutional order will be. It is a task of the art of governing to navigate these contradictions and tensions. M Wilkinson (eds), Cambridge Handbook on the Material Constitution (Cambridge, Cambridge University Press, forthcoming). 70 It seems the case that there has to be a certain level of institutionalisation. A social movement, without any form of institutionalisation would not be able to count as a bearing subject. A classic mechanism of constitutional development is the movement-party dynamic: cf B Ackerman, We the People: The Civil Rights Revolution (Cambridge MA, Harvard University Press, 2013). 71 It should be made clear that the recognition of this state of affairs is premised on the idea that politics has made a visible entrance into the realm of constitutional law in modern times: Thornhill, A Sociology of Constitutions (n 23). 72 Given that the material study is about reconstructing constitutional reality, whether values, principles or, more generally, aims are morally good or bad is irrelevant.

278  Marco Goldoni In this vein, it should also be added an essential aspect of the fundamental norms: they identify the nature of the constitutional order. Because they condense the directive idea within foundational social relations, fundamental norms become the characterising normative aspects of an order. For the modern state, for example, this means that fundamental norms identify the form of the state (the liberal state, the totalitarian state, the authoritarian state, the constitutional state) or, as other authors prefer to refer to, political regimes. The fundamental norms do not obtain their force in virtue of a chain of formal validation. Their capacity of shaping the constitutional order depends on the relative strength of the bearing subjects and on their (ie, of the norms’) grip on the constitutional imagination. In technical terms, fundamental norms are in a relation of dependence with certain social facts and at the same time they determine part of the content of the other norms (principles and rules) of the constitutional order.73 The determination of the content cannot be complete as in the creation of new norms there is always an element of adaptation to the changing circumstances. But the way the determination is carried out depends largely on the institutional arrangements and the political will of the bearers of the constitutional order. IV.  CASE STUDY: CONSTITUTIONAL CHANGE

In this section, a practical application of the material study of the constitutional order will be developed. The topic of constitutional change74 lends itself quite well to a material analysis. It is often the case that sheer reference to formal criteria is not enough to grasp the reality and scope of constitutional change. This can give rise to mistakes in the reconstruction of constitutional reality. False negatives and false positives can be the end result of a formalist analysis of constitutional mutation. A material study of the constitutional order draws a distinction between rationalisation of the constitution and structural change. It also equips the public lawyer with the instruments for detecting when an informal constitutional change has taken place or, to the contrary, a formal constitutional change has not had any impact on the constitutional order. The key criterion is to look for a change or an effort to change in dramatic ways some of the fundamental relations and objectives of the constitutional order. Though these relations and objectives might not always be easily identifiable, a fictitious example of a change that might or might not be registered in constitutional formal terms should be sufficiently clear for the purpose: the complete transformation 73 Mitchell Berman believes that this relation of dependence can be described as a grounding relation, although, as he admits, this does not explain the nature of the determination relationship: Berman, ‘Our Principled Constitution’ (n 9) 1364. Be that as it may, the nature of the dependence relation is a topic that will have to be broached in future material research. 74 Constitutional change is here understood as synonymous with transformation. It might include constitutional amendments, but only under certain conditions. The literature on the topic is burgeoning and it is impossible to mention all the relevant titles. See, among the most important and recent ones: Y Roznai, Unconstitutional Constitutional Amendments (Oxford, Oxford University Press, 2017); R Albert, X Contiades and A Fotiadou (eds), The Foundations and Traditions of Constitutional Amendment (Oxford, Hart Publishing, 2017); R Albert, Constitutional Amendments (Oxford, Oxford University Press, 2019).

The Material Study of the Constitutional Order  279 of the regime of property in a given jurisdiction has so many consequences on multiple social sectors that would almost certainly produce a substantial constitutional change. A standard case of a false positive is represented by Switzerland.75 Switzerland’s total revision of its federal Constitution in 1999 represents an example of formal constitutional change without material constitutional change. Although the 1999 Constitution replaced the previous 1874 Constitution with a complete repackaging, it did not change any of the fundamental underpinnings of the Swiss political and social system. The foundations of the Swiss federal arrangement set by the first federal Constitution of 1848 are still in place today even after two total revisions of the Constitution.76 Rather than focusing solely on the latest total revision in 1999 itself, in order to understand constitutional change in Switzerland, it is more pertinent to pay attention to what happened between the two constitutional replacements, ie between 1874 and 1999. The main constitutional developments in this period are identified as the ‘expansion of the instruments of direct democracy, progressive centralisation – still with considerable cantonal autonomy, and the strengthening of Rechtsstaatlichkeit (Rule of Law)’.77 The 1999 Constitution represented the consolidation of these developments, rather than effecting such changes to the Swiss political system. The total revision in 1999 also served to codify the previously unwritten constitutional principles such as the principle of legality, proportionality, good faith.78 This constitutional reform for total revision was first proposed by the Federal Council and the Parliament with the aim of editing the wording of the Swiss federal Constitution.79 Among the goals of adopting a ‘new’ constitution was to ‘ensure that all fundamental rights would be clearly stipulated in an up-to-date text corresponding to the latest jurisprudence of the Supreme Court and to international standards’.80 It was therefore not presented as a major change to the constitutional system but rather a systematisation of the existing constitutional essentials of Switzerland. The 1999 Constitution consequently built upon the reforms made under the 1874 Constitution and limited itself to the reorganisation of the structure of the constitutional text and introduced no major constitutional change.81 However, from a formal perspective, as this is a new text of the constitution, its adoption created a brand new constitutional order for Switzerland and marked discontinuity in the constitutional history of Switzerland.

75 I rely, for the analysis of the Swiss case, on a chapter written together with T Olcay, ‘The Material Study of Constitutional Change’ in X Contiades and A Fotiadou (eds), Routledge Handbook of Constitutional Change (Abingdon, Routledge, 2020) ch 16. 76 F Biaggini, ‘Switzerland’ in D Oliver and C Fusaro (eds), How Constitutions Change: A Comparative Study (Oxford, Hart Publishing, 2011) 303. 77 ibid 317. 78 ibid 325. 79 T Fleiner, ‘Constitutional Revision: The Case of Switzerland’, in X Contiades (ed), Engineering Constitutional Change: A Comparative Perspective on Europe, Canada and the USA (Abingdon, Routledge, 2013) 344. 80 M Haller, ‘The New Swiss Constitution: Foreign and International Influences’ (2002) 30 International Journal of Legal Information 256, 261. 81 A Barbera, Costituzione della Repubblica (Milan, Giuffré, 2016) 12–3.

280  Marco Goldoni Given that the 1999 Constitution was also adopted through political means and by respecting the rule of law, it is fair to assume that a purely formalist reading would still interpret the change as a fundamental one because it is the outcome of a supermajoritarian political consensus and it produced a whole new constitutional text. However, this would fail to examine that constitutional change from a more material perspective, which would require to check whether the most important institutions and the fundamental aims have substantially been transformed. The point of the material study is seen here in its ability to identify where real change in the constitutional order took place. In the Swiss case, it is not the 1999 Constitution that brought about constitutional change. This total revision merely updated the structure and content of the constitutional text but introduced no significant change to fundamental social relations and political institutions. Neither did it represent an agreement on future transformations supported by the main political forces. Therefore, while the adoption of the 1999 Constitution is presented as an important formal constitutional change, it is of little consequence with regard to its effect on social life in Switzerland. In fact, Switzerland’s democratic principles, federalism, and main political subjects were not substantially transformed nor dismissed by the 1999 Constitution. At best, the new document can be defined as a rationalisation of the already existing constitutional reality. False negatives tend to be more numerous than false positives. A classic false negative is represented by the constitutional change determined by the Fascist regime in Italy.82 At the time of the rise of Mussolini to the position of President of the Council of the Government in 1922, the main constitutional document in Italy was still the so-called Statuto Albertino. This was a flexible constitution, meaning that almost all its provisions could be modified by statute and without special procedure. Moreover, the Statuto, despite being the constitutional charter for the whole country, was initially an octroyée constitutional document kindly offered by the King of Piedmont and Sardinia in 1848 to appease the tensions with the rising middle classes. At the moment of the creation of Italy as a unitary state in 1861 (under the lead of the Kingdom of Piedmont), the Statuto was simply extended to the rest of the kingdom, confirming the nature of Italy as a constitutional monarchy. The rise of Fascism, first as a movement and then as a party, dates to 1919/1920, and by 1922 the Fascist party obtained to have Mussolini appointed as the President of the Council in a coalition government. The appointment was made by following the letter and the practice of the Statuto, so there was no signal of a change from established patterns. In the years between 1922 and 1926, key statutes were approved by the Mussolini government, all by following the norms of the Statuto. Formally, there had been full continuity between the liberal constitutional age of the Statuto and the constitutional approach of the Fascist government. But the statutes enacted

82 Another prominent example of an analysis that shows how a constitutional transformation has been obtained without a formal recognition is Ackerman’s interpretation of the New Deal as a new constitutional regime: We the People: Transformations (Cambridge MA, Harvard University Press, 1998). Although Ackerman would probably not agree with this, his take on the New Deal as a transformative political experience is conducted with an approach utterly compatible with the material study.

The Material Study of the Constitutional Order  281 by Mussolini re-shaped in depth key aspects of the Italian constitutional order: abolition of the right to association and of free trade unionism with the establishment of a single Fascist trade union; introduction of censorship for media; a new electoral law based on a unique list of candidates (hence, abolition of political pluralism); creation of the Great Council of Fascism through which the Fascist party elite and militia were introjected directly into the state structure. Substantially, these are not only momentous changes, but the signal that a new political and constitutional regime, incompatible with the one envisaged by the Statuto, has emerged. Already in 1928, the prominent German constitutional lawyer Gerhald Leibholz, was noting that substantially of the provisions of the Statuto only the symbolic functions attributed to the monarchy were left.83 Although there is disagreement among scholars about the extent of the transformation, there is no doubt that a constitutional transformation had taken place.84 The principles and objectives introduced by the Fascist government were mostly incompatible with the moderate liberal constitutionalism of the previous political regime, but these changes were never fully systematised in a formal constitution.85 They actually took place through the creation of a single-party system, the deep restructuring of the state structure and its relation with society86 and the re-organisation of labour in corporatist terms.87 It is impossible to appreciate the relevance of this transformation without making reference to the main tenets of the material study of the constitutional order: the form of the political unity, its bearers and the main objectives of the order. It is also evident that these changes had a profound impact on fundamental social relations of Italian society.88 One could say that at least from 1926, Italy’s political regime had changed in a profound and impactful way. V. CONCLUSION

It is time to take stock. The material study of the constitution serves the aim of retrieving how the formation of the constitutional order is deeply embedded in fundamental aspects of social organisation. This is the material angle that it has to adopt when observing the constitutional order. The epistemic fruits of this approach are manifold: an understanding of constitutional functions, an appreciation of the role of the bearing subjects, a grasp of the difference between sovereignty

83 G Leibholz, Il diritto costituzionale fascista (Naples, Guida, 2017) 125. 84 The most sceptical among influential constitutional lawyers has been L Paladin, Saggi di storia costituzionale (Bologna, Il Mulino, 2008) 35. 85 Another momentous change was the agreement, ratified in a pact with the Vatican State, between the Church and the Italian state which brought to an end the conflict initiated with the formation of Italian unity. 86 See C Mortati, La costituzione in senso materiale (Milan, Giuffré, 1998) ch 3. 87 C Thornhill, ‘The Rise and Fall of Corporatist Constitutionalism’, in A Costa (ed), Corporatism and Fascism. The Corporatist Wave in Europe (Abingdon, Routledge, 2017) 79–100; P Kjaer, ‘European Crises of Legally-Constituted Public Power: From the “Law of Corporatism” to the “Law of Governance”’ (2017) 23 European Law Journal 417. 88 For an overview see C Ghisalberti, Storia costituzionale d’Italia (Bari, Laterza 2002) ch 4.

282  Marco Goldoni and government, a capacity for detecting constitutional change. One could also imagine other important benefits: for example, the material angle might illuminate which state organs can be deemed to be of constitutional relevance or simply administrative agencies. It also provides an entry point for a much-needed study of the political economy of the constitutional order. However, a cautionary note is necessary at this point. The material study does not claim to be an exhaustive explanatory methodology. For understanding certain phenomena of the constitutional order, for example constitutional interpretation, it will have to be integrated with other approaches.

11 The British Constitution as an Improvised Order DAVID HOWARTH

I. INTRODUCTION

T

he assertion that constitutions, in common with the whole of the law, are human artefacts might now seem commonplace. But for many centuries, and still currently in parts of the world, constitutions were not for humans to determine but for God, or perhaps for Nature. Thinking about constitutions as human artefacts, and specifically as artefacts that humans can design, is a recent development. Although occasional mentions of the phrase ‘constitutional design’ occur even in the nineteenth century, the phrase began to appear regularly only from the late 1980s,1 and although a common theme in that literature is that ‘constitutional design’ can be traced back to the Federalist Papers or even to Plato, as a self-conscious activity it has barely begun. The immediate inspiration for much of the constitutional design literature lay in economics, especially in public choice theory, which amounts to treating politics as if it were reducible to economics, devoid of idealism and values and motivated purely by self-interest.2 Those origins still tend to dominate the field, but that dominance gives rise to a serious problem. The predictive power of public choice theory is poor.3 Its most celebrated prediction, that rational voters would not bother to vote, is invalidated somewhere in the world almost every day.4 As a result, for practical legal purposes, including for drafting constitutions, its value is limited. Instead, if constitutional lawyers are to take the idea of design seriously as a method of understanding and making constitutions, we could turn instead to ideas drawn not from public choice theory but from disciplines whose purpose is to make 1 See, eg, M Shugart and J Carey, Presidents and Assemblies: Constitutional Design and Electoral Dynamics (Cambridge, Cambridge University Press, 1992). 2 See generally, T Ginsburg, ‘Public Choice and Constitutional Design’ in D Farber and AJ O’Connell (eds) Research Handbook on Public Choice and Public Law (Cheltenham, Elgar, 2010). 3 J Mashaw, ‘Public Law and Public Choice: Critique and Rapprochement’ in Farber and O’Connell Research Handbook (n 2). 4 See generally, J Aldred, Licence to be Bad: How Economics Corrupted Us (London, Allen Lane, 2019).

284  David Howarth something that works in the world as it is, in all its complexity. Foremost among those disciplines is engineering, especially the part of engineering that deals with ‘human factors’ such as information systems.5 Engineers have long thought systematically about design and have produced not only formal models of how it should be done,6 but also ethnographic accounts of the messy reality of how it is done.7 The idea I take up here is from the messy reality, although the formal ideal model has its attractions too, especially as a way of understanding structures. The observation of messy reality I begin with is that in the everyday world of information systems, which are plagued not only by minor errors but also by obstacles often unwittingly created by the system’s designers, much depends on users adapting and changing the system to make it work. Users improvise and invent ‘workarounds’, which result in the system functioning better, but without, at least intentionally, changing it in any fundamental way.8 I want to explore the suggestion that much of what passes for constitutional reform in the United Kingdom is better understood as constitutional improvisation or workarounds. British constitutional change has often been criticised as piecemeal and incoherent.9 One could perhaps better see it as a series of attempts, some more successful than others, to coordinate new improvisations with older ones, or to co-ordinate different improvisations offered at the same time. In methodological terms, this chapter attempts to instantiate an approach in which a model is chosen on the basis of a broader theoretical perspective, here the perspective that treats law as a designed human artefact, and then tests it against the observed facts. Its goal is to answer a question in the form ‘to what extent is X an accurate representation of Y’ – here ‘to what extent is UK constitutional reform improvised’?10 But it also attempts to assess the success or failure of X – here improvisation – against a set of criteria developed from the implications of the model itself. The concepts in the model, such as ‘design’ and ‘improvisation’, are constructs taken from their own aspects of social reality, but here they are used as tools to investigate another aspect of social reality – constitutional systems. This kind of analogical investigation might work only in tactical terms, by providing no more than an immediate flash of recognition, but it might operate instead as a strategic analogy, in which similarities in the structure of the two activities are close enough that it works for many different aspects of the two activities, to the extent that it provides illumination not only where the parallel works but also where it does not, bringing

5 D Howarth, Law as Engineering: Thinking about what lawyers do (Cheltenham, Elgar, 2014). 6 G Pahl, W Beitz, J Feldhusen and K Grote (K Wallace and L Blessing trans. & eds) Engineering Design: A Systematic Approach 3rd edn (London, Springer-Verlag, 2007). 7 D Vinck, Everyday Engineering: an ethnography of design and innovation (Cambridge, Mass, MIT Press, 2003). 8 S Alter, ‘Theory of Workarounds’ (2014) 34(55) Communications of the Association for Information Systems 1041. Alter defines a workaround as ‘A goal-driven adaptation, improvisation, or other change to one or more aspects of an existing work system in order to overcome, bypass, or minimize the impact of obstacles, exceptions, anomalies, mishaps, established practices, management expectations, or structural constraints that are perceived as preventing that work system or its participants from achieving a desired level of efficiency, effectiveness, or other organizational or personal goals.’ (1044). 9 V Bogdanor, The New British Constitution (Oxford, Hart Publishing, 2009). 10 Cf Alison Young’s ch 3 in this volume.

The British Constitution as an Improvised Order  285 out the distinctive features of each activity in a kind of dialectical process. In structural analogies, one can potentially transfer ideas, for example about what counts as a success or a failure, from one side of the analogy to the other. The analogy between law and design, especially engineering, seems to be a structural analogy, so there is some chance that one can push it further in the way proposed here. But it is only by pushing an analogy too far that one comes to see what kind of analogy it is.11 The chapter begins with introducing the idea of improvisation, its potential theoretical importance, and its suitability for describing much legal activity, not only case law, which is perhaps its natural home, but also statute law. It proposes a way of distinguishing good improvisation from bad and suggests how that might apply in law. It then discusses some of the main UK constitutional reforms of the late twentieth and early twenty-first centuries, classifying some as improvisations, and some not, and cautiously sorting the improvisations into good or bad examples of the genre. It ends with some tentative reflections about the significance of constitutional improvisation for constitutional theory. II.  SPONTANEOUS ORDER, IMPROVISATION AND DESIGN

In The Craftsman, Richard Sennett observes that we should distinguish between the ‘spontaneous’ and the ‘improvised’.12 The ‘spontaneous’ is a ‘mindless occurrence’, an unthinking reaction to a situation. The ‘improvised’, in contrast, is a form of creativity, proceeding, even if done rapidly, from previous ‘observation and experiment’, using the tools and materials immediately at hand.13 Improvisation occurs frequently not only in crafts but also in performing arts, especially music and theatre. Organisational theorists have also noticed it in business organisations.14 The ‘improvised’ in turn contrasts with the ‘designed’, which involves mental anticipation of the creation of a new object, separated from the creator’s immediate encounter with the materials – ‘the idea of a thing made complete in conception before it is constructed’.15 In improvisation, although consisting of deliberate action, conception and execution are not separate.16 Improvisation can also be contrasted with ‘muddling through’ or ‘incrementalism’, in the sense of refraining from attempting a comprehensive analysis of all 11 See further D Howarth, Law as Engineering: Thinking about what lawyers do (Cheltenham, Elgar, 2013) vii–viii. 12 R Sennett, The Craftsman (London, Allen Lane, 2008) 235–237. 13 A Miner, P Bassoff and C Moorman ‘Organizational Improvisation and Learning: A Field Study’ (2001) 46 Administrative Science Quarterly 304, 314 define using the materials immediately to hand as a separate activity of ‘bricolage’, reserving ‘improvisation’ for the activity of conceiving and making at the same time, but distinction seems too fine – it is difficult to think of examples of ‘bricolage’ that are not also ‘improvisation’. 14 Eg, C Moorman and A Miner ‘Organizational improvisation and organizational memory’ (1998) 23(4) Academy of Management Review 698; Miner et al, ‘Organizational Improvisation’ (n 13); K Weick, ‘Improvisation as a mindset for organizational analysis’ (1998) 9(5) Organization Science 543; A Lewin, ‘Jazz improvisation as a metaphor for organization theory’ (1998) 9(5) Organization Science 539; A Hadida, W Tarvainen and J Rose, ‘Organizational Improvisation: A Consolidating Review and Framework’ (2015) 17 International Journal of Management Reviews 437. 15 Sennett, The Craftsman (n 12) 42. 16 Cf Miner et al, ‘Organizational Improvisation’ (n 13) 314.

286  David Howarth possible courses of action but simplifying a problem and limiting choice so that it becomes tractable.17 Incrementalists are still planning or designing, albeit appreciating the limits of those activities. Improvisers confine choice in a different way, by using only what it is at hand to take the next step. They focus on the process of creation more than on the desired result. ‘Muddling through’ is another form of limited planning, in which an existing object or institution is put to a new use, making planned changes that make allowance for its new environment. This can be called ‘adaptation’.18 Sometimes adaptation incorporates an element of improvisation, when the planned changes fail to work and are supplemented by workarounds. III.  THEORETICAL IMPLICATIONS

The three-way contrast between the ‘spontaneous’, the ‘improvised’ and the ‘designed’ is potentially important at a theoretical level for constitutional studies and possibly more widely. For many decades, at least since Polanyi’s The Logic of Liberty: Reflections and Rejoinders and Hayek’s The Constitution of Liberty,19 and possibly much earlier,20 lawyers and social scientists have grappled with the contrast between law as facilitating ‘spontaneous order’ and law as formally designed artificial social structure. In Hayek’s terminology, one can contrast order that is ‘grown’ (or kosmos) with order that is ‘made’ (or taxis). Advocates of spontaneous order claim that it uses the vast amount of information individuals generate about what they want (and what they can do to satisfy what others want) far more effectively than attempts at social planning.21 Sceptics about spontaneous order respond that purely evolutionary processes have no foresight and cannot anticipate even obvious future changes in the environment, so that in an unconscious evolutionary process, forms potentially useful following a future change might be selected out before the change happens. That is why people and organisations do in fact plan, consciously shaping their interactions and relationships, and not only survive but flourish.22 Improvisation, however, provides a third option, neither unconsciously spontaneous nor planned. In improvisation, unlike spontaneity, coordination happens, but coordination is an activity not a result. Improvisers consciously try to bring their actions into harmony with the actions of others, but unlike planners and designers, they are not aiming to bring about a chosen final pattern.23 17 C Lindblom, ‘Still Muddling, Not Yet Through’ (1979) 39(6) Public Administration Review 517. 18 Miner et al, ‘Organizational Improvisation’ (n 13) 314. 19 M Polanyi, The Logic of Liberty: Reflections and Rejoinders (London, Routledge, 1951); F Hayek, The Constitution of Liberty (London, Routledge, 1960). 20 See N Barry The Tradition of Spontaneous Order (Arlington, Library of Economics and Liberty, 1982). 21 D Klein, ‘Convention, Social Order, and the Two Coordinations’ (1997) 8 Constitutional Political Economy 319. 22 See, eg, JK Galbraith, The New Industrial State 2nd edn (London, Penguin, 1974). 23 Cf Klein, ‘Convention’ (n 21); T Schelling, The Strategy of Conflict (Cambridge, Mass, Harvard University Press, 1960) and T Schelling, Micromotives and Macrobehavior (New York, Norton, 1978).

The British Constitution as an Improvised Order  287 Advocates of spontaneous order might argue that improvisation will not necessarily differ from spontaneous order in its effects, because competitive processes will select from among improvisations those that favour a higher level of coordination.24 But improvisation has important features absent from processes of random variation and selection. In particular, improvisation, because it involves conscious coordination, seems more likely to lead to path dependency.25 That is, it is more likely to lead to situations in which future choices are partially determined by past choices and starting again becomes progressively more costly. To take an example from the world of information technology, creating a supplementary paper record system because a computerised one fails to work properly creates a backlog of details needing to be entered into the computerised system before it could begin to operate again by itself.26 Admittedly, some improvisations are more easily reversed, and even random variation can lead to path dependency, but because improvisation starts with coordinating the next step with what has already happened, it effectively begins with choosing a path from the past. Some have claimed that kosmos includes the common law process of building credible legal rules from individual cases, sifting the wise from the unwise by a process supposedly akin to random variation and selection of the fittest.27 Contrary to this Hayekian view, however, although the common law does not do comprehensive planning, it is not a set of mindless random reactions somehow culled by natural selection either. It is largely improvisation. Common law judgments occur only when needed, and although the practice of giving ex tempore judgment is in decline,28 so that judges now take time for reflection, they reflect almost exclusively on the process of decision itself and on the materials at hand, not on their plans for how their decision will affect the rest of the law or the world.29 Common law decision aims at coordinating the past (previous cases) with the present (the facts) and the immediate future (the result, and possibly the next case). But, also contrary to the Hayekian view, statutes can be improvisations, not just examples of social planning. As the Brexit crisis illustrated, the House of Commons 24 Cf Klein, ‘Convention’ (n 21) 330. 25 K Dooley and A Van de Ven, ‘Explaining Complex Organizational Dynamics’ (1999) 10(3) Organization Science 358; P Augsdorfer, Path dependency in unplanned R&D (Ingolstadt, Ingolstadt University of Applied Sciences, 2004); but see also Y Liu, D Lv, Y Ying and F Arndt, ‘Improvisation for innovation: The contingent role of resource and structural factors in explaining innovation capability’ (2018) 74 Technovation 32. 26 G Fitzpatrick and G Ellingsen, ‘A Review of 25 Years of CSCW Research in Healthcare: Contributions, Challenges and Future Agendas’ (2013) 22 Computer Supported Cooperative Work 609, 623–624. 27 R Sugden, ‘Spontaneous Order’ (1989) 3(4) Journal of Economic Perspectives 85; A DeVany, ‘Information, Chance, and Evolution: Alchian and the Economics of Self-Organization’ (1996) 34 Economic Inquiry 427; A Ogus ‘Law and Spontaneous Order: Hayek’s Contribution to Legal Theory’ (1989) 16(4) Journal of Law and Society 393–409. 28 But see J Dyson A Judge’s Journey (Oxford, Hart Publishing, 2019) ch 8, recording that most judgments on the criminal side of the Court of Appeal are still given ex tempore. 29 Eg, Sir John Dyson (Dyson, A Judge’s Journey (n 28)) rarely refers to wider consequences when discussing his own decisions. Admittedly judges do confess to making ‘policy’ decisions, especially in ‘technical’ or ‘lawyers’ law’ fields (see, eg, A Paterson, Final Judgment: The Last Law Lords and the Supreme Court (Oxford, Hart Publishing, 2013) 274–276), but no one could accuse judges of undertaking detailed policy analysis before making them. Cf N Simmonds, ‘Bluntness and Bricolage’ in R Harrison and H Gross (eds), Jurisprudence: Cambridge Essays (Oxford, Oxford University Press, 1992).

288  David Howarth can pass Bills near instantly with little forethought, using newly fashioned procedural means.30 Even before that crisis, governments sometimes promoted Bills that reacted quickly to events using materials immediately to hand, for example recycling the text of previous provisions.31 Some Bills that are longer in gestation can also plausibly count as improvisations where they respond to immediate problems using existing materials and without much thought about the consequences. The Fiscal Responsibility Act 2010, for example, was produced to solve a short-term political problem, the Prime Minister’s need to be seen to do something about the government’s fiscal deficit, which it achieved by transposing an existing policy announced in budget documents into instant legal form.32 One objection to saying that legislation can be improvised is that, as Sennett remarks, ‘improvisation is a user’s craft’.33 Judges, civil servants or commercial lawyers designing contracts are all plausibly users of rules created by others, but how can a sovereign Parliament, making primary legislation unconstrained by higher law, count as a user of the law? Is it therefore always a norm designer? But even a sovereign Parliament is constrained by the requirements of legal form. Statutes need to be written in a way that citizens, officials and judges recognise as imposing legal obligations. Parliament is a user of the concept of law. IV.  INTERACTION BETWEEN IMPROVISATION AND DESIGN

Improvisation and design interact in more complex ways. For example, designers can learn from and incorporate successful improvisation into their designs.34 Designers can also deliberately allow space in their designs for improvisation or adaptation by users.35 Similarly, legislatures can incorporate case law developments into statutes and deliberately leave issues for determination by the courts.36 More broadly, design can take on some of the characteristics of improvisation. Designers can, for example, reflect, as a design technique, on the success or failure of workarounds.37 One can see that process legally in a recurrent working method of the Law Commission, building general analyses of problems out of analyses of specific 30 See the proceedings on the European Union (Withdrawal) Act 2019 (https://services.parliament.uk/ Bills/2017-19/europeanunionwithdrawalno5/stages.html) and the European Union (Withdrawal) (no. 2) Act 2019 (https://services.parliament.uk/Bills/2017-19/europeanunionwithdrawalno6.html). Admittedly the latter was longer in preparation than the former, but it also depended on an entirely new procedure, a substantive motion under Standing Order 24 (see Hansard, HC (series 6) Vol 664, cols 76–77, introduced the previous day). See also the proceedings on the government Bill which became the Early Parliamentary General Election Act 2019 (https://services.parliament.uk/Bills/2019-20/earlyparliamentarygeneralelection. html). 31 Eg, Criminal Evidence (Witness Anonymity) Act 2008, the process for passing which is described in D Howarth, ‘The Criminal Evidence (Witness Anonymity) Act 2008’ (2008) 8 Archbold News 5. I had a similar experience with the Terrorist Asset-Freezing (Temporary Provisions) Act 2010. 32 Cf HM Treasury, Pre-budget Report 2008 (London, TSO, 2008) 39. 33 Sennett, The Craftsman (n 12) 236. 34 Alter, ‘Theory of Workarounds’ (n 8) 1053. 35 ibid, 1055. 36 D Howarth, ‘On Parliamentary Silence’ UK Const L Blog (13 Dec 2016) (available at https://ukconstitutionallaw.org/) (adopting an idea of Sir Stephen Laws). 37 Alter, ‘Theory of Workarounds’ (n 8) 1055.

The British Constitution as an Improvised Order  289 cases.38 Less to be admired, is the process of drafting statutes that can be deflected by political circumstances into improvisation. For example, a government might take advantage of a Bill that just happens to be under consideration in Parliament to respond ad hoc to a sudden political issue, or it might produce short-term workarounds to drafting problems or to policy misunderstandings rather than re-starting the design work from scratch.39 V.  THE CONFLICTUAL SIDE OF IMPROVISATION

We should also notice a more conflictual side of improvisation. Users can deploy their creativity not only better to do their jobs but also to frustrate someone else’s purposes, for example to resist employers’ organisational controls, giving false impressions of progress or converting organisational resources to personal use.40 The legal equivalent can be seen in suspicion of ‘judicial activism’, not so much that judges are benefiting themselves materially but that they are covertly following their own political agendas.41 The problem for those suspicious of judges improvising for their own (normative) benefit is the same as that faced by employers suspicious of their employees: it is very difficult to prevent illegitimate improvisation without also killing off improvisation that allows the job to be done at all.42 Ministers would doubtless consider the series of parliamentary manoeuvres in 2018–19 that led to the two backbench European Union (Withdrawal) Acts to constitute hostile improvisation intended to frustrate the proper purpose of the UK system of government, which, from the Whitehall perspective, is to facilitate the implementation of government policies. Conversely, members of Parliament who voted for that legislation might consider that they were improvising to undermine ministers’ attempts to evade Parliamentary control.43 38 See, eg, Law Commission of England and Wales, Aggravated, Exemplary and Restitutionary Damages: A Consultation Paper (London: HMSO, 1993). 39 Examples include: inserting the Leveson Inquiry’s recommendations into the Crime and Courts Bill 2013 and the Enterprise and Regulatory Reform Bill 2013; putting clauses about compulsory carrier bag charges into the Climate Change Bill 2008; protecting the Prince of Wales’s correspondence with ministers from freedom of information requests in the Constitutional Reform and Governance Bill 2010; and, in private law, the statutory reversal of Baker v Corus (UK) plc [2006] UKHL 20 in the Compensation Bill 2006. A more venerable private law example is the late insertion of the notoriously confusing clauses 1(3)(a) and 1(3)(b) into the Occupiers Liability Bill 1984. 40 Alter, ‘Theory of Workarounds’ (n 8) 1054. 41 J Griffith, The Politics of the Judiciary (Manchester, Manchester University Press, 1977); R Ekins, The Dynamics of Judicial Power in the New British Constitution (London, Judicial Power Project, 2017). 42 Another design problem is that legislators want judges to be able to improvise about means and not about goals, but either cannot agree on goals amongst themselves or want to avoid declaring their goals openly. The Simms ‘Principle of Legality’ in effect exploits those weaknesses, allowing judges to insert their own goals through claiming that the legislature recognises all general legal principles not expressly excluded – a neat example of the ‘civilising force of hypocrisy’ (J Elster, ‘Deliberation and Constitution Making’ in J Elster (ed), Deliberative Democracy (Cambridge, Cambridge University Press, 1998). 43 See D Howarth, ‘Westminster versus Whitehall: What the Brexit debate revealed about an unresolved conflict at the heart of the British constitution’ in O Doyle, A McHarg and Jo Murkens (eds), Constitutions Under Pressure: The Brexit Challenge for Ireland and the United Kingdom (Cambridge, Cambridge University Press, 2021) and ‘Westminster versus Whitehall: Two Incompatible Views of the Constitution’, UK Const L Blog (10 April 2019) (available at https://ukconstitutionallaw.org/).

290  David Howarth VI.  ASSESSING IMPROVISATIONS

The conflictual side of improvisation illustrates one way in which improvisations might go wrong. But they might go wrong in other ways. Most obviously a shortterm solution might create bigger problems later. Engineers point to workarounds in which users induce systems to work by turning off safety devices or alarms, with predictably catastrophic results.44 More subtly, a short-term solution might take the long-term development of the system down a path that obstructs the adoption of better solutions later. In this way, a series of improvisations, each trying to patch problems caused by its predecessor might not generate a satisfying sense of collective creativity but rather rising panic as matters hurtle out of control. As Sennett says in relation to jazz, bad improvisations cause players to ‘lose focus’ and create a ‘maze’.45 On the positive side, good improvisations simplify the pre-existing material so that later players can create new variations of their own logically related to the same theme. Translated into the legal sphere, the characteristic of good improvisation, that it facilitates future development, comes through in, for example, the idea that good judges identify principles, or reasons underlying rules, which future courts could use in as yet unanticipated cases.46 The identification of principle focuses discussion and simplifies it, but at the same time allows further development. In contrast, bad common law decisions might resolve cases in ways that strike the judge as fair or reasonable, or perhaps just decide the case as narrowly as possible, but fail to resolve the underlying issues and, by making the law more complex and less certain, in the longer term create a Sennett ‘maze’. One striking example is English law about tortious liability for psychiatric harm, described by a leading commentator as ‘an embarrassment’.47 Moving from common law to statute, one can posit a similar set of criteria. Did the reform identify a specific problem and solve it? Did it do so in a way that would allow future legislators to understand the direction of change and either develop it further or consciously reverse it? And did it later cause worse problems than it solved? VII.  DISTINGUISHING IMPROVISATION FROM NON-IMPROVISATION

The view that not only case law but also statutes can be improvisations or workarounds has great attractions. But before applying it to UK constitutional developments of the past 20 years, a word of caution is necessary. Not everything is an improvisation. Both spontaneity and design also exist. We should take care not to over-interpret. How do we tell the difference? 44 Alter, ‘Theory of Workarounds’ (n 8) 1052. 45 Sennett, The Craftsman (n 12) 236. 46 K Llewellyn, The Common Law Tradition: Deciding Appeals (Boston, Little, Brown, 1960) 37–40; K Llewellyn, The Theory of Rules (Chicago, University of Chicago Press, 2011) ch 4. Ronald Dworkin’s ‘chain novel’ metaphor contains a similar idea (R Dworkin, Law’s Empire (London, Fontana, 1986) 228–238). 47 J Stapleton, ‘In Restraint of Tort’ in P Birks (ed), Frontiers of Liability (Oxford, Oxford University Press, 1994) 95.

The British Constitution as an Improvised Order  291 A preliminary point is that as researchers we have no access to the minds of the relevant actors. All we can do is to make inferences from external evidence – official papers, Bills, debates and self-reports such as diaries and autobiographies. The inevitable result is a reduction in the degree of certainty with which we can announce our results. Another preliminary point is that improvisation and design can be achieved not only by individuals but also by groups (think of sports teams or musical ensembles) but it is perfectly possible that different individuals in a group are acting differently – some following a preconceived plan, others not. It is rarely possible to pick apart these differences and so we can only characterise the activity of groups based on the evidence of what they produced as groups rather than on individual contributions. How can we distinguish between spontaneous reaction, improvisation, and design? Sennett’s suggestion is that we should concentrate on tools and materials. Improvisation is often characterised by using tools for purposes for which they were not originally intended on materials that just happen to be at hand. On that basis, a classic example of a constitutional improvisation would be using the Standing Orders of the House of Commons to bring about ‘English Votes for English Laws’, a procedure under which only MPs elected for English constituencies participated in some stages of the passage of Bills that affected only England, and the consent of a majority of those MPs was required for such Bills to proceed.48 Even designers use tools and materials, but the characteristic of improvisation is using tools and materials that already exist and are to hand, rather than searching out or developing new tools or materials. A Bill cannibalised from bits of previous Bills is a strong candidate for being an improvisation (although it might sometimes be possible to argue that a Bill reproducing a previous proposal is merely reintroducing a fully thought through plan). Another important difference between design and improvisation is that design fully separates conception and execution, whereas improvisation to some extent combines them. For that reason, good evidence of improvisation as opposed to design is absence of an elaborate formal design process, which in turn means absence of common sub-processes of design such as refining and clarifying objectives and constraints, elaborating and assessing options and planning for implementation.49 Of course, demonstrating an absence is difficult. Absence of evidence is not evidence of absence. But lack of publicly visible design activity – Royal Commissions, green and white papers, draft Bills and so on – while not itself conclusive, is at least highly suggestive. At the same time, the existence of such activities is not in itself decisive in the other direction. One must consider both the content of any documents and the period allowed for thought. Some White Papers, for example, are produced at great speed and betray confusion of objectives and failure to think about a range of options. But their very existence is an indication that some design activity might be happening.

48 P Reid, ‘English votes on English law: Just Another Running Repair’, UK Const L Blog (28 Oct 2015) (available at https://ukconstitutionallaw.org/)). The procedure was abolished on 13 July 2021. 49 Pahl et al, Engineering Design (n 6).

292  David Howarth Subsidiary evidence of improvisation might include putting forward miscellaneous proposals, each of which might have been preceded by some planning but with no evidence of thinking through the interactions between them. In effect, conception and execution are initially separated but come together in attempts to deal with problems the proposals themselves generate. We also need to distinguish between spontaneous reaction and improvisation. Sennett’s characterisation of the former as ‘mindless’ provides a way forward. Perhaps more precisely spontaneous reaction, unlike improvisation, lacks any conscious attempt to bring one’s actions into harmony with the actions of others. The difficulty is how to identify mindlessness. One indication might be great speed in reacting to a situation. But, as Sennett points out, highly trained and experienced people, such as jazz musicians, can react so quickly that observers might see their actions as unconscious, when in fact they are merely highly skilled. This is the point at which Sennett’s idea that engagement with tools and materials comes into its own. If lack of apparent effort to coordinate is combined with lack of engagement with tools and materials, we should be confident enough that the action is spontaneous and not improvised. In constitutional (and legal) terms, the tools and materials are rules and procedures. And so, reacting to a question at Prime Minister’s Questions with an exclamation is probably spontaneous, but reacting with a proposal for a new rule or procedure would be improvisation. VIII.  CONSTITUTIONAL IMPROVISATIONS

Almost any constitutional reform statute might be a candidate to be labelled an improvisation. After all, producing even a draft Bill cannot be purely spontaneous, and since British constitutional reform has always been piecemeal, the criterion that improvisations or workarounds are not intended to change the system in any fundamental way is usually met. But we need to start by cutting down the list of possibilities in two ways. First, statutes need to be plausibly constitutional, and second there needs to be prima facie evidence that they were improvisations in the way they came about. On the first criterion, although it is notoriously difficult to decide what in the UK counts as a ‘constitutional’ statute,50 since the answer varies with the questioner’s purpose, we are here looking for statutes that affect the structure of government at a national level. On the second criterion, prima facie evidence of improvisation might appear from the circumstances in which the statute was passed. Two types of external evidence might indicate improvisation, each suggesting an absence of design or planning. First, statutes passed very quickly after the start of a new government are unlikely to have benefitted from much planning or design. Second, statutes proposed in chaotic political circumstances, such as during major economic or political crises or where policy was rapidly changing, are unlikely to have been carefully designed.

50 See, eg, D Feldman, ‘The Nature and Significance of “Constitutional” Legislation’ (2103) 129 LQR 343.

The British Constitution as an Improvised Order  293 On that basis, a list of possible prima facie constitutional improvisations might include: the Fixed-term Parliaments Act 2011, the Acts setting up the Brexit referendum in 2015, House of Lords reform after 1997, the Human Rights Act 1998, devolution, the creation of the Supreme Court and reform of the office of Lord Chancellor, and the various reforms included in the Constitutional Reform and Governance Act 2010. All are arguably constitutional in the structural sense and, except, possibly, for devolution, none was intended fundamentally to change the whole system of government. As for the circumstances in which they were introduced, the Fixed-term Parliaments Act, the 2015 referendum, the Human Rights Act, House of Lords reform and devolution were introduced very early in Parliaments. The 2010 Act arose from proposals emerging very early in the term of a new Prime Minister, Gordon Brown, but also in what were then the most chaotic political circumstances imaginable – a global economic crisis and the MPs expenses scandal, in addition to which Brown’s colleagues were constantly plotting to remove him.51 The 2005 Act arose out of a more specifically political crisis: the desire of the Prime Minister, Tony Blair, to reshuffle his cabinet and to remove a specific minister.52 We need, however, to examine each of these reforms in more detail to understand the extent to which they should count as improvisations. A.  Reforms Proposed Very Early in the Life of a Government The Fixed-term Parliaments Act 2011 is a prime example of a reform enacted very early in a government without much official preparation. Fixed-term Parliaments Bills had been proposed previously (including one by me) and two of the main party manifestos in the 2010 election supported the idea. But the Bill came to Parliament without having gone through any green or White Paper or draft Bill process. That was largely because one of its aims was to help stabilise a new coalition government, by removing the power of the leader of the larger coalition party, as Prime Minister, unilaterally to ask the monarch for a dissolution. Its rapid passage was therefore politically important, with the result that one can see very little separation between its conception and its deployment. In terms of materials, it recycled ideas from previous Bills: making the starting point the date of the next election and abolishing the power to dissolve Parliament in any other way was from my 2007 Bill,53 and the early dissolution by a vote of two-thirds of the whole number of members of the House of Commons came from the devolution legislation.54 The other route to an early election, the elapse of

51 See generally, G Brown, My Life, Our Times (London, Vintage, 2017) and D McBride, Power Trip (London, Biteback, 2013) ch 33. 52 T Blair, A Journey (London, Hutchinson, 2010) 631; A Campbell and B Hagerty, The Burden of Power (Diaries volume 4) (London, Hutchinson, 2012) 408, 410; D Blunkett, The Blunkett Tapes: My life in the bear pit (London, Bloomsbury, 2006) 430, 512–513. 53 See Fixed-term Parliaments Bill 2007 (https://publications.parliament.uk/pa/cm200708/cmbills/030/ 08030.i-i.html). 54 Scotland Act 1998, s 3; Government of Wales Act 2006, s 5. The 2011 Act resisted the devolution statutes’ rule that the relevant Presiding Officer sets the election date, empowering instead the Prime Minister, with serious political consequences in 2019.

294  David Howarth 14 days from a vote of no confidence without a new vote of confidence was also based, albeit more loosely, on the devolution statutes.55 An additional indication of the Fixed-term Parliaments Act’s improvised nature is that the proposal changed radically during its passage. The government’s Bill originally gave the Speaker a role in certifying whether any Commons resolution constituted a vote of no confidence, but an amendment in the Lords, which the government accepted, removed the Speaker’s role, following complaints that the government’s proposal lacked certainty and might require the Speaker to make controversial decisions.56 The EU Referendum Bill 2015 was introduced three weeks after the 2015 general election. It similarly had antecedents, two private members’ Bills supported by the Conservative Party in the 2010–15 Parliament, both of which had failed, and it was foreshadowed in the 2015 Conservative manifesto but no official consultation took place – no green or White Paper and no draft Bill. Again, the Bill’s rapid progress was politically important. It solved a political problem of how to keep together a Conservative Party its own leader described as ‘ungovernable’.57 The best evidence for the Bill being an improvisation is that although the party’s leadership had decided three years earlier to hold a referendum, the Prime Minister, by his own account, only started work the week after the general election on how the referendum would work, such as the franchise, the question and ‘purdah’ (the date after which the government machine must stay neutral).58 Conception and implementation were closely linked. As for material, the Act follows the Conservatives’ own 2014 Bill with the exception of a small change to the question (the result of the Electoral Commission’s comments on the 2014 Bill) and the addition of technical material, itself largely based on legislation for previous referendums. House of Lords reform, when examined in detail, looks less like improvisation. It was subject to a large number of White Papers, independent commissions and parliamentary votes.59 It even included a draft Bill scrutinised by a Joint Committee of the two Houses.60 Nevertheless, one can make an argument for counting it as at least partly improvised, since very few of the proposals advocated in the White Papers or supported by the House of Commons were implemented. Instead the process was chaotic. For example, the Blair government’s 1998 White Paper contained three specific proposals: abolishing hereditary peers, establishing a statutory independent Appointments Commission and establishing a Royal Commission to consider the Lords’ long-term future. Because of resistance from the Lords themselves, supported by an alliance of reactionaries and abolitionists in the Commons, neither of the first two happened. The Royal Commission did meet, made a few suggestions for minor

55 Scotland Act 1998, ss 45 and 46; Government of Wales Act 2006, s 47. The reconstruction period in the devolution statutes is 28 days rather than 14, and the process of government formation comes under the control of the Parliament itself and its presiding officer, rather than being left unspecified. 56 Hansard HL (series 5) Vol 727, cols 1146–1179 (16 May 2011). 57 D Cameron, For the Record (London, Collins, 2019) 339. 58 ibid 623–626. 59 For a short summary see House of Lords Library, House of Lords Act 1999: Twenty Years On (London, House of Lords, 2019). 60 Joint Committee of the Draft House of Lords Reform Bill, Report HL Paper 284–I, HC 1313–I (London, TSO, 2012).

