Complex Equality and the Court of Justice of the European Union: Reconciling Diversity and Harmonization [1 ed.] 9789004354265, 9789004299993

Citation preview

Complex Equality and the Court of Justice of the European Union

Nijhoff Studies in European Union Law Series Editors Prof. Fabian Amtenbrink (Erasmus University Rotterdam) Prof. Ramses A. Wessel (University of Twente)

volume 14 Nijhoff Studies in European Union Law is a refereed scholarly monographs series dedicated to the critical analysis of the current state and development of European Union law in a broad sense. Apart from constitutional, institutional and substantive issues of eu law, the series also embraces state-​of-​the-​art interdisciplinary, comparative law and eu policies research with a clear link to European integration. Titles in the Nijhoff Studies in European Union Law series will be of particular interest to academics, policy makers and practitioners dealing with eu law and policies, as well as national and international (non-​) governmental institutions and bodies.

The titles published in this series are listed at brill.com/​seul

Complex Equality and the Court of Justice of the European Union Reconciling Diversity and Harmonisation By

Richard Lang

with a foreword by

Mark Bell

LEIDEN | BOSTON

Library of Congress Cataloging-in-Publication Data Names: Lang, Richard (Richard A.), author. | Bell, Mark, 1974Title: Complex equality and the Court of Justice of the European Union :  reconciling diversity and harmonisation / Richard Lang ; with foreword by Mark Bell. Description: Boston : Brill, 2018. | Series: Nijhoff studies in European Union law,  ISSN 2210-9765 ; volume 14 | Includes bibliographical references and index. Identifiers: LCCN 2018023152 | ISBN 9789004299993 Subjects: LCSH: Court of Justice of the European Union. | Equality before the law—European  Union countries. | Discrimination—Law and legislation—European Union countries. |  Equality—Philosophy. | Distributive justice. | Cultural pluralism. | Walzer, Michael. Classification: LCC KJE5142 .L35 2018 | DDC 342.2408/5—dc23 LC record available at https://lccn.loc.gov/2018023152

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill.” See and download: brill.com/​brill-​typeface. ISSN 2210-​9 765 ISBN 978-90-04-29999-3 (hardback) ISBN 978-90-04-35426-5 (e-book) Copyright 2018 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi, Brill Sense and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. This book is printed on acid-​free paper and produced in a sustainable manner.

There is no progress... but at most a continuous and sublime recapitulation. UMBERTO ECO, The Name of the Rose (Harcourt 1983)

∵

Contents Foreword xi Acknowledgements xiii 1 Introduction 1 1.1 Starting Point 1 1.2 The Trouble with Aristotle: A Précis 1 1.3 The Project in Outline 9 1.4 The Search to Find a Complement 12 1.4.1 The Difficulty 12 1.4.2 Equality versus Freedom 14 1.4.3 Reasons for Choosing Walzer 18 2

Michael Walzer and Complex Equality 22 2.1 Introduction 22 2.2 Complex Equality: A Thumb-​Nail Sketch 23 2.3 Negative Dominance 26 2.4 The Concept of ‘Shared Understandings’ Investigated Further 28 2.5 An Objection: Walzer’s ‘Relativism,’ and The Need for An Override 36 2.6 The CJEU and its Override: An Answer to the Objection 42

3

The Principle of Equal Treatment of Persons Irrespective of Gender 46 3.1 Introduction 46 3.2 Outline of Analysis and Taxonomy of Results 47 3.3 Characteristics and Voluntary Choices: Is the Theory Fit for Purpose? 49 3.4 Pregnancy 54 3.4.1 The CJEU’s Case-​Law 54 3.4.2 Preliminary Analysis 59 3.4.3 A Walzerian Analysis 63 3.5 Paternity leave 68 3.5.1 The CJEU’s Case-​Law 68 3.5.2 A Walzerian Analysis 72 3.6 Pensions 74 3.6.1 The CJEU’s Case-​Law 74 3.6.2 A Walzerian Analysis 81



viii Contents

3.7 The Question of Part-​Time and Full-​Time Work 85 3.7.1 The CJEU’s Case-​Law 85 3.7.2 Preliminary Analysis 91 3.7.3 A Walzerian Analysis 94 3.8 Positive Action 99 3.8.1 The CJEU’s Case-​Law 99 3.8.2 Preliminary Analysis 103 3.8.3 A Walzerian Analysis 106 3.9 Insurance Premiums and Benefits 108 3.9.1 The CJEU’s Case-​Law 108 3.9.2 Preliminary Analysis 110 3.9.3 A Walzerian Analysis 112 3.10 Concluding Thoughts 114

4 The ‘Article 19’ Grounds: Racial or Ethnic Origin, Religion or Belief, Disability, Age, and Sexual Orientation 118 4.1 Introduction 118 4.2 Racial/​Ethnic Origin 119 4.2.1 The CJEU’s Case-​Law 119 4.2.2 A Walzerian Analysis 127 4.3 Religion and Belief 131 4.3.1 The CJEU’s Case-​Law 131 4.3.2 A Walzerian Analysis 134 4.4 Age 140 4.4.1 The CJEU’s Case-​Law 140 4.4.2 A Walzerian Analysis 154 4.5 Sexual Orientation 160 4.5.1 The CJEU’s Case-​Law 160 4.5.2 A Walzerian Analysis 174 4.6 Disability 182 4.6.1 The CJEU’s Case-​Law 182 4.6.2 A Walzerian Analysis 189 4.7 Concluding Thoughts 196 5

Nationality Discrimination 198 5.1 Introduction 198 5.2 Nationality Discrimination in EU Law 199 5.3 A Specific Example: The Free Movement of Persons 203 5.4 The Early Case-​Law 206

Contents

ix



5.5 A Walzerian Analysis 207 5.6 Cases Deriving from the Second World War 208 5.7 A Walzerian Analysis 210 5.8 The Citizenship Case-​Law: Early Days 213 5.9 A Walzerian Analysis 215 5.10 The Citizenship Case-​Law Evolves 217 5.11 A Walzerian Analysis 219 5.12 Interim Conclusion 219 5.13 The Citizenship Case-​Law Since 2005 220 5.14 A Walzerian Analysis 226 5.15 Solidarity 226 5.15.1 What is Transnational Solidarity? 226 5.15.2 Consideration and Responsibility 229 5.15.3 Asymmetric Solidarity, Pan-​European Solidarity Publics and Walzer’s Position 234 5.15.4 Solidarity and Complex Equality 237 5.16 Concluding Thoughts 238 6

Semi-​Suspect and Non-​Suspect Grounds 240 6.1 Introduction and Explanation of Terminology 240 6.2 Semi-​Suspect Grounds 245 6.2.1 Article 40(2) TFEU: Producers and Consumers 245 6.2.2 Article 106(1) TFEU: Public and Private Undertakings 251 6.3 Non-​Suspect Grounds 253 6.3.1 General Introduction: Omega 253 6.3.2 Public Procurement 256 6.3.3 Human Resources 261 6.3.4 Competition Law 269 6.4 Luck: Boundary Breach or Small Inequality? 271 6.5 Concluding Thoughts 274

7 Reflections 276 7.1 Introduction 276 7.2 Three Counter​arguments 276 7.2.1 The Need for Judicial Interventionism 276 7.2.2 The Possible Need for Expert Evidence, or Even Specialized Courts 282 7.2.3 The Need to Rewrite Legislation 287

x Contents 8 Presenting a Theory of Mediated Complexity 289 8.1 Introduction 289 8.2 Further Consideration of the Forum, and Some Possible Problems 290 8.2 . 1 The Design of the Forum 290 8. 2 . 2 The Danger of the Forum 292 8. 2 . 3 Interim Conclusion 296 8.3 Mediated Complexity 297 8.4 The Judge vis-​à-​vis Society 304 8.5 Conclusion 306 9 Evaluation 309 9.1 The Two Methods Compared: Erpelding 309 9.2 Formulating the Argument 311 9.3 In Search of Flexibility 312 9.4 But Does Flexibility Lead to Uncertainty? 313 9.5 Judge Right Now 317 9.6 Postmodernism and Precedent 322 9.6 . 1 Dworkin on Precedent, and A Critique 322 9.6.2 Postmodernism Challenges Neutrality and Positivism 326 9.6 . 3 Saying Yes to ‘We’ 327 9.7 Macro and Micro: The Equilibrium of Co-​Existence 329 9.8 Complex Equality and European Union 335 9.9 Complex Equality for European Courts? 341 9.10 Parting Words: To Thine Own Self Be True 344 10 Conclusion 347 Appendix i 351 Bibliography 352 Index 373