The British Constitution as an Improvised Order  295 changes and was ignored. There might have been a design process but no design of anything that took place. As for material, the provisions eventually enacted were based largely on a private member’s Bill, Lord Steel’s House of Lords Amendment Bill of 2010–12. One could best characterise the whole process as ‘adaptation’, planned intervention to adjust an existing institution to new circumstances, but with elements of improvisation in the way it was done. The Human Rights Act is also a borderline case. Its enactment followed a long campaign and a White Paper,61 but it came very quickly after a general election, so quickly that, anomalously, the government treated a purely party process of consultation, the Labour Party’s ‘Bringing Rights Home’ project, as part of the official process, immediately producing a White Paper without any official consultation stage.62 But as for materials, the Act differs markedly from the Bills campaigners and other parties had previously offered. Those Bills tended to allow judicial strike-down of existing legislation and an interpretive provision for subsequent legislation construing it as not authorising contraventions unless such a construction was ‘unavoidable’. The Act allowed only declarations of incompatibility and provides for an interpretive obligation to construe all legislation to make it compatible as long as it is ‘possible’ to do so.63 In effect, the range of choice of design was limited to designs posing no direct threat to parliamentary supremacy, which looks like incrementalism or muddling through, making a problem tractable by reducing the number of variables and thus the number of options available. Devolution, having been extensively debated outside Parliament, at least in Scotland, looks even more like design.64 The process that began very quickly after a new government came to power was not the reform itself but referendums on whether the reforms should go ahead in principle. The Bills appeared about eight months after the election. On the other hand, the process had come to no precise conclusions on several issues, in particular the relationship between the devolved bodies and the UK legislative process, a problem seemingly addressed on the hoof with the creation of the Sewell Convention during passage through the Lords.65 As for material, the Scotland Act was quite different from its 1970s predecessor in terms of basic structure and extent of powers,66 and although it garnered some ideas from the Scottish Constitutional Convention,67 such as the electoral system and a power to vary income tax, in other respects, especially nomenclature (‘Presiding Officer’, rather than ‘Speaker’, ‘First Minister’ rather than ‘Chief Minister’) it was quite

61 House of Lords Constitution Committee, The Process of Constitutional Change: 15th Report of 2010–12 HL Paper 177, (London, TSO, 2011) 17; Home Office, Rights Brought Home: The Human Rights Bill: CM 3782, (London, TSO, 1997). 62 Home Office, Rights Brought Home (n 61) 1. 63 See, eg, the Human Rights Bills of 1986, 1989, 1990 and 1992. The 1995 Bill merely declared the ECHR incorporated into UK law. 64 House of Lords Constitution Committee, The Process of Constitutional Change (n 61) 16–17; Scottish Constitutional Convention, Scotland’s Parliament: Scotland’s Right (Edinburgh, Convention of Scottish Local Authorities, 1995); B Winetrobe, ‘Enacting Scotland’s “Written Constitution”: The Scotland Act 1998’ (2011) 30(1) Parliamentary History 85. 65 Winetrobe, ‘Enacting Scotland’s “Written Constitution”’ (n 64) 96. 66 R Brazier ‘The Scotland Bill as Constitutional Legislation’ (1998) (19(1) Statute Law Review 12. 67 Scottish Constitutional Convention, Scotland’s Parliament (n 64).

296  David Howarth different. Devolution seems to be another mixed case but this time leaning towards design. B.  Reforms Proposed in Chaotic Circumstances We can now turn to the other category of prima facie improvisations, reforms introduced in chaotic circumstances. The most striking example is the creation of the Supreme Court and the Judicial Appointments Commission by the Constitutional Reform Act 2005. The story is one of near farce. The Prime Minister, Tony Blair, apparently frustrated that his Lord Chancellor, Derry Irvine, was unsympathetic to more authoritarian criminal justice and immigration policies,68 decided that he would reallocate the responsibilities of the Lord Chancellor to other ministers and abolish the post, replacing the Lord Chancellor’s Department with an ordinary department of state dealing with any topic left over after reallocating its functions to more active ministers.69 The problem, of which Blair seemed unaware,70 was that the Lord Chancellor was no ordinary Secretary of State whose responsibilities a Prime Minister could reassign at the stroke of a pen.71 Specific mention of the Lord Chancellor appeared in 5000 primary and secondary legislative provisions, from the Ecclesiastical Licences Act 1533 to the Nationality, Immigration and Asylum Act 2002. Moreover, the Lord Chancellor’s role in judicial appointments could not just be assigned to an ordinary political minister without gravely damaging the independence of the judiciary. Consequently, abolishing the office of Lord Chancellor required creating a new system of judicial appointment. Irvine offered to stay on to oversee a slow transition to a new system.72 Instead, Blair forced Irvine to resign and appointed Charles Falconer in his place. But Blair’s announcement, part of a wider reshuffle, did not include the immediate abolition of the office of Lord Chancellor, as Blair had originally intended, only that consultations would start about how it could be abolished. Blair’s announcement also contained a promise of consultation on creating a new Supreme Court.73 Oddly, the diaries of contemporaries within government, such as those of Blair’s director of communications Alistair Campbell and the Home Secretary David Blunkett,74 contain no reference to the Supreme Court proposal before it was made. No one in Parliament had discussed the idea before the government’s announcement, the first debates about it coming much later.75 68 See Blair, A Journey (n 52) 631; Campbell and Hagerty, Burden of Power (n 52) 408, 410; Blunkett, Blunkett Tapes (n 52) 430, 512–513. 69 Blair, A Journey (n 52). 70 D Irvine, Memorandum by Lord Irvine of Lairg to the House of Lords Constitution Committee 2009 [Online]. Available at: https://publications.parliament.uk/pa/ld200910/ldselect/ldconst/30/09070105.htm. [Accessed 26 August 2020]. 71 For example, using the Ministers of the Crown Act 1975, s 1. 72 Irvine, Memorandum (n 70). 73 P Webster, ‘Emergency surgery on the Cabinet’ The Times, 13 June 2003, 1. 74 Campbell and Hagerty, Burden of Power (n 52); Blunkett, Blunkett Tapes (n 52). 75 The first was in the House of Lords (Hansard HL (series 5) Vol 652, col 111ff (8 September 2003)). The first Commons debate (in Westminster Hall) was on 16 December 2003 (Hansard HC (series 6) Vol 415, cols 181WH ff).

The British Constitution as an Improvised Order  297 It did, however, appear in Irvine’s list of other measures consequential on ­abolishing the office of Lord Chancellor,76 possibly because Irvine thought that removing the Lord Chancellor from the House of Lords sitting legislatively entailed removing the other Law Lords as well – an argument the Law Lords themselves later declared fallacious.77 Three consultation papers duly appeared,78 but each sparked controversy and the government’s subsequent Bill ran into immediate trouble. The House of Lords, the first House to consider it, took the unusual step of referring the Bill to a select committee.79 When the Bill came back, the government proposed over 400 amendments, the main effects of which were to retain the office of Lord Chancellor but to abolish its judicial functions. That in turn made the office available to non-lawyers but, as a safeguard against holders of the office lacking any sense of legal values, the Lord Chancellor became subject to a statutory duty and an oath to uphold the rule of law and the independence of the judiciary. Most of the Lord Chancellor’s powers over the judiciary transferred to the new Judicial Appointments Commission or to the Lord Chief Justice and most parliamentary functions went to a Lord Speaker. The Supreme Court proposal survived, but with the independence of the court from government much strengthened.80 In his autobiography, Tony Blair claims:81 Painful though moving Derry was, and as bumpy and chaotic as the aftermath of the constitutional reform to the office of Lord Chancellor became, the result was absolutely right. The House of Lords got an elected Speaker and sensibly chose a woman, Helene Hayman; judges were appointed by an independent commission; and the Lord Chancellor’s Department became the Department of Constitutional Affairs, focused on driving forward the reform and improvement of the justice system, and on reforming and safeguarding the constitution.

One might read this passage as if these outcomes were planned all along. But they were not. They were fortunate outcomes of the ‘bumpy and chaotic’ process itself. No consultations occurred before the process started and when a consultation and planning phase did begin, it was fractured and incoherent, with solutions proposed for problems the earlier proposals had themselves generated. In terms of materials, the technique of requiring an oath of a public official, is a repurposing of a very old technique and the Judicial Appointments Commission proposals bear more than a passing resemblance to the system adopted in Scotland in 2002.82 Categorising the 2005 Act as improvisation looks unavoidable. 76 Irvine, Memorandum (n 70). 77 House of Commons Constitutional Affairs Committee, Judicial appointments and a Supreme Court (court of final appeal) (London, TSO, 2004) 10. 78 Department for Constitutional Affairs, Constitutional Reform: A New Way of Appointing Judges; Constitutional Reform: A Supreme Court for the United Kingdom; and Constitutional Reform: reforming the office of the Lord Chancellor (London: HMSO, 2003). 79 Hansard HL (series 5) Vol 658, col 1108 (8 March 2004). 80 House of Commons Constitutional Affairs Committee, Constitutional Reform Bill [Lords]: the Government’s proposals (London, TSO, 2005) 40–43. 81 Blair, A Journey (n 52) 631. 82 The Scottish Judicial Appointments Board was established on an administrative basis in 2002. It is now a non-departmental public body under the Judiciary and Courts (Scotland) Act 2008.

298  David Howarth Superficially, the Constitutional Reform and Governance Act 2010 looks designed: it was preceded by a Green Paper and a draft Bill.83 But the proposals covered a vast range of seemingly unconnected problems, from giving Parliament a role in authorising the deployment of armed forces to creating a ‘British statement of values.’ Moreover, possibly resulting from the political chaos of 2008–10, which might have contributed to long delays in the Bill’s progress, the menu of proposals under consideration varied wildly at different stages of the process. Only two proposals, out of more than 15 in the Green Paper that required legislation, appear in the Act. The Bill’s final stages became entangled in in the 2010 ‘wash-up’, in which Bills were truncated to allow them to pass the House of Lords before dissolution. But even as it left the Commons the Bill contained only four Green Paper proposals (putting the civil service on a statutory basis, creating a statutory process for the ratification of treaties, controlling protest in the vicinity of Parliament and further reducing the role of ministers in judicial appointments). Moreover, the Bill at that stage included more than half a dozen reforms not mentioned in the Green Paper, including a government proposal for a referendum on changing the voting system for the House of Commons. That particular reform failed to survive the wash-up, but several proposals not mentioned in the Green Paper appear in the Act. They include a set of provisions inserted very late, advertised by the government as merely implementing proposals from an obscure committee about the publication of official documents, but in fact immunising the Prince of Wales’s correspondence from the effect of the Freedom of Information Act, so that the subsequent, and celebrated, case R (Evans) v Attorney-General affects only the Prince’s ‘black spider letters’ written before 2010.84 In terms of materials, much of the civil service part of the Act, especially the duties of civil servants, were cribbed from the existing civil service code and the treaty ratification part was based on the existing Ponsonby rules. Even if, which is difficult to maintain, the Green Paper constituted a coherent design, the Act as passed, and even as it entered wash-up, is full of improvisation. IX.  GOOD OR BAD IMPROVISATIONS?

Merely characterising constitutional reforms as improvisations does not in itself make them undesirable. To assess the success or failure of constitutional improvisations, we should return to the questions we posed for assessing statutory improvisations. Did the reform identify and solve an immediate problem? Did it solve it in a way that made clear the reasons for the reform, leaving room for further development or conscious reversal? And did it cause further problems later? That is, returning to the path dependency theme, did it set future developments on a path that continued to solve problems, or did it lead to a Sennett ‘maze’? Restricting consideration to the four clearer cases of improvisation identified above (the Fixed-term Parliaments Act, the EU referendum, the 2005 reforms, and the 83 The Governance of Britain CM 7170 (London, HMSO, 2007); Draft Constitutional Renewal Bill 2007. 84 [2015] UKSC 21; [2015] AC 1787. See D Howarth, ‘Bowing and scraping’ TImes Literary Supplement 30 June 2017 32.

The British Constitution as an Improvised Order  299 2010 Act, what can we say about whether they should count as successful or unsuccessful improvisations? Assessments of this type are inherently qualitative and debatable, and judgments might shift over time, but I offer some first impressions at least as a basis for future debate. A.  Fixed-term Parliaments Act The immediate problem the Fixed-term Parliaments Act was intended to solve was that in a coalition government, if one party holds the prime ministership and the Prime Minister can ask the monarch for a dissolution and early election even if the other coalition party objects, nothing in previous practice, or in the announced rules (the Lascelles principles),85 prevented such a dissolution. Those negotiating the coalition deal believed the coalition would be more stable if the Prime Minister’s discretion were removed.86 Since the coalition lasted a full parliamentary term, one can say at least that no evidence exists that the Act failed to achieve its immediate goal. As for clear reasons for the reform, the position is less conclusive, but one can say that the Liberal Democrat side of the coalition, if not the Conservative side, was committed to the view that continuing to allow Prime Ministers an almost unencumbered power to call elections gave the governing party an unfair advantage and led to unwelcome speculation about the timing of general elections, thus contributing to political instability.87 It also continued an established line of reform aimed at reducing the electoral advantages of incumbency, such as removing the power of the chancellor of the exchequer to set interest rates.88 The principle is one of fairness in elections. The Act did contain a comprehensible direction of change. But the Act resulted in problems. Some complaints admittedly originated from its intended targets, incumbent governments, and amount merely to the objection that it worked. But a more serious problem was that the Act reformed only part of a set of rules and institutions and unsettled the unreformed parts. For example, some claimed that the Act altered the basic rule of the UK system that a government cannot remain in office unless it retains the confidence of the House of Commons, replacing it with a rule that the government can remain in office even if it loses confidence as long as the House refuses to allow a general election to be called; others denied that the Act has any such effect.89 Consequently, in the autumn of 2019 a government that had arguably lost the confidence of the House, but to which the Commons had initially

85 Senex [pseudonym of Sir Alan Lascelles], ‘Dissolution of Parliament: Factors in Crown’s Choice’ The Times, 2 May 1950, 5. 86 Cameron, For the Record (n 57) 289, D Laws, 22 Days in May: The Birth of the Lib Dem–Conservative Coalition (London, Biteback, 2010) 1604. 87 Laws, 22 Days in May (n 86) 1604; Hansard HC (series 6) Vol 515, col 621 (13 Sept 2010) (N Clegg MP (Deputy Prime Minister)). 88 Bank of England Act 1998, s 10. 89 House of Commons Public Administration and Constitutional Affairs Committee, The Role of Parliament in the UK Constitution Interim Report The Status and Effect of Confidence Motions and the Fixed-term Parliaments Act 2011 (London, TSO, 2019).

300  David Howarth refused an election, managed to hang around for a month or so without resigning or being replaced – a ‘zombie government’.90 That could in theory happen under the Lascelles principles, but the difference is that the monarch was unlikely to allow such a situation to persist more than a few days. The essential problem was that the Act solved one problem, excess prime ministerial power over election timing, without solving related problems about the formation and dismissal of governments. The optimistic way of describing the situation is that, as in many improvisations, reformers are engaged in a voyage of discovery, uncovering the system element by element and changing it sequentially in accordance with the principles they have already used. The pessimistic description is that they lacked awareness of the scale of what they were doing. Whatever the correct view, the incoming Johnson government moved to repeal the Act in its entirety.91 B.  EU Referendum The political goal of the legislation setting up the EU referendum was, according to David Cameron, the Prime Minister at the time, to stabilise both the Conservative Party and the UK’s relationship with the EU: On Europe, the Conservative Party (reflecting broader opinion in the country) was becoming increasingly ungovernable. And Britain’s current status in Europe was becoming increasingly unsustainable, as the whole project continued to mutate into something so different from what we signed up to all those years ago.92

In neither of those goals could one describe the referendum as a success. But here we are more interested in the referendum as an example of constitutional reform. On that aspect, Cameron’s explanation is more interesting. After professing a belief in parliamentary democracy, he says: However, I also believe that there are some issues – particularly where Parliament is giving up its powers, or fundamentally rewriting the rules of our democracy – where there is a strong justification for asking voters to take the decisions themselves; otherwise politicians can end up giving away powers that are not theirs to surrender. In other words, it is not for those who govern to fundamentally change the rules around governing.93

The EU Referendum Bill of 2015 was not, however, proposed in the light of any possible further transfer of competencies to the EU. The European Union Act 2011 had already ensured that no further transfers of power could occur without a referendum, and nothing had happened to trigger such a referendum. That was why the 2016 referendum required its own statutory underpinning. Moreover, by the time the referendum happened, Cameron claimed to have secured a deal that had shifted

90 See Howarth 2019 and 2021 (n 43). 91 Dissolution and Calling of Parliament Bill 2021, cl. 1. 92 Cameron, For the Record (n 57) 339. 93 Cameron, For the Record (n 57) 287. The sentiments resemble those expressed in V Bogdanor, ‘After the referendum the people and not Parliament are sovereign’, Financial Times, 9 December 2016 (perhaps not surprising given that Bogdanor taught Cameron at Oxford).

The British Constitution as an Improvised Order  301 the balance of power back towards the UK. The referendum could not, therefore, fulfil the constitutional function Cameron identified. Turning to the reasons behind the referendum, Cameron offered a justification for holding the referendum perhaps more relevant than his constitutional argument: Not only are there past examples, there is a present compulsion for referendums. We live in an age when people fear that the rules are set by an elite from which they are excluded. A device that gives them control is more important than ever.94

The difficulty, as several commentators have remarked, is that if, as seems plausible given the rhetoric of his successors, Cameron intended to include within the meaning of the term ‘elite’ Parliament itself, his approach produces a conflict between representative democracy and plebiscitary democracy, a conflict the 2015 Act failed to resolve.95 The 2015 Act, unlike, for example, the statute establishing the 2011 voting system referendum,96 sets out no legal consequences of the result going one way or the other. To describe the referendum even as ‘advisory’ goes too far, since that might imply that someone came under an obligation to be advised, an obligation that appears nowhere in the Act. The government of the day distributed an explanatory booklet to every household saying, ‘This is your decision. The Government will implement what you decide’,97 but that promise had no legal force – indeed it would be highly irregular for a government to be able by distributing a leaflet to create legal obligations it was unable to persuade Parliament to include in a statute. The main weakness of the 2015 Act as an improvisation is that its principle is unclear and so it fails to suggest where the system goes next. Indeed one view is that it leads nowhere: because Parliament’s legal supremacy remained intact and general elections have given the Commons more recent democratic credentials than the referendum, Parliament will eventually take back control. The opposite view is that, politically if not legally, the referendum will become indispensable for the legitimacy of all post-referendum politics. On this view, unless governments and parliaments ‘accept’ and ‘respect’ the referendum, their legitimacy should be rejected because they are resisting the fixed and permanent ‘will of the people’. These paths are so divergent that the referendum as a constitutional improvisation has led to confusion and uncertainty about the most basic features of the system. The ‘will of the people’ interpretation was used to justify several attempted ruptures with past practice: a long (and unlawful) prorogation, suggestions that the government could break the law, and the idea that the government could legitimately ask the monarch to veto a Bill passed by both Houses.98 All these effectively amounted

94 Cameron, For the Record (n 57) 287–288. 95 See, eg, A Weale, The Will of the People: A modern myth (Cambridge, Polity, 2018) 109–110; J Sumption, Trials of the State: Law and the Decline of Politics (London, Profile, 2019) 30–31; and M Gordon, ‘Parliamentary Sovereignty and the Political Constitution(s): From Griffith to Brexit’ (2019) 30(1) King’s Law Journal 25. For a more positive view, see V Bogdanor, Beyond Brexit (London, I. B. Tauris, 2019) 87–112. 96 Parliamentary Voting System and Constituencies Act 2011, s 8. 97 HM Government, Why the Government believes that voting to remain in the European Union is the best decision for the UK (London, TSO, 2016) 14. 98 See Howarth 2019 and 2021 (n 43).

302  David Howarth to a radical extension of ministerial power on the back of ministers’ claims to an exclusive right to interpret the referendum. On the other side, the view that Parliament can take back control has resulted in changes in the interpretation of the rules of the House of Commons and two Supreme Court cases reasserting the dominance of the Commons over ministers.99 Whichever of these views one takes, it seems clear that the referendum has caused serious consequential constitutional problems. If ever there was an example of an improvisation leading to an impenetrable maze, this is it. C.  The 2005 Act: The Lord Chancellor, Judicial Appointments, and the Supreme Court The immediate problem to be solved by what became the 2005 Act was political, not constitutional, namely removing a Lord Chancellor who was obstructing the Prime Minister’s desire to change political strategy. It then became a legal problem, how to redistribute the Lord Chancellor’s powers, and then a constitutional problem about the separation of powers. The initial problem was nevertheless solved: Lord Irvine was removed, and the government’s political priorities changed. The obvious principle that justifies the 2005 reform is that the courts should be functionally distinct from Parliament. Promoting the separation of powers formed no part of the Prime Minister’s motivation at the outset and so the principle can only be an ex post facto justification. But causes and reasons are different. The process of making sense of past events, itself often a kind of improvisation, can create order out of apparent disorder, and provide coherent reasons for future action. The 2005 Act is an example of an improvisation that groped its way to a principle. Moreover, because the principle had already found expression in the system, for example in the judges of the House of Lords’ self-denying ordinance that they would keep out of the legislative process,100 one can say that the 2005 Act took up a theme already present in the material. The 2005 reforms, although controversial at the time, have caused few ­serious problems. The new judicial appointments system was initially resisted within government,101 and some have criticised it for giving too much influence to existing judges and not enough to politicians, thus worsening the problem of lack of understanding between political and legal institutions.102 In addition, some commentators have been disappointed that the Commission’s record on diversity has been less good than they hoped.103 Nevertheless, on the one occasion when it seemed that a government was proposing to reverse the reform, a rapid and determined campaign

99 ibid and R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5; R (Miller) v Prime Minister; Cherry v Advocate General [2019] UKSC 41. 100 See C Leslie, (Parliamentary Under-Secretary, Department of Constitutional Affairs), Hansard HC (series 6) Vol 430, col 735 (1 February 2005). 101 G Gee, R Hazell, K Malleson and P O’Brien, The Politics of Judicial Independence in the UK’s Changing Constitution (Cambridge, Cambridge University Press, 2015) 165–170. 102 ibid 192–193. 103 ibid 191.

The British Constitution as an Improvised Order  303 by lawyers and the judiciary quickly defeated the proposal.104 Future attempts to restore political control of judicial appointments cannot be ruled out,105 but the Supreme Court is no less respected as a court than the former Appellate Committee and claims that it is more ‘activist’ than its predecessor seem exaggerated.106 Judges as a whole retain the trust of the public.107 Some proposals have emerged,108 seemingly in response to government defeats in the Miller cases,109 to rename the Court and give it a different set of judges, but the President of Supreme Court felt confident enough in the Court’s standing openly to denounce the ideas as ‘idiotic’ and driven by ‘spite’.110 The position of Lord Chief Justice as head of the judiciary now seems uncontroversial, just as no one seems to miss the days when the Lord Chancellor acted as Speaker of the House of Lords. The only serious qualms about the new system are that the Lord Chancellor played a role in conveying the views of judges into government, especially into Cabinet, a role which, especially when a non-lawyer holds the office, and even more a mid-career politician hoping for further advancement, has arguably become less effective.111 But this does not amount to a catalogue of serious consequential problems. A further round of populism might conceivably threaten the reforms, but the threat would be purely political. D.  The 2010 Act: Prerogatives, the Civil Service and Treaties The Governance of Britain Green Paper identified no specific problem needing to be solved, but it did say that the government’s purpose was: [T]o forge a new relationship between government and citizen, and begin the journey towards a new constitutional settlement – a settlement that entrusts Parliament and the people with more power.112

It divided the task into four categories: limiting the power of the executive; making the executive more accountable; reinvigorating democracy; and ‘Britain’s future: the citizen and the state’, which was about developing a list of ‘British values’.

104 ibid 168. 105 The 2019 Conservative Party Manifesto promised to ‘look at the broader aspects of our constitution: the relationship between the Government, Parliament and the courts’ (Conservative Party, Get Brexit Done: Unleash Britain’s potential (London, Conservative Party, 2019, 48). See also R Ekins, Protecting the Constitution: How and why Parliament should limit judicial power (London, Policy Exchange, 2019) 19. 106 B Dickson, ‘Activism and Restraint within the UK Supreme Court’ (2015) 21(1) European Journal of Current Legal Issues. 107 See www.ipsos.com/ipsos-mori/en-uk/trust-politicians-falls-sending-them-spiralling-back-bottomipsos-mori-veracity-index. 108 E Malnick, ‘Supreme Court to be overhauled to curtail its constitutional powers’ The Telegraph (14 November 2020). 109 [2017] UKSC 5 and [2019] UKSC 41. 110 Evidence to the House of Lords Constitution Committee, 17 March 2021. 111 Gee et al, The Politics of Judicial Independence (n 101) 38–40. Lawyers have also complained about the quality of some, though not all, of the non-lawyer Lord Chancellors, but some pre-Act Lord Chancellors were far from perfect (See RFV Heuston, Lives of the Lord Chancellors, 1940-1970 (Oxford, Clarendon Press, 1987). 112 The Governance of Britain 5.

304  David Howarth The 2010 Act limits the power of the executive in two ways, by converting prerogative powers over the Civil Service and treaty ratification into statutory powers. Whether they limit the power of the executive depends on whether one believes that converting a convention into law strengthens it, but one can see that it might. But the Green Paper’s wider ambitions, for example about limiting war powers, went nowhere. The Act even increases the power of the executive in one respect, to hide the Prince of Wales’s correspondence. The Act does nothing to make the executive more accountable, except for minor reforms about financial reporting, or to reinvigorate democracy. The project to produce a statement of ‘British values’ ran into the ground. As a result, it is difficult to say that the Act identified an immediate problem or solved it. The lack of any theme covering the proposals and the modesty of what reached the statute book also make it difficult to identify any reason or principle to develop further. But one can argue that a principle exists justifying both the Civil Service and treaty ratification reforms, namely that the broad discretionary prerogative powers should gradually give way to statute or at least to greater parliamentary control, a view previously expressed by parliamentarians.113 The civil service part of the Act has not yet caused major difficulties, spawning only a few contested cases,114 but the advent of a Prime Minister who has given special advisers and political appointees great power and is dismissing more Permanent Secretaries than ministers might produce serious conflicts in the future. The treaty ratification part of the Act, did, however, create a problem around possible EU withdrawal agreements. Such agreements fall within the definition of a ‘treaty’ under the Act. To begin with, ministers conceded that the 2010 Act would apply but they gave no indication that applying it would cause problems.115 But when the Johnson government eventually produced a Bill implementing its withdrawal agreement, the Bill contained a provision disapplying the 2010 Act.116 The only explanation offered was that the point of the disapplication was to remove a delay of 21 sitting days, but that was unconvincing because section 22 of the Act allows ministers to ratify immediately if they believe that ‘exceptional circumstances’ exist. One possibility is that the real purpose was to remove any prospect of legal challenge, which suggests that the 2010 Act will not be allowed to fulfil its purpose of subjecting the government to legal obligations in place of conventions. Another possibility was that the 2010 Act was never intended to apply to ‘legislative’ treaties – ones that require domestic legislation before becoming effective – since approval of the implementing legislation, which happens before ratification, counts as approval of the treaty. But that argument depends on a misunderstanding of the process of treaty ratification. The implementing legislation need not come before ratification. It does so only as a

113 See, eg, House of Commons Public Administration Committee Taming the Prerogative: Strengthening Ministerial Accountability to Parliament (London: TSO, 2004) 17. 114 The only substantive case so far appears to be Secretary of State for Justice v Betts [2017] ICR 1130. 115 Miller v Secretary of State for Exiting the European Union [2017] UKSC 5 at [163] per Lord Reed (dissenting). 116 European Union (Withdrawal Agreement) Bill 2019, cl 33, eventually European Union (Withdrawal Agreement) Act 2020, s 32.

The British Constitution as an Improvised Order  305 matter of practicality. An unscrupulous government might try to coerce Parliament by reversing the order.117 In any case, the argument looks like an improvisation to correct a previous improvisation, which is how mazes begin. X.  IMPROVISING BETTER

Improvisation and workarounds, engineers have realised, form a necessary part of everyday life.118 Legal improvisation, in case law and legislation, is also arguably necessary. Constitutional law is no exception. Countries with formal constitutions might be less prone to constitutional improvisation because the process of constitutional amendment is usually deliberately slow and forces reformers to fit their ideas into an existing text. But even constitutional amendments are subject to improvisation within the political processes that produce them.119 Some have suggested various procedural reforms to slow down the process of constitutional reform, in effect requiring more deliberate design activity (Green Papers, White Papers, draft Bills and more), and more participative design processes aimed at producing consensus. These would, reformers hope, lead to more coherence with the existing system and with other proposals for reform.120 I myself have joined calls for such reforms.121 But as long as governments come to power with manifesto commitments to introduce specific constitutional reforms and as long as their proposals include no change to fundamentals, in particular to parliamentary supremacy, one can expect improvisation not only to continue but even to predominate. Under those conditions we might want to ask a different question – not how do we reduce constitutional improvisation but, since improvisation is going to happen, how do we increase the chances of successful improvisation? The examples studied here offer some, albeit limited, guidance. The most ­successful appears to have been the 2005 Act, which arguably worked because, without intending to, it alighted on a principle already working its way through the system, a version of the separation of powers focussed on the independence of the courts. One might argue that the same was true of the least successful, the EU referendum, which repeated an innovation already attempted in the 2011 voting system referendum. The difference, however, was that the principle of the independence of the judiciary threatened no other basic principle of the system, whereas the use of referendums did pose such a threat, namely to representative democracy. The lesson is not to use improvisations to attempt too much, a lesson reinforced by the more

117 See my evidence to the House of Lords Constitution Committee inquiry on the Parliamentary Scrutiny of Treaties: Select Committee on the Constitution Corrected oral evidence: Parliamentary Scrutiny of Treaties, Wednesday 21 November 2018, Q2 (p.2). 118 Alter, ‘Theory of Workarounds’ (n 8) 1053. 119 See, eg, S Holmes and C Sunstein, ‘The Politics of Constitutional Revision in Eastern Europe’ in S Levinson (ed), Responding to Imperfection The Theory and Practice of Constitutional Amendment (Princeton, NJ, Princeton University Press, 1995) 281. 120 House of Lords Constitution Committee, The Process of Constitutional Change (n 61). 121 ibid 13.

306  David Howarth modest successes of some of the other improvisations in the 2005 Act, which have at least caused no harm. That lesson might also apply to the 2010 Act, although it might be too early to tell. The Fixed-term Parliaments Act is an intermediate case. It instantiated a principle of electoral fairness, particularly that incumbency should bring no advantage. That looks like a principle posing no threat to fundamental features of the system, but perhaps we should at least entertain the thought that electoral unfairness, in the form, for example, of the first-past-the-post system, is more fundamental to the United Kingdom’s system of government than we care to believe. More speculatively, going beyond the examples, Sennett’s account of craft skills, echoing Llewellyn’s account of the lawyers’ craft, suggests that one would expect repeated practice and above all a trained instinct for what comes next, for the normal flow of events, to be important for good improvisation. Although some people have a talent or feel for such flows, most practitioners learn it by trial and error. Another possibility, taken from improvisation in theatre but applicable to business management, is that improvisation, unlike design, works better when participants concentrate on the process itself and not on outcomes.122 It is, therefore, unfortunate that early twenty-first century attempts at constitutional improvisation have come from neophyte political leaders in an era in which one election or referendum defeat ends a party leader’s career. Rapid circulation of leading officials and reliance on inexperienced special advisers also seem unlikely to lead to successful improvisation. If we are to continue to improvise, we need improvisers who have learned from their mistakes, not those whose first, and probably last, mistake is yet to come. And we should value those, such as the clerks of the House of Commons, who are dedicated to processes and procedures rather than to policies and political outcomes.123 XI.  IMPROVISATION AND CONSTITUTIONAL THEORY

Are there lessons to draw from the prevalence of constitutional improvisation in the United Kingdom for constitutional theory more generally? That depends ultimately on what one takes constitutional theory to be. One could, for example, take constitutional theory to be concerned with giving an objective, social scientific account of constitutions as socio-cultural phenomena. On that basis, the prevalence of improvisation in Britain might provide an interesting starting point for comparisons with other constitutions, raising questions such as whether the uncodified nature of the British Constitution makes improvisation more prevalent than elsewhere. One could also raise larger comparative questions, such as whether constitutional improvisation is associated with an improvisational style in the wider culture.

122 V Dusya and M Crossan, ‘Theatrical Improvisation: Lessons for Organizations’ (2004) 25(5) Organisational Studies727, 736–739. 123 See Hansard HC (series 6) Vol 658, cols 145–146 (8 April 2019) for an illustration.

The British Constitution as an Improvised Order  307 But constitutional theory is widely understood (not least by other authors in this volume) as a normative enterprise. It is concerned with providing answers to questions such as how should judges decide constitutional cases and whether certain institutional arrangements (separation of powers, judicial review and so on) are always and everywhere desirable. The prevalence of improvisation is a matter of fact rather than value, and the relevance of facts to normative enterprises is itself controversial.124 But I would attempt two observations. The first observation is that theorists who appeal to common law processes and reasoning might consider how common law reasoning really works. Some of these theorists’ accounts of common law reasoning are highly idealised and implausible. They fail to take into account the improvised nature of much judicial practice. They fail to allow for the possibility that common law reasoning can lead to dead-ends and intractable confusion, to Sennett’s mazes. Theorists might counter that they are not saying that common law reasoning always provides a single right answer. They are saying only, with Dworkin, that common law reasoning presupposes a single unified account of what the law requires, even though we might be uncertain about what that account requires in specific cases. But the improvisatory nature of real common law reasoning is a potential challenge even to that point. Improvisation presupposes not a single unified normative framework but only that we can use existing materials, whether they are coherent or not, to produce an answer that plausibly works to generate an acceptable result in the case at hand and possibly the next case. It need not work for ever. It need not work anywhere except around here. If it fails to work, another improvisation will come along shortly. The second observation is that constitutional theorists might think about the tolerance of constitutions for incoherence. Constitutional improvisation, whether by judges or politicians, can produce a set of arrangements that make no sense. But institutions need not make sense. They can carry on for long periods coping with their contradictions by generating yet more improvisations and workarounds. Just as software generates a fatal error only when contradictory routines are run together, constitutional systems can run for a long time containing contradictory rules if the contradictory rules are kept apart from one another. The potential contradiction between devolution and parliamentary sovereignty, for example, remained inconsequential when the two elements were insulated from each other by improvisations such as the Sewel Convention, but if a crisis causes them to come into contact, as in Miller no 1,125 the result is the constitutional equivalent of the blue screen of death. Similarly, the contradictions between the ‘Whitehall’ and ‘Westminster’ views of the Constitution, on whether the government or the House of Commons is the dominant institution in the United Kingdom system,126 subsisted for decades before needing to be confronted in Miller 2/Cherry.127



124 See

further Arvind and Stirton, ch 4 in this volume. UKSC 5. Howarth 2019 and 2021 (n 43). 127 [2019] UKSC 41. 125 [2017] 126 See

308  David Howarth Many constitutional theorists appear to presuppose that contradiction should always be eliminated. For them, as for many lawyers, eliminating incoherence is not just a matter of aesthetics but also a moral duty. But attempting to eliminate contradiction has costs as well as benefits. It might expose conflicts better managed by creating distance between the conflicting elements. Coherence is neither inevitable nor is the search for it necessarily desirable. Improvisation might often be the best we can do.

Comparisons

310

12 A Proposal for Defining and Classifying Systems of Constitutional Government PAUL YOWELL*

I. INTRODUCTION

As Mary Wollstonecraft surveyed the rise of constitution making in France and the US, she said: A constitution is a standard for the people to rally around. It is the pillar of a ­government, the bond of all social unity and order. The investigation of its principles make it a fountain of light; from which issue the rays of reason, that gradually bring forward the mental powers of the whole community.1

The conception of a constitution defended in this chapter is more mundane but, it is hoped, not less important. In contrast to accounts that imbue ­constitutionalism with a set of substantive aims such as democracy, liberty, and equality,2 I give a largely formal definition. This chapter aims to answer two questions: (1) what is a ­constitution? and (2) what are the types of constitutions in the world? One method for answering these questions is to take a descriptive stance on ­question 1: every nation is assumed to have a constitution, and the term is used as an equivalent for a system or form of government. This approach might then employ normative criteria for question 2, for example by classifying some constitutions as authoritarian or totalitarian and others as democratic or liberty-respecting. Another approach holds that having a constitutional system of government is a normative achievement, a standard to which all nations can be held. Theorists have proposed

* The author is grateful to Max McGiffen for research assistance. 1 M Wollstonecraft, An Historical and Moral View of the Origin and Progress of the French Revolution (1794). 2 An example is N Barber, The Principles of Constitutionalism (Oxford, Oxford University Press, 2018), which gives an account of constitutionalism with six principles: sovereignty, separation of powers, the Rule of Law, civil society, democracy, and subsidiarity.

312  Paul Yowell numerous criteria to define constitutionalism: democracy; constraints on power (limited government); separation of powers; subsidiarity; respecting the rule of law; recognising equality before the law; protecting the kind of rights typically found in Bills of Rights, etc. Some of these are functional: they refer to how the powers of the state are structured and exercised, and how powers interact. Other criteria are substantive: they set standards that the law and exercise of power should meet, with respect to rights, liberty, equality, etc. Whatever the criteria chosen by a particular theorist, this approach holds that having a constitutional system of government is a requirement for moral legitimacy. Constitutional states are in this sense superior to non-constitutional ones. I call this the thickly normative approach. In this chapter, I take a thinly normative approach to question 1, and a descriptive approach to question 2. I define a constitution according to a set of functional criteria. I do not claim that having a constitutional government is required for moral legitimacy, or that all nations should necessarily aspire to it. However, that does not mean that the concept is normatively inert. Adopting a constitution is generally chosen as a means for restraining, limiting, or disciplining power that, in the absence of legal limits, might prove too dangerous. Constitutionalism is often the most eligible, salient means of holding those who exercise state power accountable. The thinly normative approach to question 1 will enable a classification of countries into those with a constitutional system of government and those without. The latter category includes absolute monarchies, military and other dictatorships, and singleparty states. Regarding question 2, I take a descriptive approach that focusses on the structure of executive and legislative powers. This enables a further classification of constitutional states into parliamentary and presidential systems, as well as identification of constitutional systems that do not fall neatly into one category. While the classification is neutral and descriptive, it is not without potential moral significance. There are arguments, for example, that parliamentary systems are generally better than presidential systems at meeting the goal of limiting state power and holding it accountable; and vice versa. I will lay out these arguments but not take sides between them. I do not present the methodology sketched above as superior to other approaches. It is a useful methodology, particularly for: (1) students of comparative constitutional law; (2) those who see the structure of legislative and executive powers as central to constitutionalism; and (3) those who favour the approach of political constitutionalism over legal constitutionalism. Theorists who favour legal constitutionalism, or who take judicial review of legislation and the enforcement of constitutional rights to be central, may find my methodology less useful. I will not attempt to settle the debate between legal and political constitutionalism or over judicial enforcement of a Bill of Rights. I will, however, give reasons to doubt that judicial enforceability of rights, or of constitutional law, is a sine qua non of constitutionalism. At the end of this chapter is a one-page table proposing a broad classification of constitutional and non-constitutional systems for all the members of the United Nations. Along the way, I discuss some of the considerations that go into making the chart and into classifying certain countries into one category another.

Defining and Classifying Constitutional Systems  313 II.  ON CONSTITUTIONAL GOVERNMENT AND ITS HISTORY

In this section I propose a definition of constitutional government, continue the methodological discussion begun in the Introduction, and sketch the historical rise of constitutionalism. For the latter discussion, my main interlocutor is the German political theorist Karl Loewenstein, whose work on the history and classification of constitutions deserves to be more widely known.3 A system of constitutional government exists when the executive, legislative, and judicial powers of a state are exercised in accord with, and by persons appointed pursuant to, a set of legal rules and stable conventions and principles. These rules, conventions, and principles provide for the composition and methods of operation of the main institutions of government,4 and for orderly, regular transitions of power (whether by democratic elections or otherwise). Together they comprise a constitution, to which both officials and citizens show an allegiance, transcending day-to-day politics and individual leaders, and which has a quality of enduring through time. The definition of a constitution in the above paragraph is explained further in section III below and contrasted to non-constitutional systems of personal or party rule. This section shows that constitutional government – in the thinly normative sense aforementioned – has its modern roots in the system of parliamentary government that emerged from the English Civil War and the American and French revolutions. Constitutionalism, however, is not purely a modern phenomenon. There was constitutional government in Rome, Greece, and elsewhere in the ancient world, as well as the medieval. It is said that Aristotle set out to describe and classify 158 constitutions (politeia).5 This formed the background research for his Politics and his famous six-fold classification of types of government, as well as his analysis of the ‘mixed’ constitution which combines different types. Aristotle at one point defines politeia as ‘a certain ordering of the inhabitants of the city-state [polis]’.6 He speaks of the constitution [politeai] of a community as ‘the form of the compound’ and argues that whether the community is the same over time depends on whether it has the same constitution. The constitution is not a written document, but an immanent organizing principle, analogous to the soul of an organism. Hence, the constitution is also ‘the way of life’ of the citizens.7

Politeia, in this sense, has a broader meaning than my thinly normative definition of a constitution. Methodologically, Aristotle’s usage is akin to the first approach 3 See especially K Loewenstein, Political Power and the Governmental Process (Chicago, University of Chicago Press, 1957). 4 This draws on Ivor Jennings’s definition of a constitution as ‘the document in which are set out the rules governing the composition, powers and methods of operation of the main institutions of government’, or in the absence of such a document, ‘the rules determining the creation and operation of governmental institutions’. I Jennings, The Law and the Constitution (London, University of London Press, 1959) 33, 36. 5 D Laërtius, Lives of the Eminent Philosophers, Book V. 6 Aristotle, Politics (S Everson ed, Cambridge, Cambridge University Press, 1996) 1274b32-41. 7 F Miller, ‘Aristotle’s Political Theory’, The Stanford Encyclopedia of Philosophy (Winter 2017 Edition), at https://plato.stanford.edu/archives/win2017/entries/aristotle-politics/.

314  Paul Yowell described in the Introduction, in which a theorist assumes all nations have a constitution and then goes about classifying them. Aristotle’s six-fold classification is both descriptive and normative. The three pairings reflect his well-known descriptive categorisation of rule by one, the few, or the many.8 The divisions between the pairs are normative. In the types on the left-hand side, rulers govern for the sake of the common good. On the right-hand side are the corrupted forms in which rulers govern for the benefit of those who rule: Monarchy

Tyranny

Aristocracy

Oligarchy

Politeia

Democracy

Aristotle’s usage of politeia here is narrower than the usage above to refer to a ‘form of the compound’ or ‘way of life’. Here politeia refers to the rule of many when they govern for the common good, as distinguished from a regime in which the many rule more narrowly, for the sake of their class.9 There is a connection between Aristotle’s narrow sense of politeia and the definition of a constitution as a rules-based system of government. If the many are to rule, they can hardly do so without a stable set of rules to determine who exercises decision-making power at which time. In Athens, all citizens were entitled to participate in the Assembly, which held ultimate power and met periodically. For the offices that carried out day-to-day decision-making, there was a principle of rotation in office.10 The primary means of selection was by lot from citizens who volunteered.11 In The Constitution of Athens (sometimes ascribed to Aristotle but likely by one his students), after a comparative and historical survey, there is an account of the ‘constitution [politeia] of the present day’.12 It details the various roles of the Assembly (Ekklesia), Council (Boule), magistrates, and other offices, and the rules that govern them. This is, in effect, a particular constitution in the sense I have defined above. A monarchy would not necessarily need such a framework of established rules. In Books III and IV of the Politics, Aristotle says that the best regime would be aristocracy or monarchy. The argument depends, however, on the availability of a wise and virtuous monarch or similar aristocratic group. In books VII and VIII, Aristotle argues that the best regime is a ‘mixed’ constitution, which combines elements of monarchy, aristocracy, and politeia (in the narrow sense above). Vander Waert offers a reconciliation of these two positions.13 For Aristotle, the true ideal is a good monarchy, since it enables more citizens to pursue the good of contemplation

8 Aristotle, Politics (n 6) 1279b. 9 ibid. 10 See generally MH Hansen, The Athenian Democracy in the Age of Demosthenes (Oxford, Oxford, University Press, 1991) for discussion of Athenian government and Aristotle’s account of it. 11 ibid. 12 Aristotle, The Constitution of Athens (S Everson ed, Cambridge, Cambridge University Press, 1996) 243–63. 13 PA Vander Waert, ‘Kingship and Philosophy in Aristotle’s Best Regime’ [1985] Phronesis 249.

Defining and Classifying Constitutional Systems  315 and philosophy. In a democracy or mixed constitution, people will have to spend more time on governing, which, while a worthy and noble pursuit, is less valuable than a life devoted to contemplation.14 Since good monarchs are not frequently found, the mixed constitution is a second-best ideal that should be sought in typical conditions. In comparing types of government, Aristotle employs a sound methodological principle, which informs my thinly normative approach: ‘[T]he best is often unattainable, and therefore the true legislator and statesman ought to be acquainted, not only with that which is best in the abstract, but also with that which is best relatively to the circumstances.’15 Under certain conditions, one might be warranted in preferring non-constitutional government – say, an absolute monarchy – to a constitutional system. In other circumstances as discussed in section IV, a non-constitutional government might be an appropriate intervention against a constitutional system that has deteriorated. Constitutional government, in the sense I have defined, probably tends to produce good substantive outcomes, such as promoting the common good and respecting equality and liberty; but this is not guaranteed. Constitutionalism can in this way be linked to the Rule of Law, in the understanding of scholars such as John Finnis and Nigel Simmonds, who have argued that observance of the Rule of Law tends to promote good law but may be compatible with evil law.16 One might consider the ideal of constitutional government as an extension of the ideal of the Rule of Law to the basic framework of government: the creation of rules to do with the appointment of and accountability for those who hold executive, legislative, and judicial power. Both constitutional government and the Rule of Law are ways of promoting the ideal of ‘a government of laws, and not of men’.17 On my approach, substantive aims such as respecting liberty might be expected by-products of constitutional government, but they are not part of the definition. Karl Loewenstein employs a thickly normative approach in a 1969 article that traces the historical rise of constitutionalism and contrasts the liberal democratic constitutions of the West with the constitutions of the Soviet Union and its satellites.18 He classified the former as ‘normative’ constitutions, because their devices for limiting government and securing individual rights were, in large part, implemented. The constitution of the Soviet Union, however, was a ‘semantic’ constitution, both because the individual rights and liberties guaranteed in the text were not observed in practice and because the constitution’s functional provisions provided no effective restraints on the exercise of state power. Loewenstein used the term ‘nominal’ for constitutional systems wherein the provisions for limiting power and protecting rights were observed imperfectly.

14 ibid. 15 Aristotle, Politics (n 6) 1288b. 16 See J Finnis, Natural Law and Natural Rights (Oxford, Oxford University Press, 1980) 273–4; N Simmonds, ‘Evil Regimes and the Rule of Law’ in N Simmonds (ed), Law as a Moral Ideal (Oxford, Oxford University Press, 2008). 17 Massachusetts Constitution of 1780, art 30. 18 K Loewenstein, ‘Constitutions and Constitutional Law in the West and in the East’ [1969] The Indian Journal of Political Science 203.

316  Paul Yowell Loewenstein’s historical analysis connects constitutional government to the aim of securing individual liberty against arbitrary power. He attributes the ‘significant breakthrough to constitutionalism’ to the ‘genius of the introspective Greeks’: For the sake of rationalizing and, thereby, limiting power, they devised certain political institutions and techniques by which the individual citizen could participate in the formation of the will of the community and protect his sphere of self determination against caprice and arbitrariness of the actual power-holders whose tenure of office was strictly limited. Constitutional government and, at the same time, constitutional democracy had been discovered.19

It is doubtful, however, that the democratic institutions of the Greeks were established the purpose of protecting such an individual sphere. Athenians adhered to a unity of the polis and the social order, and restrictions on individual liberty were pervasive.20 Private fortunes could be seized for state purposes, and military service was compulsory until age 60.21 The education of children was directed by the polis toward the end of forming virtuous citizens. Nicholas Aroney summarised Aristotle’s view of the matter: ‘As part of a political unity, it follows that no citizen should consider that he belongs to himself.’22 Loewenstein also identifies a constitutional form in Rome, albeit without the Greek experiments in democracy. The Roman Republic ‘was from the start constitutional government in the fullest sense of the term’; for half a millennium, the political process of the civitas observed commonly agreed conduct patterns for the exercise of public power, manifested by certain institutions and established, on occasions, even in statutory form – all intended to restrain absolute or unrestricted power of the elected magistrates.23

Recounting the later rise of constitutionalism in the early modern period, Loewenstein includes the protection of individual liberty directly in his definition of it and sees the separation of powers as a key to realising it: The subsequent history of constitutionalism since the middle ages is nothing but the quest of political man to limit the exercise of absolute power by his rulers and to substitute for the traditionally practiced or enforced submission to factual social control the moral or ethical legitimation of authority. This purpose of limiting absolute power or brute force in the common interest was deemed to be served best by an agreement on certain definite rules for the conduct of the political process. … In time these were believed to consist in arrangements by which, instead of a single power-holder – king, prince, or any other state organ monopolizing power – its exercise is being shared among several power-holders bound to cooperate in the formation of the will of the community. … Splitting the Leviathan is the essence of constitutional government.24

19 ibid 203. 20 Thanks to Pablo Ortuzar for helpful discussion on this point. 21 N Fustel de Coulanges, The Ancient City. A Study on the Religion, Laws, and Institutions of Greece and Rome (Dover Publications, New York, 2006 [1863]). 22 N Aroney, ‘Subsidiarity in the Writings of Aristotle and Aquinas’ in M Evans and A Zimmermann, (eds), Global Perspectives on Subsidiarity (Dordrecht, Springer, 2014) 14, 9–27. 23 Loewenstein, ‘Constitutions and Constitutional Law’ (n 18) 204. 24 ibid 204–05.

Defining and Classifying Constitutional Systems  317 Loewenstein’s account is similar to Charles MacIlwain’s contention that ‘in all its successive phases, constitutionalism has one essential quality: it is a legal limitation on government.’25 Nick Barber calls this ‘negative constitutionalism’.26 While this view is stressed in Loewenstein’s historical account, it does not preclude what Barber calls ‘positive constitutionalism’, that is, a concern with ‘creating an institutional structure that facilitate[s] the successful functioning of the state’.27 Modern constitutionalism, Loewenstein argues, finds its first fruition in England after the Civil War and opposition to the claims of the Long Parliament.28 The Agreement of the People (1647), drafted by Levellers and army officers, provided for frequent Parliaments and freedom of religion. This early blueprint was followed by Cromwell’s Instrument of Government (1653) ‘as the first fully articulated or written constitution’.29 England was the first nation since antiquity to attain a lasting constitutional government, according to Loewenstein, taking its final form after the Glorious Revolution (1688) as a constitutional monarchy limited by the sovereign parliament. British parliamentarianism served as the model of constitutional and subsequently constitutional-democratic government for a long time to come.30

It was the American and French revolutions, however, that led to constitutionalism becoming the dominant form of government in the world: After having passed through its formative stage in the revolutions of the 18th century, constitutional government conquered the globe. True, the process was not uniformly synchronized; relapses occurred in royal supremacy (in the form of legitimism). But in due course all old established nations of Europe had turned to constitutional government and the newly created political entities adopted it automatically as the only feasible form of self-rule. Under the pressure of the rising masses, nearly everywhere constitutionalism was converted into constitutional democracy in one form or other. In all these cases, the principal technique of limiting power was the written constitution; it has become the symbol of national sovereignty and the hallmark of independent statehood.31

While Loewenstein has a thickly normative conception of a constitution in mind, his account accurately tracks the steady rise of constitutional government, in my thinly normative sense, in the eighteenth and nineteenth centuries. The next century, however, saw setbacks for constitutional government in the lead up to World War II and during the Cold War. In 1969, the Eurasian landmass was dominated by non-constitutional government, from East Germany to Vladiovostok, mainly in the form of single-party states ruled by a Communist Party. While these states had written constitutions, they served to describe the operations of power rather than to constrain them. It was the



25 C

MacIlwain, Constitutionalism: Ancient and Modern (Indianapolis, Liberty Fund, 2008) 21. Principles of Constitutionalism (n 2) 2–6. 8. 28 Loewenstein, ‘Constitutions and Constitutional Law’ (n 18) 205. 29 ibid. 30 ibid. 31 ibid 206. 26 Barber, 27 ibid

318  Paul Yowell Party that transcended the constitution, not vice versa. As Loewenstein observed at that time: So lasting, however, was the tradition and so great the reputation of constitutional government that the new authoritarian regimes, with few exceptions, are finding it expedient to disguise the exercise of naked force by pseudo-constitutional forms and even written constitutions adroitly designed for the political needs of the actual power-holders.32

Twenty years later, the Berlin Wall fell, sparking a wave of reforms that led eventually to the fall of the Communist Party in the Soviet Union and its European satellites – in most places replaced by constitutional government. As Table 1 (see p 339) ­indicates, a large majority of the countries in the world are now constitutional. Before moving to the next section, I briefly consider the respective contributions of Britain, France, and the US to the history of constitutionalism. (I use Britain and British here, and in rest of this chapter, in a composite way to include England in the 1600s, Great Britain following union with Scotland in 1707, and today’s United Kingdom.) The circumstances of the drafting and ratification of the 1787 US Constitution are well known and do not need repeating here. Of the three countries, the US has been the most influential, at least in providing a widely copied structural form. Most constitutional systems in the world are presidential republics (see Table 1), with many closely following the US model. Most also have a codified, ‘large-C’ Constitution, a Bill of Rights, and some form of judicial enforcement of the constitution. The US has also been influential in the theoretical arena. The Federalist is possibly the most widely read work of constitutional theory of the modern era. In particular, Madison’s essays on the separation of powers have contributed to widespread adoption of the US model.33 The French experience of constitutionalism, in contrast to that of the US, has been turbulent. France has had almost twenty constitutions since the Revolution and five republics, punctuated by periods led variously by Napoleon as emperor, different royal houses, and Napoleon III.34 The first French Constitution of 1791 set up a constitutional monarchy lasting two years before the execution of Louis XVI. The second Constitution of 1793 was to install the National Assembly as the supreme power and an executive council with little power to check it; but never came into force. The Committee of Public Safety, led by Robespierre, set it aside and ruled as a revolutionary government until the Constitution of 1795, which established an executive Directory as the principal power and a subordinated (bicameral) legislature. The 1795 Constitution lasted four years, followed by a series of constitutions that respectively recognised Napoleon as First Consul (1799); Consul for Life (1802); and Emperor (1804). The house of Bourbon was restored in the Charter of 1814, on a model of constitutional monarchy similar to Britain’s.

32 ibid. 33 See The Federalist Nos 46–48, 51–53. 34 See S Boyron, The Constitution of France: A Contextual Analysis (Oxford, Hart Publishing, 2012), ch 1.

Defining and Classifying Constitutional Systems  319 I need not recount in detail the history of further Bonapartist empires, constitutional monarchies, and republics up until the constitution of 1958 that governs the Fifth Republic.35 Suffice to say that during a significant part its history, France was not under constitutional government even in the thinly normative sense of my definition, but instead had a semantic or nominal constitution, reflecting rather than constraining the will of Napoleon and other leaders. France’s chaotic experimentation with different constitutional systems, especially in the 15 years following the Revolution, has left little imprint on the constitutions in the world today, in terms of structural models. The US Constitution, in contrast, has endured for almost 250 years, surviving vicissitudes including the Civil War, Reconstruction, the Great Depression, and two World Wars, and has been widely imitated. It has even been imitated by the French, for the 1958 Constitution, which originally set up a parliamentary system, has been transformed through a series of amendments into a presidential republic.36 The main French contribution to constitutionalism has been not through providing a structural model, but in theorising the idea of a constitution. The most important and distinctive theoretical contribution of the French has been that of constituent power – pouvoir constituant,37 in the coinage of the Abbé Sieyes – holding that the framers of a constitutional order, with the support of the people, have the rightful power to create a comprehensive framework of government embodied in a written document. It was overstatement when Sieyes said that a sound and useful idea was invented in 1789; the separation of the constituent power from constituted powers. It will go down in history as a discovery that advances science, for which the French can be thanked.38

The American founders acted on a similar idea and extended it further by submitting the 1787 Constitution to popular ratification. The French, however, did more to spread the idea in Europe, and the frequency with which they created new constitutions could be seen to illustrate the nature of the modern constitution as positive law brought about by the will of its makers. The main British contribution to constitutionalism has already been noted: establishing an enduring constitutional government since the late seventeenth century, and a stable model of parliamentary democracy since the early nineteenth century. The influence of the British model can be seen in the many parliamentary systems listed in Table 1. One of the animating ideas of the British constitution stands in contrast to Sieyes’s pouvoir constituant. It is reflected in Dicey’s conception of a ‘historic constitution’ and defended by Edmund Burke in his Reflections on the Revolution in France. Burke argued that the British constitution was distinguished by a ‘principle of conservation, and a sure principle of transmission’ without ‘excluding a principle

35 ibid. 36 ibid, ch 3. I set aside the question here of whether France has a semi-presidential system: see section IV and Table 1. 37 See L Rubinelli, ‘How to think beyond sovereignty: On Sieyes and constituent power’ [2019] European Journal of Political Theory 47–67. 38 See ibid.

320  Paul Yowell of improvement’.39 He saw the Glorious Revolution of 1688 ‘not as a break with the past and a new beginning, but as a restoration – a way of securing the preservation of “our antient indisputable laws and liberties”’.40 As Grégoire Webber has noted, today’s ‘received wisdom’ strongly favours a comprehensive written constitution and regards the Burkean view with suspicion, or even as ‘heretical’.41 It would, however, be rash to dismiss it as a model, given the long and stable endurance of the British Constitution, which created a framework for democracy and for the effective protection of civil liberty. Moreover, the British parliamentary system has been a model for many of the most stable and successful democracies of the modern era, as can be seen from a glance at Table 1. The US presidential system has arguably been much less successful as an export model. The comparison of parliamentary and presidential systems is pursued further in section IV. III.  THE CHARACTERISTICS OF CONSTITUTIONAL GOVERNMENT

A difficulty in defining constitutional government arises in trying to articulate a conception that reflects the US and French ideal of a written constitution, without neglecting the historical influence of the British constitution. Sartori noted the paradox that ‘the mother country of modern constitutionalism appeared to have an obscure constitution, or even – according to some of the standards that seemed very important elsewhere – no constitution at all’.42 The definition in section I, repeated here, is meant to bridge this divide: A system of constitutional government exists when the executive, legislative, and judicial powers of a state are exercised in accord with, and by persons appointed pursuant to, a set of legal rules and stable conventions and principles. These rules, conventions, and principles provide for the composition and methods of operation of the main institutions of government, and for orderly, regular transitions of power (whether by democratic elections or otherwise). Together they comprise a constitution to which both officials and citizens show an allegiance transcending day-to-day politics and individual leaders, and which has a quality of enduring through time.

This is a ‘small-c’ conception of a constitution. It includes, but is not limited to, provisions contained in a single codified constitutional document (‘large-C’), such as the French Constitution of 1958. This section explores the difference between the small-c and large-C conceptions, comments on the definition above, and compares it to the accounts of Dieter Grimm and Joseph Raz. Grimm and Raz offer functional definitions that are comparable in some ways.

39 See G Webber, ‘Can the constitution of a fruit fly be written?’ (2020) Queen’s University Legal Research Paper No. 2020-004, at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3635487. 40 ibid. 41 ibid. 42 G Sartori, ‘Constitutionalism: A Preliminary Discussion’ [1962] The American Political Science Review 853, 853.

Defining and Classifying Constitutional Systems  321 A.  The Elements of Constitutional Government i.  Legal rules By legal rules I mean positively formulated laws, which are usually found in a large-C constitutional text but may also be found in statutes, such as ‘constitutional statutes’ in the British conception or ‘organic laws’ (lois organiques) in the French. Such legal rules may be judicially enforceable but are not necessarily so. The US Judiciary Act of 1789 is a statute that, in part, belongs to the small-c constitution. The large-C Constitution establishes only the Supreme Court while authorising Congress to establish other courts. The Judiciary Act establishes the District Courts and Circuit Courts of Appeal and regulates the jurisdiction43 of the Supreme Court and the number of its justices. Originally there were six justices; an amendment to the Act in 1869 set the number at nine. That is, by now, a matter of constitutional significance. ii. Conventions Having nine Supreme Court justices is not a matter of constitutional significance merely by virtue of appearing in the Judiciary Act. Rather, through long usage and agreement around the need for a fixed number, a Supreme Court with nine members has acquired the status of constitutional convention in the US. The reason for having a fixed number is to avoid Congress and the President adding justices to achieve desired outcomes (‘packing the Court’). While they have the sheer legal power to do so under the large-C US Constitution, the result would be to undermine the authority and legitimacy of the Court. Thus, its stable membership is part of the small-c constitution. Conventions, as a method of constitutional governance, are most closely associated with the British constitution. Several conventions provide for the composition and operation of political institutions: the appointment of the leader of the majority party in the House of Commons as Prime Minister; the Salisbury convention, providing that the House of Lords will not block a bill that arises from a manifesto commitment; the convention of royal assent to bills passed by the Commons and Lords; and many others. Given the importance of conventions to British constitutional practice, it is not surprising that British scholars have done the most to define conventions and theorise their practice. A widely accepted definition says that a convention exists where: (i) there is a settled practice; (ii) that is recognised; and (iii) there is a reason for the practice.44 Despite their association with the British constitution, conventions operate widely in other countries. Indeed, it is likely that every constitution will, after a period of time, develop conventions.45 Another example from the US helps to highlight their importance, namely, the direct democratic election of the US President. The large-C

43 Subject to limits in Art III of the US Constitution. 44 Jennings, The Law and the Constitution (n 4) 81–82. 45 See P Doudonis, ‘The Normativity of Constitutional Conventions, with Special References to the UK, USA and Greece’ (2020), DPhil Thesis, University of Oxford.

322  Paul Yowell Constitution provides for indirect election via the Electoral College, but, by the middle of the nineteenth century, there was a settled practice of political parties nominating slates of electors who pledged to vote for the party nominee rather than to exercise independent judgment when the Electoral College meets.46 This is a constitutional convention so pervasive that voters largely do not realise that when they vote Clinton, they are voting for electors who technically have power to vote Bush. Although the institution of voting for pledged electors is a constitutional convention in the US (hence part of the small-c rather than the large-C Constitution), the democratic election of the President is the key power-conferring event in the US political system. iii. Principles The legal rules and conventions of a constitution are ‘framed and supported – animated – by constitutional principles, which are “not merely a matter of general normative propositions” but rather “those elements of the common good which the law [and practice] of our constitution articulates and promotes”’.47 An example is the principle of the Rule of Law, as recognised in the constitutional traditions of countries such as Britain and Canada. While frequently informing general constitutional practice, principles are on occasion used directly in constitutional adjudication. A dramatic application of the principle of the Rule of Law was made by the Canadian Supreme Court in a case concerning the laws of the province of Manitoba, which were held to be null and void because they had not been translated into French, as required by both the provincial and federal constitutions.48 In order to avoid the ‘chaos’ that would result from a legal vacuum, the Court relied on the principle of the Rule of Law – which, though unexpressed in constitutional documents, was held to be central to the constitution – to issue a novel remedy of a suspended declaration of invalidity.49 The effect of the order was suspended for one year, during which the provincial legislature could make the French translations.50 Several constitutions express a principle of democracy in the written text, and such a principle animates many other constitutions where it is not so expressed. The US Constitution makes no mention of a democratic principle, and the founders were wary of popular democracy. While they incorporated a democratic element in the election of the House of Representatives, they warned of the propensity of frequent popular elections to be swayed by passions of the electorate inflamed by demagogues.51 They shared the worry that Tocqueville would later articulate as the

46 JG Wilson, ‘American Constitutional Conventions: The Judicially Unenforceable Rules that Combine with Judicial Doctrine and Public Opinion to Regulate Political Behavior’ [1992] Buffalo Law Review 645, 651. 47 R Ekins, ‘Constitutional Principle in the Laws of the Commonwealth’ in R George and J Keown (eds), Reason, Morality and Law: The Jurisprudence of John Finnis (Oxford, Oxford University Press 2013) 396 (quoting John Finnis). 48 Reference re Manitoba Language Rights [1985] 1 SCR 721. 49 ibid. 50 ibid. 51 The Federalist Nos 10 and 78.

Defining and Classifying Constitutional Systems  323 ‘tyranny of the majority’.52 Thus, they hedged the democratic element with i­ndirect elections for the Senate and the US President. The transformation of these key institutions into direct popular elections (by the 17th Amendment for the Senate) has resulted in a constitution that is more democratic than the founders intended or would have thought wise. However, the modern embrace of the democratic principle in the constitutional system of the US cannot be doubted. Among other things, it animates the convention of electors voting for the presidential candidate they pledged to support.53 While there have been instances of ‘faithless’ electors, any large-scale defection would be condemned as undemocratic and therefore unconstitutional. The large-C Constitution would have nothing to say against such exercise of independent judgment, and could be read to support it.54 However, if the Electoral College were to award the presidency to a candidate other than the one chosen by the people democratically, a constitutional crisis so grave would ensue that it is not certain that the large-C constitution could survive. iv.  Endurance and Transcendence As Joseph Raz has said, a constitution is, and is meant to be, of long duration: It is meant to serve as a stable framework for the political and legal institutions of the country, to be adjusted and amended from time to time, but basically to preserve stability and continuity in the legal and political structure, and the basic principles that guide its institutions.55

It is not enough that a constitution is the result of the exercise of constituent power. The series of short-lived constitutions following the French Revolution, dramatically changing the basic structure of power, were marginal cases of constitutions. A constitution transcends day-to-day politics and commands allegiance from people and officials. To the extent that this is lacking, a constitution becomes nominal or semantic. A test for whether a country has constitutional government is whether its leaders, when contemplating a significant exercise of power that is novel or contrary to precedent, must concern themselves with whether the action can plausibly be blamed by critics or the people as unconstitutional. Where a leader can safely ignore such questions – or, for the sake of appearance, change the constitution at will – the system is non-constitutional. Having explored the basic elements of a constitutional system of government, we are now able to consider the relationship between large-C and small-c constitutions. 52 Democracy in America (Henry Reeve (tr), Cambridge, Sever and Francis, 1862 [1839]) vol 1, 324 (ch XV). 53 See KE Whittington, ‘The Status of Unwritten Constitutional Conventions in the United States’ [2013] University of Illinois Law Review 101, 111. 54 Following the election of President Trump in 2016, Lawrence Lessig, professor of constitutional law at Harvard, argued that electors should vote against Trump despite their pledge and argued that the Constitution’s provisions on the Electoral College authorised the exercise of such independent judgment. See K Cheney, ‘Lessig: 20 Trump electors could flip’ (2016) Politico, at www.politico.com/story/2016/12/ donald-trump-electors-lessig-232598? 55 J Raz, ‘On the Authority and Interpretation of Constitutions: Some Preliminaries’, in J Raz (ed), Between Authority and Interpretation (Oxford, Oxford University Press, 2009) 324.

324  Paul Yowell Almost all constitutional systems of government – and non-constitutional systems too – have large-C constitutions in the form of a single, codified document, that purports to set out the legal rules that establish and govern political and legal institutions and exercises of state power. Often, they express constitutional principles as well. Constitutional conventions are not included in the large-C constitution. Despite the ubiquity of large-C constitutions, it would be a mistake to ignore the exceptions. The most obvious of these is the British constitution, which contains numerous constitutional statutes but no overarching codified document. Israel also has an unwritten constitution, whose main provisions are found in a series of nine ‘basic laws’ – statutes enacted between 1958 and 1988. And in some countries what might first appear to be a large-C constitution turns out, on closer inspection, to be a combination of documents. The Constitution of Canada, for example, is defined by the Constitution Act 1982 to include that Act and the British North America (Constitution) Act of 1867 and its amendments. Many key provisions regarding federal and provincial powers are in the 1867 Act, promulgated originally as a statute of the UK Parliament. According to the Supreme Court, the Canadian Constitution also includes pre-confederation statutes and unwritten elements.56 As Kenneth Wheare has observed, the concept of a large-C constitution is narrower than the traditional British conception, because it includes only some of the laws that govern institutions and the exercise of power on constitutional matters.57 This narrower conception excludes statutes on electoral systems, representation, and regulation of the courts and executive, which may have a significantly greater impact on how a country is governed, than certain matters in the large-C Constitution.58 As a constitutional system matures, the gap between the large-C and small-c constitution will grow. One way is through constitutional conventions, which can supplement original provisions or even alter (as in the case of the US Electoral College convention) their original meaning. The gap can also grow through practices of interpretation of legal provisions; not only courts, but also executive and legislative officials, interpret the constitution. Interpretations might be innovative, and they might respond to situations that the drafters could not have foreseen. A settled interpretation will shape the way the constitution is understood. When a constitutional system is old, the way that state powers function may become significantly detached from the large-C constitution. Both small-c and large-C constitutions should be the object of study. To the extent of divergence, there is good reason for thinking that the small-c constitution matters more. B.  On Judicial Enforcement and Bills of Rights A conspicuous omission from my definition of a constitutional system of government is the notion of the constitution as a fundamental law and, for that reason, enforceable 56 Eg, Reference re Manitoba Language Rights [1985]. 57 K Wheare, Modern Constitutions (Oxford, Oxford University Press, 1951) ch 1. 58 Some large-C Constitutions are lengthy, such as those of several US States, and include provisions on matters within the usual scope of private law.

Defining and Classifying Constitutional Systems  325 by courts. Grimm and Raz include this in their definitions discussed below. While I agree that the constitution is law, it is possible, and historically common, for constitutional law not to be judicially enforceable. Grimm also stresses the importance of an enforceable Bill of Rights, suggesting that submission of the law to fundamental rights is necessary for constitutionalism to be complete. The most prominent image of constitutionalism, for many people, is a court striking down a law because it violates a constitutional right. For most countries, both judicial review and an enforceable Bill of Rights have been a recent development in comparison to the longer history of constitutionalism. The fame of Marbury v Madison59 has spread alongside the adoption of the US model, and the force of its logic seems irresistible. Chief Justice John Marshall held that all those who have framed written Constitutions contemplate them as forming the fundamental and paramount law of the nation, and consequently the theory of every such government must be that an act of the Legislature repugnant to the Constitution is void.60

Raz and Grimm effectively reformulate Marshall’s claim in their definitions of a constitution, as seen below. The US position as represented by Marbury, however, was an outlier worldwide until the post-WWII era. It was rejected by the other two great influences on constitutionalism. The French revolutionaries, associating a powerful judiciary with the ancien regime, explicitly rejected judicial review in their constitutional documents. That intellectual tradition was maintained through the early decades of the twentieth century, when French theorists saw the Lochner era in the US as a lesson in what could go wrong under les gouvernement des juges.61 The 1958 Constitution allowed limited pre-enactment review of legislation by the quasi-judicial Conseil constitutionel, and that body unilaterally extended it to cover the Declaration of the Rights of Man in 1971. It was only in 2008 that the power was extended to a posteriori review by constitutional amendment.62 The British have never accepted the American doctrine, but have maintained the principle of parliamentary sovereignty. The Human Rights Act of 1998 allows judges to declare laws incompatible with rights, but not to annul statutes. The Westminster reluctance to embrace a judicially enforceable Bill of Rights still bears traces in Australia, which has judicial review under the federal constitution but no Bill of Rights; New Zealand, which has a statutory Bill of Rights but maintains parliamentary sovereignty; and Canada, which adopted an enforceable charter of rights in 1982, but with a power for Parliament to override judicial decisions (section 33). The Dutch constitution explicitly forbids courts to use it in judicial review (Article 120), and the Swiss constitution provides that federal law takes precedence to constitutional law in the event of a conflict. In several countries where there is a nominal power of judicial

59 Marbury v Madison 5 US 137 (1803). 60 ibid. 61 See É Lambert, Le gouvernement des juges et la lutte contre la législation sociale aux États-Unis (1921). 62 F Fabbrini, ‘Kelsen in Paris: France’s Constitutional Reform and the Introduction of a Posteriori Constitutional Review of Legislation’ (2008) 9 German Law Journal 1297.

326  Paul Yowell review of legislation, little use is made of it. These include the Scandinavian countries and Japan.63 Although most constitutions created since World War II have included a form of judicial review of legislation, nearly all are modelled on the constitutional court pioneered by Austria in its 1920 constitution. This is commonly called the Kelsenian model, because of the prominent role of Hans Kelsen in drafting the relevant articles of the 1920 constitution, serving on the constitutional court, and writing influential theoretical defences of the model.64 A crucial feature of the Kelsenian court is that the power of judicial review is centralised in the constitutional court and denied to ordinary courts.65 Though Kelsen conceived of the constitution as a superior law conditioning the validity of lower law, he did not think that a constitutional system should give ordinary courts the power to enforce constitutional law.66 The constitutional court is not a ‘mere’ court charged with adjudicating cases according to law. Rather, the court is a ‘negative legislator’, which could be considered part of the legislative branch as much as the judicial.67 In sum, the idea that the constitution is fundamental law and therefore enforceable by courts generally is historically peculiar to the US. From the late eighteenth century to World War II, it was mainly to be found there. Today only a handful of countries with the power of judicial review follow the American model.68 Most countries with such a power entrust it to a single constitutional court on the Kelsenian model – a sui generis, quasi-legislative institution. Finally, it is worth noting that a bill of rights did not figure in the theoretical design of the originators of the American and Kelsenian models of judicial review. Kelsen opposed giving a constitutional court the power to enforce a Bill of Rights.69 The idea of a Bill of Rights was discussed briefly at the Philadelphia convention and dismissed. Alexander Hamilton, who wrote the classic defence of judicial review in the Federalist No. 78, opposed adding a Bill of Rights to the constitution in No. 84 and thought that such rights were too vague to be justiciable.70 Marshall, who followed Hamilton’s reasoning in Marbury, also opposed an enforceable Bill of Rights during discussions on the Constitution of Virginia.71 The US Supreme Court did not regularly enforce the Bill of Rights until the twentieth century, and it was only after WWII that such adjudication became frequent.

63 D Law, ‘The Anatomy of a Conservative Court: Judicial Review in Japan’, in A Chen and T Ginsburg (eds), Public Law in East Asia (Farnham, Ashgate, 2013). 64 See P Yowell, Constitutional Rights and Constitutional Design: Moral and Empirical Reasoning in Judicial Review (Oxford, Hart Publishing 2018) 1–3; 148–51. 65 P Yowell ‘The negative legislator: On Kelsen’s idea of a constitutional court’ in M Belov (ed), Courts, Politics and Constitutional Law: Judicialization of Politics and Politicization of the Judiciary (London, Routledge, 2019). 66 ibid. 67 ibid. 68 The model is followed in Canada, Ireland, and India, and partially in Greece, Argentina and Brazil. 69 See H Kelsen, ‘On the Nature and Development of Constitutional Adjudication’ in L Vinx (ed), The Guardian of the Constitution: Hans Kelsen and Carl Schmitt on the Limits of Constitutional Law (Cambridge, Cambridge University Press, 2015) 48, 60, 76; and see discussion in Yowell ‘The negative legislator: On Kelsen’s idea of a constitutional court’ (n 65) 126–8. 70 See Yowell, Constitutional Rights and Constitutional Design (n 64) 149. 71 Eliot, The Debates in the Several State Conventions (1836), vol III, 561.

Defining and Classifying Constitutional Systems  327 Judicial review of the constitution, and especially a Bill of Rights, is now a very common constitutional institution of great importance. However, to say that it is a necessary element of constitutional government is to imply that constitutionalism in its full form is mainly a post-WWII phenomenon. This fails to appreciate the rapid, pervasive rise of constitutionalism from the late eighteenth century forward, and diminishes the contribution of the British and French constitutional traditions, where judicial review of a Bill of Rights has been largely absent. C.  Grimm’s and Raz’s Definitions of a Constitution Dieter Grimm gives a five-part definition of a modern constitution: (1) The constitution in the modern sense is a set of legal norms, not a philosophical construct. … (2) The purpose of these norms is to regulate the establishment and the exercise of public power as opposed to a mere modification of a pre-existing public power. … (3) The regulation is comprehensive in the sense that no pre- or extra-constitutional bearers of public power and no pre- or extra-constitutional means to exercise this power are recognized. (4) Constitutional law is higher law. It enjoys primacy of all other laws and legal acts emanating from government. Acts incompatible with the constitution cannot claim legal validity. (5) Constitutional law finds its origin with the people as the only legitimate source of power. The distinction between pouvoir constituant and pouvoir constitué is essential to the constitution.72

This definition shows the continued prominence of the idea popularised by Sieyes. The pouvoir constituant is expressed in part (2), implied in (1) and (3), and appealed to explicitly in (5). Grimm wrote that the most important idea of modern constitutionalism is that ‘the constitution claims to establish legitimate government instead of only modifying the conditions for a pre-existing government that derives its legitimacy from sources other than the constitution’.73 However, making this the central conception of constitutionalism comes at the cost of excluding the British constitution, as well as others that were historically formed and shaped, rather than written on a blank slate. Canada is just one example. Grimm’s definition, moreover, leaves little room for the notion of a small-c constitution, and is instead tilted toward the large-C Constitution Grimm’s focus on the constituent power, as well as on judicial enforceability of a constitution in his point (4), make his definition too narrow. His account of the general aims of constitutionalism, however, fits with my thinly normative definition and links made earlier with the Rule of Law: By submitting all government action to rules, a constitution makes the use of public power predictable and enables the governed to anticipate government behaviour vis-à-vis 72 D Grimm, ‘Types of Constitutions’, in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012) 104. 73 ibid 103.

328  Paul Yowell themselves and allows them to face government agents without fear. A constitution provides a consensual basis for persons and groups with different opinions and interests to resolve their disputes in a civilized manner and enables peaceful transition of power.74

Using a thinly normative methodology, he recognises that a constitution may pursue a range of substantive aims. The constitution is ‘positive law’ that regulates political decision-making, but ‘this cannot save it from being a product of the political will itself.’75 The content of a constitution is thus ‘contingent on different conditions and different ideas of a just order’.76 Grimm identifies broad types of constitutions that can reflect these different ideas: (1) liberal democratic; (2) liberal non-democratic; (3) non-liberal democratic; (4) the social or welfare constitution; and (5) socialist constitutions.77 Raz’s methodology is also thinly normative,78 and his definition of a constitution has seven functional elements. A constitution (1) structures state power; is (2) stable, (3) written, (4) superior law, (5) justiciable, (6) entrenched; and (7) includes commonly held principles.79 This definition captures a broader range of constitutional systems than Grimm’s five-part definition. In contrast to Grimm’s focus on constituent power, Raz says that most constitutions are not ‘originating’ constitutions but ‘are made by legitimate legal authorities as part of a process of legal reform’.80 He notes that ‘constitutions that stand at the birth of a new independent country are often made in pursuance of legal authority conferred on their makers by the previous legal order in force’, and this includes most countries in the British Commonwealth.81 Elements (1), (2) and (7) of Raz’s definition are tracked in mine,82 which directly incorporated his statement on stability in the discussion above. Raz’s definition, however, could exclude constitutions that mine embraces, because of the insistence in (4) and (5)83 on the constitution as higher, judicially enforceable law, and because (3) and (6) apply mainly to large-C constitutions. However, Raz does not say that these are seven necessary and sufficient elements. Rather, they are a ‘characterisation’ of a constitution; each criterion is vague in its application, together yielding a vague concept.84 This methodology fits his aim, which is not to ‘draw borderlines, but to focus discussion’. His purpose ‘is to highlight the central features of constitutions [that] give rise to theoretical questions that do not apply, at least not to the same degree, to other law’.85 Those questions include how to interpret constitutions and

74 ibid. 75 ibid 116. 76 ibid. 77 ibid 116–29. 78 Raz, however, uses different terminology. He uses ‘thin conception’ for what I described as a ‘descriptive’ approach in the Introduction, and he uses ‘thick’ similarly to my use of ‘thinly normative’. 79 Raz, ‘On the Authority and Interpretation of Constitutions’ (n 55) 324–5. 80 ibid 330–1. 81 ibid. 82 ibid 324. 83 ibid 324–5: ‘Fifth, there are judicial procedures to implement the superiority of the constitution, that is judicial processes by which the compatibility of rules of law and of other legal acts with the constitution can be tested, and incompatible rules or legal acts can be declared inapplicable or invalid.’ 84 ibid. 85 ibid.

Defining and Classifying Constitutional Systems  329 understand their authority. Unlike Raz’s, my definition is intended in part for drawing borderlines between constitutional and non-constitutional governments, and hence is more minimal. IV.  TYPES OF CONSTITUTIONAL AND NON-CONSTITUTIONAL GOVERNMENT

The essential components of my definition of a constitution are formal and functional: rules and recognised principles and conventions that establish and govern the main political institutions and provide for orderly and regular transitions of power. This might be thought too broad to be capable of drawing a line between constitutional and non-constitutional governments. However, this section uses the definition to draw such distinctions, and it classifies three types of non-constitutional government: absolute monarchy, dictatorship, and the single-party state. A.  Types of Non-constitutional Government Absolute monarchy has been a prevalent form of government in history, perhaps the most common type of government before the modern era. This is a system in which a monarch (usually hereditary) governs a realm, especially in the exercise of executive and law-making powers, without significant restraint in the form of legal rules, conventions, or oversight via a Parliament or other council. In constitutional monarchy, the monarch governs in the presence of said restraints, for example in eighteenth-century Britain, often under written constitutional rules. Absolute monarchy does not imply the absence of sources of competing power, or the ability to rule with a totally free hand. Kings have typically had to vie for control with the nobility, who could control access to armies and funds, and their perceived legitimacy to rule has historically depended, in part, on the support of religious authorities. In medieval Europe, popes could wield a significant influence over kings and even, on occasion, depose them.86 Often a king would freely adopt, or accept under pressure, the kinds of constraints that amount to a constitutional monarchy. By the twentieth century, Britain had evolved into a parliamentary system of government where the monarch is head of state but does not exercise day-to-day executive power. There is no clear line between the three categories of absolute monarchy, constitutional monarchy, and parliamentary government with a royal head of state. A given country might rest on the borderline; another might be evolving, by imperceptible degree, in one direction or another. Given the decline in absolute and constitutional monarchy today, however, we need not belabour matters. The distinctions are sufficient for offering a generally stable classification of pertinent countries (see Table 1).

86 See B Bueno de Mesquita, ‘Popes, Kings, and Endogenous Institutions: The Concordat of Worms and the Origins of Sovereignty’ [2000] International Studies Review 93–118.

330  Paul Yowell The table lists the commonly recognised absolute monarchies of Saudi Arabia, Oman, Brunei, Eswatini, and the Vatican. Except for the last, all have hereditary rules of succession, though none are simple rules of primogeniture. What chiefly qualifies a system as an absolute monarchy is the concentration of executive and legislative powers in the personal authority of the monarch. Hence, it is an instance of personal rule. Each of the countries above have a written constitution or basic law, which describes a system in which the monarch cooperates in the exercise of legislative or executive powers with councils or ministers. But the constitutions do not purport to limit monarchical powers.87 Insofar as executive or legislative powers are delegated to or shared with others, they are exercised by persons either appointed by the monarch or answerable to him. Moreover, there is nothing to prevent the monarch from changing the text of the constitution at will. It should be noted, however, that an absolute monarchy may observe a degree of operational judicial independence. In Brunei, for example, there is a working common law system inherited from its days as a British Protectorate, and retired judges from British, Hong Kong and other Commonwealth systems sit on the Court of Appeal.88 The second type of non-constitutional system is dictatorship. Some broad definitions of dictatorship include absolute monarchies and single-party states, or any system that is not a functioning democracy with regular, fair elections. Other definitions characterise a dictatorship by its authoritarian or repressive measures. I use a narrower definition. A dictatorship is a system of government in which the leader exercises executive and legislative power largely or entirely in the absence of constitutional rules and conventions, and is not effectively subject to rules governing transition of power. Many dictators today keep up an appearance of democracy by holding occasional elections, but they are sham elections in which the result is determined in advance by various means.89 These can include a readiness to resort to cheating to fix the results, manipulation of the process by a cabal who chooses candidates guaranteed to win, or repression of other parties or political opponents. Another form of dictatorship, often seen following a military coup, is when the leader suspends elections indefinitely but intimates their future return. A third form is a president-for-life who expresses an intent to govern without rules regarding transitions of power. Dictatorship, like absolute monarchy, is a form of personal rule – usually of one person, though sometimes by a small group (eg a junta following a coup). Dictatorship is distinguished from absolute monarchy by its lack of established, recognised rules of transition, which tend to make the former relatively stable. Absolute monarchies share a feature of constitutional systems, namely, that the people show an allegiance that transcends day-to-day politics. The allegiance is to a royal house or dynasty,

87 See, eg, Constitution of the Vatican, s 1: ‘The Supreme Pontiff, Sovereign of Vatican City State, has the fullness of legislative, executive and judicial powers.’ 88 See A Black, ‘Judicial Independence, Impartiality and Integrity in Brunei Darussalam’ in HP Lee and Marilyn Pittard (eds), Asia-Pacific Judiciaries: Independence, Impartiality and Integrity (Cambridge, Cambridge University Press, 2017). 89 See J Gandhi and E Lust-Okar, ‘Elections Under Authoritarianism’ [2009] Annual Review of Political Science 402–33.

Defining and Classifying Constitutional Systems  331 rather than to a constitution. Dictators may be popular as individuals, but dictatorships lack such transcendent allegiance. Dictators often operate in the shadow of a large-C constitution. They may follow it to some degree, to win perceived legitimacy, but they face no significant restraint from its rules.90 Dictators may claim the need to take unconstitutional measures temporarily to restore the constitution – and occasionally such a promise comes true. My definition of dictatorship is formal and thus consistent with claims made by some scholars that, in extraordinary circumstances, a country might need a ‘benevolent’ dictator.91 I distinguish the personal rule of absolute monarchs and dictators from the party rule that obtains in a single-party state. In the latter, society exhibits an allegiance transcending day-to-day politics not to a particular leader or to a constitution but to a political party. In my definition, a single-party state is one in which only one party enjoys legal legitimacy. I distinguish this from dominant-party systems, where one party handily wins consecutive elections, but other parties are allowed to compete, at least nominally. If the competition is a sham, or if potential competition is suppressed by illegal means, then the system is a dictatorship. However, some dominant parties win elections because they are genuinely more popular and effective than alternatives. The single-party state is associated particularly with the Communist Party. In Table 1, five out of six of the single-party states are governed by Communist or closely-linked parties – China, North Korea, Laos, Vietnam, and Cuba – and the sixth, Eritrea, by a successor92 to a Marxist-Leninist party. The state and law will wither away, according to the Marxist-Leninist theory of history.93 For the Communist Party, the political and legal institutions of ‘bourgeois’, capitalist society can be used in the transition period toward fully realised communism.94 The party also used constitutions, as Loewenstein observed in 1969, and by then the end point could be taken for a fantasy: Victorious Socialism was bound to use the machinery of the bourgeois-liberal state it had dislodged. Elections, parliaments, courts, the administrative machinery, and, with them, the constitution would have to continue for an unforeseeable time, the withering-away of the state being an Utopian fiction, never entertained in practice.95

The principal functional feature of the Communist constitution was assembly government. The 1936 Constitution of the Union of Socialist Soviet Republics (USSR) provided in Article 20: ‘The highest organ of state power in the USSR is the Supreme Soviet of the USSR’. Article 84 provided that the Council of Ministers of the USSR was the ‘highest executive and administrative organ of the state power’, while Article 65 made it responsible and accountable to the Supreme Soviet.96 90 See B Geddes, How Dictatorships Work: Power, Personalization, and Collapse (Cambridge, Cambridge University Press, 2018). 91 See C Rossiter, Constitutional Dictatorship: Crisis Government in the Modern Democracies (Princeton NJ, Princeton University Press, 1948). 92 The People’s Front for Democracy and Justice. 93 See A Lukina, ‘Opening the Pandora’s Box: Kelsen and the Communist theory of law’ (2020) 30 Jurisprudence 551, 549. 94 ibid. 95 Loewenstein, ‘Constitutions and Constitutional Law’ (n 18) 233. 96 ibid.

332  Paul Yowell Such assembly government had a precedent in the revolutionary 1793 French Constitution – drafted but never enacted – which provided for a supreme National Assembly and subordinate executive council. The Supreme Soviet, though it purported to represent the people, was not a locus of popular decision-making. Rather, as Loewenstein observed, ‘it is the nonconstitutional agency of the Communist Party and its subdivisions that dominate the allegedly omnipotent assembly’, because the constitution assigns nomination of candidates to the Party.97 ‘Behind the democratic façade of the assembly government, the Communist autocracy, represented by the Central Committee of the Communist Party … has solidly entrenched itself as the sovereign policy-making and policy-executing organ’.98 In Communist single-party states, the form of elections is often observed, and Marxist-Leninist theory claims to practice a truer democracy in comparison to the party politics of the West: a choice among candidates committed to the ideals of the people is better than a choice between self-interested parties inherently corrupted by capitalism. In this view of democracy there can be no competition among political philosophies or ideals. There is a conflation of the nation or community with the Communist Party, as in the Soviet anthem: ‘Be glorious, our free Fatherland / A reliable stronghold of the peoples’ friendship! / Party of Lenin, strength of the people, / Leads us to the triumph of communism!’ In Soviet constitutional theory, assembly government is subsumed under the Party: ‘the parliament as the supreme state organ is not subject to any control by, or intervention of, any other power-holder’.99 No judicial review of legislation is permissible, and no separation of powers can be recognised: ‘for the sovereign assembly to share power with any other organ would be a contradiction of terms’.100 The Soviet model of constitutionalism was generally followed closely by its satellite states in Europe. The collapse of the Soviet empire led to the formation of about 20 new constitutional governments, leaving only the short list of single-party states above. China, far from collapsing as many expected in last decade of the twentieth century, has, economically, succeeded where the Soviet Union failed, by adapting to the global market economy and building the world’s dominant manufacturing sector. In the early 2000s, free trade with China was promoted on the ground that economic liberalisation would lead inevitably to political liberalisation. But in this century China’s economic success has strengthened the Communist Party. As Quianfan Zhang, a leading English-speaking Chinese constitutional scholar, has said, China’s ‘increasing economic powers has reduced the weight of international pressure that can be used to force the improvement of domestic human rights’.101 He describes China as ‘with a constitution but without constitutionalism’ and ‘without constitutional mechanisms to keep its powers in check’.102



97 ibid. 98 ibid. 99 ibid.

100 ibid. 101 Q

Zhang, The Constitution of China: A Contextual Analysis (Oxford, Hart Publishing, 2012) 261.

102 ibid.

Defining and Classifying Constitutional Systems  333 Zhang’s comparison of the text of the Chinese Constitutional to institutional reality shows that nearly all that Loewenstein said of Soviet constitutional theory can be extended to China. The 1982 Constitution, Article 2, provides for a form of assembly government: ‘All power in the People’s Republic of China belongs to the people. The National People’s Congress and the Local People’s Congresses at various levels are the organs through which the people exercise their power.’ In reality, however, the Congresses are ‘rubber stamps’, Zhang says, ‘since their ordinary roles have been reduced to endorsing official acts decided by the Chinese Communist Party (CCP), the real holder of supreme power in China’.103 The power structure of the Party ‘completely subverts the power structure defined in the Constitution, which as far as possible has been interpreted to conform to the dictates of the Party and, when that is not possible, is simply ignored’.104 The 2018 amendment of the Constitution of 1982 to allow President Xi Jinping to serve beyond the previously set 10-year term illustrates the latter point. Much more could be said about China and constitutional reform efforts,105 the arguable rise of a legal culture and beginnings of an independent judiciary,106 and views of scholars that differ somewhat from those of Zhang. Larry Backer, for example, argues that China’s system could be seen as a constitutional with some observance of the Rule of Law and separation of powers, if one includes the governing documents of the Chinese Communist Party, calling this ‘state-party constitutionalism’.107 But he acknowledges that the CCP is entrenched as the party in power, and that political citizenship is limited to its members.108 For present purposes, the lesson of modern China is that non-constitutional government, in the form of a Communist singleparty state, remains a force to be reckoned with. B.  Types of Constitutional Government Since my definition of a constitutional system focuses on the rules that establish political institutions and govern the exercise of power, it lends itself to classifying constitutions according to matters such as the appointment and removal of the executive, and the relationship between prime ministers, presidents, and (constitutional) monarchs. Classifications of this kind have in recent decades been pursued more in the fields of political science and comparative politics than in constitutional law. Loewenstein, speaking of the ‘differentiation between the monarchical and the republican form of state or between the parliamentary and the non-parliamentary

103 ibid 122. 104 ibid. 105 See ibid, ch 9. 106 See ibid, ch 5. 107 L Backer, ‘The Party as Polity, The Communist Party, and the Chinese Constitutional State: A Theory of State-Party Constitutionalism’ [2009] Journal of Chinese and Comparative Law 101–169. 108 ibid.

334  Paul Yowell institutionalization of the power process’, dismissed it as ‘antiquated’.109 It referred, he said, to the ‘form of state or the pattern of government embodied by a constitution rather than to the constitution proper’.110 Albert Dicey, the leading scholar of British constitutional law in the early twentieth century, took a different view. In his lectures on comparative constitutional law, he analysed the differences between parliamentary and presidential government at length and saw the choice between them as having profound consequences.111 Walter Bagehot had taken a similar approach in The English Constitution (1867), which argues that the parliamentary system has important advantages over the presidential system of the US.112 Today, the detailed analysis of such differences is carried out largely in faculties of politics. Recent studies have responded to the variety of modern constitutions by proposing classifications with several sub-divisions, including semi-presidential and semi-parliamentary. In this section and Table 1, I use a simpler division of parliamentary and presidential as the two main ways for structuring the executive power. My focus is on the executive, because in most countries that is the branch where the most power is concentrated.113 The great majority of constitutional systems can be classified as presidential or parliamentary. The divisions are geographically concentrated, as can be seen in Table 1. With minor exceptions, all countries in Central and South America have adopted a presidential system based on the US model, as have a large majority of African nations. In Europe, though a few countries have adopted a presidential system, most have a form of parliamentary democracy broadly like that which evolved in Britain in the seventeenth and eighteenth centuries and reached its current form in the nineteenth. This model has also been adopted in many former British colonies, including Canada, Australia, New Zealand, and India. In additional to parliamentary and presidential systems, there remain today several constitutional monarchies in the sense defined above, mostly in smaller countries but also in a few larger ones. Table 1 lists Monaco, Liechtenstein, Jordan, Morocco, Bhutan, Thailand, Samoa, Kuwait, United Arab Emirates, Bahrain, and Qatar. In these countries the monarch exercises executive powers, but under constraints previously discussed. Finally, there are few countries with systems that are sui generis or difficult to classify, including three that have a plural executive: Switzerland, Bosnia and Herzogovina, and the city-state of San Marino. The Swiss Constitution is notable for its diminished executive power. It establishes an assembly government, which in formal terms is similar to the (unenacted) French Constitution of 1793 and the Soviet Constitution of 1936 in making the legislature the highest power, though in operational terms very different due to the

109 Loewenstein, ‘Constitutions and Constitutional Law’ (n 18). 110 ibid. 111 See AV Dicey, Comparative Constitutionalism (JWF Allison ed, Oxford, Oxford University Press, 2013) 137–50. 112 W Bagehot, The English Constitution (Paul Smith ed, Cambridge, Cambridge University Press, 2001) 18–9. 113 See E Posner and A Vermeule, The Executive Unbound: After the Madisonian Republic (Oxford, Oxford University Press, 2011) ch 2.