Foreword ‘Equality’ commands widespread support as a political objective. It finds strong foundations in both constitutional law and international human rights law. Its appeal as a value or concept is, however, matched by an enduring debate on what the pursuit of equality means in practice, whether in the realm of public policy or before the courts. For some, the essence of equality is treating everyone in the same way, whereas for others equality demands different treatment in order to redress engrained patterns of disadvantage and under-​representation. Confronted with the task of interpreting equality, judges in many jurisdictions have turned to the broad Aristotelian notion of treating those who are similarly situated in the same way, and treating those who are differently situated in a different way. The tendency of courts to use the Aristotelian idea of equality as a springboard for the analysis of individual cases of alleged inequality has often been subject to critique within academic commentary. Much of this literature has directed its concern towards a judicial inclination to concentrate upon ensuring the same treatment of those in a similar situation. This emphasis on consistency has been equated with the label of ‘formal equal treatment’ and lamented for its reinforcement of the existing ‘norm’, which tends to reflect advantaged groups in society, such as white, able-​bodied men. Frequently, legal scholars in Europe have exhorted courts to strive further to attain ‘substantive equality’ by giving appropriate attention to the need for different treatment to combat disadvantage and to accommodate diversity. The expansion of EU anti-​discrimination legislation since, in particular, 2000 has spawned a rapid expansion of academic commentary in this field. While this embraces a rich diversity of ideas, it seems fair to observe that the most common theoretical paradigm remains the perceived tension between formal and substantive equality. In this book, Richard Lang offers the reader a fresh and alternative perspective through drawing on the writings of Michael Walzer and his concept of ‘complex equality’. This is not presented as a rejection of the Aristotelian framework, but a complement. Walzer’s ideas are not frequently cited in legal scholarship on EU anti-​discrimination law, but this book demonstrates that his philosophy can be meaningfully transported into this area of law. The book is a theoretical project, but it avoids the comfort of remaining in the realm of abstract reasoning. Instead, Richard Lang puts the Walzerian perspective to the test. He does this by examining a broad swathe of topical themes from the case-​law of the EU Court of Justice. In so doing, this book provides the reader with an insightful overview of the case-​law, including its

xii Foreword rapid expansion in the past decade. For example, issues addressed include discrimination related to pregnancy and maternity; the scope for positive action to combat the under-​representation of women in the workplace; and the rights of workers with same-​sex partners to receive the same treatment as married couples in relation to employment benefits and pensions. This book identifies the existing critique of the Court’s case-​law, which is often premised on a perceived failure of the Court to embrace fully the ideal of substantive equality. The author takes us beyond the existing literature by reconsidering how the cases might have been decided if Walzer’s concept of complex equality were the framework for analysis. This applied testing of Walzer’s ideas exposes with great clarity how they could expand and enrich analysis of judicial decision-​ making in this field. No doubt opinions will differ on whether Walzer’s theory of complex equality offers the optimum paradigm through which to advance the application of anti-​discrimination legislation. This book demonstrates some of the strengths of Walzer’s framework. For example, it underscores the importance of not allowing advantages in one sphere of life (e.g. legal recognition of relationships) to spillover into how resources are allocated in other spheres (e.g. workplace benefits). The framework advanced offers a means of breaking free of the rigidities imposed by the exhaustive list of protected characteristics in current EU anti-​discrimination legislation. This provides more flexibility to respond to the needs of those claimants who do not fit within the current framework, such as those who become parents via surrogacy arrangements. Importantly, the book does not shy away from those areas where Walzer’s theory might prove controversial and challenging. In particular, a key role is ascribed to community deliberation in determining the ground rules for the meaning and distribution of social goods. This could, for example, provoke concern about how to ensure that communal fora are not dominated by majoritarian interests. At the same time, it prompts the reader to reflect on the role of participatory bodies in making difficult choices around the allocation of scarce resources and it suggests a need to be cognisant of the limitations to the judicial role in this respect. Above all, this book makes a compelling case for legal scholars to engage further with political philosophy and to use this to deepen analysis of how law instantiates the concept of equality. The originality of the ideas are matched by a fluent, vibrant and enjoyable style of writing. The richness of the research and the author’s skilful ability to draw upon texts well beyond the domain of law, while at the same time providing excellent legal analysis, commend this text to the reader. Mark Bell 14 August 2017

Acknowledgements This book is an updated version of a PhD dissertation which I wrote at Kings College London between 2007 and 2010. My PhD was awarded, without corrections, in February 2011. I would like to express my gratitude to my first and second supervisors at KCL, Professor Alexander Türk and Professor Aileen McColgan respectively, and to my examiners:  Professor Mark Bell, Regius Professor of Law at Trinity College Dublin, and Professor Gareth Davies, Professor in European Law at vu University Amsterdam. Especial gratitude is due to Professor Bell for his encouragement and for kindly agreeing to write the Foreword to the present volume. In 2012, an extremely truncated version of the thesis was presented at a lunchtime seminar of the Centre for Research in Law at the University of Bedfordshire, and at the Nineteenth International Conference of Europeanists, held under the auspices of the Council for European Studies in Boston. I should like to say thank you to all who attented my presentations and commented or asked questions, in particular, in respect of the Boston talk, Professor Leonard Besselink. I should also like to say thank you to Carmen Draghici, Judith Fox, Françoise Henry, Anna Myrvang, Claudia Padovani, Graham Smith, Michael Stambolis-Ruhstorfer and Mieke Verloo for their ideas and support. I should further like to thank Ingeborg van der Laan, Malathy Chandrasekaran and their colleagues at Brill, and the anonymous reviewers of my book for their helpful suggestions. One of the latter was to conduct a survey of national courts to find out what the position was there. In this regard, I would like to thank members and former members of the European Commission’s European Network of Legal Experts in the Non-​Discrimination Field who assisted with this survey: Corina Demetriou (Cyprus), Nassos Theodoridis (Greece), András Kádár (Hungary), Neža Kogovšek Šalamon (Slovenia), Vadim Poleshchuk (Estonia), Orlagh O’Farrell (Ireland), Manuel Malheiros (Portugal), Lovorka Kušan (Croatia), Tonio Ellul (Malta), Rainer Hiltunen (Finland) and Romanita Iordache (Romania), as well as Christa Tobler and Lisa Waddington. However, the kind contribution of these individuals to my work in no way represents an endorsement by them of the Walzerian theory which I am putting forward in this book, and indeed the same goes for everyone acknowledged herein. I would like to say a special thank you to the Centre for Applied Philosophy, Politics, and Ethics (CAPPE) at the University of Brighton, in particular, Professor Bob Brecher, Professor Sam Chambers (the latter in fact a visiting Leverhulme Professor from Johns Hopkins), Mark Devenney, Ian Sinclair and Clare Woodford. I was always made to feel welcome when attending CAPPE

xiv Acknowledgements seminars and events over the last two years, and I believe that the present work has been immeasurably improved by my attendance. Finally, a huge thank you to Helen Connolly, a former colleague at the University of Bedfordshire, for her friendship and for enabling me to meet Baroness Margaret Prosser, and to Baroness Prosser herself for sparing some of her most precious time to help me in my final preparations of this manuscript. The law is accurate as at 30 April 2017. A new system for the consultation and citation of case-​law in the European Union, ECLI (European Case-​Law Identifier), was launched on 29 April 2011. For that reason, judgments of the Court of Justice of the European Union and of the General Court from after that date are cited according to the new system; those from before are cited according to the old. As the new system gives a separate ECLI reference for decisions of the Court of Justice and for the Opinion of the Advocate General (if any), I have also made this distinction. The United Kingdom voted to leave the European Union on 23 June 2016, but at the time of writing little progress has been made. References will therefore continue to be made to the EU’s having twenty eight Member States. All mistakes, big and small, are entirely my own. At the end of the day, though, I am a lawyer. My treatment of the more political scientific or social scientific aspects of this inquiry, vital as those aspects are, may suffer from flaws and shortcomings:  its appeal to categories, its inevitable unconscious bias, and/​or its conscious endorsement of aspects of the status quo which others may feel need changing or even abolition. I hope I can be forgiven for these. As for the philosophy in this tome, it may be as well for the reader to regard this as being, to adapt a comment of Luc Boltanski’s about Michael Walzer, less political philosophy and more the philosophy of ordinary life. This book is for Elizabeth and Colin Lang. Richard Lang Lewes 31 May 2017

Chapter 1

Introduction 1.1 Starting Point In his History of Western Philosophy, Bertrand Russell wrote, ‘Ever since the beginning of the seventeenth century, almost every serious intellectual advance has had to begin with an attack on some Aristotelian doctrine’.1 In the case of this book, and substituting the word ‘debate’ for ‘advance’, the doctrine in question is Aristotle’s formulation (and formalization) of the principle of equality, namely that equality involves treatment proportionate to worth, as set out in the next section. The aim of the present author, however, writing exclusively in the legal context and for the most part in the context of the Court of Justice of the European Union (‘Court of Justice’ or ‘cjeu’), is not to supplant but to supplement Aristotelian equality, a black-​and-​white formula that now seems at odds with the colourful diversity of twenty first century living. It does serve some useful purposes, but where should one turn for nuance, finesse or flexibility? Must a judge do without these important things? This book presents, and tests, the thesis that Michael Walzer’s theory of complex equality could be used by the Court of Justice when dealing with cases concerning equality, as a complement to the Aristotelian ‘test’ that likes should be treated in like fashion and unlikes in unlike fashion. 1.2

The Trouble with Aristotle: A Précis

In his Nicomachean Ethics, Aristotle declared:  ‘[T]‌he origin of quarrels and complaints [is] when either equals have and are awarded unequal shares, or unequals equal shares’.2 Many courts of law around the world have adopted this principle, or at any rate its underlying logic, in dealing with present-​day ‘quarrels and complaints.’ The principle is usually restated in a slightly elongated form, such as: 1 B Russell, History of Western Philosophy (rev edn, Routledge 2004) 157. 2 Aristotle, The Nicomachean Ethics (WD Ross tr, oup 1980) 112 (Book v, Chapter 3). He made the same point (although sometimes with different wording) on other occasions. Irwin’s translation prefers ‘quarrels and accusations:’ Aristotle, The Nicomachean Ethics (T Irwin tr, 2nd edn, Hackett 1999) 71 (Book v, Chapter 3).