Defining and Classifying Constitutional Systems  335 system of multi-party democracy. The executive consists in a seven-person council which oversees the executive branch; it is chaired by a President who has an equal vote with the others. The council has no legislative veto. Neither the President, nor anyone else, functions as head of state in the usual sense. In a system of parliamentary government, the day-to-day executive power is exercised primarily by a prime minister (i) whose appointment to office arises from his or her party winning a majority of parliamentary seats or otherwise gaining the confidence of parliament (eg in a coalition government) and (ii) whose continued service depends on maintaining the confidence of a parliament that has the power to dismiss the prime minister (and his or her government) through a vote of no-confidence or similar means.114 Usually, a prime minister has a correlative power to dissolve parliament and call new elections. A dynamic tension exists between the two bases of power – the parliament and the office of the prime minister – each of which can dismiss the other. The prime minister is empowered through members of parliament but can easily lead them when they have a firm majority of seats and solid base of democratic support. No one has described this tension better than Bagehot, who famously said the British system is based upon a fusion rather than separation of powers – but also stressed the importance of parliament’s power to dismiss the prime minister.115 In a system of presidential government, the executive power is held by a president (i) who is directly elected by the people or indirectly through a legislature or electoral college and (ii) who holds office for a set term and cannot be dismissed by the legislative branch. The president is also the head of state, whereas in a parliamentary system the head of state is separate from the office of prime minister. A presidential system is based on the separation of executive and legislative powers. Thus, it is possible for the presidency and the legislative branch – or one of its houses in a bicameral system – to be under the control of opposing parties. This will depend, however, on the rules of the electoral system, the timing of elections, and the party system. In France, for example, elections to the National Assembly are held after the presidential election and usually deliver a comfortable majority for the president’s party. Thus, the French President often has the same leeway for advancing a legislative agenda as a British Prime Minister, with the added advantage of having a secure term in office not subject to dismissal. In most presidential systems, however, there is a degree of friction between the presidency and the legislature. In some, like the US, the friction often reaches the level of gridlock. A parliamentary system is designed to achieve coherence between the executive and legislature and a degree of legislative efficiency.116 The degree of efficiency depends on the electoral system and the organisation of political parties. The traditional British



114 See

A Bradley and C Pinelli, ‘Parliamentarism’ in Rosenfeld and Sajó (eds), Oxford Handbook (n 72). The English Constitution (n 112) 3–23. 116 Bradley and Pinelli, ‘Parliamentarism’ (n 114) 657–9. 115 Bagehot,

336  Paul Yowell system of first-past-the-post voting incentivises a two-party system, with periodic shifts between one major party or another holding a stable majority in parliament. This is a central feature of what became known as the Westminster model. In continental Europe, however, all countries with parliamentary systems have adopted some form of proportional representation. This incentivises voting for multiple parties, making it difficult for one party to gain an absolute majority in parliament. Thus, in European parliamentary systems, coalition governments are the norm. Some former Westminster systems, including Australia and New Zealand, have also opted for a proportional system. The significance of having a separate head of state in a parliamentary system should not be overlooked. As Table 1 shows, there are 23 parliamentary systems with royal heads of state (13 in the British Commonwealth), and twice that number of republics. Presidents of parliamentary republics are appointed or elected in a variety of ways, including direct popular elections and indirect election by an electoral college. While the role is sometimes thought to be merely ceremonial, especially when the head of state is royal, the head of state typically holds residual powers that are formally broad. These can include powers to dismiss or form governments and dissolve parliaments, and to veto or withhold assent from legislation. A recent study on monarchies in Europe shows that, while royal heads of state are less active than presidents, their residual powers are occasionally exercised.117 Whether by a royal or a president, such powers are normally used only in crises or to address political instability. But the exigencies of party politics are such that coalition governments can sometimes break down quickly. Elections do not always deliver decisive results, and political instability can be chronic in some places. Italy has had 66 governments since 1945. Many of these have been fragile coalitions of parties with different views, and many have been technocratic governments appointed by the president after a falling out of coalitions. The instability of parliamentary government in the French Fourth Republic led to scrapping the 1946 Constitution and adopting the 1958 Constitution, with a stronger role for the president. If chronically unstable government represents one extreme of dangers with a parliamentary system, the concentration of executive and legislative powers in the hands of a prime minster backed by a strong majority represents another. The British system presents a danger of ‘elective dictatorship’ to some critics; some argue that this is exacerbated by parliamentary sovereignty and urge greater judicial power to check Parliament, while others contend dangers could be ameliorated through proportional representation.118 As for presidential systems, there is a growing literature arguing that they have a tendency towards authoritarianism, corruption, and poor outcomes on economic measures, protection of civil liberties, and maintenance of a functioning

117 R Hazell and B Morris (eds) The Role of Monarchy in Modern Democracy: European Monarchies Compared (Oxford, Hart Publishing, 2020). 118 JA Zecca, ‘Avoiding Elective Dictatorship in the United Kingdom: Debate on Constitutional and Electoral Reform through Proportional Representation’ [1992] Hastings International and Comparative Law Review 425–60.

Defining and Classifying Constitutional Systems  337 democracy.119 In an influential and frequently cited article, Juan Linz argued in 1990 that ‘the vast majority of the stable democracies in the world today are parliamentary regimes, where executive power is generated by legislative majorities and depends on such majorities for survival’ and that, in contrast, ‘the only presidential democracy with a long history of constitutional continuity is the United States’.120 My aim is not to take a side on these debates but rather to argue that constitutional theorists should take heed of them.121 This chapter has only scratched the surface of the complex relationship between types of constitutional systems (presidential, parliamentary, or other) and voting systems. The aim has been to help constitutional theorists think beyond some of the more common constitutional devices for separating and balancing power – such as bicameralism, federalism, judicial review – and to think more deeply about how all the components of a constitutional system inter-relate. As Aristotle advised, we should study what is best in the abstract, but seek to implement what is best relative to the circumstances in which a country finds itself. The two may be very different, and one of the aims of constitutional theory should be to help society navigate between them. V. CONCLUSION

In 1989, writing shortly before the Tiananmen Square protests and the fall of the Berlin Wall, Francis Fukuyama observed ‘unmistakable changes in the intellectual climate the world’s two largest Communist countries’, and ‘the ineluctable spread of consumerist Western culture in the peasants’ markets and color television sets now omnipresent throughout China’.122 He sensed the ‘total exhaustion of viable systematic alternatives to Western liberalism’. What we may be witnessing, he said, is not just the end of the Cold War, but ‘the end point of mankind’s ideological evolution and the universalization of Western liberal democracy as the final form of human government’.123 Fukuyama’s vision of the end of history was the triumph of a particular form of constitutional government. One of the aims of this chapter has been to find the places in the world where, as it were, history has not reached its end. The result of that study, for descriptive purposes, is visible in Table 1. My aim, however, has not been to classify and catalogue for its own sake, but to draw distinctions and propose definitions that can be helpful to thinking at the most

119 See the collection of essays in A Lijphart (ed), Parliamentary versus Presidential Government (Oxford, Oxford University Press, 1994). 120 J Linz, ‘The Perils of Presidentialism’ [1990] Journal of Democracy, 51, 51–2. 121 For recent work addressing these questions see T Khaitan, ‘A Case for Moderated Parliamentarism’ (2021) 7 Canadian Journal of Comparative and Contemporary Law 81; G Webber, ‘Opposition’ in R Bellamy and J King (eds), The Cambridge Handbook of Constitutional Theory (Cambridge, Cambridge University Press, forthcoming). See also R Ekins, The Nature of Legislative Intent (Oxford, Oxford University Press, 2012) 169–79 (comparing the Westminster and Washington models). 122 F Fukuyama, ‘The End of History’ (1989) The National Interest. 123 ibid.

338  Paul Yowell general level about how state power should be exercised and how it can be effectively limited. Constitutional theorists often think of limits on power mainly in terms of legal, constitutional rights enforced by courts. But it would be a mistake to assume that this is an institution that can be grafted onto any constitutional system or transform a non-constitutional system. China is not waiting on a ‘Marbury v Madison moment’. Rights-based judicial review will likely prove ineffectual against a powerful president, and it cannot make up for the deficiencies of a parliamentary system unable to form stable and effective governments. A greater need of many constitutional systems is a better balance between executive and legislative power and a healthier party system. Finally, this chapter has sought to defend a thinly normative conception of a constitution, defined in functional rather than substantive terms, and neutral on great moral questions such as democracy and individual rights. My aim is not to reject thickly normative accounts of constitutionalism that stress substantive, moral criteria. Rather it is to expand our methodological horizons to embrace an understanding of constitutionalism with historical roots that reach back well before the post-war rise of rights-based constitutional review. When such a conception of constitutional government is employed, it requires that arguments about liberty, authoritarianism, and democracy be carried out directly and on their own terms. It is no longer viable to embrace Fukuyama’s historicism regarding the inevitable triumph of liberal democratic constitutionalism, especially after the pandemic crisis that began in 2020. China’s response has received applause in liberal democratic countries, and commentators have argued it was successful because of China’s ‘authoritarian capacity to quickly mobilize people’ – including through ‘coercive tools’ and ‘severe, mandatory mobility restrictions’ – and its ability to ‘pull together all of the resources of a one-party state’.124 We do not yet know what the histories of twenty-first century constitutionalism will say, or which forms of constitutional or non-constitutional government will thrive in the decades ahead.

124 See, eg, Sui-Lee Wee, C Buckley et al, ‘Power, patriotism and 1.4 billion people: How China beat the virus and roared back’, New York Times, 1 April 2021, at www.nytimes.com/2021/02/05/world/asia/ china-covid-economy.html.

Table 1 Constitutional Systems Parliamentary Royal head of state House of Windsor United Kingdom Canada New Zealand Australia

Belgium Denmark Luxembourg Netherlands Norway Spain Sweden Japan Malaysia Cambodia Lesotho

Belize Solomon Islands Plural executive: Switzerland Bosnia & Herzegovina San Marino

Presidential Republic

Germany, Italy, Austria, Greece, Portugal, Ireland, Hungary, Bulgaria, Croatia, Czechia, Andorra, Albania, Estonia, Finland, Iceland, Latvia, Malta, Moldova, Montenegro, North Macedonia, Serbia, Slovakia, Slovenia, Ukraine Armenia, Georgia, Kyrgyzstan, Iraq, Israel, Lebanon India, Pakistan. Bangladesh, Timor-Leste, Nepal Tunisia, Botswana*, South Africa* Ethiopia

Monaco Liechtenstein

France, Russia, Poland Cyprus, Lithuania, Romania

Jordan

China Laos Vietnam N. Korea

Turkey, Iran*, Kazakhstan, Azerbaijan

Morocco

Eritrea

Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Panama, Paraguay, Peru, Uruguay, Dominican Republic, Haiti

Bhutan Thailand

Cuba

Angola, Benin, Burkina Faso, DR Congo, Gabon, Gambia, Ghana, Guinea, Guyana, Ivory Coast, Kenya, Liberia, Madagascar, Maldives, Malawi, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, Suriname*, South Sudan, Tanzania, Togo, Uganda, Zimbabwe

Kuwait UAE Bahrain Qatar

United States

Singapore, Papua New Guinea, Fiji, Indonesia, Mongolia, Philippines, Sri Lanka, Marshall Islands, Nauru, Tuvalu, South Korea, Taiwan Vanuatu Dominica Trinidad & Tobago

Single-party state

Palau, Cape Verde, Comoros, Sao Tome & Principe, Seychelles Micronesia*, Kiribati*

Samoa

Dictatorship Afghanistan Belarus Uzbekistan Tajikistan Turkmenistan

Absolute Monarchy Saudi Arabia Oman Eswatini

Algeria Brunei Burundi Chad The Cameroon Vatican Congo Djibouti Eq. Guinea Egypt Lybia Sudan Mali Guinea-Bissau Nicaragua Myanmar Venezuela Syria

Explanatory note. The classifications in this table are made by the author according to the definitions stipulated in section IV and in light of a variety of sources. An asterisk indicates that a country has a functional feature that differs from others in its category (e.g. in South Africa and Botswana the Prime Minister serves as head of state unlike other parliamentary systems with a separate president). The category of presidential republic includes ‘semi-presidential’ systems, wherein a president shares some power with a prime minister, and other variations on the historical US model, such as systems where the president is indirectly elected. Decisions to classify a country as a dictatorship rely in part on rankings in the Democracy Index published by The Economist Intelligence Unit and the Election Integrity Project at the University of Sydney, insofar as those rankings reflect functional rather than substantive criteria as defined in this chapter, and in part on other sources. The definition of dictatorship in this chapter is relatively narrow and is applied so as to err on the side of a smaller list. In many countries the relevant circumstances are in a state of flux.

Defining and Classifying Constitutional Systems  339

Antigua & Barbuda Bahamas Barbados Grenada Jamaica St Kitts & Nevis St Lucia St Vincent & the Grenadines

Other

Non-constitutional Systems Constitutional Monarchy

340

13 The View from Nowhere in Constitutional Theory: A Methodological Inquiry SILVIA SUTEU

I. INTRODUCTION

T

his chapter begins from the premise that constitutional theory is still on a quest of self-definition as a field of inquiry and consequently is ripe for methodological reassessment. The chapter explores the methodological ­ assumptions – mostly unstated but widely shared – underpinning much constitutional theoretical work by focusing on the role of comparison in constitutional theorising. Insofar as it finds itself at the intersection of, or in close conversation with, other disciplines such as political and legal theory, constitutional theory has developed hybrid (or, less generously, hodgepodge) methodological tools. It often finds it difficult to occupy the space between ideal and non-ideal theory. It may employ tools such as analogy and extrapolation, but frequently decontextualises or generalises from cherry picked or ‘usual suspect’ case studies. This chapter seeks to mine the rich methodological advances in cognate fields such as comparative constitutional law and comparative politics for insights on how constitutional theory can embrace contextualism and robust comparison. Our aim should be to avoid a ‘view from nowhere’ approach in constitutional theory that mistakes for objectivity what are ultimately contextual impoverishment and comparative blindness. The inspiration for the chapter’s title, of course, comes from Thomas Nagel’s eponymous work.1 His insight that we are constantly grappling with the reality of interconnected subjectivities and objectivity in the world and that these sometimes will be irreconcilable has farther-reaching consequences for scholarly inquiry than will be explored here. For my purposes, most relevant will be Nagel’s insight that recognising our own subjectivity is directly connected to recognising our possibility of detaching from it. In other words, recognising that we as scholars are



1 T

Nagel, The View from Nowhere (Oxford, Oxford University Press, 1986).

342  Silvia Suteu always contextually situated and individually subjective does not preclude us from engaging in robust scholarship – quite on the contrary, it is a prerequisite for it, especially when drawing inferences we assume have wide application. I would not overemphasise the inspiration drawn from Nagel’s work, however. I use it, instead, as a reminder to resist lawyerly ‘amateurish … pseudo-philosophical efforts’ as Mark Tushnet once put it.2 In other words, as a call for self-reflection and self-awareness when doing constitutional theory, aided by a comparative outlook. The chapter proceeds by, first, discussing the comparative turn in constitutional theory and its potential for correcting old assumptions and revealing new insights. I illustrate these benefits by revisiting two well-trodden debates in constitutional theory and comparative constitutional law, one around the legitimacy of constitutional review and the other around citations of comparative material in constitutional adjudication. The chapter then turns to investigating recent advances in comparative constitutional change literature, in particular new work on constituent power and on unconstitutional constitutional amendments. These are used as examples of the type of interdisciplinary, comparative, and also deeply contextual scholarship that has the capacity to shed new light on old core questions of constitutional theory. I ultimately argue that this is demanding work, but that much more rewarding for it. II.  THE COMPARATIVE TURN IN CONSTITUTIONAL THEORY

The type of engagement with comparative work I am interested in is not entirely alien to constitutional theorising. Recent work has indeed proceeded precisely on the assumption that there is scope for innovative thinking by approaching questions of constitutional theory, old and new, from a comparative vantage point.3 The authors of a recent collective volume on the topic note their own precursors among Weimar-era constitutional theorists grappling with questions of democratic institution-building in comparative historical perspective.4 They also remark on a familiar conundrum in comparative work in constitutional and human rights studies: the tension between the universal and the particular. The question is whether we are dealing with sufficiently similar constitutional questions that necessitate common solutions, or whether differences in culture and history preclude such universalising.5 Their answer is that this is a false paradox and the tension can be reconciled, while constitutional principles can be divided between those that remain polity-specific and others that ‘can make a claim of universality’ but nevertheless require local adaptation.6

2 M Tushnet, ‘Constitutional Scholarship: What Next?’ (1988) 5 Constitutional Commentary 28, 31. 3 GJ Jacobsohn and M Schor (eds), Comparative Constitutional Theory (Cheltenham, Edward Elgar, 2018). 4 ibid 4–5. 5 ibid 5. See also C McCrudden, ‘A Common Law of Human Rights?: Transnational Judicial Conversations on Constitutional Rights’ (2000) 20(4) Oxford Journal of Legal Studies 499 and V Jackson, ‘Comparative Constitutional Law: Methodologies’ in M Rosenfeld and A Sajo (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012), 54–74. 6 Jacobsohn and Schor, Comparative Constitutional Theory (n 3), 6–7.

The View from Nowhere in Constitutional Theory  343 The volume in question is a welcome scholarly addition that sees comparative constitutional studies and constitutional theory as a unified field of study, or at least as two bodies of literature that are in close correspondence.7 Its chapters cover important constitutional theoretical topics that can only be enriched by comparative engagement, including the separation of powers, federalism, and constitutional amendment. Valuable though such collective work is, it suffers from two limitations, one of design and one of scope. The latter is the usual problem of edited collections on comparative themes being unavoidably limited in their scope, in terms of the number both of substantive topics and of jurisdictions covered. An objection that comparative constitutional scholarship is not exhaustive would be disingenuous, as authors themselves readily acknowledge the limits inherent in their projects. Choices must be made, of course, and topics and themes cut for principled or pragmatic reasons. The range of jurisdictions covered, however, risks skewing results (to employ a more scientific terminology) in a way that should concern us more. This is where the research design-based objection comes in. Insofar as we extrapolate, on the basis of a limited number of jurisdictions, to infer that our constitutional theorising is sound, we had better make sure that those jurisdictions are representative of the questions tested and are robustly comparable. We should also make efforts to ensure no serious omissions remain that could call into question our findings. A.  The Nature and Legitimacy of Constitutional Review Two examples illustrate these problems and constitutional scholarship’s repeated attempts to grapple with them. The first is the much-explored debate in constitutional theory surrounding the nature and proper scope of judicial review in a constitutional democracy. The Dworkin-Waldron disagreement,8 with its implications for our understanding of rights, judicial interpretation, and the judicial function, has since been re-evaluated, reconciled, and restated by constitutional theorists.9 Some of the most valuable contributions to that debate have been the ones identifying comparative gaps and faulty assumptions about the universality of the strong versus weak judicial review debate. For example, work on transitional democracies has revealed that, for institutional capacity and political stability reasons, the choice

7 On comparative constitutional studies as itself an interdisciplinary body of literature bridging comparative constitutional law and comparative politics, see R Hirschl, Comparative Matters: The Renaissance of Comparative Constitutional Law (Oxford, Oxford University Press, 2014). See also G Frankenberg, Comparative Constitutional Studies: Between Magic and Deceit (Cheltenham, Edward Elgar, 2018). 8 See, inter alia, R Dworkin, Freedom’s Law: The Moral Reading of the American Constitution (Cambridge, Mass, Harvard University Press, 1996); J Waldron, Law and Disagreement (Oxford, Oxford University Press, 2002). 9 See, inter alia, A Kavanagh, ‘Participation and Judicial Review: A Reply to Jeremy Waldron’ (2003) 22(5) Law and Philosophy 451–486; W Sinnott-Armstrong, ‘Weak and Strong Judicial Review’ (2003) 22 Law and Philosophy 381–392; D Kyritsis, Where Our Protection Lies: Separation of Powers and Constitutional Review (Oxford, Oxford University Press, 2017); A Latham-Gambi, ‘Political Constitutionalism and Legal Constitutionalism – an Imaginary Opposition?’ (2020) 40(4) Oxford Journal of Legal Studies 737–763.

344  Silvia Suteu of constitutional review model is at best secondary and at worst counter-productive in many jurisdictions.10 Waldron’s conditions for trusting the legislative process to guard rights, for example, will be of limited value to the many constitutional systems plagued by electoral instability, cronyism in legislative seat allocation, and weak Parliaments. Equally, Dworkin’s trust in the judge to dispense her constitutional role with integrity and fairness will seem misguided to one in a constitutional system where the judiciary has never enjoyed the perception of such integrity, whether due to political co-optation or, more mundanely, because its judges do not actually engage in the type of discursive, reason-giving, non-formalist judging Dworkin assumes them to. The point here is not simply to show that neither of their models fits the whole spectrum of possible case studies. It is instead to highlight that their operating assumptions or stipulated conditions were, from the onset, ignorant of the vast number of jurisdictions that operated under different constraints and along different parameters. At the very least, this awareness would have qualified and narrowed down the purported general import of both approaches. Nor does this comparative awareness need to highlight only the pathological and marginal. Take comparative work on the so-called ‘weak’, ‘dialogic’, or ‘Commonwealth’ model of judicial review. This work has shown the complex operation and benefits of systems without judicial strike-down of legislation.11 Indeed, the same rich contextualism has revealed the subtle differences between these three labels,12 as well as challenged the assumption that the model is confined to the common law world.13 This scholarship has reminded us that foundational constitutional concepts such as ‘constitution’, ‘basic law’, or ‘constitutional law’ carry very different meanings across jurisdictional boundaries. We may think we speak of similar legal institutions, in other words, but we should know better. It has also challenged the too easily accepted dominance of the strong judicial review model, which had been assumed – especially during the post-1989 wave of new constitution-making – to be part of a sort of ‘end of history’ set package of liberal constitutionalism.14

10 For an argument that this should result in constitutional designers reorienting their attention to mechanisms to guarantee judicial independence in such contexts, with an accompanying weaker model of constitutional review to avoid destabilising constitutional conflicts, see S Gardbaum, ‘Are Strong Constitutional Courts Always a Good Thing for New Democracies?’ (2015) 53 Columbia Journal of Transnational Law 285–320. 11 S Gardbaum, The New Commonwealth Model of Constitutionalism: Theory and Practice (Cambridge, Cambridge University Press, 2013). 12 A Kavanagh, ‘What’s So Weak about ‘Weak Form Review? The Case of the UK Human Rights Act 1998’ (2015) 13(4) International Journal of Constitutional Law 1008–1039; R Gargarella, ‘‘We the People’ Outside of the Constitution: The Dialogic Model of Constitutionalism and the System of Checks and Balances’ (2014) 67(1) Current Legal Problems 1–47. 13 J Husa, ‘Guarding the Constitutionality of Laws in the Nordic Countries: A Comparative Perspective’, (2000) 48 American Journal of Comparative Law 345–381; T Gyorfi, Against the New Constitutionalism (Cheltenham, Edward Elgar, 2016). 14 For a challenge to this, asking whether traditions of political constitutionalism were ignored and then sidestepped during constitution-making in Central and Eastern Europe, see P Blokker, New Democracies in Crisis? A Comparative Constitutional Study of the Czech Republic, Hungary, Poland, Romania and Slovakia (London, Routledge, 2013).

The View from Nowhere in Constitutional Theory  345 At the very least, these reassessments of the debate surrounding the legitimacy of judicial review show that earlier contributors to it would have benefitted from deeper comparative knowledge. This would have allowed them to consider the applicability and limits of their respective theories beyond the narrow common law world (and really, beyond a narrow sub-set of jurisdictions even within that world). Waldron’s belief in Parliaments as the quintessential arena for democratic lawmaking would have seemed immediately suspect with more familiarity with countries where majoritarianism has led to exclusion, abuse, and corruption. Dworkin’s belief in judges as moral agents might equally have appeared alien to a scholar aware of the numerous societies where legal education, legal inheritance (for example, authoritarian), and even simply legal culture and tradition (for example, a commitment to legal formalism) have created a judge figure very different from that he presents. It might be argued that theirs are theories that presuppose certain conditions, that they rest on certain assumptions about the inner workings of the legal system that need to be met before the theories hold. However, from the point of view of constitutional theory as a midway theory between jurisprudence and applied constitutional law, this would be problematic: if, applying their criteria, we find that their theories fit only the narrowest of cases, the theory itself may need to change. Moreover, even their implicit case studies may not meet these criteria, if they ever did. We could indeed ask ourselves how Dworkin would assess the extreme politicisation of judicial appointments in the US and its impact on the legitimacy of the Supreme Court there.15 Or how Waldron would assess the erosion of political accountability mechanisms so crucial in ensuring the operation of the UK’s political constitution.16 What might also be argued, in broader terms, is that their intention was closer to ideal theory and as such that it would be disingenuous to attack them for failing to capture the full complexity of comparative experience in the world. This possible retort reveals the different understandings of the aims of constitutional theory embedded in scholarship. We could envision the field as bringing together a spectrum of work, some indeed closer to ideal theory such as political or moral theory and other more interested in applied questions of constitutional design and constitutionbuilding.17 The problem remains that this is not a neat distinction, and insofar as we are seeking answers to concrete constitutional problems, whatever our starting point, our solutions need to be at least plausibly workable in the real world. This is what differentiates constitutional theory from other forms of theorising. Its existence in the liminal space between political theory and constitutional practice means it is harder to define its contours. But we somehow understand that a good work of constitutional theory might be equally attractive to the theorist looking for conceptual refinement

15 He had expressed deep concern over the growing politicisation of judicial appointments and the ­possibility of young ideologues being appointed to the Supreme Court and tilting the ideological balance on its bench. See, R Dworkin, Is Democracy Possible Here?: Principles for a New Political Debate (New Jersey, Princeton University Press, 2006), 158. 16 For an overview of these, see, eg, ‘The Political Constitution at 40’ special issue in (2019) 30(1) King’s Law Journal. 17 For examples of the latter, see T Ginsburg (ed), Comparative Constitutional Design (Cambridge, Cambridge University Press, 2012).

346  Silvia Suteu as to the practitioner looking for deeper analysis, systematisation, and critique of real world problems she has encountered in her practice. B.  The Use of Comparative Material in Constitutional Adjudication A second example comes from the rich scholarship on the use of comparative material in adjudication.18 Here too the terminology employed will itself reveal much about the jurisdictional milieu: it is no coincidence that American constitutional scholarship refers to it as the use of ‘foreign law’, with ‘foreign’ citations timidly embraced and then quickly renounced following backlash to the comparative references in the case of Lawrence v Texas.19 Nor is American constitutional law alone in this. ‘Foreign jurisprudence’ citations are also viewed sceptically in UK human rights adjudication, for example, which reflects a widespread assumption that the common law’s long history of recognition of civil liberties is tantamount to a certain epistemological superiority in all rights matters.20 The label is not determinative of a given constitutional system’s openness to comparative material – section 39(1)(c) of the South African Constitution indicates that courts ‘may consider foreign law’ when interpreting the Bill of Rights, as indeed happened quite extensively in the country’s early post-apartheid years.21 What this example also reveals is the importance not just of broad comparative awareness, but also of a theory of comparison itself.22 The appeal of looking for judicial answers comparatively is great, whether it fully embraces a universalist perspective on constitutional questions or is simply an indirect source of judicial inspiration.23 Accompanied by jurisdictional openness, the potential for innovation and radical thinking increases. For example, it has been argued that the Delhi High Court’s incorporation of comparative citations in its decision decriminalising homosexuality

18 T Groppi and M-C Ponthoreau (eds), The Use of Foreign Precedents by Constitutional Judges (Oxford, Hart Publishing, 2013). 19 M Tushnet, ‘When Is Knowing Less Better than Knowing More? Unpacking the Controversy over Supreme Court Reference to Non-U.S. Law’, (2006) 90 Minnesota Law Review 1275; M Rosenfeld, ‘Comparative Constitutional Analysis in United States Adjudication and Scholarship’ in M Rosenfeld and A Sajo (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford, Oxford University Press, 2012) 38. 20 For an overview of such foreign citations in UK human rights adjudication, see H Tyrrell, Human Rights in the UK and the Influence of Foreign Jurisprudence (Oxford, Hart Publishing, 2018). 21 C Rautenbach, ‘The Influence of Foreign Judgments on the Development of Post-apartheid Constitutional Law in South Africa: Judicial Law-making in Action?’ (2020) 7(1) Journal of International and Comparative Law 99; U Bentele, ‘Mining for Gold: The Constitutional Court of South Africa’s Experience with Comparative Constitutional Law’ (2009) 37(2) Georgia Journal of International and Comparative Law 219; DM Davis, ‘Constitutional Borrowing: The Influence of Legal Culture and Local History in the Reconstitution of Comparative Influence: The South African Experience’ (2003) 1(2) International Journal of Constitutional Law 181. 22 For an attempt to put forward such a theory in the American context, see R Dixon, ‘A Democratic Theory of Constitutional Comparison’ (2008) 56 American Journal of Comparative Law 947. 23 M Tushnet, ‘The Possibilities of Comparative Constitutional Law’ (1999) 108(6) Yale Law Journal 1225; S Issacharoff, ‘Comparative Constitutional Law as a Window on Democratic Institutions’, New York University Public Law and Legal Theory Working Papers, Working Paper No. 17-14, May 2017, http:// lsr.nellco.org/nyu_plltwp/582/.

The View from Nowhere in Constitutional Theory  347 was a positive step, allowing it not only to better ground its doctrinal solution but also to increase the moral legitimacy of its conclusion.24 By looking to comparative doctrinal developments in places like Nepal and Fiji, therefore, the Court revealed the possibilities of ‘an inclusive approach towards constitutional comparison [that] can be both useful and responsible’, one that ‘can allow for our work to be not merely comparative but truly global’.25 This comparative enthusiasm risks overlooking potential pitfalls in the comparative turn, especially in adjudication. First, comparative citations can be misused, such as by being selective and acontextual – a point often raised by judges and scholars resisting reliance on comparative material.26 These qualms are valid, however, even while they leave open the question of whether selection bias and insufficient attention to context can be overcome. Robust comparative methodology is specifically designed to help prevent such misuse. Rejecting comparison wholesale because methodological rigour is needed when engaging in it means renouncing its benefits because doing the comparison well is difficult. Second, the politics of comparison in constitutional law matter. Methodologically, this takes us back to the problem of shoddy research design. Insofar as we rely on the usual suspects in our comparative studies, or simply selecting comparators whose legal tradition or language we may be familiar with, we are inadvertently importing our biases and epistemic limitations into our comparative research design. Even more worryingly, it leads us to leave unchallenged disciplinary assumptions and frameworks of analysis that perpetuate the same elisions and injustices. In one influential formulation, the endurance of this western-centric political imagination amounts to a form of ‘epistemicide’.27 Or, to quote a comparative constitutional law scholar and theorist having grappled with the difficulties of bridging the gap between the two fields: Preoccupied with judicial review and, with time, methodological questions about the migration of constitutional ideas comparative constitutionalism has systematically neglected issues such as poverty, imperialism, or socioeconomic emancipation which have otherwise been high on TWAIL’s agenda. Though a number of scholars associated with the comparative constitutionalist movement contributed systematic and innovative theoretical reflections on the nature and functions of constitutional government, none of them challenged the liberal-democratic ‘software’ of Western constitutionalism.28

It is not just that the dominant paradigm remains unchallenged. It is also that these same assumptions inform the work of constitutional designers and their (often international) advisers. The ‘significant methodological limitations’ of a young field

24 M Khosla, ‘Inclusive Constitutional Comparison: Reflections on India’s Sodomy Decision’ (2011) 59(4) American Journal of Comparative Law 909. 25 ibid 934. 26 C Saunders, ‘The Use and Misuse of Comparative Constitutional Law’ (2006) 13(1) Indiana Journal of Global Legal Studies 37. 27 B de Sousa Santos, Epistemologies of the South: Justice Against Epistemicide (London, Paradigm Publishers, 2014). 28 Z Oklopcic, ‘The South of Western Constitutionalism: A Map Ahead of a Journey’ (2016) 37(11) Third World Quarterly 2080, 2082.

348  Silvia Suteu (comparative constitutional law) are often unreflectively imported into constitutionmaking processes.29 This brief foray helps illustrate the numerous dangers entailed by comparative legal citations. While most of the scholarship in this area focuses on such citations in constitutional adjudication, I would argue the same self-awareness and methodological rigour required of the good adjudicator when citing comparative material should inform the good constitutional theorist. While disciplinary boundary policing serves no one, I think constitutional theorising carries with it a certain duty of care on the part of those engaging in it. Our work informs constitutional processes in the real world, even if indirectly, and as such we should strive to know that world and to do so beyond our own backyard. The following section will illustrate these points by focusing on concepts in constitutional theory that have migrated into constitutional adjudication: constituent power and the notion of unconstitutional constitutional amendments. III.  CONSTITUTIONAL THEORY AND COMPARATIVE CONSTITUTIONAL CHANGE

I propose to further illustrate what advances in comparative constitutional theory can illuminate by revisiting two problems with which the theory and practice of democratic constitutional change has long grappled.30 The first, broad set of questions asks how to ensure the legitimacy of both the process of constitutional reform and its substantive outcome. Even under ideal conditions, these aims demand that we clarify our notions of democratic constitutional legitimacy and authority, as well as the no less thorny relationship between constitutional process design and its results. Constituent power, understood as the power to constitute and reconstitute the constitutional order, has often been relied on in pursuit of answers to such questions of constitutional authorship, legitimacy, and endurance. The second but related set of questions, raised with ever greater frequency today, asks how to prevent abusive constitutional change such as that pursued by would-be autocrats and populists in power. This can take the form of constitutional or legislative amendments and even replacement that undermine core democratic constitutional commitments, or else pack and subordinate courts so as to remove the possibility of legal accountability.31 As we will

29 A Pastor y Camarasa, ‘The Limitations of International Expertise when Drafting New Constitutions’, The Conversation (17 November 2020), https://theconversation.com/the-limitations-of-internationalexpertise-when-drafting-new-constitutions-149504. See also M Malagodi, ‘International Assistance to Constitution Making between Principle and Expediency’, I-CONnect Blog (13 January 2021), www.iconnectblog.com/2021/01/international-assistance-to-constitution-making-between-principle-and-expediency/. 30 When it comes to the study of constitutional change, scholarly attention has turned to the need to bridge comparative constitutional law and theory. See, inter alia, R Albert, Constitutional Amendments: Making, Breaking, and Changing Constitutions (Oxford, Oxford University Press, 2019); R Albert, X Contiades, and A Fotiadou (eds), The Foundations and Traditions of Constitutional Amendment (Oxford, Hart Publishing, 2017); and X Contiades and A Fotiadou (eds), Routledge Handbook of Comparative Constitutional Change (London, Routledge, 2020). 31 There are, of course, other forms that abusive constitutional change can take. For example, changes to electoral and media laws have been used to entrench incumbents’ hold on power and undermine fair electoral

The View from Nowhere in Constitutional Theory  349 see, the two strands are interrelated, as how constituent power is understood will have a direct impact on what we deem democratically acceptable constitutional reforms or replacements, and what we consider illegitimate change. A.  The Revival of Constituent Power Theory For a long time, and to an extent still, discussions of constituent power embraced Sieyes’s notion distinguishing it from constituted powers and saw it as both extralegal and extinguished at the time of constitution-making. This understanding in turn led to a ‘sanitised’, apolitical view of constitutional authority that could, for the most part, ignore the messy dynamics of radical politics, coups and revolutions. They were simply outside the duly constituted legal order, even while they could fundamentally alter it. More recently, scholars have insisted on revisiting earlier understandings of the notion of constituent power in order to better grasp the complexities of the concept.32 One such scholar, Lucia Rubinelli, has provided a rich intellectual history of the concept of constituent power through close reading of Sieyes’s work and its reception.33 Through careful reading of the original historical sources and comparative material, Rubinelli seeks to illuminate the many lives the concept of constituent power has lived over the years, and thus frames hers as a history of the theory of constituent power rather than its practice. This close engagement with a range of key thinkers (albeit an unavoidably selective sample) allows her to reveal the multiple different ideas often subsumed to the concept of constituent power, some theorists having collapsed it into sovereignty, some viewing it as inherently democratic, constitutional, or juristic, and others as pre-political and leading to decisionism. In Rubinelli’s words, this is ‘a story that portrays constituent power as one amongst other ways of framing the principle of popular power over time’.34 Rubinelli’s is not just a pursuit to recapture the origins of constituent power as an idea, but also to track the uses of the language of constituent power over time and expose those instances when it has been misrepresented, reinterpreted, and even abused. In so doing, she seeks to expose those authors who operate a misguided appeal to history to defend their definition of constituent power as the one and true.35 Such circular appeals to history – positing a ‘true’ definition of a constitutional concept in order to justify one’s reading of ‘real’ constitutional practice – are to be resisted. According to Rubinelli, ‘[c]onstituent power is … in the eyes of the beholder and

competition. On the issue of ‘abusive constitutionalism’, see D Landau, ‘Abusive Constitutionalism’ (2013) 47 UC Davis Law Review 189. On the diversity of populist attacks on the separation of powers, see S Gardbaum, ‘The Counter-playbook: Resisting the Populist Assault on Separation of Powers’ (2020) 59(1) Columbia Journal of Transnational Law 1. 32 For an earlier collection of contributions discussing constituent power as embedded in different constitutional traditions, both national and supranational, see M Loughlin and N Walker (eds), The Paradox of Constitutionalism: Constituent Power and Constitutional Form (Oxford, Oxford University Press, 2008). 33 L Rubinelli, Constituent Power: A History (Cambridge, Cambridge University Press, 2020). 34 ibid 3. 35 ibid 229.

350  Silvia Suteu not, as some seem to believe, a positive reality from which one can straightforwardly abstract a concept’.36 Here is, then, one of the most valuable contributions the historian can make to constitutional theoretical debates: unearthing original meaning, retracing legacies of ideas, and exposing misuses, oversimplifications, and claims to exclusive interpretation of a complex idea. There are also important limitations to such an approach, especially from the perspective of a constitutional theorist seeking to understand contemporary processes of constitution-making and constitutional change. By explicitly excluding a history of the practice of constitution-making from her account, Rubinelli’s analysis operates on the basis of a distinction that is difficult to maintain and, I would argue, artificial from the onset.37 This is not only because constitutional ideas and meaning are revealed, shaped, and transformed by practice, but also because at least some of the constitutional theorists she reads in the book view constituent power itself as a practice, rather than a grand narrative.38 Theirs are not just attempts to recapture the ‘true’ meaning of constituent power but, especially when reacting to new problems in constitutional practice or to answer old questions in a new key, attempts at constructing novel theories of constituent power. As Rubinelli herself admits, ‘contemporary theorists are fully part of the process of negotiation of the meaning and implications of the principle of popular power’.39 In other words, they are ‘fully part of the history … narrated’ in the book.40 This does not detract from the value of such a robust historical re-evaluation of the idea of constituent power. Quite on the contrary: it should further emphasise the sense of responsibility for the constitutional theorist relying on notions of constituent power to read, justify, and denounce contemporary processes. The theorist must remain aware not only of the partiality of her standpoint on the meaning of the concept, but also of the fact that she is developing and enriching its meaning herself. A second strand of recent work on constituent power that has significantly advanced our understanding of the concept comes from scholars directly engaged in explaining and integrating contemporary constitutional practice into broader narratives of constitutional legitimacy. Andrew Arato and Joel Colon-Rios are both, in distinct ways, excellent examples of such richly comparative theorising that both clarifies and adds to existing understandings of constituent power.41 36 ibid 16. 37 To be fair to her, she acknowledges the difficulty of the distinction even while defending it. ibid 18, fn 40. 38 This would lead to a more charitable reading of Arendt’s understanding of constituent power for example. Arendt is discussed in ibid ch 5. 39 ibid 30. 40 ibid 222. 41 See, inter alia, A Arato, ‘Multi-Track Constitutionalism Beyond Carl Schmitt’ (2011) 18(3) Constellations 324; A Arato, Post Sovereign Constitution Making: Learning and Legitimacy (Oxford, Oxford University Press, 2016); A Arato, The Adventures of the Constituent Power: Beyond Revolutions? (Cambridge, Cambridge University Press, 2017); J Colon-Rios, ‘Carl Schmitt and Constituent Power in Latin American Courts: The Cases of Venezuela and Colombia’ (2011) 18(3) Constellations 365; J Colon-Rios, Weak Constitutionalism: Democratic Legitimacy and the Question of Constituent Power (London, Routledge, 2012); J Colon-Rios, ‘Five Conceptions of Constituent Power’ (2014) 130 Law Quarterly Review 306; J Colon-Rios, Constituent Power and the Law (Oxford, Oxford University Press, 2020).

The View from Nowhere in Constitutional Theory  351 Arato argues that constitution-making today is post-sovereign, in the sense that no single agency drives the process; post-revolutionary, in the sense that it transcends the reform/revolutionary binary; and multi-stage, as it involves periods of transitional legality, such as governed by interim constitutions.42 With examples from Spain and South Africa to Poland and Hungary, these constitution-making processes have developed in a staggered way, sometimes constrained by pre-agreed principles, and have involved a multitude of actors negotiating over time before a permanent constitution was finally agreed upon. It is the combination of this multiplicity of the who of constitution-making, coupled with a protracted when and a multi-step how that, together, afford democratic legitimacy.43 Arato is not the only or indeed the first to note that constitutionalism, as a ‘process of becoming’, necessitates a concept of the constituent power that is able to capture this cross-temporal dimension.44 However, what Arato’s work does very well is to showcase, through dense, comparativelyinformed and at the same time critical and theoretically rigorous research, what comparative constitutional theory can look like. Arato’s work, grounded in Arendtian ideas, seeks to avoid the pernicious idea of a fully embodied constituent power (always appealing to autocrats) and to explore models of diffuse authority generated by a multi-track process with a plurality of actors. His, then, is a constitutional theory that is a normatively engaged project to resist anti-democratic claims to constitutional authority and legitimacy. His general scepticism of innovations such as constitutionalising constituent assemblies or citizens’ assemblies stems from a comparatively-informed distrust that participatory democracy will not be abused.45 He thus sees developments such as the political co-optation of the Venezuelan constituent assembly and constitutional court in 1999 as directly deriving from a flawed understanding of constituent power, one that should be resisted for its propensity to legitimise constitutional abuse. Arato’s post-sovereign model has not been without its critics of course, and the author himself acknowledges gaps in it. For example, he has admitted that the case studies he has built his theory upon did not have the benefit of double differentiation: they did not have a separately elected legislature and constitution-making body, the former acting as procedural limitation on the latter.46 However, the fact that the post-sovereign model developed in transitional countries without legitimate and competitively elected legislatures means they cannot provide a model for a double differentiated amendment process.47 He speculates that looking to already existing democracies could further improve the idea of post-sovereign constitution-making.48 A potential shortcoming of the comparative sample underpinning his theoretical model, therefore, may not necessarily be its undoing, but simply a reason for enlarging the model to include non-transitional constitutional orders. 42 Arato, The Adventures of the Constituent Power (n 41), ix. 43 This idea of democratic legitimacy not emanating from a single source, moment, or actor is increasingly being recognised in comparative constitutional scholarship. See V Jackson, ‘‘Constituent Power’ or Degrees of Legitimacy?’ (2018) 12(3) Vienna Journal of International Constitutional Law 319. 44 Loughlin and Walker, The Paradox of Constitutionalism (n 32). 45 Arato, Post Sovereign Constitution Making (n 41) 9. 46 Arato, The Adventures of the Constituent Power (n 41) 416–417. 47 ibid 417. 48 ibid.

352  Silvia Suteu Most interesting for our purposes here is the very different position adopted by Colon-Rios to the question of whether constituent power can be institutionalised in the existing constitutional order. Drawing on Latin American examples in particular, Colon-Rios argues that such institutionalisation is possible and a worthy addition to the constitutional design arsenal. Drawing on cases such as Colombia’s and Venezuela’s constituent processes, Colon-Rios makes the case that constituent assemblies procedurally mandated under existing constitutions can nevertheless exercise constituent power insofar as they remain unbound by substantive positive law.49 Insofar as their outcome (whether constitutional amendment or replacement) is indistinguishable from that of an original constitution-making, and also as they are often just as if not more participatory than the latter, Colon-Rios defends these institutionalised constituent processes on democratic grounds. He does so, however, careful to note that they are defensible only insofar as they do not purport to be sovereign, as opposed to constituent. In other words, as long as they are performing their task of producing new constitutional content – on the basis of an imperative mandate given by the people through a referendum or popular initiative – and not as omnipotent sovereigns claiming to embody the nation. Colon-Rios’s argument relies on a particular set of beliefs about the nature of constituent power, as well as about constitutional theory and law more generally. His analysis in many ways echoes Rubinelli’s, in that he also retraces different understandings of constituent power over time and space with a view to better understanding this multifaceted idea. Where he differs, however, is that he is very much drawing on constitutional practice – notably in continental Europe and Latin America – to argue that the concept of constituent power should be understood as inherently juridical. If we are to ‘grasp the full potential of the theory of constituent power’, he argues, we must squarely face the concept’s legal and institutional implications.50 While not part of positive law, constituent power can play an important role in the assessment of the legality and illegality of different forms of political action, as well as in the types of institutions that should be present in a constitutional order whose creation is attributed to a constituent people.51

An entity authorised to exercise constituent power, such as a constituent assembly, is unlike a true sovereign in that it ‘cannot transform any will into law, but only produce constitutional law’.52 To believe the constitutional order closed off to constituent power once instituted is both to ignore the juridical nature of this power and to confuse constituent power and sovereignty.53 Whereas the latter is unbound and may give rise to any legal content, constituent power should be viewed as inherently tied to constitution-making. As such, for Colon-Rios, a constituent process institutionalised as part of the existing constitutional framework is neither illogical nor unavoidably leading to abuse by political forces. For seemingly substantively unbound, such



49 Colon-Rios,

Constituent Power and the Law (n 41), 15 and chs 9 and 10. 2. 51 ibid 28. 52 ibid 260. 53 ibid 3 and ch 9. 50 ibid

The View from Nowhere in Constitutional Theory  353 a constituent process is nevertheless not sovereign as such: it is procedurally bound (especially in terms of how it is to be initiated) and only mandated to engage in constitutional reform. Moreover, when bound by a popular constitutional mandate (which would normally also contain an implicit requirement to respect the separation of powers), the constituent may be subject to judicial control seeking to enforce this mandate.54 The contrast with Arato is clear. The emphasis shifts for Colon-Rios from merely seeking to prevent constitutional abuse justified in the name of constituent power (conflated with sovereignty) to retaining constituent power’s enabling role as a generative, juridical power to be accommodated within the constitutional order. Both are committed to democratic constitutionalism and to the usefulness of the concept of constituent power, but arrive at different conclusions about the role it should play in contemporary constitutional discourse and institutional design. Yet I view these two theorists as having more in common than might initially appear. Not only are their analyses grounded in comparative, deeply contextual work, they are also fluent in debates in political and constitutional theory spanning centuries. Both resist the fusion of constituent power and sovereignty, even while assessing differently the institutional implications of seeking to retain limits on constituent processes. Both also resist oversimplified distinctions between law and politics that, in the case of constituent power, were too quick to relegate it to the extra-legal realm. This resistance is no doubt tied to their familiarity with comparative case studies that directly challenge the idea of constitutional creation ex nihilo: post-authoritarian constitution-making in Europe and South Africa in Arato’s case, and twenty-first century constituent processes in Latin American constitutional orders in Colon-Rios’s. Whereas Rubinelli’s work helps us understand how we got to speak the language of constituent power as we do today, Arato’s and Colon-Rios’s illustrate the potential of a constitutional theory that starts from and is continuously checked against the messy and dynamic reality of constitutional practice. B.  Constituent Power Theory and Unconstitutional Constitutional Amendments From a niche subject, unconstitutional constitutional amendments and their related cousins, basic structure doctrines, have exploded as topics of scholarly interest in recent years.55 Their appeal to constitutional theorists is irrefutable. The question at the centre of these doctrines is whether constitutional amendment can be limited substantively, either explicitly in the constitutional text (in the form of an eternity clause) or implicitly on the basis of principles of democratic constitutionalism 54 ibid 294. 55 See, inter alia, PJ Yap, ‘The Conundrum of Unconstitutional Constitutional Amendments’ (2015) 4(1) Global Constitutionalism 114; Y Roznai, ‘Unconstitutional Constitutional Amendments – The Migration and Success of a Constitutional Idea’ (2013) 61 American Journal of Comparative Law 657; Y Roznai, Unconstitutional Constitutional Amendments: The Limits of Amendment Powers (Oxford, Oxford University Press, 2017); Albert, Constitutional Amendments (n 30); S Suteu, Eternity Clauses in Democratic Constitutionalism (Oxford, Oxford University Press, 2021).

354  Silvia Suteu (developed by courts as unamendability doctrines), to prevent the subversion or replacement by stealth of an existing constitutional framework. This ties back to problems of constitutional precommitment, insofar as altering existing arrangements is seen as undoing the original political settlement. Also related is the notion of militant democracy and how to calibrate its tools to prevent constitutional democracies being rendered defenceless against their enemies. Equally relevant are questions about constitutional courts as guardians of the constitutional order and how far they can go in exercising their guardianship, ie whether it includes preventing procedurally legitimate exercises of the amendment power for being substantively incompatible with entrenched values, principles, or provisions. Yaniv Roznai, in his comprehensive study of the rise and migration of unconstitutional constitutional amendment doctrines worldwide, has shown that, rather than a fringe or anomalous development, judicially enforced substantive limits on amendment have been embraced across the globe and continue to be on the rise.56 Moreover, he has carefully constructed a theory that helps explain the doctrinal move of courts enforcing substantive limits on procedurally sound constitutional amendments. Relying on the distinction between primary (original) and secondary (derived) constituent power, Roznai describes the amendment power institutionalised in a constitution’s amendment rule as embodying the latter type.57 As such, the amendment power is a delegated power, held in trust by the amendment authority and constrained by the limits delineated by the constitution. As guardians of the constitution, constitutional courts are also guardians of the original, or primary, constituent power and empowered to review amendments materially. Only a renewed, revolutionary constitution-making moment could overcome these restraints. A series of important consequences flow from this understanding of constituent power as it operates in the contest of constitutional amendments. One is that this guardianship role is incumbent upon courts regardless of whether the constitutional text formally grants them the power to substantively review amendments – in fact, this power is taken to exist even contrary to the constitution explicitly forbidding this type of review. Another has to do with the nature of the amendment process itself. Roznai describes a sliding scale of amendment legitimacy, wherein limits on amendment are more or less legitimate depending on the features of the amendment process: the more inclusive, participatory, multi-procedural, or deliberative it is, the more legitimate.58 Additionally, the legitimacy of such far-reaching judicial intervention relies on judicial self-restraint. The understanding is that these will be exceptional interventions, aimed at preventing egregious abuses of the constitutional process and blocking constitutional dismemberment.59 The comparative breadth of Roznai’s study makes the underpinning theory that much more appealing. For we see distant and disparate jurisdictions from

56 Roznai, Unconstitutional Constitutional Amendments (n 55). 57 ibid 105–134. 58 ibid 158, 219–220. 59 On doctrinal interpretive techniques to engender such self-restraint, see Roznai, Unconstitutional Constitutional Amendments (n 55) 213–225. On the notion of ‘constitutional dismemberment’, see Albert, Constitutional Amendments (n 30) 78.

The View from Nowhere in Constitutional Theory  355 Germany to India and Belize to Taiwan developing rich case law invalidating amendments on substantive grounds that include fears of resurgent authoritarian political parties, threats of executive aggrandizement under the cover of a state of emergency, restrictions on judicial review feared to undermine the rule of law, and more. As is unavoidable in such large sample studies, however, there is a risk that important local context is missed and consequently that the macro-level theory omits important nuances. I will highlight three such potential omissions.60 One important addition is recognising that unamendability, especially formal eternity clauses, often play a very different role in post-conflict and transitional constitution-making.61 Frequently conducted under the threat of violence and in conditions of deep societal divisions, the outcome of constitutional negotiations is never a sure thing in such contexts. The resulting political settlement is often a hard-fought pact enshrining uneven and sometimes even contradictory compromises. Far from the lofty ideals of constitutionalism that precommitment understandings of unamendability assume underpin constitutional eternity clauses, such provisions may concern themselves with much more pragmatic topics. An example would be the entrenchment of amnesties for previous political opponents whose buy-in is needed in the new political dispensation, or of executive term limits in contexts plagued by executive overstay. The point here is important insofar as a great number, if not the majority, of constitution-making processes today occur in precisely such fraught contexts. A second typology of understudied cases involves otherwise democratic constitutional frameworks that nevertheless incorporate unamendable constitutional provisions of an explicit but more often implicit exclusionary nature.62 Such exclusionary entrenchment is insidious in that it often coexists with liberal democratic constitutional commitments. An unamendable constitutional commitment to an official language, for example, may appear innocuous and is typically defended by its drafters as a mere statement of fact: the constitutional recognition of a majority language. However, constitutional practice may reveal, as it has done in Romania and Turkey for example, that such seemingly harmless constitutional unamendability is anything but and will be relied on by constitutional courts to block linguistic minorities’ attempts at gaining greater recognition or entitlements. The exclusionary aims of constitutional entrenchment, including unamendability, are difficult to spot absent deep contextual analysis and will frequently only be revealed as part of constitutional adjudication. Importantly, such outcomes can be present in constitutional democracies otherwise committed to fundamental rights, equality, and non-discrimination.

60 I develop these in much greater detail in Suteu, Eternity Clauses (n 55) chs 2–4. 61 See further discussion in S Suteu, ‘Eternity Clauses in Post-conflict and Post-authoritarian Constitution-making: Promise and Limits’ (2017) 6(1) Global Constitutionalism 63. 62 See further discussion in S Suteu, ‘The Multinational State That Wasn’t: The Constitutional Definition of Romania as a National State’ (2017) 11(3) Vienna Journal on International Constitutional Law 413 and S Suteu, ‘Eternity Clauses as Tools for Exclusionary Constitutional Projects’ in R Abeyratne and Ngoc Son Bui (eds), The Law and Politics of Unconstitutional Constitutional Amendments in Asia (London, Routledge 2021, forthcoming). See also M Masri, ‘Unamendability in Israel: A Critical Perspective’ in R Albert and B Emrah Oder (eds), An Unamendable Constitution? Unamendability in Constitutional Democracies (Cham, Springer, 2018) 169.

356  Silvia Suteu Thus, when defending unamendability as the ‘lock on the door’ protecting liberal constitutionalism, comparative knowledge should have us question whose values and principles are actually being protected and at what price.63 A third corrective to constitutional theoretical understandings of unamendability directly challenges the understanding of constituent power usually taken to underpin it. As Vicki Jackson has argued, the delegation model – wherein amendment power is a secondary or derived form of constituent power bound by the constitutional framework, with courts tasked with defending the original constituent’s will as enshrined in the constitutional framework – is deeply problematic.64 The constituent power is assumed to be an ‘it’ instead of a ‘they’, reified as a univocal people where instead there is plurality of voices. This in turn risks conflating constitutional legitimacy with the expression of a supposed unified original will (‘the people’). Instead, Jackson proposes we assess ‘the legitimacy bases of the existing constitution’ on a case by case basis and have them ‘weigh in the calculus of whether a subsequent amendment should be understood as unconstitutional’.65 While popular consent is necessary, it is not a sufficient condition to establish constitutional legitimacy, which can only be ascertained through a holistic evaluation over time of both the process of adoption and the substantive commitments within a constitution.66 As I have argued elsewhere, drawing on work by Arato and others, ‘[t]he risk is there to derive legitimacy from identity, ascribing it solely to the who of constitution-making without taking into account the when, what and how of the process.’67 This conflation also carries with it the danger of masking contestation, including of the very existence and boundaries of ‘the people’ in whose name the constitution is adopted.68 As cases such as Bosnia and Herzegovina have shown, no amount of constitutional engineering and entrenchment can overcome the deeply unstable contours of the constitutional subject. There, an uneasy peace and federal structure were meant to be buttressed by power-sharing arrangements and far-reaching international human rights commitments, entrenched in the constitutional eternity clause (Article X(2) of the Constitution of Bosnia and Herzegovina).69 Yet decades later, Bosnian peoplehood remains fragmented and as contested as ever.70

63 Here again, a robust analysis by a political theorist of the history of constitutional entrenchment since antiquity reveals its propensity for entrenching elite privilege. See M Schwartzberg, Democracy and Legal Change (Cambridge, Cambridge University Press, 2009). 64 Jackson, ‘Constituent Power’ (n 43). 65 ibid 327–328. 66 ibid 338. 67 Suteu, Eternity Clauses (n 55) 171, citing Arato, The Adventures of the Constituent Power (n 41) 31. 68 Z Oklopcic, ‘Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments’ in R Albert, X Contiades, and A Fotiadou (eds), The Foundations and Traditions of Constitutional Amendment (Oxford, Hart Publishing, 2017) 51, 60. 69 Moreover, the case of Bosnia and Herzegovina is rendered even more complex by the highly internationalised nature of the peace process that resulted in the Constitution being annexed to the Dayton Accords in 1995. This further challenges our received notions of constituent power, raising the prospect that it is no longer (if it ever was) nationally bound and inward looking. For a lengthier discussion, see Suteu, Eternity Clauses (n 55) ch 5. 70 The constitutional preamble itself remains indicative of this fragmented peoplehood, referring to ‘Bosniacs, Croats, and Serbs, as constituent peoples (along with Others), and citizens of Bosnia and Herzegovina’.

The View from Nowhere in Constitutional Theory  357 These observations, informed by comparative constitutional experience, show that our theoretical understanding of unamendability must accommodate more varied contexts. These are contexts where it is not so much the high point of constitutionalism, but rather plays the role of political insurance in constitutional negotiations, of entrenching elite interests that may be exclusionary, and of entrenching a purportedly unified will where a fragmented constitutional subject exists. We could further question the notion of constituent power underpinning unamendability doctrines by considering a constitutional system grounded in parliamentary sovereignty such as the UK’s. There, not only does the language of constituent power tend to be absent, but also the distinctions between original and derived constituent power and between constituent and constituted powers are difficult to map onto constitutional practice. At most, we could say about the UK Parliament, with Dicey, that it is acting as a legislative (constituted) power when adopting ordinary laws but as a constituent power when its acts ‘shift the basis of the constitution’.71 This is not to say that the notion of supra-constitutional limits or de facto unamendable commitments is entirely alien to the British constitutional tradition. As recent scholarship is proposing to show, building on a century of development of the principle of legality at common law, we may speak of a ‘theory of legislative legitimacy’ that allows for invalidation of acts of the sovereign UK Parliament in extreme, as yet hypothetical scenarios.72 This example shows how implied unamendability may yet operate while the language of constituent power remains alien to the UK system. Such comparative insights again remind us not to be too quick in assuming the inevitability of any given concept in constitutional theory. IV. CONCLUSION

In many ways, my aim in this chapter has been to take seriously Andrew Arato’s observation that ‘[w]hat is missing [in scholarship] are serious synthetic works linking theory and comparative analysis that are capable of prudentially orienting actors engaged in constitution-making processes all over the world’73 and to see how this might be remedied. In particular, I sought to illustrate the pitfalls of doing constitutional theory without such comparative awareness. In a nutshell, I sought to show that this type of theorising risks being impoverished and of limited applicability, especially when we take seriously the need to decolonise constitutional theory. Rendering it truly reflective of the myriad constitutional experiences out there will only be possible by engaging in difficult (though rewarding) comparative work. I have also sought to illustrate the important work already done in this direction in the field of comparative constitutional change, particularly looking at the rise

71 See discussion in Colon-Rios, Constituent Power and the Law (n 41), 11–12. See also, generally, M Loughlin, ‘Constituent Power Subverted: From English Constitutional Argument to British Constitutional Practice’ in Loughlin and Walker, The Paradox of Constitutionalism (n 32) 27. 72 H Hooper, ‘Legality, Legitimacy, and Legislation: The Role of Exceptional Circumstances in Common Law Judicial Review’ (2021) 41(1) Oxford Journal of Legal Studies 142. 73 Arato, The Adventures of the Constituent Power (n 41) vii.

358  Silvia Suteu of new scholarship on constituent power and unamendable constitutional amendment. This scholarship is invariably interdisciplinary, richly comparative, and as such more representative of a variety of constitutional experiences across time. We need not agree with the conclusions reached by the authors working in this way to be able to appreciate the fruitfulness of their method. We will continue to differ on whether constituent power can be institutionalised and constrained by the existing constitutional order. However, we will do so more competently with the benefit of a cross-temporal understanding of the idea of constituent power, as well as of the practice of institutionalisation of constituent processes in countries in Latin America. Similarly, we will go on disagreeing over the benefits and drawbacks of unamendability and constitutional entrenchment more generally. However, appreciating the different roles such entrenchment will play in different contexts will allow us to do so with more nuance. After all, whether unamendability is relied on in a given system as an instrument of militant democracy, as political insurance for political elites, or as a tool for majoritarianism and exclusion should surely influence our assessment of its constitutional protective role. At the same time, the possible emergence of implied limits on legislative authority in the UK should have us question whether it is only through the lens of constituent power theory that we can understand unamendability. The type of constitutional theory I have discussed in this chapter is no doubt laborious. It requires not just interdisciplinarity but also comparative fluency that is difficult both to obtain and to sustain. My intention is not to suggest that this is the only type of constitutional theory worth engaging in, of course. It is, instead, to highlight and praise this type of scholarship for the new insights it can offer and for the correctives to received notions it can encourage. It is also to suggest that more comparative awareness will help us enrich our conceptual arsenal, avoid too easy generalisations, and overall make us better constitutional theorists.

Administrative Law

360

14 Common Understandings of Administrative Law MATTHEW LEWANS*

I. INTRODUCTION

A

dministrative law can be a mysterious subject. When constitutional lawyers construct a theory about legislation, they can lean on the familiar doctrine of parliamentary sovereignty to explain why statutes impact our understanding about what the law requires. Likewise, when they develop a theory about the common law, they can rely on a rich history of common law adjudication to explain how unwritten legal principles impact legal interpretation. By contrast, constitutional lawyers tend to approach the topic of administrative law indirectly, by asserting that the legality of administrative decisions is grounded either by a formal statutory delegation of discretionary power or by substantive alignment between the merits of an administrative decision and judicial interpretations of law.1 Thus, instead of generating a theory of administrative decisions qua law to complement theories of legislation and the common law within an institutionally complex constitutional order, theoretical debates about administrative law revolve around whether judicial review of administrative decisions is necessarily justified by parliamentary intent or common law principles. In this chapter, I will identify and critique different conceptions of administrative law, and the different methods scholars use to understand this phenomenon. When viewed in isolation, each of these perspectives can be characterised as reflecting either descriptive or normative modes of scholarship; however, all of them can be situated within a broader, overarching intellectual history of constitutional theory. Thus, in order to identify and explain different methods for understanding administrative law, I will employ an historical method of exposition to better understand and critique these different perspectives. By highlighting some of the latent thoughts, arguments, beliefs, and attitudes that motivate these different understandings, I hope to

* Thanks to David Dyzenhaus, Dimitrios Kyritsis, Hillary Nye, Charles-Maxime Panaccio, and Kristen Rundle for their feedback on previous drafts. 1 C Forsyth (ed), Judicial Review and the Constitution (Oxford, Hart Publishing, 2000).

362  Matthew Lewans alleviate the sense of mystery that lingers over the subject, identify productive points of ­engagement, and prepare the ground for a theory of administrative law to complement theories of legislation and the common law.2 In section I, I will briefly explain why administrative law presents a puzzle for constitutional theorists, because it simultaneously raises general questions about institutional design and more particular questions about the legitimacy of specific administrative decisions and decision-making processes. In sections II and III, I argue that the complex problems posed by the phenomenon of administrative law are reflected by deep tensions in the intellectual history of constitutional theory, as illustrated by theoretical frameworks proposed by William Blackstone and Albert Venn Dicey. In section IV I will briefly compare and contrast four different perspectives of administrative law that have gained prominence at various points over the past century. Interestingly, while each of these has distinct conceptual, libertarian, egalitarian, and procedural underpinnings, they reflect the deeper institutional and normative tensions that constitutional theorists have been struggling to address over the past three centuries. In section V, I will conclude by suggesting if one engages constructively with these different understandings, one might discover a more robust theory of administrative law. II.  THE PUZZLE OF ADMINISTRATIVE LAW

One central puzzle posed by the emergence of the modern administrative state is to explain how law both empowers and constrains administrative institutions within a constitutional culture committed to the rule of law.3 If administrative law concerns ‘the field of legal control exercised by law-administering agencies other than courts, and the field of control exercised by courts over such agencies’,4 the subject of constitutional law shares the same focus. Both constitutional and administrative law theorists grapple with problems associated with the distribution of political power amongst different (but related) legal institutions as well as the legality of more specific decisions rendered by those institutions. Furthermore, the fundamental legal principles that animate the practice of administrative law – especially principles of natural justice, substantive reasonableness, and proportionality – reflect many of the same principles that animate constitutional law and theory more broadly as it relates to legislation.5 Put simply, both administrative and constitutional law scholarship are 2 M Taggart, ‘Prolegomenon to an Intellectual History of Administrative Law in the Twentieth Century: The Case of John Willis and Canadian Administrative Law’ (2005) 43 Osgoode Hall Law Journal 224, 225. 3 D Rodriguez, ‘Administrative Law’ in G Caldeira, D Kelemen, and K Whittington (eds), The Oxford Handbook of Law and Politics (Oxford, Oxford University Press, 2008) 340–1. 4 F Frankfurter, ‘The Task of Administrative Law’ (1926–1927) 75 University of Pennsylvania Law Review 614, 615. 5 T Poole, ‘Back to the Future? Unearthing the Theory of Common Law Constitutionalism’ (2003) 23 OJLS 435; M Walters, ‘The Unwritten Constitution as a Legal Concept’ in D Dyzenhaus and M Thorburn (eds), The Philosophical Foundations of Constitutional Law (Oxford, Oxford University Press, 2016) 33; M Walters, ‘Deliberating about Constitutionalism’ in H Kong, R Levy, G Orr and J King (eds), The Cambridge Handbook of Deliberative Constitutionalism (Cambridge, Cambridge University Press, 2018) 167.

Common Understandings of Administrative Law  363 inspired by the values of constitutionalism, legality, and the rule of law – values that evoke wholesale questions about institutional design (particularly matters relating to institutional checks and balances) as well as retail questions regarding the merits of more particular interactions between governmental institutions, officials, and individuals. But while a general outline of administrative law can be sketched in this way, framing the subject in purely abstract terms risks obscuring more than it illuminates. Without more, the abstract concepts of constitutionalism, legality, and the rule of law cannot hope to dissolve the complex and controversial questions it evokes: whether the rule of law entails a particular formal, procedural, or substantive legal constraints on legislators, judges, and administrative officials; whether those relevant legal constraints operate on administrative processes and decisions ex ante or ex post; and how one ought to navigate the complex relationship between legislatures, courts, and administrative institutions within a constitutional democracy in which multiple legal institutions share responsibility for creating, interpreting, implementing, and enforcing the law. When one considers the array of problems thrown up by administrative law, one encounters immediately deep-seated disagreements regarding the best way to tackle them. On top of it all, despite the theoretical overlap between constitutional and administrative law, there are significant differences between the intellectual, institutional, and historical contexts that shape juridical approaches to these subjects. While the constitutional principles of parliamentary sovereignty and judicial independence were secured through the Glorious Revolution of 1688, the Bill of Rights 1689, and the Act of Settlement 1701, the relatively incremental (but no less revolutionary) development of the modern administrative state has been overshadowed by a persistent cloud of constitutional controversy. The legal foundation for the administrative state was established through episodic legislative reforms, but those reforms did not culminate in a new constitutional settlement.6 Instead, administrative lawyers in the UK, like their counterparts elsewhere in the common law world, have been racked by perennial debates regarding the constitutional legitimacy of the administrative state and administrative law since at least the late nineteenth century.7 While constitutional controversies like universal suffrage were resolved through new constitutional settlements, debates about administrative law have continued to simmer – like recurrent controversies regarding devolution, the role of local government, the incorporation

6 See, eg, B Ackerman, We the People: Foundations, vol. 1 (Cambridge MA, Belknap Press, 1991) 34–57. 7 See, eg, Lord Hewart, The New Despotism (New York, Cosmopolitan Book Corp, 1929); J Landis, The Administrative Process (New Haven, Yale University Press, 1938); CK Allen, Law and Orders: An Inquiry into the nature and scope of Delegated Legislation and Executive Powers in England (London, Stevens and Sons Ltd, 1945); J Willis, The Parliamentary Powers of English Government Departments (Cambridge MA, Harvard University Press, 1933); D Ernst, Tocqueville’s Nightmare: The Administrative State Emerges in America 1900–1940 (Oxford, Oxford University Press, 2014); P Hamburger, Is Administrative Law Unlawful? (Chicago, University of Chicago Press, 2014); E Posner and A Vermeule, The Executive Unbound: After the Madisonian Republic (Oxford, Oxford University Press, 2010); A Vermeule, Law’s Abnegation: From Law’s Empire to the Administrative State (Cambridge MA, Harvard University Press, 2016).

364  Matthew Lewans of the European Convention on Human Rights into domestic law, or the UK’s present and future relationship with the European Community.8 Over the past 150 years, the British administrative state has evolved in response to episodic economic crises, two world wars, the creation of the National Health Service, public utilities, and recurrent political wrangling about the role of the state in a market-based economy.9 While orthodox constitutional doctrine appeared to remain static throughout this period, this apparent stasis masks subtle but significant shifts in legal perspectives regarding the constitutional legitimacy of administrative law. If the story of administrative law begins at the turn of the twentieth century, the narrative begins with Diceyan denial; proceeds through modest restraint in the exercise of official power, then a judicial retreat in the face of wars, depression, and the creation of the modern welfare state; is followed in the second half of the century by an awakening of judicial concern for administrative justice and evolves in the century’s death throes into a concern for fundamental rights or ‘common law constitutionalism.10

Remarkably, many points along this historical continuum can be marked by different concepts, doctrines, and methods for grappling with the central problem posed by the phenomenon of administrative law. However, it is only when one begins to excavate these different understandings of the subject, that one can engage critically with their intellectual foundations.11 III.  LEGALITY AND CONSTITUTIONAL FORMALISM

Some of the mystery surrounding administrative law can be dispelled by sifting through ambiguities in constitutional law theory and doctrine that emerged during the eighteenth and nineteenth centuries. Among other things, this period marks a renaissance in both constitutional law and jurisprudence as academic disciplines that could be detached from professional courses of training and folded into a liberal curriculum alongside theology, classics, philosophy, history, mathematics, and the natural sciences. However, this academic renaissance reveals an enduring tension between formal conceptual frameworks regarding the distribution of power between governmental institutions (on the one hand) and jurisprudential reflections on the relationship between the rule of law and moral legitimacy of state action (on the other). In the introductory chapter to his Commentaries on the Laws of England in 1765, William Blackstone sets out an argument for including academic instruction in law as part of the university curriculum.12 At the time, gentlemen were destined to serve 8 V Bogdanor, ‘Introduction’ in V Bogdanor (ed), The British Constitution in the Twentieth Century (Oxford, Oxford University Press, 2003) 1. 9 J Jowell, ‘Administrative Law’ in Bogdanor, The British Constitution in the Twentieth Century (n 8) 373. 10 Taggart, ‘Prolegomenon’ (n 3) 231. 11 Jowell, ‘Administrative Law’ (n 9) 373. 12 W Blackstone, Commentaries on the Laws of England Book I (Oxford, Oxford University Press, 2016) 10.

Common Understandings of Administrative Law  365 on juries, as justices of the peace, or as Members of Parliament, whom Blackstone considered the ‘guardians of the English constitution; the makers, repealers, and interpreters of the English laws’.13 But he also noted that professionals and men of inferior rank would also profit from academic legal instruction, especially those destined for a career at the bar. A student ‘educated to the bar, in subservience to attorneys and solicitors, will find he has begun at the wrong end’, because ‘if he be uninstructed in the elements and first principles upon which the rule of ­practice is founded, the least variation from established precedents will totally distract and bewilder him’.14 For Blackstone, the common law was not an obscure heap of pleadings and cases – it was a deep well of social knowledge and experience, which reflected the history, genius, and moral character of the English nation. The challenge was to articulate a conceptual framework for constitutional law, which could identify ‘the elements and first principles’15 in order to render that body of knowledge ‘intelligible to the uninformed minds of beginners.’16 In order to facilitate this theoretical project, Blackstone sought to ground his exposition of English law in ‘a solid scientific method’.17 However, when explaining what ‘an academical expounder of the laws should do’ Blackstone waxed lyrical, saying18 [h]e should consider his course as a general map of the law, marking out the shape of the country, its connexions and boundaries, its greater divisions and principal cities: it is not his business to describe minutely the subordinate limits, or to fix the longitude and latitude of every inconsiderable hamlet.

The choice of this cartographical metaphor was not accidental. Blackstone was making a deeper metaphysical claim about the nature of constitutional theory – although a constitutional theorist could report a conceptual framework to explain the legal relationship between governmental institutions, the reality of that institutional framework was fixed and immutable. One could not critique or revise the institutional framework for creating, interpreting, and enforcing English law any more than one could change the location of Westminster or London. Thus, the role of a constitutional theorist to describe the relevant sources of English law and the connections between them via a systematic ordering of formal, power-conferring rules.19 For Blackstone, the Crown and Parliament were the dominant landmarks on his constitutional map, and he sought to trace their historical lineage and delineate the extent of their respective jurisdictions. But while Blackstone devoted the lion’s share of volume I of his Commentaries to laying out a map defining the relative scope of parliamentary and executive power, the introductory chapters to that same volume focused on the jurisprudential relationship between legality and the moral legitimacy of state action.

13 ibid

12. 28. 28. 16 ibid 30. 17 ibid 29. 18 ibid 30. 19 HLA Hart, The Concept of Law 2nd edn (Oxford, Clarendon Press, 1994) ch 3. 14 ibid 15 ibid

366  Matthew Lewans As a result, his Commentaries reveal an important tension between Blackstone’s constitutional and jurisprudential insights. As a constitutional theorist, Blackstone provided a map of formal power-conferring rules, which empower governmental institutions, but do not entail any procedural or substantive moral limits on how governmental power is exercised; as a jurisprude, Blackstone expounded on the essential relationship between procedural and substantive principles of legality, which condition and legitimise particular legislative and executive decisions. This deep tension between Blackstone’s jurisprudence and constitutional theory – between formal, procedural, and substantive conceptions of constitutionalism – ­underscores a ­perennial tension in contemporary constitutional and administrative law scholarship. The formal elements of Blackstone’s constitutional doctrine are drawn from an eclectic array of foreign and domestic sources: John Locke’s political philosophy,20 Montesquieu’s sociological exegesis of English liberty,21 and Blackstone’s own Old Tory sympathies.22 Following Locke, Blackstone constructed his doctrine upon a foundation of natural rights and a social contract that constitutes sovereign legislative authority23 – a theory that could explain and justify the constitutional settlement wrought through the Glorious Revolution of 1688. Because James II had violated an implied social contract through ‘his own infatuated conduct’, Parliament was entitled to declare the throne vacant and appoint his successor.24 Subsequently, the Petition of Right, the Habeas Corpus Act, the Bill of Rights 1689, and the Act of Settlement 170125 rendered the natural rights contained within the implicit social contract explicit through legislation.26 Thus, Blackstone concluded that Parliament ‘can change and create afresh even the constitution of the kingdom and of parliaments themselves; as was done by the act of union, and the several statutes for triennial and septennial elections’.27 However, as we will see, Blackstone thought Parliament’s extraordinary legislative jurisdiction would be exercised extraordinarily, leaving quotidian matters of law and social policy to be determined through the exercise of the Crown’s substantial prerogative powers. Further support for Blackstone’s reading of the Glorious Revolution was drawn from Montesquieu’s sociological account of English liberty. In 1748, Montesquieu argued that liberty under the law stemmed from the formal structure of the constitution, which provided institutional checks to guard against the concentration and abuse of political power. By investing executive power in a hereditary monarch, legislative power in a hereditary nobility, and judicial power in ‘persons taken from the body of the people’, Montesquieu perceived a harmonious counterpoise between monarchical,

20 J Locke, The Second Treatise of Civil Government and A Letter Concerning Toleration (Oxford, Blackwell, 1948) ch 8. 21 Montesquieu, The Spirit of Laws (Edinburgh, T Ruddiman, 1793) Book XI. 22 W Prest, William Blackstone: Law and Letters in the Eighteenth Century (Oxford, Oxford University Press, 2008) chs 5–7. 23 Blackstone, Commentaries (n 13) ch 1. 24 ibid 136–141. 25 ibid 87. 26 ibid c 6. 27 ibid 107.

Common Understandings of Administrative Law  367 aristocratic, and democratic forms of government.28 Blackstone ­likewise celebrated a formal balance of power in this constitutional arrangement, but attributed it to the different constituencies represented in Parliament. By reserving the executive power for the Crown and assigning the legislative power to the King, Lords, and Commons, Blackstone argued constitutional liberty was secured through a formal distribution of governmental power. As Blackstone put it, ‘where the legislative and executive authority are in distinct hands, the former will take care not to entrust the latter with so large a power, as may tend to the subversion of its own independence, and therewith of the liberty of the subject’.29 While Montesquieu and Blackstone seemingly parted ways regarding the constitutional role of the judiciary, this was a relatively minor difference of opinion. Montesquieu conceded that ‘the judiciary is in some measure next to nothing’ relative to the substantial powers of the legislative and executive branches of government.30 Likewise, Blackstone thought the Crown and Parliament were more vital constitutional agents, a position which aligned with his Old Tory sympathies for traditional political institutions.31 While Parliament possessed ‘sovereign and uncontrolable authority in making, confirming, enlarging, restraining, abrogating, repealing, r­eviving, and expounding of laws’,32 the Crown was Parliament’s ‘caput, principium, et finis’.33 It possessed exclusive power to convene, prorogue, and dissolve Parliament,34 to refuse royal assent to Bills passed by the lords and commons,35 to appoint magistrates and establish courts of judicature to dispense justice in its name,36 to conduct foreign relations, and was immune from suits in domestic courts.37 As Blackstone puts it, the Crown ‘may reject what bills, may make what treaties, may coin what money, may create what peers, may pardon what offences he pleases: unless where the constitution hath expressly, or by evident consequence, laid down some exception or boundary’.38 When viewed as a package, Blackstone’s constitutional theory consists primarily of second order power-conferring rules that apportion power between Parliament and the Crown.39 These rules are formal because they provide a system of institutional checks on legislative and executive power without incorporating first order moral principles that condition the exercise of governmental power or hold the

28 Montesquieu, The Spirit of Laws (n 21) 178. 29 Blackstone, Commentaries (n 12) 41 and 98. 30 Montesquieu, The Spirit of Laws (n 21) 181: ‘Of the three powers above mentioned, the judiciary is in some measure next to nothing: there remain, therefore, only two; and as these have need of a regulating power to moderate them, the part of the legislative body composed of the nobility is extremely proper for this purpose.’ 31 Prest, William Blackstone (n 22) chs 5–7. 32 Blackstone, Commentaries (n 12) 107. 33 ibid 102. 34 ibid 100, 121–123. 35 ibid 102. 36 ibid 171–175. 37 ibid ch 7. See particularly at 162. 38 ibid 162. 39 D Dyzenhaus, ‘The Idea of a Constitution: A Plea for Staatsrechtslehre’ in D Dyzenhaus and M Thorburn (eds), Philosophical Foundations of Constitutional Law (Oxford, Oxford University Press, 2016) 9, 10–11.

368  Matthew Lewans governmental officials accountable through the legal process. According to this formalist perspective, constitutional liberty is guaranteed so long as Parliament retains unlimited law-making power to check the Crown’s broadly defined prerogative powers and immunities. But there are no procedural or substantive constraints on the use of parliamentary or executive authority. Nevertheless, Blackstone the jurisprude invokes a thicker conception of constitutionalism in the chapters that preface his map regarding the institutional distribution of governmental power. After noting in his introductory lecture that a competent understanding of English law is an essential aspect of ‘liberal and polite education’, he paused to reflect ‘on the singular frame and polity of that land, which is governed by this system of laws’.40 In doing so, he observed:41 A land, perhaps the only one in the universe, in which political or civil liberty is the very end and scope of the constitution. This liberty, rightly understood, consists in the power of doing whatever the laws permit; which is only to be effected by a general conformity of all orders and degrees to those equitable rules of action by which the meanest individual is protected from the insults and oppression of the greatest.

Later, when elaborating on individual rights, he states that ‘in vain would these rights be declared, ascertained, and protected by the dead letter of the laws, if the constitution had provided no other method to secure their actual enjoyment’.42 While parliamentary sovereignty remained the primary means for safeguarding personal liberty, Blackstone argued that ‘courts of justice must at all times be open to the subject, and the law be duly administered therein’.43 And when expounding on the nature of law in the preface to his constitutional doctrine, he famously set out a series of equitable principles for judicial interpretation of legislation.44 Curiously, Blackstone the constitutional theorist does not consider the broader implications of these principles when delineating the formal scope of legislative or executive power. As a result, there is an unresolved tension between Blackstone’s jurisprudence and constitutional doctrine – the procedural and substantive considerations in his ­philosophical reflections on the nature of law fade out of view in his constitutional theory, which is dominated by an executive branch subject to relatively loose parliamentary oversight. IV.  DICEY ON THE ABSENCE OF ADMINISTRATIVE LAW For the term droit administratif English legal phraseology supplies no proper equivalent. The words ‘administrative law,’ which are its most natural rendering, are unknown to English judges and counsel, and are themselves hardly intelligible without further explanation.



40 Blackstone, 41 ibid

11. 42 ibid 95. 43 ibid. 44 ibid 45–47.

Commentaries (n 12) 10.

Common Understandings of Administrative Law  369 This absence from our language of any satisfactory equivalent for the expression droit administratif is significant; the want of a name arises at bottom from our non-recognition of the thing itself. In England, and in countries which, like the US, derive their civilisation from English sources, the system of administrative law and the very principles on which it rests are in truth unknown.45

While subsequent constitutional theorists employed expository methods similar to Blackstone’s, they drew strikingly different conclusions regarding the conceptual framework that governs the distribution of governmental power. Like Blackstone, AV Dicey sought to separate the legal academy from the bar to remedy deficiencies in professional legal training, which he regarded as ‘fragmentary’, ‘unsystematic’, and ‘wasteful of time and labour’.46 Students needed a clue or map to guide them through the labyrinth of an unwritten constitution.47 ‘[T]he proper sphere of professorial activity’, Dicey declared, is ‘to set forth the law as a coherent whole – to analyse and define legal conceptions – to reduce the mass of legal rules to an order series of principles, and to aid, stimulate, and guide the reform or renovation of legal literature’.48 In this manner, Dicey set out to articulate ‘rules which directly or indirectly affect the distribution or the exercise of the sovereign power in the state’.49 Like Blackstone, Dicey tended to portray constitutional law in both formal and metaphysical terms. A constitutional theorist might scour the common law so as to render an intelligible conceptual framework for understanding the different sources of English law, the reality of that conceptual framework was fixed and did not incorporate moral values into its structure. Nevertheless, Dicey argued that Blackstone’s map was deeply flawed, arguing that ‘[t]he present generation must of necessity look on the constitution in a spirit different from the sentiment either of 1791 or of 1818’.50 He criticised Blackstone’s habit of ‘applying old and inapplicable terms to new institutions, and especially of ascribing in words to a modern and constitutional King the whole, and perhaps more than the whole, of the powers actually possessed and exercised by William the Conqueror’.51 Following Walter Bagehot,52 Dicey located the executive power in the Cabinet, a committee chaired by the Prime Minister but accountable to Parliament and the ordinary courts through the principle of ministerial responsibility.53 The only constraints on Parliament’s sovereignty were extra-legal political constraints54 – an ‘external’ limit that legislation might provoke popular resistance,55 and an ‘internal’ limit that legislators’ opinions would tend to mirror public opinion.56 45 AV Dicey, Introduction to the Study of the Law of the Constitution 10th edn (London, Macmillan & Co, 1960) 330. 46 AV Dicey, Can English Law be Taught in Universities? (London, Macmillan & Co, 1883) 10–12. 47 Dicey, Introduction to the Study of the Law of the Constitution (n 45) 7. 48 ibid 18. 49 ibid 23. 50 ibid 3. 51 ibid 7–8. 52 ibid 19. 53 ibid c XI. 54 ibid 73. 55 ibid 76–79. 56 ibid 80–81.

370  Matthew Lewans Put simply, Dicey rejected Blackstone’s map that ascribed substantial prerogative powers and immunities to the Crown, and instead proposed one that described how ‘royal supremacy has now passed into that sovereignty of Parliament’.57 Therefore, Dicey rejected Blackstone’s characterisation of executive power, because it distorted ‘the true position of the Crown’ by ‘fictitious ascription to the Sovereign of political omnipotence’.58 While Dicey endorsed Blackstone’s definition of parliamentary sovereignty that entailed ‘the right to make or unmake any law whatever’, he excluded the possibility that the Crown or another any other institution might ‘override or set aside the legislation of Parliament’.59 But despite Parliament’s unbounded legislative authority, Dicey discounted the possibility that Parliament might delegate authority to the executive branch to cultivate a body of administrative law. Thus, he notes that:60 In foreign countries the legislature generally confines itself to laying down general principles of legislation, and leaves them with great advantage to the public to be supplemented by decrees or regulations which are the work of the executive. The cumbersomeness and prolixity of English statute law is due in no small measure to futile endeavours of Parliament to work out the details of large legislative changes. This evil has become so apparent that in modern times Acts of Parliament constantly contain provisions empowering the Privy Council, the judges, or some other body, to make rules under the Act for the determination of details which cannot be settled by Parliament. But this is only an awkward mitigation of an acknowledged evil, and the substance no less than the form of the law would, it is probable, be a good deal improved if the executive government of England could, like that of France, by means of decrees, ordinance, or proclamations having the force of law, work out the detailed application of the general principles embodied in the Acts of the legislation. In this, as in some other instances, restrictions wisely placed by our forefathers on the growth of royal power, are at the present day the cause of unnecessary restraints on the action of the executive government.

In this passage, Dicey seems to regret constitutional restraints that inhibit the growth of administrative law. However, elsewhere Dicey celebrates the exclusive jurisdiction of ordinary courts to preserve the rule of law, which for him precludes the emergence of a domestic analogue to droit administratif. This begs an important question: why did Dicey believe that the constitutional jurisdiction of the ordinary courts (which Montesquieu and Blackstone deemed to be of marginal constitutional significance) precluded the emergence of administrative law? A tentative response to this question can be inferred from the work of Alexis de Tocqueville, which crops up repeatedly in Dicey’s work. Tocqueville was equally fascinated and troubled by the emergence of an egalitarian, democratic form of government in the US after the American Revolution. While he thought the American experience predicted a similar transition to constitutional democracy in European countries, he worried that this transition would erode the moral foundation of a political order premised upon the noblesse oblige and sense of civic virtue endemic



57 ibid

183. 10–11. 59 ibid 40. 60 ibid 52–53. 58 ibid

Common Understandings of Administrative Law  371 amongst the traditional ruling classes. Whereas Montesquieu and Blackstone argued that parliamentary sovereignty served as a bulwark against monarchical ­despotism, Tocqueville believed that the omnipotence or sovereignty of public opinion in a constitutional democracy entailed the risk of a new form of tyranny or despotism – that majority interests would overwhelm individual liberty and stifle civil discourse about the common good.61 Like Montesquieu, Tocqueville argued that this risk could be mitigated by revising the formal distribution of governmental power. In the case of the US, Tocqueville argued that the ‘judges of ordinary law’62 served as ‘the sole counterweight of ­democracy’63 by determining the scope of both legislative and executive jurisdiction through constitutional and statutory interpretation.64 Because ‘[m]en who have made law their specialty have drawn from this work habits of order, a certain taste for forms, a sort of instinctive love for the regular succession of ideas’, Tocqueville concluded that judges would remain ‘strongly opposed to the revolutionary spirit and to the unthinking passions of democracy’.65 While judges of ordinary law ‘cannot force the people to make law … at least he compels them not to be unfaithful to their own laws and to remain consistent’.66 Like Tocqueville, Dicey was both fascinated and troubled by the democratisation of government in the UK, as well as the relationship between shifting currents of public opinion and the dramatic pace of legislative reform from 1850 onward.67 Writing in 1914, Dicey noted how the expansion of the franchise ‘revealed to the electorate the extent of their power, and has taught them that political authority can easily be used for immediate advantage, not of the country but of a class’.68 After lamenting the rapid expansion of the administrative state, Dicey noted that ‘[d]espotism may continue to be tyranny, even though it may have become both popular and benevolent’.69 By contrast, judicial legislation prioritised logical consistency, certainty, and ‘the convictions of an earlier era’ to ‘the ideas represented by parliamentary legislation’.70 Most importantly, he concluded that the legislative action of the Courts represents a peculiar cross-current of opinion, which may in more ways than one modify the effect of that predominant opinion of the day which naturally finds expression in a representative assembly such as the House of Commons.71

The profound innovation in Dicey’s constitutional doctrine was to move the judiciary from the margins of the map of English constitutional law to its center by vesting

61 A de Tocqueville, Democracy in America vol1, pt 1, c 1 (Indianapolis, Liberty Fund Inc, 2010) 278. 62 ibid 245. 63 ibid 439. 64 ibid, vol I, pt II, 430–442. 65 ibid 432. 66 ibid 440. 67 AV Dicey, Lectures on the Relation between Law and Public Opinion in England during the Nineteenth Century 2nd edn (London, Macmillan, 1962) Lecture I. 68 ibid lxv and Lecture III. 69 ibid lxxxi. 70 ibid 364–369. 71 ibid 370.

372  Matthew Lewans it with exclusive jurisdiction to interpret parliamentary legislation. The principal constitutional constraints in Blackstone’s model derived from parliamentary legislation (the Petition of Right, the Habeas Corpus Act, the Bill of Rights, and the Act of Settlement),72 but in Dicey’s model the general principles of the constitution (as for example the right to personal liberty, or the right of public meeting) are … the result of judicial decisions determining the rights of private persons in particular cases brought before the courts.73

Thus, he argued that no one could ‘be lawfully made to suffer in body or goods except for a distinct breach of law established in the ordinary legal manner before the ordinary courts of the law’,74 and that ‘every man, whatever be his rank or condition, is subject to the ordinary law of the realm and amenable to the jurisdiction of the ordinary tribunals’.75 Many administrative law scholars attribute the current state of administrative law theory to Dicey’s highly influential constitutional doctrine. Unlike Blackstone’s Commentaries, Diceyan constitutionalism espouses a conception of the rule of law that emphasises the institutional authority of the judiciary.76 By investing common law judges with a monopoly to interpret legislation, the theory ensures ‘the authority of the judges and the fixity of the law’77 which ‘constantly hampers (and sometimes with great injury to the public) the action of the executive’.78 According to this reading, the rule of law excludes the ‘the existence of arbitrariness, of prerogative, or even of wide discretionary authority on the part of government’,79 and guarantees that common law courts determine what the law requires. V.  COMMON UNDERSTANDINGS OF ADMINISTRATIVE LAW

Even though Dicey’s extravagant claim that the concept of droit administratif is utterly unknown in England has been largely discredited, his constitutional paradigm still serves as an intellectual touchstone for contemporary administrative law scholarship. As a result, theoretical debates in administrative law are often framed by the apparent tension Dicey noted80 between parliamentary sovereignty and the rule of law.81 However, Dicey’s reflections upon the relationship between the principles of 72 Dicey, Introduction to the Study of the Law of the Constitution (n 45) 28. 73 ibid 195. 74 ibid 188. 75 ibid 193. 76 See, eg, W Robson, ‘The Report of the Committee on Ministers’ Powers’ (1932) 3 Political Quarterly 346; J Willis, ‘Three Approaches to Administrative Law: The Judicial, the Conceptual, and the Functional’ (1935) 1 University of Toronto Law Journal 53; F Frankfurter, ‘Foreword’ (1938) 47 Yale Law Journal 515; HW Arthurs, ‘Rethinking Administrative Law: A Slightly Dicey Business’ (1979) 17 Osgoode Hall Law Journal 1; M Loughlin, ‘Functionalist Style in Public Law’ (2005) 55 University of Toronto Law Journal 361; M Lewans, Administrative Law and Judicial Deference (Oxford, Hart Publishing, 2016) ch 2. 77 Dicey, Introduction to the Study of the Law of the Constitution (n 45) 408. 78 ibid 411. 79 ibid 202. 80 ibid ch XIII. 81 Forsyth, Judicial Review & the Constitution (n 1).