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/9789004354265_002

2 Chapter 1 [C]‌omparable situations must not be treated differently and different situations must not be treated in the same way unless such treatment is objectively justified.3 The Court of Justice, from a judgment of which this version is taken, usually refers to this as Community law’s ‘general principle of equality’ which is used by the Court (inter alia) to review national rules which fall within the scope of the ec Treaty.4 However, in the fifty years since the general principle was first invoked,5 the cjeu’s handling, or indeed mishandling, of it has provoked confusion among practitioners, dismay among litigants, and quite often anger among critics. It must be stressed at the outset that it is not the intention of this book to repeat work already done by others. The critique of Aristotle, the test named after him, and the myriad problems which its use has caused and continues to cause in equality litigation at the cjeu (and other courts), has been prolonged and thorough; there is little or nothing to add.6 The failure or at least partial 3 This exact wording taken from Case C-​ 56/​ 94 scac Srl v Associazione dei Produttori Ortofrutticoli [1995] ecr I-​1769, para 27. 4 Equality’s original role in eu law was as a driver for market integration. More recently, equality has been set free from the purely economic concerns of the single market to develop into a free-​standing, autonomous fundamental right. This process has been described fully in other works. See N Bernard, ‘What are the purposes of ec discrimination law?’ in J Dine and B Watt (eds), Discrimination Law:  Concepts, Limitations and Justifications (Longman 1996); G de Búrca, ‘The Role of Equality in European Community Law’ in A Dashwood and S O’Leary (eds), The Principle of Equal Treatment in ec Law (Sweet & Maxwell 1997); A Evans, ‘Union Citizenship and the Constitutionalization of Equality in eu Law’ in M La Torre (ed), European Citizenship:  An Institutional Challenge (Kluwer 1998). Also useful on this point is S Fredman, ‘The Age of Equality’ in S Fredman and S Spencer (eds), Age as an equality issue: legal and policy perspectives (Hart 2003) 43. 5 The first use of the principle is thought to have been in Case 1/​54 French Republic v High Authority of the European Coal and Steel Community [1954–​1955] ecr 1, 8.  However, the caveat concerning objective justification did not become an explicit component of the ‘test’ until 1977: Joined Cases 117/​76 and 16/​77 Albert Ruckdeschel & Co. and Hansa-​Lagerhaus Ströh & Co. v Hauptzollamt Hamburg-​St. Annen; Diamalt AG v Hauptzollamt Itzehoe [1977] ecr 1753. 6 In terms of books, and in the specific context of the cjeu, the reader is referred first and foremost to the two leading works, by Ellis and Bell: E Ellis, eu Anti-​Discrimination Law (Oxford ec Law Library, oup 2005); M Bell, Anti-​Discrimination Law and the European Union (Oxford Studies in European Law, oup 2002). However, see also (for example) Dagmar Schiek, Lisa Waddington and Mark Bell (eds), Cases, Materials and Text on National, Supranational and International Non-​Discrimination Law (Ius Commune Casebooks for the Common Law of Europe, Hart 2007), as well as the many learned articles referenced in this book; a good starting point would be the works cited immediately below, notes 20 to 30 inclusive.

Introduction

3

failure of Aristotle’s like-​for-​like test is almost universally acknowledged, and its flaws and inconsistencies need only the briefest summary here. In the field of gender discrimination, for example, two situations have been regarded as both comparable and different from one case to the next.7 Furthermore, the Court has been able to classify the alleged discriminator’s behaviour as both different treatment and identical treatment, depending solely on how it chose to phrase this behaviour,8 or whether it chose to take into account or ignore certain details.9 Even within one and the same case, the 7

8

9

For example, the situation of a man suffering an illness and the situation of a woman suffering a pregnancy-​related illness. These two situations were held to be comparable in Case C-​179/​ 88 Handels-​og Kontorfunktionaerernes Forbund i Danmark [Acting for Birthe Vibeke Hertz] v Dansk Arbejdsgiverforening [Acting for Aldi Marked K/​S] [1990] ecr I-​3979 and in Case C-​400/​ 95 Handels-​og Kontorfunktionærernes Forbund i Danmark, acting on behalf of Helle Elisabeth Larsson v Dansk Handel & Service, acting on behalf of Føtex Supermarked A/​S [1997] ecr I-​2757. However, they were held to be different in Case C-​394/​96 Mary Brown v Rentokil Ltd. [1998] ecr I-​4185 and in Case C-​66/​96 Handels-​og Kontorfunktionærernes Forbund i Danmark, acting on behalf of Berit Høj Pedersen v Fællesforeningen for Danmarks Brugsforeninger and Dansk Tandlægeforening and Kristelig Funktionær-​Organisation v Dansk Handel & Service [1998] ecr I-​7327. The matter is discussed further at section 3.4 below. Another example would be the situation of a part-​time worker and the situation of a full-​time worker. These two situations were held to be different in Case 170/​84 Bilka –​Kaufhaus GmbH v Karin Weber von Hartz [1986] ecr 1607, but comparable in Joined Cases C-​399/​92, C-​409/​ 92, C-​425/​92, C-​34/​93, C-​50/​93 and C-​78/​93 Stadt Lengerich v Angelika Helmig and Waltraud Schmidt v Deutsche Angestellten-​Krankenkasse and Elke Herzog v Arbeiter-​Samariter-​Bund Landverband Hamburg eV and Dagmar Lange v Bundesknappschaft Bochum and Angelika Kussfeld v Firma Detlef Bogdol GmbH and Ursula Ludewig v Kreis Segeberg [1994] ecr I-​5727. The matter is discussed further at section 3.7 below. For example, the case-​law on pensions in the uk, where a number of different companies granted pensions or pension-​related payments in a way that could not help but reflect the unequal retirement ages, as between men and women, in the uk at that time. In the Burton case, the Court used a relative formulation in describing British Rail’s behaviour; it was then able to conclude that, as between male and female workers, this had been an instance of identical treatment: Case 19/​81 Arthur Burton v British Railways Board [1982] ecr 554. But in the Birds Eye Walls case, the Court used a non-​relative formulation in describing Birds Eye Walls’ behaviour; it was then able to conclude that, as between male and female workers, this had been an instance of different treatment: Case C-​132/​92 Birds Eye Walls Ltd v Friedel M Roberts [1993] ecr I-​5579. The matter is discussed further at section 3.6 below. For example, the case-​law on positive action, and a number of cases concerning the rules governing job applications. In the Kalanke case, a rule whereby, in a tie-​break situation, all female candidates would be given automatic precedence was held to be, as between men and women, different treatment: Case C-​450/​93 Eckhard Kalanke v Freie Hansestadt Bremen [1995] ecr I-​3051. However, in the Marschall case, an almost identical provision, but with a ‘saving clause’ allowing a male candidate the possibility of

4 Chapter 1 Aristotelian ‘test’ for equality is often able to produce more than one result. That a case of ‘likes being treated in like fashion’ is not a case of ‘unlikes being treated in unlike fashion’, is frequently little more than a question of how closely the people or entities are inspected,10 and a case of likes being treated in unlike fashion, but for an objectively justifiable reason, could on many occasions be read as a case of unlikes being treated in unlike fashion, if the objective justification is, as it were, front-​loaded.11 As Advocate General Sharpston put it in her Opinion in Lindorfer, In practice … there may be some blurring between the assessment of characteristics which differentiate situations and the assessment of objective justification for differentiated treatment of otherwise comparable situations.12