Common Understandings of Administrative Law  373 constitutional and administrative law are richer and more nuanced than many of these debates suggest. Because Dicey’s constitutional theory is relatively abstract, complex, and rife with subtle tensions, it can be construed in ways that are more or less congenial to attempts to develop a theory of administrative law qua law to complement existing theories of legislation and the common law within a constitutional democracy. Each of these interpretive approaches conceives the point or purpose of constitutional theory differently and implies different political premises and analytical methods. What is striking is that, while each strain of thought reflects different aims, each enlists many of the argumentative strategies and methods pioneered by Blackstone and Dicey in their discussions regarding the constitutional character of executive power. A.  Constitutional Formalism and Administrative Lawlessness The first approach assumes that the purpose of constitutional law is to maintain a conceptually coherent, formal distribution of power between Parliament, the judiciary, and the executive. As I noted earlier, this approach crops up in both Blackstone’s and Dicey’s work. Blackstone argued that, while a constitutional theorist might render an intelligible and harmonious conceptual framework for understanding the different sources of English law, that framework did not incorporate first order constraints on the exercise of governmental power nor was it subject to revision or normative critique. Thus, when seeking to elucidate the principal sources of law under the English Constitution, he identified a set of power-conferring rules that explained Parliament’s extraordinary legislative jurisdiction while leaving ordinary matters of law and policy to be determined through the exercise of the Crown’s prerogative powers, which were immune from judicial review.82 While Dicey was attracted to Blackstone’s expository method and definition of parliamentary sovereignty, he articulated a revolutionary conception of the rule of law that gives the judiciary an exclusive licence to interpret and determine questions of law – a distribution that subordinates the executive to the legislative and judicial branches of government and excludes, by definition, the emergence of a domestic analogue to droit administratif.83 Thus, both Blackstone’s and Dicey’s constitutional theories employ similar rhetorical strategies, which claim a ‘scientific’ basis for the conceptual definitions and logical distinctions they deemed to be constitutionally significant.84 Most importantly, these definitions and logical distinctions are couched in descriptive, at times metaphysical, terms which eschew the incorporation of any moral or political values into the division of institutional labour.85 In this vein, Dicey argues

82 See above nn 23–45 and accompanying text. 83 See above nn 57–83 and accompanying text. 84 D Sugarman, ‘The Legal Boundaries of Liberty: Dicey, Liberalism and Legal Science’ (1983) 46 MLR 102. 85 D Dyzenhaus, ‘Formalism’s Hollow Victory’ [2002] New Zealand Law Review 525.

374  Matthew Lewans that constitutional law concerns those ‘rules which directly or indirectly affect the distribution … of the sovereign power in the state’,86 and that ‘[t]he duty … of an English professor of law is to state what are the laws which form part of the constitution, to arrange them in their order, to explain their meaning, and to exhibit where possible their logical connection.’87 While this approach continues to attract adherents,88 it gained prominence throughout the first half of the twentieth century during the emergence of the modern administrative state. In 1929, Lord Hewart adapted Tocqueville’s observations regarding the tyrannical tendency of democratic government by arguing that the administrative state was a ‘new despotism’, which vaunted ‘Government departments above the Sovereignty of Parliament and beyond the jurisdiction of the Courts.’89 In order to distinguish between constitutionally legitimate and illegitimate forms of law, Hewart posited an essential conceptual distinction between ‘judicial’ and ‘administrative’ modes of decision-making – a distinction that mapped onto the distinction between law and politics. Judicial decision-making aligned with a declaratory theory of law defined by a responsible, transparent, fair, and impartial decision-making process reflected by the common law tradition. Administrative adjudication was the lawless antithesis of the judicial method: it was creative, in the sense that discretionary power was not governed by statutory or common law standards; it was secretive, in the sense that administrative decisions were not rendered in public; and it was unfair, in the sense that those subject to the decision had no right to be heard. As Hewart put it, ‘[t]here are no rules or principles which can be said to be rules or principles of this astonishing variety of administrative “law”, nor is there any regular course of procedure for its application’.90 Curiously, even though Lord Hewart thought the administrative state undermined the English Constitution, he authored a decision that immunised broad swathes of administrative decisions from judicial scrutiny.91 In Rex v Legislative Committee of the Church Assembly, he held that administrative decisions that affected the legal rights of subjects were not necessarily subject to judicial review. In his view, ‘there must be superadded to that characteristic the further characteristic that the body has the duty to act judicially’.92 The upshot of Hewart’s conceptual distinction is that administrative officials operate in a political realm beyond the rule of law.93 Hence, neither an administrative decision-making process nor the substance of an administrative decision are constrained by the principles of legality. This view, which

86 Dicey, Introduction to the Study of the Law of the Constitution (n 45) 23. 87 ibid 32. 88 See, eg, C Forsyth, ‘Showing the Fly the Way out of the Flybottle: The Value of Formalism and Conceptual Reasoning in Administrative Law’ (2007) 66 CLJ 325; J Grant, ‘Reason and Authority in Administrative Law’ (2017) 76 CLJ 507. 89 G Hewart, The New Despotism (London, Ernest Benn Ltd, 1929) 14. 90 ibid 44. For a similar analytical approach, see DM Gordon, ‘Administrative Tribunals and the Courts’ (1933) 49 LQR 94. 91 Rex v Legislative Committee of the Church Assembly [1928] 1 KB 411, 415. 92 ibid. 93 D Dyzenhaus, The Constitution of Law: Legality in a Time of Emergency (Cambridge, Cambridge University Press, 2006) 123.

Common Understandings of Administrative Law  375 resurfaces in subsequent decisions from the House of Lords,94 the Judicial Committee of the Privy Council,95 and the Court of Appeal96 portrays administrative decisions as a species of uncontrollable political power akin to Blackstone’s portrayal of the Crown’s ­prerogative powers or Dicey’s understanding of droit administratif. The only recourse for abuse of administrative power is to the principle of Ministerial responsibility through a parliamentary process that is political and, therefore, non-justiciable.97 B.  Liberty from Administrative Law The second approach to the subject celebrates the judiciary’s exclusive right to determine and enforce statutory parameters in order to preserve individual freedom from administrative law. As I noted earlier, Blackstone argued that liberty was ‘the very end and scope of the constitution’, which ensured ‘a general conformity of all orders and degrees to those equitable rules of action by which the meanest individual is protected from the insults and oppression of the greatest’.98 While Dicey censured Blackstone’s characterisation of executive power, he expounded on the theme of constitutional liberty that crops up in Blackstone’s work. At times, it seems that Dicey’s revolutionary account of the rule of law contained a reactionary political edge motivated by a sense of rugged individualism and laissez-faire economic policy that held sway during the Victorian era.99 Following Tocqueville, Dicey argued that ‘judicial legislation’ provided an important cross-current of p ­ olitical opinion, which tempers the influence of public opinion on the substance of the law.100 Whereas Dicey regarded the judiciary as ‘the best and soundest of English institutions’101 and judicial legislation as ‘the best part of the law of England’,102 he was driven to a state of despair by the state of the parliamentary process and policy at the turn of the twentieth century.103 In less guarded passages noting the declining influence of nineteenth century individualism and the rising tide of collectivist reforms, he railed variously against legislative policies that have since become core competencies of the administrative state – public education, workers’ compensation, unemployment insurance, and social assistance.104 Thus, when Dicey argues that ‘judges … are disposed to construe statutory exceptions to common law principles in a mode

94 R v Halliday, ex parte Zadig [1917] AC 260 (HL); Liversidge v Anderson [1942] AC 206 (HL); Franklin v Minister of Town and Country Planning [1947] AC 87 (HL); Smith v East Elloe Rural District Council [1956] AC 736 (HL); Secretary of State for the Home Department v Rehman [2003] 1 AC 153 (HL). 95 Nakkuda Ali v Jayaratne [1951] AC 66 (JCPC). 96 R v Secretary of State, ex parte Hosenball [1977] 1 WLR 766. 97 For additional illustrations of the formal and conceptual mode of analysis see Report of the Committee on Ministers’ Powers, Cmd. 4060, 73–81; Gordon, ‘Administrative Tribunals and the Courts’ (n 90). 98 Blackstone, Commentaries (n 12) 11. 99 See, eg, Dicey, Lectures on the Relation between Law and Public Opinion (n 67), esp the Introduction to the Second Edition and Lecture XI. 100 ibid 369–370. 101 AV Dicey, ‘Judicial Policy’ (1874) 29 Macmillan’s 473. 102 Dicey, Law and Public Opinion (n 67) 361. 103 ibid 370 (fn 1). 104 ibid Lecture VIII.

376  Matthew Lewans which would not commend itself either to a body of officials, or to the Houses of Parliament’,105 his theory can be construed as incorporating a conception of judicial review that safeguards negative liberty and freedom of contract from administrative law.106 Friedrich Hayek’s conception of the rule of law amplifies this theme in Dicey’s work.107 Hayek couched his conception of the rule of law within a broader libertarian argument that ‘[t]he fundamental principle that in the ordering of our affairs we should make as much use as possible of the spontaneous forces of society, and resort as little as possible to coercion’.108 To this end, he argued the legitimate scope of state action was confined to ‘co-ordinating human efforts’ by providing a framework of general rules to protect freedom of contract and preserve a free market for entrepreneurial competition.109 Thus, he argued the law serves to provide notice ‘what action the state will take in certain types of situation, defined in general terms without reference to time and place or particular people’.110 However, [a]ny attempt to control prices or quantities of particular commodities deprives competition of its power of bringing about an effective co-ordination of individual efforts, because price changes then cease to register all the relevant changes in circumstances and no longer provide a reliable guide for the individual’s actions.111

Conversely, Hayek warned that state intervention, especially in the economic domain, undermined the moral foundation of western civilization, noting that ‘[a]lthough we had been warned by some of the greatest political thinkers of the nineteenth century, by Tocqueville and Lord Acton, that socialism means slavery, we have steadily moved in the direction of socialism’.112 Like Hewart, Hayek thought that enabling legislation, which sets out a general framework of principles and objectives to be interpreted and applied by administrative officials, was the moral equivalent to arbitrary despotism ‘because the matter in hand cannot be regulated by general rules but only by the exercise of discretion in the decision of particular cases’.113 Citing Dicey, Hayek argued that the rule of law is intrinsically opposed to the notion of administrative law, regardless of the procedural or substantive propriety of particular administrative decisions.114 In his words, the rule of law ‘means that government in all its actions is bound by rules fixed and announced beforehand’,115 which entails that ‘the discretion left to the executive organs wielding coercive power should be reduced as much as possible’.116 In short, both Hewart and Hayek conclude that administrative law 105 Dicey, Introduction to the Study of the Law of the Constitution (n 45) 413–4. 106 In this respect, some of Dicey’s reflections on political history and constitutional law at the turn of the twentieth century resonate with political and constitutional debates occurring in the United States of America during this same period. See Lewans, Administrative Law and Judicial Deference (n 76) 4. 107 F Hayek, The Road to Serfdom (Chicago, University of Chicago Press, 1944). 108 ibid 17. 109 ibid 36–39. 110 ibid 74. 111 ibid 37. 112 ibid 13. 113 ibid 66. 114 ibid 77–78. 115 ibid 72. 116 ibid.

Common Understandings of Administrative Law  377 is (by definition) constitutionally illegitimate and constitutionally irredeemable. The difference is that Hayek mounts an explicitly political, as opposed to conceptual, assault on enabling legislation that provides the legal foundation of administrative law and the administrative state. Roberts v Hopwood illustrates how Hayek’s conception of the rule of law bolsters judicial review at the expense of administrative law. In that case, the court held that even though Parliament had given the borough of Poplar the authority to determine what wages to pay its employees, the council’s minimum wage policy was illegal.117 The members of the court were alarmed by the council’s decision, because it facilitated an involuntary and illegitimate transfer of wealth from ratepayers to public employees.118 Lord Buckmaster stated that the council’s policy ‘took an arbitrary principle and fixed an arbitrary sum, which was not a real exercise of the discretion imposed upon them by the statute’;119 and Lord Atkinson complained that the council had ‘allowed themselves to be guided in preference by some eccentric principles of socialistic philanthropy, or by a feminist ambition to secure equality of the sexes in the matter of wages in the world of labour’.120 The unifying theme of the unanimous decision resonates with Hayek’s antipathy for price and wage controls. While Roberts v Hopwood is the most striking example of this line of thought, other cases provide additional evidence for this perspective of administrative law.121 As JAG Griffith famously argued, the case reports abound with cases reflecting a judicial conception of the public interest that prioritises established order, vested interests, and private property rights while taking a relatively dim view of organised labour, minority interests, and redistributive policies.122 C.  Formal Equality Through Judicial Review The third approach also celebrates the judiciary’s exclusive right to determine what the law requires, but on egalitarian grounds. Although Parliament possesses sovereign legislative power, it can only exercise that power by passing legislation that is ultimately subject to judicial interpretation. The point of ensuring that judges determine questions of law is to secure the formal equality of legal subjects through the judicial process that determines what the law requires.123 This perspective builds upon Blackstone’s argument that the ‘courts of justice must at all times be open to the subject, and the law be duly administered therein’,124 which supplies the germ for Dicey’s conception of legality: (1) that no one ‘can be lawfully made to suffer in body or goods except for a distinct breach of the law established in the ordinary legal



117 Metropolis

Management Act 1855 (18 & 19 Vict c 120) s 62. The Road to Serfdom (n 107) ch VI. 119 ibid 590. 120 ibid 594. 121 JAG Griffith, The Politics of the Judiciary 5th edn (London, Fontana, 1997). 122 ibid 336. 123 TRS Allan, ‘Dworkin and Dicey: The Rule of Law as Integrity’ (1988) 8 OJLS 266. 124 Blackstone, Commentaries (n 13) 95. 118 Hayek,

378  Matthew Lewans manner before the ordinary courts of the land’;125 (2) that ‘every man, whatever be his rank or condition, is subject to the ordinary law of the realm’;126 and (3) that constitutional principles are ‘the result of judicial decisions determining the rights of private persons in particular cases brought before the courts’.127 This egalitarian reading portrays Dicey’s constitutional theory as a forerunner to Ronald Dworkin’s interpretive theory of law organised around legislative and adjudicative principles of political integrity.128 The former holds that legislators ought to enact laws that reflect a morally coherent system of public policy;129 the latter holds that judges ought ‘to identify legal rights and duties, so far as possible, on the assumption that they were all created by a single author … expressing a coherent conception of justice and fairness’.130 A political community flouts integrity whenever it ‘enacts and enforces different laws each of which is coherent in itself, but which cannot be defended together as expressing a coherent ranking of different principles of justice or fairness or procedural due process’.131 Framed in this way, the value of integrity is egalitarian in that it ‘assumes that each person is as worthy as any other, that each must be treated with equal concern according to some coherent conception of what that means’.132 Although Dworkin’s jurisprudential theory focuses on the substance of common law adjudication, it clearly has broader institutional implications for the prospect of administrative law and administrative law theory. Elsewhere Dworkin argues that, while legislatures convert policy arguments into law, the resolution of interpretive disagreements regarding the law is a matter of principle that judges should decide.133 Without more, Dworkin’s theory of law ignores the basic crux of administrative law, which is that legislatures delegate authority to administrative officials to determine what the law requires on behalf of the political community.134 The institutional upshot of Dworkin’s theory is apparent in cases like Anisminic v Foreign Compensation, where the House of Lords held that common law judges are entitled to superimpose their interpretation of the law whenever it deviates from the substantive merits of an administrative decision.135 But to reach that conclusion, the

125 Dicey, Introduction to the Study of the Law of the Constitution (n 5) 188. 126 ibid 193. 127 ibid 195. 128 R Dworkin, Law’s Empire (Oxford, Hart Publishing, 1998) ch 6. 129 ibid 217. 130 ibid 225. 131 ibid 184. 132 ibid 213. 133 R Dworkin, Taking Rights Seriously (London, Duckworth, 1977) 110. See also Dworkin, Law’s Empire (n 128) 244. 134 See, eg, M Bayles, ‘The Justification of Administrative Authority’ in J Roland Pennock and J Chapman (eds), Authority Revisited (1987) 29 Nomos 287 (New York, New York University Press, 1987); H Richardson, Democratic Autonomy: Public Reasoning about the Ends of Policy (Oxford, Oxford University Press, 2002) ch 10; J Waldron, ‘Authority for Officials’ in Meyer, S Paulson and T Pogge (eds), Rights, Culture, and the Law: Themes from the Legal and Practical Philosophy of Joseph Raz (Oxford, Oxford University Press, 2003) 45. 135 Anisminic v Foreign Compensation Commission [1969] 2 AC 147 (HL). See also R v Hull University Visitor, ex parte Page [1993] AC 682 (HL); Huang v Secretary of State for the Home Department [2007] 2 AC 167 (HL).

Common Understandings of Administrative Law  379 judges had to explain why the court was entitled to intervene despite the presence of a privative clause declaring that the Commission’s decision ‘shall not be called into question in any court of law’.136 Privative clauses pose a dilemma for both Dicey’s and Dworkin’s theories, because they direct judges to refrain from reassessing the substance of administrative decisions while the adjudicative principle of integrity instructs judges to interpret and enforce the law as a coherent system of principle. In Anisminic, Lord Reid’s solution to this dilemma was to construe the privative clause strictly so as to preserve ‘the ordinary jurisdiction of the court’ to determine questions of law.137 In his view, the effect of the Order in council ‘was to confer legal rights on persons who might previously have hoped or expected that in allocating any sums available discretion would be exercised in their favour’.138 In other words, the Order had converted an argument of public policy into a matter of principle that entitled the court to consider whether the Commission’s decision was nullified by failing to observe principles of natural justice, misconstruing the provisions of the enabling legislation, failing to take relevant considerations into account, or considering irrelevant considerations.139 But the practical effect of this line of reasoning is to eviscerate the privative clause in order to preserve the ability of the judiciary to determine whatever substantive questions of law might arise in the course of the Commission’s statutory mandate. D.  Natural Justice and Administrative Law The fourth approach draws attention to the relationship between the fairness of the ordinary legal process and the constitutional legitimacy of administrative decisions. According to this interpretation, when Blackstone and Dicey insisted that no one should be subjected to the coercive power of the state unless the law is interpreted and administered equitably they were anticipating the mid-century renaissance in English administrative law ushered in by Ridge v Baldwin.140 Mark Walters argues persuasively that Dicey’s antipathy towards administrative law waned over the course of his academic career as he was prompted to develop a conception of legality premised upon the legitimacy of due process as opposed to judicial supremacy over substantive legal interpretation.141 As Walters puts it, the constitutional conundrum posed by the administrative state is not that it changes the formal distribution of political power per se, but that it raises new questions about ‘the distinctive interpretive process adopted in the articulation of law’.142 Beginning in 1889, Dicey responded to academic critics who argued that his 1885 description of droit administratif was a grotesque caricature of French public law. 136 Foreign Compensation Act 1950, s 4(4). 137 Anisminic (n 135) 170. 138 ibid 172. 139 ibid 171. 140 Ridge v Baldwin [1964] AC 40 (HL). 141 M Walters, ‘Public Law and Ordinary Legal Method: Revisiting Dicey’s Approach to Droit Administratif’ (2016) 66 University of Toronto Law Journal 53. 142 ibid 75.

380  Matthew Lewans Eventually, he conceded that droit administratif ‘is utterly different from the capricious prerogatives of despotic power’,143 because French administrative tribunals had been separated from the executive branch of government and adopted judicial modes of argumentation, adjudication, and justification.144 Thus, he concluded that droit administratif was apt to become ‘as completely a branch of the law of France (even if we use the word ‘law’ in its very strictest sense) as Equity has for more than two centuries become an acknowledge branch of the law of England’.145 A similar shift can be detected in Dicey’s reflections on the development of English administrative law.146 The decision in Local Government Board v Arlidge, Dicey noted, had established the principle that administrative decision-makers were ‘bound to act with judicial fairness and equity’ even though they were ‘not in any way bound to follow the rules of procedure which prevail in English courts’.147 While he noted that this development marked ‘a considerable step towards the introduction among us of something like the droit administratif of France’, he remained confident that judicial review of administrative decisions was sufficient to preserve the rule of law.148 This idea that administrative decisions are legally legitimate insofar as they are rendered in a manner which is consistent ‘the spirit of judicial fairness and equity’ was one that Dicey struggled to develop in his later work,149 and provides fertile ground for a theory of administrative law that explains and justifies why administrative decisions ought to impact both legal subjects’ and legal officials’ understanding about what the law requires.150 When interpreted in light of more recent, procedurally-oriented conceptions of legality, this approach leaves meaningful space for administrative law in an institutionally complex constitutional democracy. For example, Lon Fuller’s conception of the rule of law does not hinge on a particular division of institutional labour or substantive conception of justice. Rather, he portrays the tasks of legislating, adjudicating, and enforcing the law as different aspects in a broader, seamless project of maintaining the legitimacy of the legal order – a project he refers to compendiously as ‘the morality that makes law possible’.151 The integrity of the legal order, according to Fuller, is a modus vivendi whereby fidelity to law accrues by virtue of state action that comports with principles of legality – generality, publicity, prospectivity, intelligibility, consistency, practicability, stability, and congruence in the administration

143 Dicey, Introduction to the Study of the Law of the Constitution (n 45) 369. 144 ibid 367–8. 145 ibid 382–3. 146 AV Dicey, ‘The Development of Administrative Law in England’ (1915) 31 LQR 148. 147 ibid 149. 148 ibid 152. 149 Dicey, ‘Development’ (n 146) 149. See also AV Dicey, ‘Droit Administratif in Modern French Law’ (1901) 17 LQR 302. 150 See, eg, J Mashaw, Due Process and the Administrative State (New Haven, Yale University Press, 1985); Bayles, ‘The Justification of Administrative Authority’ (n 134); D Dyzenhaus, ‘Law as Justification: Etienne Mureinik’s Conception of Legal Culture’ (1998) 14 South African Journal of Human Rights 11; D Dyzenhaus, ‘Hobbes and the Legitimacy of Law’ (2001) 20 Law and Philosophy 461; J Waldron, ‘The Rule of Law and the Importance of Procedure’ in J Fleming (ed), Getting to the Rule of Law (New York, New York University Press, 2011) 3. 151 L Fuller, The Morality of Law (New Haven, Yale University Press, 1964) ch II.

Common Understandings of Administrative Law  381 of the law.152 The important point is that the character of this project leaves open the possibility that administrative processes and decisions are capable of generating law while maintaining to the legitimacy of the legal order writ large. When viewed in this light, one gains a deeper appreciation for the theoretical implications of Ridge v Baldwin – the landmark case in which Lord Reid held that the decision of the Brighton Watch Committee was unenforceable because it had dismissed Constable Ridge without a hearing.153 Over the course of his opinion, Lord Reid noted that while the UK did not ‘have a developed system of administrative law’, he saw ‘nothing in that to justify our thinking that our old methods are any less applicable today than ever they were to the older types of cases’.154 After noting that ‘[t]he principle of audi alteram partem goes back many centuries in our law’ and carefully tracing the lineage of case law to substantiate that claim, he roundly rejected Lord Hewart’s conceptual presupposition that the duty applied to judicial (but not administrative) decisions. Referring to Nakkuda Ali v Jayaratne, a recent decision that employed Hewart’s logic, Lord Reid stated that it ‘was given under a serious misapprehension of the effect of the older authorities and therefore cannot be regarded as authoritative’.155 Furthermore, he rejected the notion that the legality of the Watch Committee’s decision should be assessed according to the private law of contract, which held that a ‘master can terminate the contract with his servant at any time and for any reason or for none’.156 If freedom of contract provided the constitutional basis for judicial review, the Watch Committee was legally entitled to terminate Ridge’s employment unfairly, but efficiently. But Lord Reid rejected this argument as well, ‘because a Chief Constable is not the servant of the Watch Committee or indeed of anyone else’.157 Instead, he appealed to an alternative line of cases, holding that the Watch Committee could not terminate Ridge unless it ‘informed the constable of the grounds on which they propose to proceed and have given him a proper opportunity to present his case in defence’.158 Moreover, the requirements of natural justice were not fixed precisely by the model of rules set out in the Police Act or the related regulations, but were common law principles that had to be weighed and calibrated according to the context of the particular case.159 Finally, despite concluding that the Watch Committee’s decision was void, Lord Reid resisted the temptation to opine on whether Ridge should be terminated based on the allegations that had been made at his criminal trial. Thus, he declared that ‘I do not think that this House should do more than declare that the dismissal of the Appellant is null and void and remit the case to the Queen’s Bench Division for further

152 ibid. 153 Ridge v Baldwin [1964] AC 40 (HL). 154 ibid. 155 ibid. 156 ibid. 157 ibid. 158 ibid. 159 SH Bailey, ‘Ridge v Baldwin [1964]: “Nuff Said”’ in S Juss and M Sunkin (eds), Landmark Cases in Public Law (Oxford, Hart Publishing, 2017) 11, 26.

382  Matthew Lewans procedure.’ Shortly afterwards, Ridge tendered his resignation, which preserved his pension benefits and concluded the matter. VI. CONCLUSION

Juxtaposing these different approaches enables one to identify various perspectives on administrative law and illustrates how those perspectives have been shaped by the historical development of constitutional theory. As we have seen, the constitutional theories developed by Blackstone and Dicey were heroic insofar as they rendered an inchoate common law constitution intelligible. And, unsurprisingly, to a large extent their constitutional theories replicated the political zeitgeist of the legal order they inhabited or desired to inhabit. Nevertheless, they both misrepresented the character of constitutional law insofar as they employed metaphysical rhetoric to describe the institutional framework as static and immutable. Such constitutional metaphysics glosses over significant, but interesting, methodological and theoretical tensions. Thus, it seems that even the most orthodox constitutional theories can be enlisted in support of strikingly different approaches to the same subject. However, once the metaphysical veneer is stripped away, we can better appreciate the intellectual inspiration for those diverse mindsets. In addition to deepening our historical understanding of constitutional theory, clearing the ground in this way provides room for a more cogent theory of administrative law – one that explains and justifies the extent to which administrative decisions impact both subjects’ and legal officials’ understanding about what the law requires. As we have seen, that question has been given rather short shrift based on arguments and modes of analyses that appear anachronistic in light of the substantial role that administrative institutions have for creating, interpreting, and enforcing the law in an institutionally complex constitutional order. But by grappling with the underlying premises of those different approaches, we might develop a new way of understanding the constitutional legitimacy of administrative law.

15 Methodology in Constitutional Theory: The Case of the Administrative State KRISTEN RUNDLE*

I. INTRODUCTION

T

he term ‘administrative law’ might be understood in two senses. The first describes administrative law in its supervisory aspect: the body of legal rules and principles that have developed in association with the judicial review of administrative action. The second describes administrative law in its facilitative aspect: the legal and institutional frameworks through which administrative power comes to be constituted and exercised in the first place. Few would disagree that the first, supervisory, species of administrative law has been the primary focus of administrative law scholarship for generations. The rules and principles of judicial review have accordingly also been the primary object of inquiry when the relationship between administrative law and constitutional law comes up for consideration. Subject to local variations, administrative law scholars have dedicated much effort to debating the implications of the constitutional separation of legislative, executive and judicial power for the ambit of supervisory jurisdiction over administrative action, the source (statutory or common law) of the legal principles through which this judicial review function is discharged, and the constitutional limits of the legislature’s ability to remove such supervision altogether. As one would expect, different ways of thinking about the structure of constitutional arrangements produce different answers to these questions.1

* Sincere thanks to Mairead O’Connor for her insightful research assistance, and to Dimitrios Kyritsis, Will Bateman, Scott Stephenson and Vanessa MacDonnell for helpful discussions en route. 1 As indeed is most obviously portrayed in the longstanding ‘ultra vires’ debate that has long preoccupied administrative law scholars in the UK. See C Forsyth (ed), Judicial Review and the Constitution (Oxford, Hart Publishing, 2000) for an overview of major theoretical contributions to that debate. A doctrinal treatment of the Australian version of the same debate can be found in S Gageler, ‘The Underpinnings of Judicial Review of Administrative Action: Common Law or Constitution?’ (2000) 28 Federal Law Review 303, and B Selway, ‘The Principle behind Common Law Judicial Review of Administrative Action – The Search Continues’ (2002) 30 Federal Law Review 217.

384  Kristen Rundle It has been historically less common for administrative law scholars to engage with the constitutional dimensions of administrative law in the second, facilitative, sense: the actors and activities that might loosely be grouped under the label the ‘administrative state’. An innovative trajectory of inquiry has, however, begun to bring this more unruly domain of study into constitutional focus. The motivations behind these new engagements are invariably tied to particular political and legal developments in particular jurisdictions. Nonetheless, a deeper theoretical impulse is discernible, as also is a distinctly elevating aim. To frame study of the administrative state in a constitutional register is to urge greater appreciation of the significance of its actors and activities to conditions of contemporary constitutionalism. My aim in this chapter is to show how and why these efforts to ‘constitutionalise’ discussion of actors and activities of the administrative state present a provocative challenge to contemporary constitutional theory. The invitation presented in such efforts is not simply to revive debates about whether the administrative state is constitutional.2 Rather, the invitation is to reconsider whether the orthodox separation of powers account of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by mechanisms of judicial oversight, remains descriptively and normatively serviceable to contemporary conditions. By exploring new ways of articulating the relation between the actors and activities of the administrative state and the constitutional order, contributions to this novel ‘constitutionalising’ project necessarily demand examination of the explicit and implicit methodological commitments that have determined the location of the administrative state within constitutional thought and practice to this point. Looking ahead, these interventions also prompt fresh thinking on how efforts to bring the administrative state more squarely into constitutional focus might best be pursued, and why the characteristics of constitutional government in contemporary conditions make it timely to do so. The discussion proceeds as follows. I begin by addressing the threshold methodological challenge that has attended all theoretical engagement with the administrative state so far: namely, the problem of delineating the object of inquiry itself. I then turn in section III to outline the innovative efforts of Gillian Metzger3 and Kate Glover Berger4 to argue for the constitutional ‘status’ of the administrative state in the US and Canada respectively. At the heart of these ‘status’ interventions is a refusal to approach the administrative state solely through the received constitutional orthodoxy of the separation of legislative, executive and judicial power, without more. The disruptive methodological challenge presented in arguments for the constitutional

2 A debate that has an especially long history in the US. See further n 16 below. 3 GE Metzger, ‘Foreword: 1930s Redux: The Administrative State Under Siege’ (2017) 131 Harvard Law Review 1. 4 KG Berger, ‘The Constitutional Status of the Administrative State’ [forthcoming]. This piece further develops arguments articulated by the author in KG Berger, ‘The Constitution of the Administrative State’, in D Jutras and M Moore (eds), Canada’s Chief Justice: Beverley McLachlin’s Legacy of Law and Leadership (2018) 86 Supreme Court Law Review (2d) 161; KG Berger ‘Dunsmuir and the Constitutional Status of the Administrative State’ in P Daly and L Sirota (eds) [2018] Canadian Journal of Administrative Law and Practice Special Issue – A Decade of Dunsmuir 141; and KG Berger ‘The Administrative Demands of Unwritten Constitutional Principles’ [forthcoming].

The Case of the Administrative State  385 status of the administrative state is therefore immediately evident in how these interventions seek reappraisal of the units of constitutional analysis itself. In section IV I turn to consider how this wider provocation might be directed towards my own jurisdiction, Australia, where the orthodox constitutional account of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by mechanisms of judicial oversight holds firm. I begin by examining how that account informs the prevailing technique for determining whether actors and activities of the administrative state are answerable to constitutional demands: namely, by reference to the amenability of such activities to supervision by the judiciary within the boundaries of its constitutional relationship to the executive and legislature respectively. The necessarily limited gaze of this ‘supervisibility’ analysis, along with the associated question of whether the relevant administrative power or action is ‘valid’, has ensured that Australian constitutional engagement with actors and activities of the administrative state remains tightly bound to the orthodox separation of powers account. The question thus becomes whether something important is lost to constitutional theory, in its engagement with the administrative state, through this strict adherence to the dominant account and the patterns of constitutional thought and practice that it sustains. Australian examples suggest that the significance of this question becomes especially acute in how the orthodox separation of powers analysis tends to sideline acknowledgment of the central role played by actors and activities of the administrative state in giving concrete expression to relationships between government and the governed. My aim in section V is then to draw on the questions and insights raised in sections III and IV to reflect on why the attempt to elevate engagement with the administrative state to the constitutional register presents such a provocative challenge to contemporary constitutional theory. The three specific interventions in focus – Metzger’s American analysis, Berger’s Canadian analysis, and my own contribution from an Australian perspective – might each have different catalysts and targets in view. What cuts across all three, however, are concerns about descriptive inaccuracy and normative inadequacy in relation to how (if at all) the actors and activities of contemporary administrative government are represented in constitutional thought. Descriptively, positioning the administrative state at the starting point of constitutional analysis, rather than turning to it as an afterthought, reveals that the larger complex of actors and relationships that comprise the actual workings of constitutional government is often more remote from interactions between the executive, legislature and judiciary than the dominant preoccupations of constitutional theory would have it. Commencing inquiry in constitutional theory from the administrative state thus invites an approach capable of encompassing not simply overarching constitutional structure but also infrastructure, even if – to borrow Martin Loughlin’s formulation – this means that ‘in place of a clear, symmetrical, rule-based constitutional structure, we are obliged to examine a complex arrangement of government’.5



5 M

Loughlin, Foundations of Public Law (Oxford, Oxford University Press, 2010) 416.

386  Kristen Rundle Commencing inquiry from the administrative state equally invites new directions for contemporary constitutional theory in its normative guise. Endorsing Metzger’s point that to assign constitutional status to the administrative state is to promote needed engagement with the ‘scope and nature of constitutional obligations to govern’,6 my argument in section V is that engaging with the administrative state in the constitutional register promotes greater attention to how its actors and activities stand at the forefront of the constitutional relationship that we each experience directly: the relationship between government and the governed. Indeed, as I highlight in my concluding comments in section VI, attention to the kinds of administrative relationships between government and the governed that are supported within a constitutional order is poised to become all the more important in an era in which a range of emergencies present unprecedented challenges to the ambit of twenty-first century administrative government, and in which faith in the relationally impoverished techniques of neoliberal government appears to be in decline. The overarching message of efforts to elevate engagement with the administrative state to the constitutional register, as I read it, is accordingly this. To accept the standard constitutional account of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by robust mechanisms of judicial oversight is to place a limit on our understanding of what actually comprises the framework of constitutional government in contemporary conditions, and, by extension, on the questions with which contemporary constitutional theory might engage. This is precisely why efforts to ‘constitutionalise’ discussion of actors and activities of the administrative state are ultimately best understood as provocations directed to the methodology of constitutional theory. Irrespective of whether those efforts seek to challenge, defend or develop the received separation of powers account of the administrative state, and irrespective of their jurisdiction-specific motivations, each invites us to see how choices about the units of analysis considered appropriate to constitutional theory lie at the heart of constitutional orthodoxy as much as they do the prospects of its reconsideration. II.  CONSTITUTIONALISING THE ADMINISTRATIVE STATE: DELINEATING THE OBJECT OF INQUIRY

To inquire into where and how the administrative state fits within theories of constitutionalism is to confront a number of methodological challenges at the outset. These begin with the task of delineating the object of inquiry itself: the sprawling range of actors and activities involved in the administration of government power that might be caught by the description ‘the administrative state’. The ambit of this term could refer narrowly to a particular public bureaucratic order, such as designated government departments or agencies, or much more widely to ‘all the actors and activities

6 Metzger, ‘Foreword’ (n 3) 95. See further Metzger’s arguments that the administrative state ‘empowers and provides the means for effective governance’, and that ‘[m]aking government effective is one of the administrative state’s most important constitutional functions’. (ibid 85, 86).

The Case of the Administrative State  387 involved in fashioning and implementing national regulation and administration – including that which occurs in hybrid forms and spans traditional public–private and nation–state boundaries’.7 Within as inclusive a description as the latter might conceivably be placed everything from systems for the administration of welfare benefits to contracted-out prison management: activities so disparate as to raise the question of whether they can or should be amenable to treatment within the same analysis. Scepticism about whether ‘the administrative state’ can be examined as a single unit of analysis is thus surely sound. But when the inquiry in view concerns the ambit of contemporary constitutional theory, it becomes apparent that much rests on how we respond to that scepticism. This follows from what the otherwise highly differentiated actors and activities of the administrative state have in common. Each is implicated in the project of constitutional government in its direct relation to those who are subject to the authority of government, as well as to those who stand to benefit from the discharge of its duties and responsibilities. The myriad ways in which actors and activities of the administrative state render these relationships between government and the governed into concrete form is of crucial significance to everyday conditions of constitutional government in action. And yet, attention to such questions has been an at best marginal concern of constitutional theory, at least to this point. The analysis to follow accordingly proceeds from the conviction that a certain peace must be made with the inherent amorphousness of the term ‘the administrative state’, and the equally amorphous institutional reality that this term purports to describe, if the view is taken that something is lost to the study of contemporary constitutional thought and practice through its neglect. Despite its shortcomings, the term ‘the administrative state’ is a ‘helpful analytical conceit’ to mark out the distinctive domain of actors and activities implicated in the enterprise of administering government power.8 It helps to direct attention to who wields the powers of government that bear directly upon the lives of constitutional subjects, and how those powers are exercised, as possible sites of constitutional theoretical engagement in their own right. Moreover, it is only by framing the administrative state as a single unit of analysis that any such engagement can extend beyond the study of single instances to become part of a sophisticated reappraisal of the serviceability of the standard theoretical account of the administrative state within the constitutional order. On its face, little seems amiss in the logic of that ‘quotidian and uncontroversial’ account in which the administrative state is presented as ‘an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by robust mechanisms of judicial oversight’.9 This account accepts the operational fact of the administrative state and allocates clear functions to each of the three core institutions of modern constitutional government with respect to it. That such an account



7 Metzger, 8 ibid.

9 Berger,

‘Foreword’ (n 3) 8.

‘The Constitutional Status’ (n 4) 5.

388  Kristen Rundle is needed at all is of course a product of how constitutions rarely if ever entrench the administrative state – in whatever form that idea might take – as a constitutional institution in its own right.10 The question prompted by the efforts to elevate engagement with the administrative state to a constitutional register reviewed below, however, is whether this account continues to hold in contemporary conditions. As the analyses of the next sections demonstrate, the actors and activities that comprise the workings of the contemporary administrative state are in a constant state of flux and renewal. Moreover, the present era is one in which even the most standard artefacts of administrative power – administrative decisions and delegated legislation – increasingly find expression in unfamiliar forms.11 At the same time, the actors who do the administrative work of government have both increased in number and defy neat division between public and private. To take these and other features of contemporary constitutional government in action as a prompt for questioning where the administrative state belongs within contemporary constitutional thought is necessarily to invite the possibility of new directions of constitutional theory. In particular, to do so is to take seriously Berger’s astute observation that while traditional constitutional narratives have largely regarded the administrative state as ‘constitutionally permissible’, little consideration has been given to ‘whether this constitutional character may have changed over the arc of constitutional time’.12 Efforts to ‘constitutionalise’ engagement with the administrative state reside in the space set by this question, and demand a re-examination of the relation between the administrative state and the contemporary constitutional order. III.  CONSTITUTIONALISING THE ADMINISTRATIVE STATE: THE ‘STATUS’ INTERVENTION

The project of ‘constitutionalising’ the administrative state owes its current traction to Gillian Metzger’s provocative response to the trends of ‘anti-administrativism’ that

10 It is equally true, at least in the common law tradition, that few constitutions enshrine a conception of ‘the state’ from which the apparatus of the administrative state might be regarded as a key vehicle of its practical expression: see J McLean, Searching for the State in British Legal Thought: Competing Conceptions of the Public Sphere (Cambridge, Cambridge University Press, 2012) for a sophisticated account of the at best immanent presence of a concept of the state in the British public law tradition. The question whether the administrative state would be drawn more explicitly into the ambit of constitutional theory if the concept of ‘the state’ were also positioned more squarely within that enterprise is outside the scope of the present analysis, though surely a generative direction for future scholarship. 11 For example, the rise of automation and artificial intelligence in systems of administrative decisionmaking raises the conceptual question of whether the outputs of such systems plausibly fall within the idea of a ‘decision’ around which much of the conceptual architecture of the theory and practice of modern administrative law has been built. The same can arguably also be said for the wide range of executive-enacted instruments that presently fall within the concept of ‘delegated legislation’. Both of these developments have been seen across jurisdictions, and raise novel and pressing questions for research in administrative law theory. 12 Berger, ‘The Constitutional Status’, (n 4) 5.

The Case of the Administrative State  389 took root during the presidency of Donald Trump in the US.13 The features of this apparent siege of the American administrative state included attacks on its entities in the tweets of the President, the refusal of the Trump administration to fill some of the most senior roles in government agencies, and funding cuts that rendered some agencies unable to function at all. Metzger points out that although some features of this ‘political anti-administrativism’ are a product of their time, the attitude itself is not: a comparable stance found expression in the opposition of conservative political forces to the redistributive social programs of the 1930s ‘New Deal’ reforms.14 What is of primary interest to Metzger, however, is the distinctive species of ‘judicial antiadministrativism’ reflected in certain recent judgments of the US Supreme Court. The catchcry of this contemporary judicial anti-administravitism is that the administrative state is a threat to liberty and democracy, at odds with the constitutional structure that the framers created and, for these reasons, ‘unconstitutional’.15 The task Metzger sets for herself is not simply to meet this ‘unconstitutional’ objection with a ‘constitutional’ reply. Structuring analysis around the question of whether the administrative state is ‘constitutional’, she suggests, ‘injects hesitancy about its constitutional status from the outset’.16 Metzger’s aim is instead to reframe the debate about the administrative state by arguing that judicial antiadministrativism misdiagnoses its constitutional status ‘almost exactly backward’.17 As she argues it, anti-administrativism obscures the fact that the administrative state brings important constitutional ‘benefits’ in support of ‘an accountable, constrained, and effective executive branch’.18 Indeed, on specific features of American ‘administrativism’, such as the contemporary reality of extensive delegation of authority by Congress to the executive, Metzger goes much further. This accepted constitutional practice, she argues, makes core features of the administrative state constitutionally ‘obligatory’,19 because such is ‘what the separation of powers requires in a world of substantial delegation of policymaking authority to the executive branch’.20 Thus, on Metzger’s account, the outlines of a constitutionally ‘mandated’ administrative state emerge simply from the proposition that power so delegated must be faithfully executed.21 Metzger’s arguments for the constitutional status of the administrative state have surely generated much interest among scholars of American constitutional and

13 Metzger, ‘Foreword’ (n 3). 14 ibid 68. 15 ibid 3. Metzger mentions Justice Thomas, Chief Justice Roberts, Justice Alito, and Justice Gorsuch as each having expressed some version of ‘judicial anti-administrativism’ in their judgments so far, noting however that ‘rarely – with the marked exception of Justice Thomas – do they develop this originalist argument with any rigour’: ibid. 16 ibid 72, fn 421. Metzger further notes that this question has generated an extraordinary body of ­scholarship that ‘addresses whether the administrative state can accord with the Constitution’s text, structure, and history’, and a debate that ‘has been waged as ably and exhaustively as it can be’. 17 ibid 71. 18 ibid 71–72. 19 ibid 8, 87 (emphasis added). Metzger alternately uses the language of the administrative state as constitutionally ‘required’ by delegation. 20 ibid 89. 21 ibid 90.

390  Kristen Rundle administrative law alike.22 The interest of the project for present purposes, however, lies primarily in how those arguments operate. It is clear that Metzger does not intend her analysis necessarily to translate into justiciable claims of American constitutional law.23 For example, she readily acknowledges that Congress’s apparent duty ‘to provide the resources necessary for the executive branch to adequately fulfil its constitutional functions’ depends on the political branches for its realisation, and is unlikely to be judicially enforceable.24 She equally observes that the trends of judicial anti-administrativism that provide the catalyst for her analysis have ‘a notably rhetorical air’, and that its advocates seem unwilling to follow through on ‘the radical implications of its constitutional complaints’.25 This is why Metzger’s effort to ‘constitutionalise’ the status of the American administrative state is best understood as an innovative methodological intervention in (American) constitutional theory, from which we might learn a number of things. Descriptively, Metzger takes her cues for this endeavour from what is essential to the operation of the American constitutional order in practice, giving her analysis a distinctly socio-legal flavour. But what is equally clear is that Metzger does not want to limit her provocation to the descriptive register. At her lead, assigning constitutional status to the administrative state is also a normative intervention specifically aimed at safeguarding the security of a well-functioning administrative state within the system of American constitutional government.26 An attempt to secure the status of the American administrative state by elevating how it is conceptualised within American constitutional thought thus lies at the heart of Metzger’s intervention. This normative agenda obviously rests upon, and promotes, a positive conception of the administrative state as something compatible with, rather than a threat to, conditions of constitutionalism. But more importantly for Metzger, advancing the status argument is a way of highlighting the stakes of neglecting the administrative state within American constitutional thought. As she puts the point in her conclusion, assigning constitutional status to the administrative state promotes needed engagement with ‘questions too long excluded’ from constitutional discourse about the ‘scope and nature of constitutional obligations to govern’.27 North of the border, Kate Glover Berger’s effort to argue for the constitutional status of the Canadian administrative state makes clear its debt to Metzger’s lead.28

22 It is notable that Metzger does not argue that delegation itself is constitutionally obligatory, only widely done and accepted. The most provocative aspect of Metzger’s analysis is therefore her suggestion that it follows from this non-mandatory but practically established feature of the American public order that the administrative state is constitutionally ‘mandated’. 23 See also Berger, ‘The Constitutional Status’ (n 4) 4, who summarises Metzger’s claim about the constitutional ‘status’ of the administrative state as essential to, indeed mandatory within, the American constitutional order as ‘a matter of operational reality and good public administration not as a matter of constitutional law’. 24 Metzger, ‘Foreword’, (n 3) 90. 25 ibid 95. 26 See, eg, ibid 89. 27 ibid 95. 28 Berger, ‘The Constitutional Status’ (n 4).

The Case of the Administrative State  391 Yet unlike Metzger’s, the motivations behind Berger’s project are anything but defensive: the prevailing attitude towards the administrative state in Canadian jurisprudence, she suggests, ‘would fairly be described as steadfast and respectful, at times bordering on romantic’.29 Berger’s motivations lay instead with a desire to disrupt the characteristic style of engagement with the administrative state on the part of Canadian public law scholars, which she suggests has been preoccupied primarily with the fact of its expansion to the expense of sufficient attention to the implications of that growth for the place of the administrative state within the constitutional order.30 The ultimate aim of Berger’s project, therefore, is to redescribe the structure of the contemporary Canadian constitutional order as inclusive of the administrative state. As an intervention in Canadian constitutional theory, Berger’s project invites and indeed relies upon a fundamental methodological shift in how relations between the units of Canadian constitutional order are conceived. The organising idea is one of constitutional ‘architecture’ within which the separation of legislative, executive and judicial power is obviously located but not necessarily assigned the dominant place. Berger borrows this analytical framework from the effort of the Supreme Court of Canada to explain and defend its own constitutional ‘status’ in an important case that addressed the place of the Court within the Canadian constitutional order.31 The argument developed was that through a series of important historical legal and political moments within which the Court had been centrally positioned, it had ‘become central to the flourishing of Canada’s federal system’ and thus ‘an essential part of Canada’s constitutional architecture’.32 The foundational premise of the architecture analysis is accordingly that the structure of a constitutional order ‘evolves over time as institutions are established and carry out their mandates, as priorities and institutional relationships evolve, and as paradigms of governance shift’.33 The task that Berger sets for herself is to bring this methodology to an analysis of the status of the administrative state within the Canadian constitutional order, pursued through an examination of structural relationships between the administrative state and other constitutional institutions over time.34 The materials for this analysis are drawn from judgments of the Supreme Court of Canada which, Berger suggests, together flesh out a narrative in which the administrative state ‘is not merely an addition to the existing structure of the Canadian constitutional house’, but rather has ‘become integrated into the design in ways that renovate the existing structure’.35 Three features of that narrative are emphasised: the posture of deference towards administrative decision-makers in the judicial review of administrative action, the considerable jurisdiction of Canadian administrative decision-makers over

29 ibid 4. 30 ibid 4–5. 31 Reference re Supreme Court Act, ss 5 and 6, 2014 SCC 21. The question arose because the Supreme Court of Canada is not explicitly mentioned in either of the two Canadian constitutional texts. 32 Berger (n 4) 7. 33 ibid 8. 34 ibid. 35 ibid 6.