10

11

12

pleading  ‘specific’ reasons why he should be accorded precedence, was held to be, as between men and women, identical treatment: Case C-​409/​95 Hellmut Marschall v Land Nordrhein-​Westfalen [1997] ecr I-​6363. For example, in the field of nationality discrimination, the case of Vigier:  Case 70/​80 Tamara Vigier v Bundesversicherungsanstalt für Angestellte [1981] ecr 229. This could be a case of the same law applying to two victims of persecution (‘likes being treated in like fashion’), or, if one takes into consideration the facts, firstly, that one of these has paid a contribution to a certain German institution while the other has not, and, secondly, that the one can gain retroactive admission to an old-​age insurance scheme while the other cannot, it could be a case of the same law applying in two different ways to two different types of persecution-​victim (‘unlikes being treated in unlike fashion’). The matter is discussed further at section 5.6 below. As Isiah Berlin puts it, ‘unequal treatment of various members of class A  can always be represented as equal treatment of them viewed as members of some other class B’: I Berlin, ‘Equality as an Ideal’ in fa Olafson, Justice and Social Policy: A collection of essays (Prentice-​Hall 1961) 130. For example, Joined cases C-​27/​00 and C-​122/​00 The Queen v Secretary of State for the Environment, Transport and the Regions, ex parte Omega Air Ltd (C-​27/​00) and Omega Air Ltd, Aero Engines Ireland Ltd and Omega Aviation Services Ltd v Irish Aviation Authority (C-​122/​00) [2002] ecr I-​2569, and Case T-​48/​89 Fernando Beltrante and others v Council of the European Communities [1990] ecr ii-​493. The former case is discussed further at section 6.3.1 and the latter case at section 6.3.3 below. Lenaerts gives the very good example of Case 8/​82 Kommanditgesellschaft in der Firma Hans-​Otto Wagner GmbH Agrarhandel v Bundesanstalt für landwirtschaftliche Marktordnung [1983] ecr 371: K Lenaerts, ‘L’égalité de traitement en droit communautaire:  un principe unique aux apparences multiples’ [1991] cde 3, 11. The danger of such front-​loading is that it serves to ‘hide [the] discriminatory behaviour’ altogether: K Koldinská, ‘Case law of the European Court of Justice on sex discrimination 2006–​2011’ (2011) 48 CML Rev 1599, 1610. Case C-​227/​04 P Maria-​Luise Lindorfer v Council of the European Union [2007] ecr I-​6767, Opinion of AG Sharpston, para 23.

Introduction

5

And Somek makes the point even more forcefully: [T]‌he justification of equal treatment is often tacitly and implicitly addressed when pondering the right comparator … Saying that two groups of persons are not in the same situation is tantamount to saying that there is an objective factor justifying unequal treatment.13 Lenaerts, writing extrajudicially in 1991, spends some time on the front-​ loading point, and does not seem to consider it all that tacit or implicit. Rather he refers to it –​twice –​as the Court’s ‘tendency’ ‘trend.’14 Fallon and Martin go further and allege that, on some occasions, the absence of comparability becomes the justification.15 Bammens, meanwhile, refers to the comparability  test’s being ‘controlled’ by the justification test (‘when the justification grounds are convincing, the situations are incomparable’),16 and Brillat to the Court’s ‘hiding’ the reason for the discrimination in the comparison.17 13

14 15

16 17

A Somek, Engineering equality:  An essay on European anti-​discrimination law (oup 2011) 122, 125. One might wonder if the distinction makes any difference, but certainly in eu Law it does, as the former situation is regarded as one in which no discrimination obtains at all, while in the latter case discrimination has taken place (although the discriminator has later been, as it were, exonerated). However, there is some divergence of views on the point. The Human Rights Committee, for example, has declared that ‘[t]‌he right to equality before the law and to equal protection of the law without any discrimination does not make all differences of treatment discriminatory. A differentiation based on reasonable and objective criteria does not amount to prohibited discrimination within the meaning of [the provision of the International Covenant on Civil and Political Rights guaranteeing equality]’ (emphasis added): LG Danning v The Netherlands, Communication No 180/​1984, un Doc CCPR/​C/​O P/​2 at 205 (1990), para 13 of the decision. “[T]‌endance” in the French, English translations author’s own: Lenaerts (n 11) 11. M Fallon and D Martin, ‘Dessine-​moi une discrimination’ (2010) 18 Journal de droit européen 165, 170. The particular context in question is that of the free movement of capital. A case in which the absence of comparability was to all intents and purposes used as the justification is considered at section 6.3.3 below: Case F-​11/​14 Dupré v eeas [2015] oj C213/​42. For an occasion when the Court ‘did not confound the justification for the impugned practice (…) with the reason for discrimination itself’, see Judgment of 16 July 2015, chez Razpredelenie Bulgaria ad v Komisia za zashtita ot diskriminatsia, C-​83/​14, ECLI:EU:C:2015:480, noted by Atrey: S Atrey, ‘Redefining frontiers of eu discrimination law’ [2017] pl 185, 187 (emphasis added). The case is considered at section 4.2.1 and 4.2.2 below. N Bammens, The principle of non-​discrimination in international and European tax law (ibfd 2012) 541. M Brillat, Le principe de non-​discrimination à l’épreuve des rapports entre les droits européens (Fondation Varenne 2015) 275 (‘occulter le motif de discrimination dans la comparaison,’ translation author’s own).

6 Chapter 1 But aside from these dry and somewhat technical problems with the Aristotelian test, there is another, bigger problem, for the explanation of which a distinction is usually drawn between ‘formal’ and ‘substantive’ equality.18 It is felt that Aristotle’s test, with its love of bisection and its inability to see special cases or to admit of exceptions, suffers from an excess of formality, so that the results it produces do not correspond with actual, real-​world equality (or so-​called substantive equality).19 Put another way, the test cannot differentiate between situations. A woman, for example, may wish to be treated unlike her male colleagues when it comes to a benefit such as paid maternity leave, but like them in all other respects of workplace life. A disabled person might wish to be treated like his fellows in, say, a job interview, but unlike them (should the situation arise) in terms of prison conditions.20 The ‘equality’ produced when the Aristotelian test fails to differentiate between these situations may be equality in form, but it is not equality in substance.21 The phenomenon whereby the same person would wish for their differences to be disregarded in one context, but taken into account in another, in order to achieve equal treatment in both, has been described as the ‘dilemma of difference’.22 It is a dilemma on the horns of which the Aristotelian test is severely impaled.

18 19

20

21 22

This ‘bigger problem’ concerns only equality cases involving human comparators. Some critics condemn only the first limb of the Aristotelian test (likes to be treated like) as formalist, preferring to see in the second limb (unlikes to be treated unlike) the beginnings of a substantive approach; see, for example, C Tobler, Indirect discrimination: a case study into the development of the legal concept of indirect discrimination under ec law (Intersentia 2005) 25ff. Many others, however, regard the entire test as excessively formal, and substantive equality as something entirely different. This is the position adopted in this book. See G Moon and R Allen, ‘Dignity discourse in discrimination law: a better route to equality?’ (2006) 6 ehrlr 610, for further discussion of this example. As Bell and Waddington succinctly put it, the ‘[c]‌overed ground is sometimes relevant’: M Bell and L Waddington, ‘Reflecting on inequalities in European equality law’ (2003) 28 ELRev 349, 360. For one among the very many who have looked at the formal/​substantive equality problem, see PA Cain, ‘Feminism and the limits of equality’ (1989–​1990) 24 Ga L Rev 803. D Schiek, ‘A New Framework on Equal Treatment of Persons in ec Law?’ (2002) 8(2) Eur LJ 290, 310. Schiek attributes the phrase to Martha Minow. The phenomenon is also referred to simply as the sameness/​difference debate. In a fascinating exploration of discrimination on grounds of disability in (inter alia) eu law, Joly takes Minow as a starting point and proceeds to explain how the differential approach of formal equality leads to the stigmatization of the disabled who become trapped in their –​accentuated –​difference:  L Joly, L’emploi des personnes handicapées entre discrimination et égalité (Dalloz 2013), for example, at 11–​12 and 233.

Introduction

7

In respect of its equality case-​law, Carey has said that the Court of Justice is incoherent, and criticized its ‘illogical application of tests.’23 Holtmaat and Tobler describe the Court’s approach as ‘misguided.’24 Prechal talks of ‘methodologically puzzling’ approaches and ‘mixed messages,’25 and Somek of ‘interpretative liberties.’26 The Court has also been repeatedly decried as inconsistent.27 These two criticisms of the Court’s approach –​that of incoherence and that of inconsistency –​are simulataneously levelled at the Court by 23

24

25 26 27

N Carey, ‘From obloquy to equality: in the shadow of abnormal situations’ (2001) 20 yb of Eur L 79, 99. There have been further complaints of incoherence, or confusion, in the Court’s equality case-​law from Cordewener, Kofler and van Thiel (A Cordewener, G Kofler and S van Thiel, ‘The clash between European freedoms and national direct tax law: public interest defences available to the Member States’ (2009) 46 CML Rev 1951, at 1952 and 1997) and from Bammens (n 16) 538. R Holtmaat and C Tobler, ‘Cedaw and the eu’s Policy in the Field of Combating Gender Discrimination’ (2005) 12(4) mj 399, 416. This was in the particular context of Case C-​220/​02 Österreichischer Gewerkschaftsbund, Gewerkschaft der Privatangestellten v Wirtschaftskammer Österreich [2004] ecr I-​5907. Sacha Prechal, ‘Equal Treatment, Non-​Discrimination and Social Policy: Achievement in Three Themes’ (2004) 41 cml Rev 533, 537, 543 and 538. A Somek, ‘Solidarity Decomposed: Being and Time in European Citizenship’ (2007) 32(6) EL Rev 787, 801. Christopher Brown, ‘The race directive: towards equality for “all” the peoples of Europe?’ (2001–​2002) 21 yb of Eur L 195, 205 (‘Indirect discrimination is, however, a rather sensitive issue, and one which has not received consistent interpretation by the European Court of Justice’); TK Hervey and J Shaw, ‘Women, work and care: women’s dual role and double burden in ec sex equality law’ (1998) 8(1) Journal of European Social Policy 43, 47 (‘there remains a sufficient degree of inconsistency in the interpretations which the Court of Justice has given over the years’); H Meenan ‘Introduction’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 3, 11 (‘[The] development [of equality and non-​discrimination] has been … lacking in uniformity’); J Calderón and A Baez, ‘The Columbus Container Services ecj case and its consequences: a lost opportunity to shed light on the scope of the non-​discrimination principle’ (2009) 37(4) Intertax 212, 216 (‘the real problem is not about lack of rulings but about lack of consistency in the case law’); J Schwarze, European Administrative Law (rev 1st edn, Sweet & Maxwell and Office for Official Publications of the European Communities 2006) 564 (‘Not all the Court of Justice’s decisions have applied these criteria consistently and logically, however’). But, contra, J Croon, ‘Comparative institutional analysis, the European Court of Justice and the general principle of non-​discrimination –​or –​alternative tales on equality reasoning’ (2013) 19(2) Eur LJ 153. This paper does a good job in increasing one’s understanding of the Court’s approaches to equality, maybe even exonerating the Court of the charge of inconsistency, but what one comes to understand is how little actual concern for equality there is in said approaches. Rather, according to Croon, the primary concern is institutional function –​or malfunction. Thus the need for reform very much remains.

8 Chapter 1 Fallon and Martin in a masterful exegesis entitled, ‘Dessine-​moi une discrimination.’28 The cjeu’s currency as an arbiter in equality matters becomes devalued as a result of these criticisms.29 Under these circumstances, the Aristotelian test becomes little more than a figleaf for the Court’s own view of what is or is not equal. As More has put it, ‘a court can achieve any result it chooses by simply changing the measure it uses’.30 Aristotle’s cruel callipers are widened or narrowed at the whim of the person holding them; measurements are taken and lines drawn. If the litigant is lucky, these lines will be drawn so accurately that they will delineate perfectly the relative positions of him or her and his or her fellows. If the litigant is unlucky, one of the lines will be drawn on the wrong side of him or her (principally because the measurer cannot see him or her), forcing him or her and his or her fellows into a patently false opposition on one side and/​or a patently false kinship on the other. It is hard not to be reminded of the movement of counters in a board game, Snakes and Ladders, for example. Those making use of the Aristotelian test employ a strange left-​or-​right, up-​or-​down mentality, which seems to see people in two dimensions rather than three. But an overly pliable test can quickly become vulnerable to abuse by judges, spurred on by extraneous political, or even non-​political, concerns. Worse still, equality, now owing its existence merely to judicial whim, can be portrayed by its enemies as nothing more than a deceptive slogan,31 fit for discontinuation, if not abolition.32 28

29

30 31 32

Fallon and Martin (n 15). The authors describe the Court’s jurisprudence as ambiguous (165, 170 and 173), contradictory (171), fluctuating (165) and causing perplexity (172). Elsewhere, they describe how the role of different terms and provisions has become uncertain (165 and 172) or problematic (170), or their meaning confusing (169, 170 and 171) or unclear (172), how the Court will from time to time change its approach without explanation (166) or justification (173), and how the Court struggles to maintain coherence (173). All translations author’s own. Caracciolo di Torella and Masselot talk of the ‘weakness of Aristotle’s concept’ when used in the eu legal context, particularly as regards pregnancy and maternity:  E Caracciolo di Torella and A Masselot, ‘Pregnancy, maternity and the organisation of family life: an attempt to classify the case law of the Court of Justice’ (2001) 26(3) EL Rev 239, 241. More, “Equal Treatment” of the Sexes in European Community Law: What Does “Equal” Mean?’ (1993) 1 Feminist Legal Studies 45, 51. This is how Lenin saw it: Vladimir Il’ich Lenin, The deception of the people by the slogans of equality and freedom (Lawrence and Wishart 1940). Caracciolo di Torella, who asserts that the Aristotelian formula is still the Court’s mainstay in 2012, warns that the test ‘ignores the fact that the very essence of human beings resides in their diversity and thus risks reducing the concept of equality to mere rhetoric’:  E Caracciolo di Torella, ‘Gender equality after Test Achats’ (2012) 13 era Forum 59, 67.

Introduction

9

A unified, easy-​to-​use theory of equality for eu law, while desirable, is not plausible. There are too many justiciable fields of endeavour inhabited by today’s Union, each with its own particular nuance, inevitably affecting what goes on, including how equality is viewed, within the field.33 Besides, the requirement to treat likes in like fashion, and unlikes in unlike fashion, is itself an across-​the-​ board rule, its very universality stretching it until all meaning is lost, with the results described above. What is needed is a system which is capable of taking into account each field’s unique qualities. It is submitted that Michael Walzer’s ‘complex equality’, or at least a variation thereon, would fit the bill. 1.3

The Project in Outline

The aim of the project is to investigate, via a critical case-​study subjected to counterfactual analysis, whether Walzer’s theory could act as a credible complement to Aristotle for the Court of Justice, when dealing with cases on equality. It is a library-​based book, very much doctrinal in nature (‘black-​letter law’). It is a contribution to the existing literature on equality in the cjeu’s case-​law, with the difference that it makes a tentative suggestion as to how improvements might be made in the future. After a brief consideration of the theoretical underpinnings of equality, Walzer’s theory is set out in detail in Chapter 2. The next four chapters constitute a case-​study, with the Court’s jurisprudence in different cases examined first from the Aristotelian, and then –​counterfactually –​from the Walzerian point-​of-​view. Chapter  7 provides an opportunity to reflect on some of the common themes emerging from the case-​law. In Chapter 8, some problems to do with the use of a Walzerian-​style ‘Forum’ are considered, and an alternative theory –​mediated complexity –​is put forward. In Chapter 9, a final evaluation is undertaken as to whether complex equality, or mediated complexity, might serve as a possible complement to Aristotle. The purpose of the case-​study is to consider the Court’s decisions both as they are, and as they might be under a Walzerian regime. In the cases of gender, the ‘Article 19’ grounds34 and nationality, a number of cases on a given 33

34

As Bernard puts it, ‘Even when the non-​discrimination principle apparently focuses on the same “evil,” as in cases on discrimination on the grounds of nationality, it may in fact cover very different purposes, each with their own internal logic’: N Bernard, ‘What are the purposes of ec discrimination law?’ in J Dine and B Watt (eds), Discrimination Law: Concepts, Limitations and Justifications (Longman 1996) 98. So named after Article 19 tfeu –​prior to the entry into force of the Treaty of Lisbon on 1 December 2009, this was Article 13 ec.

10 Chapter 1 topic are summarized before Walzer’s theory is applied to them in a separate section. It is felt that, in these areas where the Aristotelian analysis itself raises a large number of issues and polemics, often spanning an entire sequence of decisions, it is better to separate the two analyses. However, in the case of the the semi-​suspect and non-​suspect grounds, an attempt has been made to integrate the two types of analysis to some degree, as it is felt that this can be done without undue confusion, allowing the argument to flow more smoothly. Even in that chapter, though, the basic order –​Aristotle first, Walzer second –​is for the most part still observed. Only in Chapter 9 is a case analysis undertaken wherein the two theories are considered, as it were, simultaneously. A few words are perhaps required about the parameters of the project. Needless to say, of the thousands of decisions handed down over five decades by the European court, only a representative sample can be used, in any one section, to illustrate the points being made; where there is a choice, however, an attempt is made to make use of the Court’s more recent judgments, in order to keep the arguments up-​to-​date and also to avoid unnecessary repetition of the choices made by earlier writers. Due to pressure of space, certain interesting aspects of eu law in which the principle of equality also plays a role have regrettably had to be omitted. Consequently, the book contains no discussion of Article 110 tfeu (formerly Article 90 ec, on internal taxation imposed on foreign and domestic products), or of price discrimination within the context of Article 102 tfeu (formerly Article 82 ec). However, other forms of discriminatory taxation are looked at in Chapter  5 (on nationality discrimination), and competition law, albeit a different aspect, is represented in Chapter 6 (on semi-​suspect and non-​suspect grounds). It must be emphasised that, in this book, Walzer’s theory is being mooted solely as a possible new arrow for the judicial quiver. An entirely new, Walzerian society is not being called for. Walzer’s own project may have been to ‘describe a society,’35 but that does not prevent his theory being applied in more specific, discreet contexts.36 Another important point that must be made is that, as Walzer’s theory turns on the shared meanings of goods arrived at by the relevant distributive communities, none of the meanings which are suggested in the course of the book are intended to be final, and by extension none of the outcomes which are postulated should be read as being the outcomes which would definitely be arrived at if a Walzerian system were put in place. The suggested meanings,

35 36

M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) xiv. See the long list of such applications at the beginning of Chapter 2.