392  Kristen Rundle constitutional matters,36 and the way that constitutional conceptions of Canadian judicial power have placed relatively few limits on legislative power to establish administrative decision-makers.37 Taken together with the growth of the administrative state, Berger argues, these developments tell a story in which the administrative state ‘has become embedded within the system of government that the constitution seeks to implement’.38 An equally significant feature of Berger’s intervention is how it challenges the hierarchical flavour of the received separation of powers analysis of the place of the administrative state within Canadian constitutional thought. To her eyes, actual conditions of government – the ‘shifting network of institutions that carry out the rule of law each day’39 – call into question understandings of constitutional structure within which the judiciary is seen as the institution primarily responsible for advancing and protecting the rule of law. Indeed, and resonant with moves made by Metzger, Berger goes so far as to suggest that the constitutional structure through which contemporary conditions of the rule of law are actually maintained is ‘flattening’. That is, the judiciary is no longer necessarily positioned at the top of a hierarchy of institutions charged with maintaining conditions within which actors and ­activities of the administrative state have become ‘constitutionally essential’.40 By her own presentation, Berger’s project is primarily intended to be descriptive. Larger normative questions about how the claimed constitutional status of the Canadian administrative state might bear upon its design and operation, or its relationships with the other branches of government, are expressly deferred for later consideration.41 Nonetheless, it is clear that Berger’s claim that the standard account of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by robust mechanisms of judicial oversight needs to be rearticulated is motivated by a desire to render Canadian constitutional thought more capable of illuminating and responding to the features of Canadian constitutional practice.42 Starting from status, she contends, ‘brings to the fore shifts in the allocation and exercise of public power and forces reflection on the demands and commitments of the rule of law’,43 and compels a sharpened understanding of ‘the challenges of institutional pluralism, the pathologies of legislative and executive power, and the possibilities and limits of change’.44 Both Metzger’s and Berger’s arguments in favour of the constitutional status of the administrative state are obviously jurisdiction-specific, with the consequence

36 Here Berger’s analysis concentrates especially on ‘the capacity of administrative actors to interpret the constitution, account for the Charter in the exercise of statutory discretion, and carry out obligations of the Crown to consult Indigenous peoples whose rights and interests are affected by the decisions and actions of the Canadian state’: ibid 9. 37 ibid 9. 38 ibid 16–17. 39 ibid 16. 40 ibid 17. 41 ibid 17. 42 ibid 5. 43 ibid 1. 44 ibid.

The Case of the Administrative State  393 that certain features of their respective analyses might not transfer readily to other contexts. But as innovative exercises in contemporary constitutional theory, these ‘status’ interventions offer illuminating insight into what might be opened up by taking the administrative state as the starting point of inquiry into contemporary conditions of constitutionalism, rather than positioning it as an outlier to that inquiry. From that starting point, both projects can be seen to highlight, and to seek to address, problems of descriptive inaccuracy as well as normative inadequacy in how actual conditions of contemporary government are represented in constitutional thought. In both analyses, moreover, changing the priority of the units of analysis that are brought into focus is crucial. For example, in both interventions the oversight role of the judiciary that has long dominated the imagination of constitutional engagement with actors and activities of the administrative state is just one element of the theoretical picture, and ultimately secondary to the higher objective of engaging with what the administrative state is and does within the contemporary constitutional order as an object of inquiry in its own right. This means that, at most, examination of administrative or constitutional law doctrine serves as a porthole for inquiry, rather than being its principal object. An appeal to redescribe the core units of constitutional order and constitutional analysis is accordingly an end unto itself in both analyses. But, if more overtly in Metzger’s case than Berger’s, this exercise in redescription also operates as a necessary prelude to how each intervention challenges the ambit of contemporary constitutional theory in its normative aspect. For Metzger that challenge consists of an explicit call for greater attention to the relation between the administrative state and ‘constitutional obligations to govern’ within constitutional discourse.45 For Berger, it is conveyed more implicitly through her effort to promote a deeper understanding of the contribution of the administrative state to meeting overarching demands of the rule of law within the complex architecture of contemporary constitutional government.46 I return to both points in section V. IV.  CONSTITUTIONALISM AND THE ADMINISTRATIVE STATE: REFLECTIONS FROM AUSTRALIA

To this point at least, there has been no comparable effort on the part of scholars in my own jurisdiction, Australia, to argue for the constitutional ‘status’ of the administrative state in the manner promoted by Metzger and Berger. Nonetheless, their ‘constitutionalising’ provocations invite reflection on a number of features of Australian constitutional thought and practice that also call the descriptive and normative serviceability of the received constitutional account of the administrative state into question. Australian practice reveals some stark instances of overt mismatch between an understanding of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by



45 Metzger 46 Berger

(n 3) 95. I return to this point in section V. (n 4) 17.

394  Kristen Rundle mechanisms of judicial oversight, and the way that certain activities of administrative government actually proceed.47 One striking example is the now ubiquitous technique of contracting-out the delivery of government services to private actors, which for the most part has been authorised through non-statutory constitutional executive power without more.48 Whatever else might be said about the appropriateness of the practice of contractingout per se,49 the executive-led legal framework for its advent and expansion alike marks a significant departure from the theory that actors and activities of the administrative state are authorised primarily by Parliaments through legislation. This departure is especially dramatic within a constitutional system, like Australia’s, that is designed around the convention of ‘responsible government’ and its premise that executive activity must be answerable to the legislature.50 The mismatch between constitutional theory and practice in the Australian context can also be highlighted in other ways. For example, the appeal of veteran Australian administrative law scholar, John McMillan, to ‘rethink’ the separation of powers to better reflect the contribution made by non-judicial mechanisms such as the Ombudsman to the accountability of government action resonates in several respects with efforts to argue for the constitutional ‘status’ of the administrative state.51 In particular, McMillan’s arguments in favour of reconceptualising the constitutional order to accommodate an ‘integrity’ or ‘fourth’ branch52 within which various entities involved in ensuring the accountability of the administrative state might more coherently reside are prompted by observations on how the actual workings of the Australian constitutional order have changed over time.53 His efforts to highlight the contribution of this larger network of institutions to how the constitutional

47 An increasing interest in the administrative state on the part of Australian administrative law­ scholars is clearly revealed in contributions to the recent collection edited by J Boughey and LB Crawford, Interpreting Executive Power (Alexandria, Federation Press, 2020). 48 For example, in current Australian practice, little or no legislative activity accompanies the approval of even the most major government contracts. How the answerability of such practices to constitutional demands might be framed in terms of a justiciable argument of constitutional law is far from clear: see KM Hayne, ‘Government Contracts and Public Law’ (2017) 41 Melbourne University Law Review 155. 49 For my own treatment, see K Rundle, ‘Office and Contracting-Out: An Analysis’ (2020) 70 (Issue supplement 2) University of Toronto Law Journal 183. 50 See C Mantziaris, ‘The Executive: A Common Law Understanding of Legal Form and Responsibility’ in R French, G Lindell and C Saunders (eds), Reflections on the Australian Constitution (Alexandria, Federation Press, 2003) 125. This important earlier study shares much in common with the questions raised in the present discussion, directed specifically to the implications of a range of neoliberal governmental techniques for what Mantziaris calls ‘constitutional responsibility relationships’. 51 J McMillan, ‘Re-Thinking the Separation of Powers’ (2010) 38 Federal Law Review 423, 425. 52 The ‘fourth’ or ‘integrity’ branch debate of course also has international reach, in large part due to the influence of Bruce Ackerman’s framing of the American context in ‘The New Separation of Powers’ (2000) 113 Harvard Law Review 633. For further representative works in the Australian discussion, see S Gageler, ‘Three is Plenty’ in G Weeks and M Groves (eds), Administrative Redress In and Out of the Courts: Essays in Honour of Robyn Creyke and John McMillan (Alexandria, Federation Press, 2019) ch 2, and JJ Spigelman, ‘The Integrity Branch of Government’ (2004) 78 Australian Law Journal 724. 53 Inquiry into the possibility of recognising ‘government accountability’ as a constitutional ‘value’ is another angle through which these questions have been considered by Australian scholars: see J Boughey and G Weeks, ‘Government Accountability as a “Constitutional Value”’ in R Dixon (ed), Australian Constitutional Values (Oxford, Hart Publishing, 2018).

The Case of the Administrative State  395 imperative of the rule of law is maintained – a contribution too often overlooked within the narrative which assigns that rule of law contribution exclusively to the judiciary – resonates strongly with Berger’s emphasis on the same.54 These perceptions of mismatch between constitutional theory and constitutional practice are not mere matters of academic interest in the Australian context or indeed elsewhere. The standard constitutional account of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by mechanisms of judicial oversight is, after all, an account of the constitutional legitimacy of the administrative state as much as it is a description of its apparent location within the constitutional scheme. My aim in bringing Metzger’s and Berger’s ‘constitutionalising’ provocation to the Australian context is thus to explore another example that invites reflection on what might be at stake in uncritical acceptance of that constitutional orthodoxy. The example I have in view is the ascendant method in Australian constitutional law for framing engagement with the possible constitutional implications of innovative forms of administrative power. Assuming that the constitutional validity of the legislation within which an administrative power is sourced is not in question, constitutional scrutiny of administrative power in Australia is ultimately limited to its implications for institutional relationships between the legislature, executive and judiciary respectively. According to this framework, the only means through which to challenge some aspect of the design or performance of administrative power is through an inquiry into whether either is capable of being supervised by a court for compliance with the statutory and common law parameters of its grant. At the federal level this exercise is conducted through the supervisory jurisdiction guaranteed by section 75(v) of the Australian Constitution and its statutory equivalents.55 This standard inquiry of Australian administrative law in what I earlier described as its supervisory aspect can take on an explicitly constitutional pitch if the objection is raised that some aspect of the putative authorising regime encroaches upon the ‘minimum provision of judicial review’ said to be constitutionally entrenched by operation of section 75(v).56 The orthodox circumstance in which this question arises is when the legislation through which the relevant administrative action is authorised contains a ‘privative’ or ‘ouster’ clause that purports to limit or remove the capacity of courts to supervise the legal limits of that administrative action. Though different jurisdictions have taken different approaches to this circumstance, in Australia the High Court has avoided the strongest assertion of its judicial power in relation to such clauses. Rather than declaring provisions of this kind constitutionally

54 McMillan’s analysis specifically reacts to what is in his view an overstated emphasis on the contribution of the judiciary to maintaining the rule of law within the Australian constitutional order in relation to the accountability of actors and activities of the administrative state: McMillan ‘Re-Thinking’ (n 51) 425–6. See also LB Crawford, ‘Between a Rock and a Hard Place: Executive Guidance in the Administrative State’ in Boughey and Crawford (eds), Interpreting Executive Power (n 47) 7. 55 The division of labour with respect to judicial review of federal administrative action is shared by the High Court with the Federal Court of Australia through statutory jurisdiction (Judiciary Act 1903, s 39B) that essentially replicates the structure and ambit of the s 75(v) constitutional jurisdiction. 56 Plaintiff S157/2002 v Commonwealth (2003) 211 CLR 476.

396  Kristen Rundle invalid, the Court has determined that they will be interpreted to have an at most limited operation.57 The constitutional relationship between the legislature and the judiciary with respect to such clauses has been thought to be clarified and balanced accordingly. The key feature of this jurisprudence on the implications of the constitutionally entrenched minimum provision of judicial review for the ambit of legislative power is that it has focused on the constitutional function of the judiciary in its relation to the other two branches. The appropriateness of this orientation has been uncontroversial when the legislative provision at issue has recorded, on its face, an intention to limit the supervisory capacity of the courts.58 But this is precisely also why the section 75(v) jurisprudence as presently developed indicates little about the nature and scope of constitutional demands that might be placed on the design and performance of administrative powers, in their own right.59 The position of courts within the constitutional separation of legislative, executive and judicial power will not necessarily be the primary question in view when what is being agitated for is the recognition and enforcement of constitutional demands on administrative power, in and of itself. That interventions in this vein find themselves in an under-developed zone of Australian constitutional thought is clear from the reception that has met such arguments so far. One example is the constitutional challenge mounted against non-compellable administrative discretions that are reposed personally in an administrative official and conditioned on amorphous considerations of the ‘public’ or ‘national’ interest. Arguments advanced against such provisions have been framed in terms of the threat that their structure poses to wider constitutional commitments to the rule of law and the prevention of arbitrary power that this commitment entails. The response of the High Court to that challenge has, however, not only been swift but analytically confined to a focus on the ‘supervisibility’ of these innovative forms of administrative power.60 The Court’s view has been that so long as the challenged provision reveals sufficient limits to be capable of supervision, no further question as to its constitutionality will arise. This reduction of constitutional inquiry to the question of the ‘supervisibility’ of administrative power typically concludes in a determination of its validity.61 Validity is, of course, the orthodox end point of constitutional reasoning in its engagement

57 For an excellent analysis of the normative structure of the ascendant methods of Australian judicial review see W Bateman and L McDonald, ‘The Normative Structure of Australian Administrative Law’ (2017) 45 Federal Law Review 153. 58 See Plaintiff S157/2002 v Commonwealth (2003) 211 CLR 476; Graham v Minister for Immigration and Border Protection (2017) 263 CLR 1. 59 For a doctrinal analysis of constitutional provisions and principles thought to be relevant to this question, see LB Crawford, ‘The Entrenched Minimum Provision of Judicial Review and the Limits of ‘Law’” (2017) 45 Federal Law Review 569; LB Crawford, ‘Can Parliament Confer Plenary Executive Power? The Limitations Imposed by Sections 51 and 52 of the Australian Constitution (2016) 44 Federal Law Review 287. 60 See especially Plaintiff M61/2010E v Commonwealth (2010) 243 CLR 319. 61 An exception in Australian public law is the equitable remedy of the declaration. This remedy can be used to declare an administrative action ‘unlawful’ rather than ‘invalid’, and in doing so open up the ­possibility of continuing legal effect in a way that the latter cannot.

The Case of the Administrative State  397 with the legal status of legislative provisions and administrative actions alike. As a straightforward application of positivist legal philosophy in its provision of sourcebased tests for determining the existence of valid and therefore binding legal acts, the key virtue of this ‘validity analysis’ lies in how it protects the core value of legal certainty. Yet what it also promotes and supports is a necessarily top-down understanding of actors and activities of the administrative state as instruments of authorised government power. The problem with this understanding is that it offers little, if anything, by way of standards against which to assess compliance with constitutional demands less specific in their expression than those which govern basic conditions of legal authorisation.62 The worry at issue here is not that attention to constitutional conditions for the authorisation of government power is misplaced: quite clearly this question is central. Rather, the worry lies in how the narrow ambit of this question can forestall inquiry into the myriad ways in which actors and activities of the administrative state might contribute to or undermine the structure and values of a constitutional order. Extremely broad forms of administrative power of the kind described above have passed the ‘supervisibility’ and ‘validity’ tests with little alarm, despite leaving those subject to them with few parameters by which to anticipate the patterns of their exercise. This is instructive. It suggests that there is little scope in either the ‘supervisibility’ or ‘validity’ analyses for wider conversations about whether and how actors and activities of the administrative state create, transform, sustain, or undermine the quality of relationships between government and the governed, and why this might be of constitutional significance on some other basis than the bald requirement of legal authorisation. Those who have attempted to challenge administrative powers for their ‘arbitrary’ or other characteristics have sought to place precisely these kinds of questions on the constitutional agenda, albeit with little success so far. To problematise these implications of constitutional orthodoxy in its engagement with actors and activities of the administrative state is not to doubt the contribution of the supervisory jurisdiction of courts to upholding the foundational commitments of liberal democratic constitutionalism.63 In the common law tradition of constitutional thought, the standing of courts to discharge this function is regarded as the key manifestation of a commitment to protect the rights of the individual against a potentially overreaching administrative state. Nonetheless, the worry remains that something is lost when the range of possible constitutional questions that might be asked in relation to the design and performance of administrative power is reduced to the narrow parameters of the ‘supervisibility’ and ‘validity’ analyses. The point to notice for present purposes is the role played by a particular theory of the administrative state within both. In Australia at least, it would appear that understanding the administrative state as nothing more than an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by robust mechanisms

62 For my analysis of this ‘authorising’ mindset of Australian administrative law, see K Rundle, ‘Administrative Discretion and Governing Relationships: Situating Procedural Fairness’, in D Meyerson, C Mackenzie and T MacDermott (eds), Procedural Justice and Relational Theory: Empirical, Philosophical, and Legal Perspectives (London, Routledge, 2021), 248. 63 McMillan, ‘Re-Thinking’ (n 51) 425.

398  Kristen Rundle of judicial oversight operates to limit the constitutional answerability of administrative law in its facilitative aspect to the declared ambit of administrative law in its supervisory aspect. V.  CONSTITUTIONAL THEORY REVISITED? THE PROVOCATION OF THE ADMINISTRATIVE STATE

These questions at the intersection of Australian constitutional and administrative law might appear at first glance to lie at some distance from the ambitions underscoring Metzger’s and Berger’s efforts to ‘constitutionalise’ engagement with the administrative state in their respective jurisdictions. These American and Canadian analyses are primarily directed to features of administrative law in its facilitative aspect, while the concerns just canvassed are ostensibly preoccupied with the scope of constitutional engagement with administrative law in its supervisory aspect. Nonetheless, I would argue that the common motivation cutting across all three analyses is to raise, and to take seriously, the question of whether the core units and methods of constitutional theory ought to be reconsidered in light of how actors and activities of the administrative state actually operate in conditions of contemporary constitutionalism, positive or negative as that contribution might be thought to be. For Metzger, this question is ultimately tied to the very fate of a well-functioning administrative state within the system of American constitutional government. For Berger, its posing helps to bring to the fore shifts in the contemporary allocation and exercise of public power, and to force reflection on what this means for overarching constitutional commitments to the rule of law.64 In my own case, it is way of asking whether something needs to be said, in a constitutional voice, about the implications of certain features of the design and performance of contemporary administrative power for constitutional subjects. If the reason for ‘constitutionalising’ discussion of the administrative state differs according to local questions in this way, we might also expect that the gains or benefits of these purported analytical shifts will play out in ways specific to context. Even so, all three efforts to elevate engagement with the administrative state to the constitutional register just reviewed can be seen to share a common direction. Cutting across all three is a concern to bring the relationships between government and the governed that are rendered concrete through actors and activities of the administrative state into constitutional focus, and to give these relationships priority as a site of attention within constitutional theory and practice alike. As already noted in relation to Metzger’s and Berger’s interventions, it is clear that any such theoretical development will remain inseparable from the foundational provocation to redescribe the standard separation of powers account of the administrative state in a manner more capable of grasping the contemporary constitutional contribution of its actors and activities. This might occur by promoting a



64 Berger,

‘The Constitutional Status’ (n 4) 1.

The Case of the Administrative State  399 change in emphasis in how the relations between the three primary units of constitutional analysis should best be described in light of evidence from contemporary practice, or indeed by questioning whether a separation of powers analysis ought to be given primacy in the inquiry at all. Metzger’s argument that ‘contemporary anti-administrativism forestalls development of a separation of powers analysis better tailored to the reality of current government’65 and Berger’s redescription of the ‘flattening’ structure of the institutional order of Canadian constitutionalism66 are examples of the former, while the way that Berger’s analysis of the architecture of Canadian constitutionalism assigns conceptual priority to the idea of the rule of law over the idea of the separation of powers is an example of the latter. In terms of the normative impulse underlying each of these ‘constitutionalising’ efforts, it is again helpful to recall the point made in connection with Metzger’s unapologetically normative motivations for assigning constitutional status to the administrative state: namely, that doing so promotes needed engagement with questions too long excluded from constitutional discourse about the ‘scope and nature of constitutional obligations to govern’.67 The reason why this argument is so important for present purposes is because by addressing neglect of the administrative state in constitutional thought within a push to give greater attention to ‘the scope and nature of constitutional obligations to govern’, Metzger’s appeal necessarily invites greater attention to what the administrative state is and does in relation to such constitutional ‘obligations’. My own view is that the normative potential of such an inquiry can and should be developed further. Constitutional theory and doctrine are not just important to our descriptive and normative understanding of the character and demands of administrative government. They are also relevant to our experience of both.68 To worry about what might be missed when the administrative state is sidelined as a unit of analysis of constitutional theory is thus to worry about something more than whether constitutional theory is in appropriate descriptive order vis-à-vis the realities of contemporary constitutional practice. It is also to worry about whether the gaze of constitutional theory is attending sufficiently to our experience of actual constitutional conditions, as these are rendered concrete in our relationship(s) to actors and activities of the administrative state. The point being agitated here is not that institutional relationships between the legislature, executive and judiciary are not a crucial part of how fundamental constitutional obligations to govern are discharged. The point is instead that directing constitutional inquiry solely to these dominant units of constitutional analysis may operate to divert attention from questions about how or whether those obligations are being discharged. For example, the kinds of concerns canvassed in the present discussion are at best marginal to the preoccupations of the leading debate

65 Metzger ‘Foreword’ (n 3) 7. See also at 95: ‘Repeatedly voicing [constitutional anti-administrativism’s] claims … precludes developing accounts of the separation of powers that accept and build on the administrative state’s essential role in our constitutional order.’ 66 Berger, ‘The Constitutional Status’ (n 4) 17. 67 Metzger ‘Foreword’ (n 3) 95. 68 McMillan, ‘Re-Thinking’ (n 51) 423.

400  Kristen Rundle of contemporary constitutional theory between defenders of ‘legal’ versus ‘political’ constitutionalism.69 Although we might expect both sides of that ever lively debate to encounter the administrative state as part of their engagement with the executive branch, it does not follow that either position stands to offer insight into its constitutional character in contemporary conditions. The forms, actors, activities and relationships of the contemporary administrative state demand if not a different, then at least an augmented, analytical approach. The foundational methodological impulse underlying the three theoretical interventions just canvassed is a call to reappraise the core units of analysis of contemporary constitutional theory. This call has of course registered elsewhere in the field. Recent contributions in this regard include an effort to develop a conception of ‘positive constitutionalism’ focused more directly on the role played by a competent set of state institutions in making government effective.70 In a similar vein, newer constitutional theories of rights protection emphasise the importance of productive relations between the three arms of government if the state is to be a competent and effective actor in the realm of rights specifically.71 At the intersection of administrative and constitutional law, we have also seen scholars take a critical stance on the utility of orthodox separation of powers analysis to the resolution of contemporary questions of judicial review.72 Moreover, and particularly resonant with concerns just canvassed, worries have been raised about how the orthodox units of analysis of constitutional theory tend to neglect demands on relationships between government and the governed.73 Finally and not least, public law scholars have also begun to pay

69 See M Loughlin, Public Law and Political Theory (Oxford, Oxford University Press, 1992) for an account of the development of the positions from political constitutionalism to legal constitutionalism. For a leading contemporary contribution to the theorisation of political constitutionalism, see R Bellamy, Political Constitutionalism: A Republican Defence of the Constitutionality of Democracy (Cambridge, Cambridge University Press, 2007), and for legal constitutionalism, see TRS Allan, Constitutional Justice (Oxford, Oxford University Press, 2001). 70 See NW Barber, The Principles of Constitutionalism (Oxford, Oxford University Press, 2018) especially at 9–10, 18–19. The way that the function of judicial review is conceived within prevailing constitutional orthodoxy as a check on the project of government is a classic expression of the ‘negative’ conception of constitutionalism that Barber’s ‘positive’ constitutionalism seeks to challenge. It is therefore unsurprising that, other than discussing the contribution of the ‘welfare state’ to conditions of the rule of law (see especially ibid 100–4), Barber’s treatment of administrative law is largely limited to its supervisory dimensions. One might however surmise that greater engagement with the administrative state would be a natural extension of a theoretical enterprise principally concerned with what is needed to facilitate, rather than to constrain, the project of constitutional government. 71 See especially A Kavanagh, ‘Recasting the Political Constitution: From Rivals to Relationships’ (2019) 30(1) King’s Law Journal 43; S Stephenson, From Dialogue to Disagreement in Comparative Rights Constitutionalism (Annandale, Federation Press, 2016) ch 6; E Carolan, ‘Dialogue Isn’t Working: The Case for Collaboration as a Model of Legislative–Judicial Relations’ (2016) 36(2) Legal Studies 209; and V MacDonnell, ‘The Constitution as Framework for Governance’ (2013) 63(4) University of Toronto Law Journal 624. Although these theoretical contributions emphasise the facilitative role of the state in protecting rights, they are yet to give close attention to the role of the administrative state specifically in relation to their object of inquiry. 72 See Eoin Carolan’s effort to reconceptualise the separation of powers analysis through attention to what practices of judicial review in England and Ireland reveal about assumptions about the administrative state as an aspect of contemporary constitutionalism in The New Separation of Powers: A Theory for the Modern State (Oxford, Oxford University Press, 2009). 73 One recent intervention in this vein is J Weinrib, Dimensions of Dignity: The Theory and Practice of Modern Constitutional Law (Cambridge, Cambridge University Press, 2016). Though Weinrib’s units of

The Case of the Administrative State  401 more attention to the administrative state with a view to evaluating its ‘rule of law’ credentials.74 Contemporary theoretical scholarship in public law is accordingly replete with refreshing new directions of inquiry that might be seen to carry affinities with the work of those who seek to ‘constitutionalise’ discussion of the administrative state. Yet what is distinctive about the latter effort is the way that it straddles each of the descriptive and normative ambitions just recounted, and how it does so as part of a challenge to the analytical ambit of constitutional theory, pursued from the starting point of the administrative state, that keeps its eye on the question of what it means to be governed within the administrative conditions actually supplied by a constitutional order. VI. CONCLUSION

The foregoing analysis commenced from the proposition that to frame study of the administrative state in a constitutional register is to urge greater appreciation of the significance of its actors and activities to conditions of contemporary constitutionalism. The argument put was that who wields the powers of government that bear directly on the lives of constitutional subjects, and how they wield those powers, deserves greater attention within constitutional theory than has been the case to date. Why might such an intervention have particular salience at this historical juncture?75 Three observations in particular might be offered in answer to this question. The first is that jurisdictions across the world have witnessed an increase in the use of constitutional executive power as the source of legal authority for government action, without recourse to parliamentary processes. There are many things that might be said about what is signalled in this trend. But from an administrative law standpoint, it is important to notice the correlation between increased use of nonstatutory power as the legal framework for administrative relationships and broad discretions that carry increased potential for arbitrariness within those relationships. Sourcing the authority of administrative action in executive power, or in extremely broad forms of statutorily authorised power, is commonly also accompanied by the prevalence of ‘soft law’ guidelines and other instruments that tend to evade judicial

analysis remain tied to orthodox objects of constitutional theory (for example, human rights guarantees and an independent judiciary), the priority given to relationships between ‘rulers and the ruled’ within his inquiry is noteworthy. 74 The leading example is the recent book by C Sunstein and A Vermeule, Law and Leviathan: Redeeming the Administrative State (Cambridge MA, Harvard University Press, 2020). Framed primarily in a ‘rule of law’ rather than constitutional register, like Metzger, Sunstein and Vermeule set out to defend the administrative state, and to assign it importance as a site of theoretical analysis in its own right. For an Australian example of the use of a rule of law theoretical framework to engage with developments in contemporary administrative practice, see Crawford, ‘Between a Rock and a Hard Place’ (n 54). 75 I have made the same plea with respect to the ambit of contemporary rule of law theory in its engagement with actual features of contemporary administrative government. See K Rundle, ‘Fuller’s Relationships’, Special Edition on ‘The Rule of Law and Democracy’ (2019) Archiv für Rechts- und Sozialphilosophie (The Journal for Legal and Social Philosophy) 17, especially 32–36.

402  Kristen Rundle scrutiny, despite their considerable normative influence on actual administrative practice. It bears emphasising that the great majority of administrative powers associated with governmental responses to the COVID-19 global pandemic have been sourced in various forms of executive-made law. The second observation is that the systems through which contemporary administrative power is exercised are themselves experiencing a moment of radical change as governments across the world incorporate artificial intelligence and other technologies into systems of administrative decision-making. How the increasingly central location of these new technologies within systems of administrative decision-making will play out remains to be seen. But experience so far suggests, at the very least, that problems in the intelligibility of relationships between entities of administrative government and those with whom they interact can be expected.76 Conceived primarily as a question of technique rather than a matter of foundational concern to the constitutional character of relationships between government and the governed, it might be some time before engagement with this technological shift within contemporary administrative government is expressed in a specifically constitutional register. Nonetheless, and quite apart from the worry of intelligibility just signalled, the fact that in many jurisdictions these reforms have received little parliamentary scrutiny points again to the gap between the orthodox theoretical account of the administrative state as an extension of the executive branch exercising authority delegated to it by the legislature and what actually unfolds in practice. The third observation is that, again across jurisdictions, the privatisation of government functions and services continues apace. At the same time, the COVID-19 pandemic and the exponential increase in government action and reliance on government services it has generated has revealed significant concerns about the appropriateness of some of these techniques in the context of government responsibilities that carry high stakes for human safety and security. During a moment in which the presence of government in the lives of citizens looms larger than it has for decades, the stakes of inefficacy within such practices as contracted-out government service delivery have never been more apparent to an increasingly interested audience of observers. But, again, a framework for (re)appraising these techniques that takes seriously their contribution to how the conditions of contemporary constitutionalism actually play out has largely been absent, at least to this point. There are thus many possible reasons why the ideas and arguments canvassed in the foregoing analysis have emerged at this historical juncture, and why they might also make a timely contribution to framing discussions about developments still unfolding. But this also means that, as the less developed limb of wider trends in the ‘constitutionalisation’ of administrative law, the promise and limits alike of these efforts to elevate engagement with the administrative state to the constitutional register may take some time to be revealed. In the meantime, it remains important to recognise what this intervention is, and what it is not.

76 For an Australian angle on these challenges, see W Bateman, ‘Algorithmic Decision-Making and Legality: Public Law Dimensions’ (2020) 94 Australian Law Journal 1, and T Carney, ‘Robo-debt illegality: The seven veils of failed guarantees of the rule of law?’ (2019) 44 Alternative Law Journal 4.

The Case of the Administrative State  403 Efforts to ‘constitutionalise’ discussion of the administrative state, at least in their current expression, do not necessarily demand structural change to the constitutional order itself. Rather than proposing an alternative to the separation of powers model around which received constitutional texts and traditions are primarily organised, these interventions are more concerned to encourage us to think beyond the components of the separation of powers model when articulating constitutional demands on relationships between government and the governed. The provocation contained in efforts to elevate engagement with actors and activities of the administrative state to the constitutional register is thus best understood as challenging, rather than refusing, the ‘quotidian and uncontroversial’ account of the administrative state as an extension of the executive branch, exercising authority delegated to it by the legislature, and overseen by robust mechanisms of judicial oversight.77 This is precisely why that challenge needs to be recognised, at its core, as a challenge to the methodology of constitutional theory. By questioning the serviceability of received patterns of constitutional thought to the conditions and challenges of contemporary government, the ultimate appeal of the project of ‘constitutionalising’ the administrative state is to notice, and where relevant to overcome, the tendencies of thought and practice that lie in the way of taking its actors and activities seriously as units of analysis within constitutional theory, in their own right. To frame study of the administrative state in a constitutional register is, above all, to appeal for a greater appreciation of its significance – positively and negatively – to the character and demands of contemporary constitutionalism.



77 Berger,

‘The Constitutional Status’ (n 4) 5.

404

Index A Acton, Lord  105 Adams, T et al  159 administrative law See also administrative state adjudication, importance of context  171 adjudication, statutory schemes and  170–171 analytic and synthetic truths  164–165 anatomical analysis  153, 170 common law and  13, 168 common law constitutionalism  364 common law development  54–55 common understandings of  13, 361, 372–382 computer science analytical methods  153, 159, 173–175 conflicting pairs in philosophical debate  163–164, 165 constitutional law and  153, 154–159, 363, 366 constitutional theory and  13–14, 361–382 constructive interpretation  154, 165, 166–167, 175 content analysis  153, 171–173 delegation of authority  378 descriptive conceptual analysis  169 development  43, 363–364 Dicey  368–380, 382 doctrinalism  154, 168–171 empirical research methods  159–160, 162–165, 166, 168–169, 171 enabling legislation  376–377 facilitative aspect  383, 384 formal equality through judicial review  377–379 functionalist approach  171–173 generalist and specialist approaches  171–173 Independent Review (IRAL)  154, 156–157, 158, 160–161, 163, 164 interpretive process  379–380 interpretivism  154, 166–168 judicial development  54–55, 156–157, 172, 375–376 judicial review See judicial review legality  361, 363 legitimacy  377 liberty from  375–377 methodological pluralism  153–176

moral legitimacy  364 natural justice and  362, 379–382 ‘new administrative of Wales’  166 New Public Management  166 normative approach  172 organic methodology  153, 169–170, 171 parliamentary sovereignty and  374 pluralistic origins  171 political power, distribution  362–364 proceduralism  166 proportionality  362 purpose  14 quantitative analysis  160 reasonableness  362 reforms  165 rights-based approach  155, 156–157 rule of law and  363, 364, 376–377, 380–381 secondary legislation  171 social construction  162–166 socio-legal theories  159–162, 165 ‘soft law’ sources  171, 401 subordination to constitutional law  153, 154–159 supervisory aspect  383 supervisory function  14 traditional methodological categories  153 ultra vires doctrine  50 United Kingdom  363–382 values and facts  162–168, 169 administrative state anti-administrativism  388–389, 399 Australia  385, 393–398 Canada  384, 385, 390–393 constitutional legitimacy  395 constitutionalising  386–403 contemporary constitutional theory  384, 398–403 COVID-19 pandemic  402 delineating  384, 386–388 generally  384–386 increasing use of executive power  401 separation of powers  389, 396, 398–399 United States  384, 385, 388–390, 393 Ajzenstat, J  116 Alexy, R  165

406  Index Allan, TRS  116, 166 account of UK Constitution  131–152 common law constitutionalism  64, 131, 132, 133–137, 148, 150–151 Constitutional Justice  107 interpretative method  7, 58–60, 64–65, 131, 133–134, 137, 167 Law, Liberty, and Justice  107–108 moralised constitutional theory  213, 225, 228–230 ‘Political Obligation and Public Law’  58–60, 64 The Sovereignty of law  58–60, 64, 124, 125, 134, 137 Ambedkar, BR  120 Anderson, C  108 Anisminic decision  49–50, 66, 112, 139, 140, 378–379 anti-positivism See also interpretivism generally  3 Loughlin  10, 233–234, 236 positivism compared  10 Arato, A  350–351, 353, 356, 357 Arendt, H  351 Aristotle Politics  313–316, 337 Aroney, N  316 Aronson, M  68 Artemus visualisation tool  174–175 Arvind, TT and Stirton, L  2, 7, 8, 162–163, 166, 171 assembly governments  331–332 assurance model assurance legitimacy  9–10, 212, 216–222, 223, 225, 226 constitutional law and  9–10, 212, 216–222, 225, 226 constitutional theory  225 epistemic model compared  228–230 government efficacy and  226 moral authority and  225–228 unjust rules and  182–183, 197–198, 222–225 Ataturk  113 Atkinson LJ  377 audi alteram partem principle  381 Austin, G  117 Austin, John  181, 206, 207–208 habit of obedience  196–197, 207 Australia administrative state  385, 393–398 Constitution  52, 53, 395 constitutional interpretation  53–54 judicial review  325 moderate originalism  53–54 parliamentary democracy  334, 336

parliamentary sovereignty  44 statutory interpretation  52 written constitution  44 B Backer, L  333 Bagehot, W The English Constitution  334, 335, 369 Banakar, R and Travers, M  115 Barber, N  3, 5, 79–80, 85, 86–87, 92, 135, 317 Barnett, R  116 basic law meaning  344 basic structure doctrine  112 Becker, H  77 Bell, J  8, 11, 161, 175 Anatomy of Administrative Law  168, 170–171, 172, 173 Bentham, Jeremy command theory of law  181, 207–208 positivism  238, 243 rules of law  60 Berger, KG  13–14, 384, 385, 390–393, 395, 398, 399 bias entrenched  2 epistemic  12–13 methodological  6–7, 8 Bills of Rights constitutions and  324–327 judicial review and  326 legal authority and  202, 207 Bix, B  23, 83–84, 86, 87–88 Blackstone, Sir William Commentaries on the Laws of England  13, 57, 112, 362, 364–370, 372, 377, 379 common law  365, 382 constitutional liberty  366–368, 375 natural rights  366 parliamentary sovereignty  365, 367–370, 372, 373 Blair, Tony  293, 294, 296–297 Blunkett, David  296 Bolingbroke, Henry St. John, 1st Viscount  112 Bondy, V  154, 165 Bosnia and Herzogovina  334, 356 Boyle, K and Camps, D  158, 160 Brennan J  52, 53 Brexit constitutional methodology and  6, 72, 89–99 Cooperation and Trade Agreement  72 EU Referendum Ac 2015  11, 293, 294, 298, 300–302, 305–306 European Union (Withdrawal) Acts  289, 304 generally  364

Index  407 impact on UK Constitution  72 improvisatory model of statute law  287–288, 289 Miller cases  6, 71, 90, 254–256, 303, 307 parliamentary sovereignty and  72, 89–99 parliamentary sovereignty challenged by  72 UK Constitution and  99 Withdrawal Agreement  71 British Constitution See United Kingdom Constitution Brown, Gordon  293 Brown, W  271 Buckland, R  165 Buckmaster LJ  377 Burke, Edmund  112 Reflections on the Revolution in France  319–320 C Calhoun, J  105 Cameron, David  300–301 Campbell, Alistair  296 Canada administrative state  384, 385, 390–393 charter of rights  325 Confederation  116 Constitution  52, 324, 327 constitutional interpretation  53 moderate originalism  53 parliamentary democracy  334 parliamentary sovereignty  325 rule of law principle  322 statutory interpretation  52 canonical status elevation of texts to  109, 111–114, 128 linguistic canons  46 Carnwarth LJ  66 Cart v Upper Tribunal  161 Chambers, Sir Robert  57 Charles I  112 Coleman, J  243 Colon-Rios, J  350, 352–353 command theory of law  181, 187, 192, 207–208 common good Aristotle’s Politeia  314 constitutional government and  315, 322 state policy and  218 common law administrative law and  13, 168 Blackstone  365–366 common law constitutionalism See common law constitutionalism common law pragmatism  58, 66, 67–69 common law principles in statutory interpretation  52, 55–58

constitutionalism, generally  1, 3, 7–8 constitutions and  6, 42, 43–44, 52–53, 55–56 customary norms  41 development  42–43, 51–52 Dworkin  307 improvisatory nature  285, 287, 307 interpretive method  5–6, 7, 41–43 judicial development  42–43, 51–52, 53 judicial review, theory of  54–55 legislative supremacy doctrine  6, 44–45 linguistic canons  46 methodology for ascertaining  41–42 moral principles and  7 orthodox understanding of  42–44 other forms of law, confusion with  51–55 parliamentary sovereignty and  6, 42, 44, 55–58 presumptions of intentions  48 principle of legality  357 proof of legislative intention  46 rationes decidendi and obiter dicta  43 standing commitments within legislature  47–48, 50 statute law and  42, 43, 44, 45–47, 48, 49–50, 51–54, 65 statutes and constitutions distinguished  52–53 United States  44 written constitutions and  6, 42, 43–44, 52–53 common law constitutionalism administrative law  364 Allan’s account of  64, 131–152 equal citizenship principle  134 Goldsworthy’s orthodox account compared  131–152 interpretative method  42, 53, 58, 63–67, 131, 133–152 legislative supremacy  133–134 legitimacy of statute law  134 moral principles  7, 135, 136, 137, 138, 140 parliamentary sovereignty and  55–59, 63–67, 74, 133–134, 136 rule of law  133–134, 136 statutory interpretation  42, 53, 58, 63–67, 133–134, 136–137 United Kingdom  42, 58, 63–67, 74, 131, 133–152, 364 United States  53 written constitutions  53, 63–65 common understanding of administrative law  13, 361, 372–382 common understanding of legal authority  236 common understanding of legal theory adherence to  18–25 clear common understanding  32–38

408  Index concepts, nature of generally  19–20 concepts falling outside  21–22 Condorcet Jury Theorem  28–29, 35–36, 37–38 critical approach and  80 descriptive sociology  24 disputed or unclear  18, 25, 29–32, 33 exemplar model of concepts  19–20, 87, 88, 127 fear of  25 generally  5 groupthink, distortion by  36–37 Hart’s The Concept of Law  21–22, 23–24, 37 heuristic model, as  80 historically-minded approach and  80 implied primacy  18 individuals, understandings held by  19, 20–21 information cascades  36–37 interpretivist method  5, 80, 86, 87, 88 invocation in legal theory  21–25 legal positivism  5 limitations  29–38, 39 meaning and nature of  19–21 model-theoretic approach compared  87–89 moral values and  5, 38 Oakeshott  24–25, 34–35, 81 positive methodology and  21–23 principles, concept of  5, 17, 21, 24–25, 33–35, 39 prototype model of concepts  19, 20, 87 purpose  17–18, 21, 22–23 Raz  22, 23, 30 relativism  38 relevance  18, 21, 25–29, 38–39 rendering an account intelligible  18, 23, 26, 33–35, 38 reputational bias  36, 37 role of theorist in shaping  23, 39 significance  17–39 social institutions, nature of  17, 18, 21–24, 26–27, 35, 39 socially-created rules  27 subjectivism  38 systemic moral errors  38 test of correctness, as  18, 20–21, 28–29, 33–34, 35–38 wisdom of crowds  18, 27–29, 35–38 Communist single-party states  329, 331–333, 337–338 comparative law constitutional theory  12–13 epistemic bias  12–13 use in adjudication  346–348

computer science analytical methods Artemus prototype  174–175 computer-assisted visualisation  159, 174–175 corpus linguistics  173 generally  153, 159, 173–175 Natural Language Processing  173, 174 structural topic modelling  173–174 topic modelling  173–174 concepts common understanding of legal theory  20, 30–31 Condorcet Jury Theorem  28–29, 35–36, 37–38 descriptive conceptual analysis  169 descriptive sociology  24 disputed or unclear  18, 25, 29–32 examples as focus of heuristic approach  127 exemplar models  19–20, 87, 88, 127 explanation of  22 groupthink, distortion by  36–37 individuals, held by  19, 20–21 information cascades  36–37 model-theoretic approach  83–84, 85, 87–88 nature of generally  19–20 positivist methodology  21–23 principles, concept of  17, 21, 24–25 prototype model  19, 20 relativism  38 rendering an account intelligible  18, 23, 26, 33–35, 38 reputational bias  36, 37 social institutions, concept of  17, 18, 21–24, 26–27, 35 socially-created rules  27 subjectivism  38 systemic moral errors  38 Condorcet Jury Theorem common understanding of legal theory  28–29, 35–36, 37–38 conflicting pairs in philosophical debate  163–164, 165 constituent power theory  348, 349–357, 358 delegation model  356 unamendable constitutional amendment  354–357, 358 constitution assurance model  225–228 balanced  112 basic norm  179, 182 Bills of Rights  324–327 binding force  9, 231 characteristics  317, 320–329 codified  318 common law and  55–56 common law pragmatists  58, 66, 67–69

Index  409 community and  223, 270–271, 313 conception of  311 constitutional government See constitutional government constitutional law See constitutional law constitutional principles  56–57, 322–323, 324, 328 constitutional theory See constitutional theory cornerstone of social and political life  115, 117–118, 119 definitions  313, 327–329, 344 eternity clauses  355 improvisatory model See improvisatory model interpretivism  56–62 judge-made  55–58, 63–69 judicial enforceability  312, 318, 324–327 large-C and small-c constitutions  320–325 legal authority and  182, 202, 208–209, 211–231 living constitutionalism  53 methodology for ascertaining  41–42 mixed  313, 314–315 moral footprint of settlement, as  225–228 moral importance  9–10 moral legitimacy and  312, 313, 364 moral principles  138 moralised constitutional theory  147, 211–231 originating  328 parliamentary sovereignty and  55, 57–61, 95 political creation, as  262 political legitimacy and  211–231 political settlements, as  9 positive and political approach  10, 59–61, 233–257 post-sovereign  351 power, restraining or limiting  312 protective shield, as  115–117, 119 public law as political jurisprudence  235 rulebooks, as  115, 118–121 settlement aspect  225–228 truce, as  115, 121–122 unamendable provisions  354–357, 358 United Kingdom See United Kingdom Constitution unwritten  52, 60, 182, 202, 226, 320, 324 written See written constitutions constitutional authority public law generating  237, 242 constitutional change See also improvisatory model abusive  348–349, 353–357 adaptation  286 delegation model  356 designed  285–286, 288–289, 291 detecting  278, 282

eternity clauses  355 formal  279 generally  11, 95, 231 incrementalism  285–286 kosmos and taxis  286, 287 majoritarianism  355–356, 358 material study of  278–281 path dependency  287 spontaneous  285–287, 291, 292 unamendability doctrines  354–357, 358 constitutional design concept of  283 human factors  284 information systems  284 public choice theory  283 constitutional government Bills of Rights  324–327 characteristics  317, 320–329 classification, generally  12, 311–312, 339 codified  318 common good and  315, 322 constitutional conventions  321–322, 324 constitutional monarchies  333 constitutional principles  56–57, 322–323, 324, 328 constitutional statutes  56, 321, 324 definition  313–320, 329 democracy and  312, 316, 322–323 endurance and transcendence  323–324 equality and  312, 315, 316 functional criteria  312, 313 history of  313–320, 337–338 judicial enforceability  312, 318, 324–327 judicial review and  325, 342, 343–346 large-C and small-c constitutions  320–325 legal rules as element of  321 liberty and  312, 315–316, 320, 366–368, 375 limitation on government, constitutionalism as  312, 315–317 mixed constitutions  313, 314–315 moral legitimacy and  312, 313, 364 negative constitutionalism  317 normative  311, 312, 315, 338 parliamentary systems  312, 313, 320, 330, 333, 334–337 positive constitutionalism  317 pouvoir constituant  319, 327 presidential systems  312, 318, 320, 322, 334, 335, 337 republican  333–334 rule of law and  315, 322, 327–328 rules-based system, as  313, 314, 315 substantive criteria  312, 315 types  333–337 UK influence  319–320, 334