Introduction

11

and postulated outcomes, are discursive only; other meanings and outcomes are entirely possible. The final decision would be that of the distributive community.37 Throughout the analysis, and almost at the risk of monotony, a similar caveat to this one is given before a meaning is proposed; sometimes the expression ‘Standard Contingent Reply’ is used as shorthand for the fact that the meaning cannot be known until the distributive community makes it known.38 Where assumptions are made about what the meaning would be –​and they are made fully consciously –​this is simply to drive the argument forward. Alternative meanings and outcomes –​where considered interesting –​ are given, but it must be appreciated that there could never be room to give them all. Some goods have an almost infinite number of possible meanings; to consider every one, and the outcomes that each would lead to, would be an impossible task, and would make for extremely tedious reading. However, it must be stressed that at no point is the intention to ‘speak for’ other persons, groups or communities. Finally, a small mention should be made of the problem of indirect discrimination. Whether indirect discrimination is covered by the Aristotelian test at all is an unresolved issue, but, for the purposes of this book, it is. To put it another way, the second limb of the test (unlikes treated like) is taken to be a formulation, albeit a very rough one, of the idea of indirect discrimination, whereby identical treatment is applied to two or more parties who are, in fact, differently situated, with the result that the identical treatment in fact affects them differently. However, whether Aristotle’s test is regarded as a failure with respect to indirect discrimination for the ‘traditional’ reasons (difficulties in identifying the comparators, problems of establishing sameness, and so on), or whether it is regarded as a failure with respect to indirect discrimination for not covering the concept at all, is really only a detail. The main issue, for the purposes of this book, is whether Walzer’s method can overcome the failure, producing better and more consistent results for those to whom the discrimination relates. Of the great physicist George Gamow it was said, ‘Even when he’s wrong, he’s interesting.’39 The reader of this book is urged to adopt a similar attitude towards Walzer. What is offered here is food for thought. If the theory is flawed, 37 38 39

At least under the orthodox theory; an alternative theory, ‘mediated complexity,’ wherein the final decision is left to the Judge, is proposed in Chapter 8. See below, Chapter 3, section 3.2. Gamow’s important role in particle physics is described in Atom, a bbc series produced by Paul Sen and presented by Jim Al-​Khalili, 2007, from which this assessment of Gamow –​ by his contemporaries –​is drawn.

12 Chapter 1 or even out-​and-​out ‘wrong’, either in and of itself or as applied to the cjeu’s equality case-​law, then it is at least interesting. If it is felt to be unhelpful in one area –​gender discrimination, for ­example –​then it may still be of interest in another. If it is thought to be inoperable in practice, either in its pure form or in the less cumbersome alternative version –​mediated complexity –​discussed in Chapter 8, then it may at least inspire, or embolden. Furthermore, it is hoped that its presentation here might kickstart a much-​needed discussion, which in turn might pave the way towards a future theory which does ‘work’. Above all, what is presented here is not intended to be a replacement for the Aristotelian test, but at most a complement, or perhaps less even than that: a mere tool, of the many, in the cjeu judges’ tool-​box. Ideally, it should not simply be confirmed as ‘right’ or condemned as ‘wrong’; this belittles the exercise. A complement for a 2,300-​year-​old test is unlikely to be found overnight. 1.4

The Search to Find a Complement

1.4.1 The Difficulty In searching for a potential complement to Aristotle for the Court of Justice, one is confronted by a dizzying array of equality theories; rather like Pablo Neruda’s critics, theorists have effectively stabbed equality with nibs and drowned it in ink.40 One theorist may be ruled out of the search at a fairly early stage, though, and that is Karl Marx. Article 120 TFEU (formerly Article 98 EC) declares that ‘[t]‌he Member States and the Union shall act in accordance with the principle of an open market economy’ (emphasis added). The phrase is also used at Articles 119(1) tfeu and 119(2) tfeu.41 The open market economy is naturally at odds with the marketless economy envisaged by Marx.42 Thus, prima facie, for the Court’s judges to adopt a Marxist approach to equality would not be in keeping with the European project overall. Such an approach might not even be possible, as cases usually only involve one or two individuals, and it would be hard to apply Marxist principles on such a small scale. Nevertheless, if it were possible, commentators would no doubt argue that this was ‘the thin end of the wedge,’ and would wonder why the cjeu, like a turkey voting for

40 41 42

See Pablo Neruda, ‘Oda a la crítica’. Formerly Articles 4(1) ec and 4(2) ec. This has led one commentator to talk of ‘the hard-​wiring of neo-​liberal policies into the eu’: at Callinicos, Equality (Polity 2000) 107.

Introduction

13

Christmas, had decided to advocate an adjudicatory method which would in the long run lead to the unravelling of the entire Union. Meanwhile, at the other end of the political spectrum, the outlook is no less severe. The famous ‘libertarian’ vision of Robert Nozick holds freedom to be considerably more important than equality. It is hard to believe that there would even be nation states (in the traditional sense) in Nozick’s world, still less a multi-​State entity like the eu. He envisions a ‘framework’ of entirely voluntary ‘communities,’ which will ‘wax and wane:’43 ‘[S]‌ince any … community may be established within the framework, it is compatible with all particular utopian visions, while guaranteeing none.’44 Community members are free to enter and leave communities as they wish, or indeed start their own. Such central authority as there is would be a protective association, and nothing more. No ‘paternalistic restrictions’45 may be imposed, and enforced redistribution between communities is also forbidden: ‘none have the right to impose their vision of unity upon the rest.’46 Under these circumstances, it again appears that the cjeu judges would be to say the least shooting the European project in the foot if they endorsed a Nozickian modus operandi. What does that leave? The situation is reminiscent of that in economics, where some favour one hundred percent regulation, à la Marx, while others favour hardly any regulation at all, à la Friedman. Reality forces most States into some sort of compromise-​position between the two. Competition law, of which the eu has perhaps fittingly become something of a standard-​bearer, is an example of how this works on a day-​to-​day basis: there is little or nothing to stop innovators setting up companies and prospering, but as they grow bigger and bigger, increased regulation is brought to bear on them in order to give the other runners in the race a ‘fair chance.’47 Is there a similar mid-​way point for equality? One could argue that a very small example might be inheritance tax: there is nothing to stop individuals prospering and thus creating disparities of wealth within their communities but, on their deaths, the State acts to redress the balance slightly by taking a share of the estate for itself. Given the wide disparities of wealth in evidence in today’s society, it could be argued that this measure hardly makes any difference at all. More extreme is the biblical Sabbatical, mentioned favourably 43 44 45 46 47

R Nozick, Anarchy, State, and Utopia (Basic Books 1974) 312. ibid 320 (emphasis added). ibid 324. ibid 325. A chance which Friedman and the Chicago School hold is undeserved, if not positively unfair on the market-​leader!

14 Chapter 1 by Walzer, which required land to be returned to its original owner after fifty years.48 This was undoubtedly a more drastic plan, effectively sending everyone back to square one after a half-​century, but it has also been described as a ‘utopian plan that never was put into practice.’49 However, to find a half-​way house should involve more than just locating the middle of the continuum-​of-​regulation. What Marx and Nozick highlight is the ongoing battle between equality and freedom, with Marx’s socialists being one hundred percent equal but not particularly free, and Nozick’s libertarians being (almost) one hundred percent free, but not in the least bit equal. 1.4.2 Equality versus Freedom In Aldous Huxley’s Brave New World, the point was made that excessive standardization actually leaves some people unhappy. Robbed of their ability to live an individualistic way of life (however that may have expressed itself), these people are in a very real sense less free. Thus, opponents of egalitarianism are sometimes called libertarians, or hyperindividualists. The elimination of individualism poses a very difficult, and dangerous, problem for equality theorists. As Vlastos puts it, If A is valued for some meritorious quality, m, his individuality does not enter into the valuation. As an individual he is then dispensable; his place could be taken without loss of value by any other individual with as good an m-​rating.50 This reasoning (which, it must be pointed out, is by no means Vlastos’ conclusion, or even close to it) could have come straight out of Huxley’s ‘World.’ Indeed, the Director of Hatcheries and Conditioning at one point declares: Murder kills only the individual –​and, after all, what is an individual? […] We can make a new one with the greatest ease –​as many as we like.51 Narveson takes equality’s tendency to discount the individual’s own attributes further. Critiquing socialism, he rejects the question, ‘Are freedom and 48 49 50 51

Leviticus, Chapter 25, Verses 8–​10. Lieber, Etz Hayim:  A Torah Commentary (Jewish Publication Society of America 2001) 738. G Vlastos, ‘Justice and Equality’ in Richard Brandt (ed), Social Justice (Prentice Hall 1962) 31, 44 (Vlastos’ emphasis). Aldous Huxley, Brave New World (Vintage 2004) 128.

Introduction

15

equality compatible?’ in favour of ‘whether freedom and compelled equality are compatible.’52 He answers this in the negative, and then comes down heavily in favour of freedom. One of his most powerful arguments against any form of redistribution is that goods are useless without the technical wherewithal to exploit them, and that, given that in any supposedly equal distribution there will likely be an unequal preponderance of the relevant technical ability amongst the recipients, no distribution pretending to equality can ever represent equality in fact: It is impossible, in principle and not just as a matter of operational procedure, to say what would constitute an ‘equal’ division of ‘the world’s resources,’ absent any notion of what they would be used for and how … The claim that Jones and Smith have now been made ‘equal’ by ‘giving’ each of them $6,472.60 comes to shipwreck when it is noted that Jones can realize his modest purposes, in his circumstances, quite well with that amount, whereas Smith, whose interests and skills are different, will not get to first base with it.53 Other theorists recognize the conflict between equality and freedom, but appear to see freedom as the source of the trouble, not equality: Equality lays down how we are to treat people: but Liberty entitles us to act as we choose, not as some rule lays down. If I have any Liberty then there are some decisions I am allowed to make on my own; I am free in some cases to act arbitrarily. And if that is so, I may in such cases arbitrarily choose one person rather than another, without there being any ground to justify discrimination. […] This is what it is to be free. Freedom is inherently unfair.54 The tussle between the two concepts is probably as old as philosophy itself.55 Theorists such as Matson note how the traditional settlement between the two 52 53 54 55

J Narveson, ‘Libertarianism v Marxism:  Reflections on GA Cohen’s Self-​Ownership, Freedom and Equality’ (1998) 2(1) Journal of Ethics 1, 3 (Narveson’s emphasis). ibid 15–​16 (Narveson’s emphasis). JR Lucas, ‘Against Equality’ (1965) xl Philosophy 296, 307. De Tocqueville believed that equality and freedom were ‘two different things’: Alexis de Tocqueville, Democracy in America (GE Bevan tr, Penguin Classics 2003) 584. Even further back, Cicero seems also to have been of this view, noting that ‘real freedom can only exist when it is equal for all’: Torkel Opsahl, ‘Equality and non-​discrimination’ in T Opsahl,

16 Chapter 1 ideas usually takes the form of an ‘agreement’56 (some variant on the Rousseauian social contract), which in most societies replaces the anarchic system, so famously summed up by Hobbes, which is very much based on the notion that ‘might is right’ (maximum freedom, minimum equality). The contract57 demands a surrender of freedom (or ‘sovereignty’),58 in return, principally, for a guarantee of security. However, a side-​effect of the newly-​secure society will of course be that its citzens can now cohabit peacefully, the mighty alongside the weak, the clever alongside the foolish, and so on. The contract thus balances the loss of freedom with a gain in equality, at least in the short term. But ‘might is right’ creeps back into play in new, disguised ways, and this balance is quickly imperilled once again. Contractualism on its own cannot therefore be the answer. Theorists like Dworkin,59 Arneson60 and Williams61 have investigated solutions involving complicated redistributions (even of abstracts like intelligence) and rectifications,62 Williams at least spotting the ‘inhuman’63 aspect of mixing and matching human characteristics and body parts like a game of dress-​up dolls; as in Brave New World, extreme equality threatens the annihilation of individuality.64 Real-​life examples also confirm this. The socialism

56 57 58 59 60 61 62

63 64

Law and equality: Selected articles on human rights (Ad Notam Gyldendal 1996) 165, 169, citing Nowak. Wirszubski gives it as ‘all can be free only in a State in which power is distributed between all’ and attributes it to Cicero’s de Republica: C Wirszubski, Libertas as a Political Idea at Rome During the Late Republic and Early Principate (cup 1968) 82. But cf. Laski, who, while acknowledging the view that the two concepts were ‘antithetic things,’ held that this was a ‘drastic conclusion:’ Harold J Laski, A Grammar of Politics (George Allen & Unwin 1925) 152. Wallace Matson, ‘Justice: A Funeral Oration’ (1983) I Social Philosophy and Policy 94, 99. Or its more recent variants, such as Rawls’ ‘principles’. Locke refers to resigning one’s natural power ‘into the hands of the community:’ John Locke, The Second Treatise of Civil Government (Basil Blackwell1946) 43. See, for example, R Dworkin, Sovereign Virtue: The Theory and Practice of Equality (Harvard University Press 2001), which is a compendium of all his writings on equality. See, for example, Richard J Arneson, ‘Equality and Equal Opportunity for Welfare’ (1989) 56 Philosophical Studies 77. See, for example, BAO Williams, ‘The Idea of Equality’ in P Laslett and WG Runciman (eds), Philosophy, Politics and Society, Series ii (Basil Blackwell 1972). Some of these tales of supposed rectification seem a bit far-​fetched and even offensive, particularly where disability is concerned. But for a possible real-​life example, referring to the Olympic runner Oscar Pistorius, see J Clayton, ‘Running without legs? That’s not fair on all the other athletes’ The Times (London 15 January 2008) 30. Williams (n 61) 130. Some would see it more as a promise than a threat, and would view the elimination of relationality and plurality which equality may bring about not as an unwanted side-​effect,

Introduction

17

of the Soviet Union, while (supposedly) providing equality of resources, led almost immediately to exceptional inequality of power, with a concomitant loss of freedom for those left powerless.65 As Frankfurt has put it, although without expressing a view on the ‘merit of this argument,’ ‘an egalitarian distribution of money can be achieved and maintained only at the cost of repression’.66 Space forbids a consideration of some of the more interesting modern theories of equality available, including Derrida’s suggested replacement of the concept of equal treatment with that of ‘friendship,’67 and Rancière’s theory of equal intelligence.68 The latter seems to rule itself out in the same way that Marx did earlier; Rancière’s equality calls for the ‘undo[ing of] the supposed naturalness of orders.’69 The courtroom may not be the best place for this. However, May develops Rancière’s theory into a more general one of ‘equal worthiness,’70 and this is interesting as it reconceptualises equality as a non-​comparative appeal to a single, independent standard. This reconceptualization, which is already familiar to eu free movement law as the entrave-​or obstacle-​based approach to discrimination, is increasingly being written about in the academic literature on equality, although opinion varies

65

66 67 68

69 70

but as the whole point. Dean, for example, inspired by ‘political crowds’ (those ‘authorised by neither capital nor the state’), such as the ones in the us (Zuccotti Park/​Occupy Wall Street), Greece or Ukraine, has reinvestigated the idea of the crowd, even its less appealing manifestations (English football supporters in the 1980s), as a ‘ “state of absolute equality” supplant[ing] individuating distinctions’ and ‘a blessed moment of joyous belonging:’ J Dean, Crowds and Party (Verso 2016)  ch 3, but in particular 121–​3. The quote within a quote is from E Canetti, Crowds and Power (C Stewart tr, Victor Gollancz 1962) 29. See, for more on freedom versus equality, DT Rodgers, ‘Contesting equality’ (2014) 33(4) Raritan 19, 29. It was ‘a staple of Cold War rhetoric’ according to Allen: Danielle Allen, ‘Equality and American democracy: why politics trumps economics’ (2016) 95(1) Foreign affairs 23, 24. Harry Frankfurt, ‘Equality as a Moral Ideal’ (1987) 98 Ethics 21, 22. J Derrida, The Politics of Friendship (G Collins tr, Verso 1997); J Derrida, ‘Force of Law: the “Mystical Foundations of Authority” ’ (1990) 11 Cardozo Law School 919. This may be found, inter alia, in J Rancière, The ignorant schoolmaster:  five lessons in intellectual emancipation (K Ross tr, Stanford University Press 1991). It is accepted that Walzer’s theory, however creative, is still liberal egalitarianism of the Rawlsian sort which Rancière finds too passive, even if Ricoeur described Walzer’s work as ‘one of the most brilliant rebuttals’ of Rawls’ formalism: P Ricoeur, The Just (D Pellauer tr, University of Chicago Press 2003) 76. J Rancière, On the Shores of Politics (L Heron tr, Verso 1995) 33. T May, Nonviolent resistance: a philosophical introduction (Polity 2015) 144.