410  Index US influence  318, 319, 320, 334 Westminster model  336 constitutional law See also constitution administrative law and  153, 154–159, 363, 366 assurance legitimacy and  9–10, 212, 216–222, 225, 226 comparative  346–348 compromise and  214–215 computer science analytical methods  173–175 fundamental rights, status  256 liberalism of fear  224 meaning  344 methodological challenges  1–2 moral aspirations  2, 212–213, 231 political law and  2 constitutional monarchy  329 constitutional order bearer or subject of  273–274, 276–277 fundamental norms  273–274, 277–278 political objectives  277–278 constitutional practice positivist and political critique  251–253 constitutional principles generally  322–323, 324 constitutional reform legitimacy  348 Constitutional Reform Act 2005  293, 296–297, 302–303, 305–306 Constitutional Reform and Governance Act 2010  293, 298–299, 303–306 constitutional theory See also constitution administrative law  13–14, 361–382 administrative state  383–403 analytic and synthetic truths  164–165 assurance model  225 basic structure doctrine  112 canonical status, elevation of texts to  109, 111–114, 128 common law constitutionalism See common law constitutionalism common law pragmatism  48–51, 58, 66, 67–69 comparative constitutional change  348–357 comparative law  12–13 comparative turn  342–348 conflicting pairs in philosophical debate  163–164, 165 constituent power theory  348, 349–357, 358 constitutional continuity  109–110, 111, 113, 128 constructed-type models  6, 76–78, 81, 83, 98 constructing, generally  105, 125 core precepts  109, 111–114, 128

critical approach  79 customary norms  41 deontic power ascribed to facts  106, 107, 108–110 dialectical  104 discord and concord perspectives  115, 119–122 doctrinal approach  10, 11 Dworkin  42–43, 58–63 economy of trust  62 epistemic model  225 explanatory approach  10, 11, 12 factual evidence  105, 106–108, 109, 132, 162–163 formal/substantive distinction  114–115, 118 as framework for future development  8, 108, 123–124, 125–126 functional criteria  12 Gardbaum’s new commonwealth model  73, 75–76 goal  104, 229 Grundnorm  182–183 heuristic approach to See heuristic models historically-minded See historically-minded approach improvisatory model See improvisatory model interpretivism See interpretivism judges, constitution-making by  55–58 legal constitutionalism  74, 106, 107, 110, 156–159, 247, 277, 312 legal philosophy and, generally  58–63 legal and political theories  8 legitimacy enhanced by  9–10, 211–231 material study of constitutional order  10–11 memorialisation, role  110, 111, 128 methodological assumptions  341–342 methodologies  105, 114–122, 216, 225–231 model-theoretic approach See model-theoretic approach moral principles and, generally  214–215 moralised  147, 211–231 normative  77, 78, 307 normative criteria  12 orthodox understanding of  41–48 political constitutionalism  74, 106, 107, 110, 119–120, 156–159, 246–248, 277, 312 political morality  211 positivism See positive method practical consequences  105 practitioners’ understandings and  61 rationalism, use of  81 recension-like methodology  125–126 rhetorics of intransigence  122–125 role of morality  8–10, 61, 62 rule of law  66–67

Index  411 rule of recognition  65, 68, 132–133, 181–183, 184, 187 social epistemology of  105, 106–114 society, relation to  11 traditions compared  108–109 unamendability doctrines  354–357, 358 unconstitutional constitutional amendments  348–349, 353–357 Young’s theory of democratic dialogue  72, 93 constraints on power constitutionalism and  312 constructed-type models generally  76–78, 81, 83, 98 interpretivist method compared  6 constructive interpretation administrative law  154, 165, 166–167, 175 Dworkin  235–236, 245 content analysis administrative law  153, 171–173 automated  173 conventionalism arbitrariness, condition of  270 constitutional theology and  271 dependency, condition of  270 material study of constitutional order  269–270 core precepts axiomatisation  109, 111–114, 128 cornerstone of social and political life constitutions as  115, 117–118, 119 corpus linguistics  173 Cotterrell, R  165 courts See judiciary Craig, P  93, 120–121, 155, 156, 157, 161, 166 interpretivism  167, 168 Cram, I  76 critical approach to constitutional theory  79 common understandings and  80 model-theoretic approach compared  85–86, 91–92 models, use of  80–81 parliamentary sovereignty and  91–92 Cromwell, Oliver  317 customary norms common law  41 methodology for ascertaining  41–42 D Daly, P  166, 167–168, 170, 171, 175 declaratory theory of law  42 delegation model  356 democracy administrative institutions  363 common understanding of  21, 34

constitutional  312, 316, 322–323, 342, 343–344, 363 equality and  157 judicial review and  342, 343–346 Kelsen’s political theory  198–199, 200–201 legal authority and  198–199, 200–202 legal positivism and  250 legitimacy and  222, 250, 251, 348 militant  354, 358 moral aspiration  2 multi-party  335 parliamentary  319–320, 334–336 parliamentary sovereignty and  93–94, 254 public law and  250 Tocqueville  322–323, 370–371, 374 transitional  343–344 tyranny of the majority  323 United Kingdom  319–320 voting systems  336–337, 338 Westminster model  336 deontic power ascribed to facts  106, 107, 108–110 descriptive sociology Hart  24 Dicey, AV  13, 57, 357 administrative law  368–380, 382 conception of legality  377–378 constitutional liberty  375 constitutional theory  362, 371–372 executive power  373–374, 375 judiciary  371–372, 375, 377–379 parliamentary sovereignty  64–65, 67, 93, 94, 95–96, 334, 370, 373 rule of law  107, 375 dictatorships  312, 329, 330–331 Dindjer, H  91–92 Dixon J  52 doctrinalism administrative law  154, 168–171 dominant-party systems  331 Douglas, M grid-group cultural theory  114–115, 128 Dutch Constitution  325 Orangists and Patriots  103, 113 Dworkin, R  2, 34, 107, 141, 207 common law reasoning  307 constructive interpretation  235–236, 245 Dworkin-Waldron disagreement  343–344, 345 Freedom’s Law  343 intepretitive thesis  42–43, 58–63, 64, 167, 235–236, 245, 378–379 judges as moral agents  345, 378 originalist approach to constitutional law  43, 228

412  Index parliamentary sovereignty  63 theories of law and of adjudication  190–191, 193–194, 206 theory of political legitimacy  212 Dyzenhaus, D  3, 6, 9, 10 E economy of trust constitutional theory  62 Ekins, R  56, 91, 92, 156 Elliott, M  166 empirical analysis administrative law  159–160, 162–165, 166, 168–169, 171 model-theoretic approach  2, 73, 75–77, 78, 82–83, 85, 87–88, 98, 127 positivist methods  131–132, 135 empty place theory  269 enabling legislation  376–377 engineering human factors in  284 epistemic model assurance model compared  228–230 constitutional theory  225 equality constitutional government and  312, 315 democracy and  157 equal citizenship principle, statutory legitimacy and  134 equality before the law, constitutionalism and  312 formal equality through judicial review  377–379 moralised constitutional theory  211–212 equitable interpretation statute law, of  46–47 eternity clauses  355 European Convention on Human Rights UK accession  364 European Union Act 2011  300 European Union (Withdrawal) Acts  289 executive parliamentary sovereignty and  96 prerogatives, customary norms  41 F factual considerations constitutional theories, generally  162–163 criteria of importance  6–7, 8 deontic power ascribed to facts  106, 107, 108–110 empirical analysis and  132, 162–165 as framework for future development  108, 123–124, 125–126 heuristic approach to  7, 125–128

methodological bias  6–7, 8 moral principles and  6–7 positivist method and  41, 132, 186 selection of facts, factors driving  7 ‘faithful agents’ doctrine legislative supremacy and  45, 47, 56 Falconer, Charles  296 Faulks, Lord  158 Federalist Papers  112, 283, 318, 326 Feynman, R ‘Ode to a Flower’  170 Finnis, J  33, 91, 92, 120–121, 315 Fiscal Responsibility Act 2010  288 Fisher, L  170 Fixed-term Parliaments Act 2011  11, 293–294, 298, 299–300, 306 formal/substantive distinction in constitutional theory  114–115 Forsyth, CF  166 Fraassen, BC van  73 France Conseil constitutionel  325 constitutions  118, 230, 311, 313, 317, 318–319, 323, 325, 332, 336 lois organiques  321 presidential system  335, 336 Freedom of Information Act  298 Fukuyama, F  337, 338 Fuller, LL  198, 199, 380–381 functionalist approach administrative law  171–173 fundamental rights constitutional rulebook  120 discord perspective on society  116, 120–122 Gardbaum’s new commonwealth model  75–76 moral aspiration  2 positive and political approach  10, 256 status within constitutional law  256 statutory infringement, assurance legitimacy and  220–222 statutory interpretation and  47–48 US Constitution  228 G Gardbaum, S new commonwealth model  73, 75–76 Gardner, J  244 Gee, G and Webber, GCN  73–74, 75, 81 Giere, R  73 Girard, R  270 Glorious Revolution  317, 320, 363, 366 Goldoni, M  10, 11 Goldsworthy, J  3, 5–6 account of UK Constitution  131–152

Index  413 historical methodological argument  142–146 legal positivism  7, 41–69, 131–133, 137 philosophical methodological argument  142–146 Gordon, M  3, 9, 10, 94, 96–98 Greece, Athenian democracy  313–316 Greenberg, M moral impact theory  146–147, 149 grid-group cultural theory  114–115, 128 Griffith, JAG ‘The Political Constitution’  144, 247 The Politics of the Judiciary  377 Grimm, D  320 definition of constitution  325, 327–328 Grundnorm legal authority and  182–183 H Habermas, J  165 Halbertal, M  271 Hale LJ  45, 66, 90 Halliday, S  114, 159–160 Hamilton, Alexander  326 Harlow, C and Rawlings, R  171 Harris, JW  197 Hart, HLA The Concept of Law  21–22, 23–24, 37, 181–182, 188, 207, 208, 269 descriptive sociology  24 ‘internal point of view’  180, 182, 190, 191, 200, 205–206 judicial acceptance, principle of  181, 182–184, 187–188, 191–192, 197, 203 legal positivism  3, 5, 9, 21–22, 180, 184, 188, 198, 205–207 rule of recognition  144–145, 181–183, 184, 187, 269 rules of adjudication  181 rules of change  181, 192 Separation Thesis  180, 183, 186–187, 191–192, 197, 203–204, 206, 235, 237 theory of legal authority  9, 180, 181–193, 199, 200, 204–209 Hauriou, M  263, 264, 266 Hayek, F The Constitution of Liberty  286, 287–288 rule of law  376–377 Held, D  75 Heller, H  263 heuristic models case for  103–129 common understandings, as  80 examples, focus on  127 meaning of heuristic  126–127 model-theoretic approach  7, 75–76, 77, 80, 81

non-heuristic conceptions compared  126–127 rationalism and  81 rhetoric and  122–128 Hewart, Lord  374, 376–377, 381 Himma, KE  30 Hirschman, AO  123, 124 historically-minded approach aim  79 common understandings and  80 Dworkin  145 Goldsworthy  142–146 method  79–80 model-theoretic approach compared  79–80, 85, 90–91 moralised constitutional theory and  229–230 parliamentary sovereignty and  90–91, 144–145 Hobbes, Thomas Leviathan  120, 189, 202, 226 positivism  238, 243 on protection and obedience  189, 222–223 social contract theory  201 Hood, C  126 Hooker, C  169–170 Howarth, D improvisatory model of constitutions  2, 11–12 Hozumi Yatsuka  113 HS2  56 Human Rights Act 1998  293, 295, 325 human rights See fundamental rights Hume, David  206, 268–269 I ideal theory  345 implied repeal statute law  56 implied terms ‘reading in’  47 improvisatory model adaptation and  286 assessing improvisations  290 British Constitution as  11–12, 283, 292–308 case law  285, 287 characteristic of good improvisation  290 common law  285, 287, 307 conflictual side  289 coordination and  286–287 designed change, interaction with  285–286, 288–289, 291 generally  2, 283–285 non-improvisations distinguished  290–292 path dependency  287 spontaneity compared  285–287, 291, 292 statute law  285, 287–288, 289–308

414  Index incrementalism constitutional change  285–286 Independent Review on Administrative Law (IRAL)  8 India comparative material, use in adjudication  346–347 Constitution  106, 110, 112, 117, 120, 334 parliamentary democracy  334 information cascades conceptual understanding and  36–37 information systems constitutional design and  284 institutions See social institutions international treaties parliamentary sovereignty and  95 interpretivism abstraction  79 administrative law  154, 166–168 Allan  7, 58–60, 64–65, 131, 133–134, 137 Barber  80, 86–87, 92 Bentham’s rules of law  60 common law  5–6, 41–43 common law constitutionalism See common law constitutionalism common understandings, development  5, 80, 86, 87, 88 compromise and  62 constructed-type models compared  6 constructive interpretation  166, 235–236 contestable value judgments  64 declaratory theory  42 disagreements, consequences of  60–61, 62 Dworkin  42–43, 58–63, 64, 236, 245, 378–379 empirical sources  166 future constitutional development and  8 generally  3, 5, 6, 79 judges, constitution-making by  58, 63–69 Lakin  131–152 Loughlin  235–237, 245–246 model-theoretic approach compared  86–87 models, use of  80 moral principles and  5, 7, 41–42, 58–63, 132, 137, 147–149, 245 natural law theory and  236 non-moral  61, 62, 245–246 normative positivism compared  244–246 parliamentary sovereignty and  58–62, 64, 68, 92 positivist method compared  7, 10, 59–61, 131–152 statute law and  42, 43, 58 UK Constitution  131, 133–134 weakness  8, 60–62 written constitutions  62

Irvine, Derry  296–297 Italy Fascist constitutional change  280–281 Ito Hirobumi  112 J Jackson  65, 112–114 Jackson, V  356 Jamaica a Charter of Fundamental Rights and Liberties  120 Constitution  119–120 James II  366 Johnson, Boris  304 Judicial Power Project  8, 156–157, 159 judicial pragmatism constitutional theory, generally  48–51, 58 parliamentary sovereignty and  66, 67–69, 139 judicial review administrative decisions, generally  361, 377 administrative law, generally  383, 391 Anisminic decision  49–50, 66, 378–379 approaches to  153, 154–155 Australia  325 Bills of Rights and  326 Cart judicial reviews  161 common law theory of  54–55 Commonwealth model  344 constitutional democracy and  342, 343–346 constitutional government and  324–327 constitutional legitimacy  170, 342–346 Costs Capping Orders  156 dialogic model  344 Dworkin-Waldron disagreement  343–344, 345 expansion  155–157 formal equality through  377–379 generally  8, 342–346 grounds  171 immigration and asylum claims  154 individual/good administration dichotomy  171 Kelsenian model  326 methodological pluralism  166–168 parliamentary sovereignty and  377 quantitative analysis of caseloads  160–162 reforms  155–159, 165 rule of law and  377 social construction  162–166 strong versus weak model  343–344 subordination to constitutional law  154 success rates  161–162 Totally Without Merit claims  156 ultra vires theory of  54–55 United States  112, 325, 326

Index  415 judiciary absolute monarchies  330 administrative law, development  54–55, 156–157, 172, 375–376 common law, judicial development  42–43, 51–52, 53 constitution as rulebook  120 constitution-making by  55–58, 63–69 constitutional government and  312 Dicey  371–372, 375–376, 377–379 Dworkin  345, 378 equitable interpretation by  46–47 ‘faithful agents’ doctrine  45, 47, 56 interpretative theory and  42–43, 58 judicial activism  172, 289, 303 judicial discretion  251–252 judicial review See judicial review legal authority  181, 183–184, 185, 188–189, 191, 196, 199–200 legal positivism and  251–252 Miller cases  254–256, 303, 307 parliamentary sovereignty and  55–58, 63–69, 133, 254 politicisation  345 rationes decidendi and obiter dicta  43 ‘reading down’  46 ‘reading in’  47 rectification of linguistic meaning by  46 statutes, interpretation See statutory interpretation United Kingdom constitution  367, 368, 369, 371–372, 375 written constitutions, interpretation  43 justice community’s collaborative quest for  229, 250 Hart’s Separation Thesis  180, 183, 186–187, 191–192, 197, 203–204, 206 legitimacy and  213, 216–226 moralised constitutional theory  213, 216–226, 227, 229 positive political law and  9, 10 Rawls’ A Theory of Justice  34 unjust rules  133, 182–183, 197–198, 222–225 Waldron’s Law and Disagreement  220, 221, 238 K Kavanagh, A  49, 51 Kelsen, Hans Grundnorm  182–183 on law/morality relationship  192–193 natural law theory and  194 political theory  198–199, 200

positivism  3, 9, 179–180, 194 theory of legal authority  179–180, 191–203, 205–206, 208–209 Kelsenian model of judicial review  326 Kerr LJ  45, 66 Kesavananda Bharati v State of Kerala  112 Kirkham, R and O’Loughlin, E  8, 172 Kluit, Adriaan  103–104, 124 Kuhn, Thomas  127 Kuklick, B  118 Kyritsis, D  3, 6, 9–10, 147 L Lakin, S  3, 5, 7–8, 257 interpretative method  131–152 moralised constitutional theory  213 law administrative See administrative law binding legal obligations  9, 149, 231 constitutional government, legal rules as element of  321 constitutional See constitutional law customary norms  41 desirability of governance by  241 fact, matters of  41, 132 inevitable indeterminacy  252 judicial discretion  251–252 legal philosophy and  58–63 methodologies for ascertaining  41–69 political role  236 positivist approach  41, 239, 251–252 positivist and political approach  233–257 public See public law role, generally  241 social or political impact, critique of  251–253 Lascelles principles  299, 300 Law, DS  173–174 Law Commission working method  288–289 Lawrence v Texas  346 Laws, Sir John  50, 56, 116 Lefort, C  269 legal authority See also legitimacy Bills of Rights, role  202, 207 command theory of law  181, 187, 192, 207–208 conflicting interest groups and  184, 195–196 conflicting moral rules  182–183, 197–198 constitution  179–209 constitutions and  202, 208–209 democracy and  198–199, 200–202, 348 Dworkin’s theory of  190–191, 193–194, 206, 212 effectiveness as requirement for validity  195

416  Index establishing  9, 134, 199, 236–237 foundational law  180, 181–182 habit of obedience  184, 196–197, 207 Hart’s Separation Thesis  180, 183, 186–187, 191–192, 197, 203–204, 206, 235, 237 Hart’s theory of  9, 180, 181–193, 199, 200, 204–209 Hobbes on  189, 201, 202, 222–223 ‘internal point of view’  180, 182, 190, 191, 194, 196, 200, 205–206 judicial acceptance, principle of  181, 182–184, 187–188, 191–192, 197, 203 Kelsen’s theory of  179–180, 191–203, 205–206, 208–209 legal norms  12, 179, 182, 199–201 legal officials  181, 183–184, 185, 188–190, 191, 196, 199–203 legitimacy  9, 134, 198, 199–200, 250 natural law theory  179–180, 185, 194, 204, 205, 207–208 normal justification thesis  204 order of peace  195, 199 parliamentary sovereignty and  202 political terms  191, 198–199, 200–202 positive and political approach  241 positivism and  179–180, 188, 190–191, 194, 235 public law as political jurisprudence  235 Raz’s ‘legal man’ concept  185–186, 191–192, 197, 200, 205 Raz’s theory of  185–188, 191, 196, 204–205, 212 relativized authority  198 role of juristic interpretation  192–193 rule of law and  203 rule of recognition  181–183, 184, 187 rules of adjudication  181 rules of change  181–182, 192 social contract theory  179–180, 201, 206 sovereignty and  182 legal constitutionalism  74, 106, 107, 110, 156–159, 247, 277, 312 legal positivism See positivism legal realism  251 legal reasoning distinctiveness  251 positivism and  251–252 legal theory ascertaining customary norms  41 common understanding See common understanding of legal theory constitution as political creation  262 declaratory  42 institutionalists  265–266, 267 legal institutionalism  264

normative theory of constitutional order  262 organisation as central concept  263–264, 267 parliamentary sovereignty and constitutional order  262 purpose  22 rendering an account intelligible  18, 23, 26, 33–35, 38 societal constitutionalism  262–263, 264–267 legality principle of  49, 50 legislative authority establishing  9 legislative history admissibility  46 legislative intention generally  133, 146, 150 inferring  46, 47, 48, 49 judicial pragmatism and  48–51 presumption of  48 primacy of  49, 133 proof of  46 statutory presumptions  49 legislative supremacy challenges to doctrine  47, 55–56, 63–69 common law and  6, 44–45 common law constitutionalism  133–134 generally  3, 44, 146 judicial pragmatism and  48–51, 58 parliamentary sovereignty and  94 purposivist methodology  45 rule of law and  133 statutory interpretation and  3, 6, 44–45, 47 textualist methodology  45 United Kingdom  55–56, 58–59 United States  44–45 legitimacy See also legal authority administrative law  377 administrative state  395 assurance legitimacy  9–10, 212, 216–222, 223, 225, 226 constitution necessary to  9–10, 12, 211–231, 312, 313 constitutional law and assurance legitimacy  218–220, 226 constitutional reform, of  348 constitutionalism of fear  222–225 decision-making procedures and  221–222 delegation model and  356 democracy and  222, 250, 251 ‘drastic deterioration’ clause  217, 219 equal citizenship principle  134 justice and  213, 216–225 legal positivism and  9, 250, 252–253 moralised constitutional theory  147, 211–213 normative criteria  12, 199

Index  417 political  211–231 public law  250, 252–253 statutory infringement of fundamental rights  220–222 what constitutes  9, 134 Leibholz, G  281 Leiter, B  59 Lenin, V  33, 35 Lewans, M  13, 165, 166 Lewis, D  269 Lewis, WA  121–122 liberal-democratic orders conventionalism  269–270 liberalism of fear  223–224 liberty constitutional government and  312, 315–316, 320 limited government constitutionalism and  312 linguistic canons statutes, interpretation  46 Linz, J  337 living constitutionalism  53 Llewellyn, K  306 Lloyd-Jones LJ  66, 67 Local Government Board v Arlidge  380 Lochner  325 Locke, John  116, 117, 118, 366 Loewenstein, K  313, 315–318, 331–332, 333–334  Loughlin, M  3, 94, 128, 385 anti-positivism  10, 233–234 empowering character of public law  249 on Hobbes  238 interpretative method  235–237, 245–246 law, legality and superlegality  236, 239–240, 242 legal authority as common understanding  236 political jurisprudence  10, 233–237, 242, 257 Waldron’s normative positivism and  10, 233–241, 242, 248, 257 Loughlin, M and Tierney, S  254 Luhmann, N  265 M McDonald v City of Chicago  108 MacIlwain, C  317 McKinney, JC  77 McMillan, J  394–395 Madison, James  318 majoritarianism  355–356, 358 Manley, Norman  119–120 Manning, J ‘faithful agents’ doctrine  45 Marbury v Madison  112, 325, 326

Marmor, A  269–270 Marshall CJ  111–112, 325, 326 Marsons, L  167 Mason, Sir Anthony  49 Masterman, R and Wheatle, S  68 material study of constitutional order aim  273, 281 bearer or subject of constitutional order  273–274, 276–277 case studies of constitutional change  278–281 constitutive elements  261 conventionalism  269–270 detecting constitutional change  278, 282 empty place theory  269 fundamental norms of constitutional order  273–274, 277–278 fundamental traits of social organisation  272 internal formation of society, as  263 legal institutionalism  264, 267 legal organisation, constitutional order as  273–278 legal theory of  261–267 normative concepts  273 political objectives of constitutional order  277–278 political theology  270–272 political theory of  261, 267–273 social contract  267–269 social and legal dimensions  261 societal constitutionalism  262–263, 264–267 sovereignty and  270–271, 275, 281–282 unity of constitutional order  273–275 memorialisation role in constitutional theory  109, 111, 128 mens rea new criminal offences  48 Metzger, GE  13–14, 384, 385, 386, 388–390, 392–393, 395, 398, 399 Michelman, F  220 Mill, John Stuart  105 Miller I  10, 44, 55, 64–69, 71, 254–256, 303, 307 Miller II  6, 10, 55, 64–65, 71, 90, 120–121, 254–256, 303, 307 Millett LJ  69 Minobe Tatsukichi  113 model-theoretic approach advantages of use  81–89 aim  85–86, 89 analogue models  74–75 Barber  79–80, 85, 86–87 Bix  83–84, 86, 87–88 Brexit and parliamentary sovereignty  72, 89–99 common understandings and  87–89 conceptual claims  83–84, 85, 87–88

418  Index constructed-type models  6, 76–78, 81, 83, 98 correspondence rules  88 critical method and  80–81, 85–86, 91–92 defining  73–78 distinctive feature  75, 78–81 empirical claims  2, 73, 75–77, 78, 82–83, 85, 87–88, 98, 127 exemplar models  80, 87, 88, 127 Gardbaum’s new commonwealth model  73, 75–76 generally  72, 78–81 heuristic models  7, 75–76, 77, 80, 81 historically-minded approach compared  79–80, 85, 90–91 ideal-type models  76, 77, 80–81, 127 interpretative method and  80, 86–87 normative constitutional theory, development  77, 78–81 objective possibilities  80 objective probabilities  80, 81, 98 origins  73, 81–82 other approaches distinguished  79 parliamentary sovereignty and  89–99 prototypes  80, 87 set-theoretic entities  73 sovereignty, models of  75 theoretical models  74–75, 83 UK Constitution  74 Young’s theory of democratic dialogue  72, 93 moderate originalism constitutional interpretation  53–54 statutory interpretation  53–54 monarchies absolute  312, 315, 329–331 Aristotle  314–315 constitutional  329, 333, 334 number of monarchies  336 United Kingdom  329, 366–368, 369–370 Montesquieu  366–367 moral aspirations constitutional law  2, 212–213, 231 intepretative method and  42–43 moral principles See moral principles moral authority assurance model and  225–228 moral footprint of settlement constitution as  225–228 moral impact theory  146 moral legitimacy administrative law  364 constitutional government and  312, 313, 364 moral obligations assurance legitimacy and  217–218, 220–225 Hart’s Separation Thesis  183, 186–187, 191–192, 197, 203–204, 206, 235

Hobbes’s Leviathan  189 Kelsen on  192–193 natural law theory  185–186 positivism and  186 moral principles Allan’s constitutional methodology  213, 225, 228–230 allegiance to political regimes and  3, 7 assurance legitimacy and  217–218, 220–225 authority, establishing  9 benchmarks for exercise of state power  12 binding  7, 231 common law constitutionalism  7, 135, 136, 137, 138, 140, 148–150 common understanding  5, 17, 21, 24–25, 33–35, 39 constitutional theory, role in  8–10, 214–215 contested nature  3 criteria of importance and  6–7 equality  211–212 evaluation of constitutional law  8 generally  2 interpretivist method and  5, 7, 41–42, 58–63, 132, 137, 147–149, 245 Lakin’s constitutional methodology  213 law conflicting with  182–183, 197–198 legitimacy, constitutional theory enhancing  9–10 moralised constitutional theory  147, 211–213 political legitimacy and moralised constitutional theory  211–231 political morality  211 positivism and  132, 235 positivist and political approach to  250 public law and  249–250 rationalism and  81 rule of law  211–212 statutory infringment of fundamental rights  220–222 statutory interpretation and  47–48, 137, 146, 149–150 systemic moral errors  38 Mortati, C  263, 264 Moyle, W  112 Mussolini, B  280–281 N Nagel, T The View from Nowhere  341–342 Nakkuda Ali v Jayaratne  381 Nason, S  2, 8, 10, 11 ‘new administrative of Wales’  166 Reconstructing Judicial Review  155, 156, 157, 163, 165, 167, 172

Index  419 natural justice administrative law and  362, 379–382 Natural Language Processing (NLP)  173, 174 natural law theory  179–180, 185, 204, 205, 207–208 foundational law  180 interpretativism and  236 Kelsen  194 positivism and  185–186, 194, 238, 245, 246 natural rights Blackstone  366 Neuberger LJ  69 New Public Management  166 New Zealand Bill of Rights  325 constitution  41 customary norms  41 parliamentary democracy  334, 336 parliamentary sovereignty  44, 325 Nkrumah, Kwame  121 non-constitutional systems absolute monarchy  312, 315, 329–331 dictatorships  312, 329, 330–331 generally  313, 315, 338 single-party states  312, 317–318, 329, 331–333, 337–338 normative analysis  2 normative constitutional theory model-theoretic approach  77, 78 O Oakeshott, M common understanding of ethical principles  24–25, 34–35, 81 organic naturalist interactivist-constructivist approach  153, 169–170, 171 originalist constitutionalism Dworkin  43, 228 generally  227–228 moderate  53–54 political morality and  228 P Paine, Thomas Agrarian Justice  118 Rights of Man  118 parliamentary democracy  319–320, 330, 334–335 prime minister  325 voting systems  336–337, 338 Waldron  345 parliamentary sovereignty See also legislative supremacy; sovereignty administrative law and  374 Blackstone  365, 367–370, 372, 373

Brexit and  72, 89–99 Canada  325 common law and  6, 42, 44, 55–58, 68–69 common law constitutionalism  55–59, 63–67, 74, 133–134 constitution-making and  55 constitutional principles and  56–57 courts and  254 critical approach  91–92 customary norm, as  41 democracy and  93–94, 254 Dicey  64–65, 67, 93, 94, 95–96, 334, 370, 373 Dworkin  63, 68 executive and  96 Gordon  94, 96–98 historically-minded approach  90–91, 144–145 international treaties and  95 interpretivist method  58–61, 64, 68, 92 Jackson  65, 112–114 judicial review and  377 judiciary and  55–58, 63–69, 133, 254 legal authority and  202 legal pragmatism  66, 67–69 legal theory of constitutional order  262 legislative legitimacy theory  357 legislative supremacy and  94, 133 limitation  66–68, 95, 112–113, 114 Loughlin  94 Miller I and II  44, 55, 64–69, 90, 120–121, 254–256, 307 model-theoretic approach  89–99 New Zealand  44, 325 political constraints  254 positive and political approach  10, 59, 68, 254 recent developments in judicial reasoning  55–58, 63–69 rule of law and  133, 254 self-correcting unitary democracy  93, 95, 98 United Kingdom  44, 55–59, 63–69, 89–99, 112–113, 133, 155, 325, 367–368, 370 Young’s theory of democratic dialogue  93, 94–95, 97–98 parliamentary systems  312, 313, 320, 330, 334–337 Westminster model  336 Peoples’ Republic of China Constitution  332–333 non-constitutional system  338 Pettit, P  217 Plato  283 Platt, L  154, 165 Polanyi, M The Logic of Liberty  286 political authority approaches to  236

420  Index political constitutionalism  74, 106, 107, 110, 119–120, 156–159, 246–248, 277, 312 criticisms of  247 positivism and  156–157, 247 political ideologies as product of political activity  24 political jurisprudence legal authority and  235 legal positivism and  10, 233–257 Loughlin  10, 233–237 public law as  233–257 political law constitutionalism  1, 2 positive law and  10 political legitimacy See legitimacy political morality assurance legitimacy  9–10, 212, 216–222, 223 moralised constitutional theory  211 originalist constitutionalism  228 political regime See state political theology material study of constitutional order  270–272 order, producing  271 sacrificing for  270–272 sovereignty and  270–271 political will formation of constitutional order  265 Ponsonby rules  298 popular understanding generally  3 positivism anti-positivism  3, 10, 233–234, 236 Bentham  238, 243 command theory of law  181, 187, 192, 207–208 concepts, explaining  21–23 democracy and  250 empirical analysis  131–132, 135, 137 generally  3, 5, 41, 63 Goldsworthy  7, 41–69, 131–152 Hart  3, 5, 9, 21–22, 180, 184, 188, 198, 205–207 historical methodological argument  142–146 Hobbes  238, 243 interpretative method compared  7, 10, 59–61, 131–152 judicial discretion and  251–252 justice and  9, 10 Kelsen  3, 9, 179–180, 194 law as matter of fact  41, 132, 186 ‘law may have any content’ thesis  188 legal authority and  179–180, 188, 190–191, 235, 241 legal positivism  243

legal reasoning, impact on  251–252 legal validity, separation  242, 249–250 legitimacy and  9, 250, 252–253 moral principles and  132, 235 natural law theory and  185–186, 194, 238, 245, 246 normative positivism  10, 233–234, 237–241, 243–246, 250 parliamentary sovereignty  10, 59–61, 68 philosophical methodological argument  142–146 political authority, approach to  236 political constitutionalism and  156–157 political jurisprudence and  10, 233–257 political positivism  238 Raz  22, 186, 198 rule of recognition  65, 68, 132–133, 181–183, 184, 187 UK Constitution  41–69, 131–133 Waldron  10, 233–234, 237–241 weakness  7 Postema, G  60 Pound, R  50 pouvoir constituant  319, 327 presidential systems  312, 318, 320, 322, 334, 335, 337, 338 principles See also moral principles common understanding of  5, 17, 21, 24–25, 33–35, 38 constitutional  56–57, 322–323, 324, 328 rationalism and  81 Privacy International  66, 139, 140, 150 proportionality administrative law  362 administrative law reforms  155 protection constitutions as protective shields  115–117, 119 public choice theory constitutional design and  283 predictive power  283 public good constitutions and  116, 117 public law common understanding and  236 community’s collaborative quest for justice  229, 250 constitutional authority, generating  237, 241, 242 democracy and  250 empowering character  249, 249241 function of  236 interpretivism and normative positivism compared  244–246 legitimacy  250, 252–253

Index  421 morality, inherent connection with  249–250 normative positivism  233–257 political authority  236 political constitutionalism  246–248 positivist and political approach  10, 233–257 role, generally  241, 249 sovereignty See parliamentary sovereignty; sovereignty values and facts  235, 237 purposivist methodology legislative supremacy doctrine  45 Q Quine, WVO  164–165 R R (Evans) v Attorney-General  298 R v Legislative Committee of the Church Assembly  374 rationalism heuristic models, development  81 use in constitutional theory  81 Rawls, J  112 A Theory of Justice  34 Raz, J  9, 107, 112, 320 conceptual analysis, approach to  22, 23, 30 on constitutions  320, 323, 325, 328–329 ‘legal man’ concept  185–186, 191–192, 197, 200, 205 legal positivism  22, 186, 198 normal justification thesis  204 theory of legal authority  185–188, 191, 196, 204–205, 212 ‘reading in’ implied terms  47 reasonableness administrative law  362 recognition, rule of  65, 68, 132–133 conventionalism  269–270 Hart  144–145, 181–183, 184, 187, 269 Reed LJ  66, 67, 90 Reid LJ  379, 381 republican states  333–334 number of republics  336 rhetorics of intransigence constitutional theory and  122–125 Ridge v Baldwin  379, 381–382 rights-based approach See also fundamental rights discord perspective on society  116, 120–122 growth  155, 156–157 Roberts v Hopwood  377 Roederer, C  73, 82 Roman Republic  316 Romano, S  263, 264, 265, 266

Roznai, Y  354–355 Rubinelli, L  349–350, 352, 353 Rucho v Common Cause  119 rule of law administrative law and  363, 364, 376–377, 380–381 common law constitutionalism  133–134 conceptual analysis of  24, 34 constitutional government and  315, 322, 327–328 constitutionalism and  312 Dicey  107, 375 evil law, compatibility  315 Hayek  376–377 judicial review and  377 legal authority and  203 legislative supremacy and  133 moralised constitutional theory  211–212 parliamentary sovereignty and  133, 254 statutory interpretation and  47, 66–67 United Kingdom  66–67, 138–139, 140, 372, 375 rulebooks, constitutions as  115, 118–121 Rundle, K  13–14, 168, 170, 175 S Sales J  64 San Marino  334 Sartori, G  214, 320 Scalia J  176 Schauerman  76  Schmitt, C  121, 270 Scotland Act 1998  293–294, 295–296 Scottish Constitutional Convention  295–296 self-understandings common understanding in legal theory  17–39 generally  5–6 Sennett, R The Craftsman  285, 288, 290, 291, 292, 306, 307 separation of powers administrative state  389, 396, 398–399 Australia  394–395 conceptual analysis of  21, 24, 34 constitutionalism and  312 judges, constitution-making by  55–58, 63–69 moral aspiration  2 United Kingdom  56–58 United States  318 Vile’s account  34 Separation Thesis  183, 186–187, 191–192, 197, 203–204, 206, 237, 242 Sewell Convention  295, 307

422  Index Shapiro, S  32, 62 Shklar, J liberalism of fear  223–224 Sieyes, Abbé  319, 327, 349 Simmonds, N  315 single-party states  312, 317–318, 329, 331–333, 337–338 Smart, Carol  234, 244 social contract assurance legitimacy  9–10, 212, 216–222, 223 Blackstone  366 constitutional theology and  271 constitutionalism of fear  222–225 legal authority and  179–180, 201, 206 material study of constitutional order  267–269 natural law theory  179–180 social forces material study of constitutional order  11 social institutions common understanding of  17, 18, 21–24, 26–27, 35, 39 constitution  17, 26–27 law as a social institution  22 role of theorist in shaping  23, 39 socially-created rules  27 social order bearer or subject of constitutional order  277 social organisation legal order and  263 material study of constitutional order  272 societal constitutionalism functional systems  266–267 generally  262–263, 264–265 political will and  265 state and  266 socio-legal theories of administrative law  159–162 South Africa Bill of Rights, interpretation  346 Constitution  122 sovereign authority See also parliamentary sovereignty constitutions as rulebooks  115, 118–121 sovereignty See also legislative supremacy; parliamentary sovereignty conceptual analysis of  24 doctrine of  44 legal authority and  182 material study of constitutional order  270–271, 275, 281–282 political theology and  270–271 public law and  249

state See also legitimacy assurance legitimacy  9–10, 212, 216–222, 223 assurance model  225–228 bearer or subject of constitutional order  273–274, 276–277 collaborative quest for justice  229 common good, state policy and  218 common understanding of  18, 39 constitutionalism of fear  222–225 decision-making procedures  221–222 fundamental norms of constitutional order  277–278 government efficacy  226 monarchies See monarchies moral bond with  3, 7, 10, 217–218, 220–225 non-constitutional systems  329–333 political legitimacy  211–231 republican  333–334 statutory infringement of fundamental rights  220–222 unitary order  273–274 statute law common law and  42, 43, 44, 45–47, 48, 49–50, 51–54, 65 common law distinguished  52–53 conflicts between two Acts  90 constitutional improvisations  292–308 constitutional statutes  56, 321, 324 equal citizenship principle  134 implied repeal  56 improvisatory  285, 287–288, 289–308 interpretation See statutory interpretation judicial review  54–55 legislative intention See legislative intention legislative supremacy See legislative supremacy legitimacy  9, 134 new criminal offences  48 parliamentary sovereignty See parliamentary sovereignty positive and political approach  10, 59–61, 64–65 rule of recognition  65, 68, 132–133 unjust  133, 182–183, 197–198, 222–225 statutory interpretation common law and  42, 43, 44, 45–47, 48, 49, 51–54 common law constitutionalism  42, 53, 58, 64–67, 133–134, 136–137 common law pragmatism  58, 66, 67–69 common law principles  52 criminal and civil law  146 drafting errors  46, 48

Index  423 equitable interpretation  46–47 ‘faithful agents’ doctrine  45, 47, 56 generally  52, 198 implied terms  47 indeterminacies, resolving  46, 47, 48 interpretativism See interpretative method judicial  43, 45–55, 133, 146 judicial pragmatism  48–51, 58, 66, 67–69, 139 legislative history, admissibility  46 legislative intention, inferring  46, 47, 48, 49, 133, 150 legislative supremacy doctrine  3, 6, 44–45, 47 linguistic canons  46 linguistic meaning, rectification  46 methodology  41–42, 43, 45–47 moderate originalism  53–54 moral principles and  47–48, 137, 146, 149–150 orthodox view of  47 parliamentary intent  148 parliamentary sovereignty and  44, 55–56, 63–69 positivism See positive method presumptions of intentions  48 principle of legality  49, 50 proof of legislative intention  46 ‘reading down’  46 ‘reading in’  47 rectification  46–48 rights and  47–48 rule of law and  47, 66–67 spurious  50 standing commitments within legislature and  47–48, 50 statutory presumptions  49 supplementing interpretation  146 unanticipated social or technological developments and  46 Stavropoulos, N  43 Steel, David  295 Steyn LJ  50, 112–114 structural topic modelling  173–174 subsidiarity constitutionalism and  312 Sumption LJ  66, 67 Sunkin, M  155, 165 Sunstein, C  28, 36 ‘faithful agents’ doctrine  45 Suppe, F  73 Suppes, P  73 Suteu, S  12–13 Switzerland Constitution  334–335 constitutional change  279–280 federal law, precedence  325

T Taiwan status, generally  18, 39 Teubner, G  265 textualist methodology generally  228 legislative supremacy doctrine  45 Thiruvengadam, A  127 Thomas Justice C  108 Thomas Justice EW  50 Tocqueville, Alexis de  376 judicial legislation  375 tyranny of the majority  322–323, 370–371, 374  Tomkins, A Our Republican Constitution  107–108, 110 Tomlinson, J  167, 171 Tomlinson, J and Pickup, A  161 topic modelling  173–174 Touré, Sekou  121 traditions, development  108–109 truce, constitutions as  115, 121–122 Trump, Donald  389 Tsuzuki Keiroku  121 Turkey Constitution  113, 122 Tushnet, M  342 U ultra vires doctrine administrative law  50 ultra vires theory judicial review, of  54–55 Union of Socialist Soviet Republics (USSR)  337–338 Constitution  331–333 unitary order constitutional order as  273–275 United Kingdom administrative law, development  361–382 audi alteram partem principle  381 Brexit See Brexit Cabinet  369 case law  139–140, 144 comparative material, use in adjudication  346 conflicts between two Acts  90 constitution See United Kingdom constitution constitutional continuity  109–110, 111, 113 constitutional conventions  135, 321 constitutional principles  56–57 Constitutional Reform Act 2005  293, 296–297, 302–303, 305–306 Constitutional Reform and Governance Act 2010  293, 298–299, 303–306 constitutional statutes  324

424  Index customary norms  41 democracy  319–320, 334 electoral system  306 English Civil War  313, 317 English Votes for English Laws  291 European Convention on Human Rights  364 European Union Act 2011  300 European Union Referendum Act 2015  11, 293, 294, 298, 300–302, 305–306 European Union (Withdrawal) Acts  289, 304 ‘faithful agents’ doctrine  56 Fiscal Responsibility Act 2010  288 Fixed-term Parliaments Act 2011  11, 293–294, 298, 299–300, 306 Freedom of Information Act  298 House of Lords reform  293, 294–295 HS2  56 Human Rights Act 1998  293, 295, 325 improvisatory statutes  287–288, 289 judicial pragmatism  51, 58, 66, 67–69, 139 Lascelles principles  299, 300 Law Commission, working method  288–289 legislative intention  133 legislative legitimacy theory  357 legislative supremacy  55–56, 58–59 monarchy  329, 366–368, 369–370 parliamentary sovereignty  44, 55–61, 63–69, 89–99, 112–113, 114, 133, 155, 254, 325, 357, 367–368, 370 parliamentary Standing Orders  71 parliamentary system  313, 320, 330, 334–336 political accountability, erosion  345 political and legal constitutionalism  106, 107 Ponsonby rules  298 post-Union Parliament  109–110 Prime Minister  335, 369 principle of legality at common law  357 R (Evans) v Attorney-General  298 R v Legislative Committee of the Church Assembly  374 Roberts v Hopwood  377 rule of law principle  66–67, 138–139, 140, 372, 375 rule of recognition  65, 68, 132–133 Salisbury convention  321 Scotland Act 1998  293–294, 295–296 separation of powers  56–58 Sewell Convention  295, 307 statutes  133 Supreme Court, Miller I and II  44, 55, 64–69, 71, 90, 120–121, 254–256, 303, 307 Supreme Court, Privacy International  66 Westminster model  336

United Kingdom constitution See also United Kingdom Act of Settlement 1701  363, 366, 372 Agreement of the People (1647)  317 Bagehot  334, 335, 369 balanced constitution  112 Bill of Rights 1689  363, 366, 372 Blackstone  364–370, 372, 373, 375, 377, 379, 382 Burke  319–320 changes to  95, 112–113, 114, 284 common law constitutionalism  42, 58, 63–67, 74, 131–152, 364 constitutional liberty  366–368, 375 Dicey  13, 57, 95, 319, 334, 357, 368–380, 382 executive power  366–367, 368, 369–370, 373–374, 375 generally  4 Glorious Revolution  317, 320, 363, 366 Habeas Corpus Act  366, 372 improvised order, as  11–12, 283–308 influence  319–320 Instrument of Government (1653)  317 interpretative method  42, 58, 60–61, 131, 133–134 judge-made constitutional principles  55–58, 63–69 judiciary  367, 368, 369, 371–372, 375 legal authority  182 legal constitutionalism  74 legal or political concept  164 legislative power  366–367, 368 Locke  116 Paine  118 Petition of Right  366, 372 political constitutionalism  74 positivist approach to  41–69, 131–152 unwritten  52, 60, 202, 320, 324, 369 United States administrative state  384, 385, 388–390, 393 Bill of Rights  326 common law constitutionalism  53 comparative material, use in adjudication  346 Constitution  116, 228, 313, 317, 318, 319, 320 constitutional conventions  321–322 democracy  322–323 ‘faithful agents’ doctrine  45, 47 Federalist Papers  112, 283, 318, 326 judicial review  112, 325, 326 Judiciary Act 1789  321 Lawrence v Texas  346 legislative supremacy  44–45 Lochner  325 Marbury v Madison  112, 325, 326 moderate originalism  53

Index  425 politicisation of judiciary  345 presidential system  322, 334, 335, 337 Rucho v Common Cause  119 Second Amendment  108 separation of powers  318 sovereignty, concept of  44 Supreme Court  321, 345, 389 written constitution  44 V Vander Waert, PA  314 Vermeule, A  36 Vile, MJC  34 W Wade, Sir William  50 Waldron, Jeremy  3, 60 Dworkin-Waldron disagreement  343–344, 345 Law and Disagreement  220, 221, 238, 343 Loughlin’s political jurisprudence and  10, 233–234, 237–241, 242, 248, 257 normative positivism  10, 233–234, 237–246, 257 parliamentary democracy  345 Wales, Prince of ‘black spider letters’  298, 304 Walters, M  379 Webber, G  320 Weber, Max ideal-types  77, 81

Weinert, F  77 Weinrib, J  219 Western liberalism  337, 338, 344 Wheare, K  324 Wilberforce LJ  56 Williams, B  222 Wilson LJ  66 Windeyer J  53 Wollstonecraft, Mary  311 written constitution absolute monarchies  330 written constitutions common law and  6, 42, 43–44 common law constitutionalism  53, 63–65 common law distinguished  52–53 customary norm, as  41 generally  320, 324, 328 interpretation  43, 62 living constitutionalism  53 X Xi Jinping  333 Y Young, A  3, 6 theory of democratic dialogue  72, 93, 94–95, 97–98 Yowell, P  12 Z Zhang, Q  332–333

426