18 Chapter 1 as to what the standard is or should be. Respect for/​violation of human dignity has been mooted, although not without some reservations, by Moon and Allen.71 McCrudden has suggested that the cjeu may be moving towards ‘the avoidance of humiliation and stigma’ as its standard, or driver, in anti-​ discrimination cases.72 In a similar vein, Hellman writes about acts of subjugation or demeaning as the ones that discrimination law ought to target.73 As will be seen in Chapter 3, below, Ellis has criticised the non-​comparative approach as the simple absence of ‘nasty treatment.’74 Rochel has written about ‘equality qua minimal moral worth,’ drawing on both Hellman and on Moreau’s ‘freedom-​based account,’ wherein ‘[deliberative] freedoms are not the result of an interpersonal comparison in terms of opportunities or rights, but reflect what is due to the person in terms of recognising his entitlements’.75 1.4.3 Reasons for Choosing Walzer From the last two sections, it can be seen that the search to find a complement for Aristotle is not an easy one. The first theorists looked at either offered too much freedom and not enough equality, or too much equality and not enough freedom. Some required too much action on the part of the judge, who can only ever react passively to things which have already happened. In the last paragraph of the previous section, consideration was given to theories which avoid comparison via the setting of a single threshold criterion by which to judge the alleged discriminator’s behaviour. This reconceptualisation of equality, not universally endorsed even by those describing it, was interesting from the point of view of this book, given that the comparison stage can be one of the Aristotelian test’s biggest weaknesses. However, with its in-​out methodology,

71 72

73 74 75

Moon and Allen (n 20). C McCrudden, ‘The new architecture of eu Equality Law after chez: Did the Court of Justice reconceptualise direct and indirect discrimination?’ [2016] 1 European Equality Law Review 1, 10. D Hellman, When is discrimination wrong? (Harvard University Press 2011). Evelyn Ellis, ‘The definition of discrimination in European Community sex equality law’ (1994) 19(6) el Rev 563, 571–​2. J Rochel, ‘Welcome to the eu: Immigration and Non-​discrimination’ in S Besson and AR Ziegler, Egalité et non-​discrimination en droit international et européen: Equality and non-​ discrimination in international and European law (Schulthess 2014) 79, 90–​91; the reference is to S Moreau, ‘What is discrimination?’ (2010) 38(2) Philosophy & Public Affairs 143 and S Moreau, ‘In defense of a liberty-​based account of discrimination’ in D Hellman and S Moreau (eds), Philosophical foundations of discrimination law (oup 2013).

Introduction

19

one wonders if it can provide the adjudicatory subtlety sought here. Better, it is submitted, would be a theory which was more particularist, allowing for the ascertaining of the appropriate standard on a case by case basis, rather than forcing the judge to accept a single, monolithic standard at the outset.76 Arguably such a variable approach also tackles the equality/​freedom problem as it judges each situation afresh, and equalizes only where it is appropriate to do so,77 letting freedom prevail otherwise.78 Walzer’s conception of a world of spheres, rigorously separated one from another, achieves this equilibrium. As he himself has written: The art of separation doesn’t make only for liberty but also for equality … Religious liberty annuls the coercive power of the political and ecclesiastical officials. Hence it creates, in principle, the priesthood of all believers … Academic freedom provides theoretical, if not always practical, protection for autonomous universities, within which it is difficult to sustain

76

77 78

Special mention should be made here of Manuela Brillat’s superb 2014 thesis looking at the principle of non-​discrimination as interpreted at both the cjeu and the European Court of Human Rights (‘ECtHR’): M Brillat, Le principe de non-​discrimination à l’épreuve des rapports entre les droits européens (Fondation Varenne 2015). Having found fault with the existing test for discrimination at the cjeu, Brillat’s project too has been to redetermine this test, by better defining the concept of discrimination in the first place and by making sure that the test better corresponds to the definition, leading to better coordination between the two courts. Her new version of the test among other things involves merging the comparability stage with the objective justification stage, and sharing the test between the two parties instead of the current situation, wherein the applicant has responsibility for the comparability test and the defendant has responsibility for the justification. New procedural rules will impose much greater disclosure obligations on both sides, but particularly with regard to disclosure from the defendant to the applicant, and with adverse inferences drawn by the Court from any silences, as happens at the ECtHR. Naturally, the bringing of the objective justification stage of the procedure, which can be the stage at which the more context-​specific arguments are made, into the often abstract comparability stage, will have the effect of integrating the particularity, as she puts it (ibid 268), and of recognising the ‘lien indissoluble’ between juridical problems and social/​ political problems (ibid 415). It is in this regard that her proposal is most Walzerian, and it is clear that she favours, as does the present book, what Sussmann has called, in relation to Walzer, the ‘prioritization of practice:’ N Sussmann, ‘Introduction: The substantive unity of Michael Walzer’s pluralism’ in Y Benbaji and N Sussmann (eds), Reading Walzer (Routledge 2014) 1, 9. Inter-​sphere situation. Intra-​sphere situation.

20 Chapter 1 the privileged position of rich or aristocratic children. The free market is open to all comers, without regard to race or creed […]; and though it yields unequal results, these results never simply reproduce the hierarchy of blood or caste or, for that matter, of ‘merit’ … Under the aegis of the art of separation, liberty and equality go together.79 The very fact of the spheres ensures the freedom, but it is their separation which ensures the equality. However, as Walzer hints in the quotation above when talking about the market, this is not simple equality; within a single sphere, disparities may be perfectly permissible. Whether equality gives way to freedom, or whether freedom gives way to equality, must be determined from the specific context: [E]‌ach freedom entails a specific form of equality or, better, the absence of a specific inequality –​of conquerors and subjects, believers and infidels, trustees and teachers, owners and workers  –​and the sum of the absences makes an egalitarian society.80 But, crucially, Walzer’s theory provides the means for making this important determination, as will be explained further in the next chapter. This means that society need not abandon redistribution altogether, the conclusion which Narveson reaches,81 but merely redistribute in a more context-​specific manner, taking the individual’s own –​relevant –​attributes into account. Walzer’s spheres, then, rather like those of Shakespeare’s time,82 allow for collective harmony while still acknowledging individual variation. Complex

79 80 81

82

M Walzer, ‘Liberalism and the Art of Separation’ (1984) 12(3) Political Theory 315, 320–​1. ibid 326. For example, ‘The idea that “wealth” is just there, lying around, needing only to be “distributed,” so that the fundamental subject of political economics is how to “distribute” that steadily diminishing pile of stuff, is totally unreal … [N]‌o rationale can be produced for whatever distributions a theorist may wish to specify …’ Narveson (n 52) 21. It is submitted in this book that Walzer does produce such a rationale, or at least a method for arriving at it. See, for example, Twelfth Night, IIIi:  ‘I had rather hear you to solicit that/​Than music from the spheres’. It was held that, in their rotations, the crystalline spheres containing the planets and the fixed stars created ‘a ravishing harmony inaudible to mortal ears’: Shakespeare, Twelfth Night (The New Cambridge Shakespeare, cup 1985) 99.

Introduction

21

equality is thus the true half-​way house between the two extremes,83 offering the prospect of a genuine complement for the cjeu without requiring the dismantling of its parent organization, and for these reasons has been chosen over the other available theories as the appropriate one to apply to the problem under consideration. 83

In a way, Walzerian equality is not just the half-​way point on the continuum, it is the continuum. Thus it is not bourgeois equality, or Marxist equality, but in theory it could be either. In his critique of bourgeois equality, which he describes as ‘measurement in terms of a common standard,’ Marx expresses his dissatisfaction that that standard is workplace productivity and ‘nothing else about [the workers] is taken into account, all else being disregarded’. But it would be open to the members of a Marxist Forum to (for example) include physical fitness, mental agility, individual talent, marital status and/​or number of children in their shared understanding of workplace productivity, as it would be equally open to the members of a bourgeois Forum to exclude them. See K Marx, Marx: Later political writings (T Carver ed, cup 1996) 214; these observations are to be found in the ‘Critique of the Gotha Programme.’

Chapter 2

Michael Walzer and Complex Equality 2.1 Introduction Michael Walzer launched his theory of complex equality in 1983 in a book entitled Spheres of Justice.1 The theory, as will be seen, is innovative in that it focuses on distribution, where distribution is defined very broadly to include not just tangible goods, but also abstract goods such as rights,2 and even characteristics.3 Walzer’s writing style in Spheres of Justice is also innovative; instead of arguing from first principles, he chooses to take a more artistic approach. He produces what some have called a ‘portrait’4 (or even an entire ‘portrait gallery’),5 others an ‘impressionist painting.’6 In this book, Walzer’s lead will be followed. The emphasis will be on working through ‘accounts of distributions’ (of which the court reports provide an almost limitless supply), letting these ‘accounts’ (or even ‘stories,’ as Walzer puts it) ‘suggest’ any necessary principles for themselves –​‘stand[ing] … in the city,’ as it were, rather than ‘climb[ing] the mountain,’ interpreting rather than describing.7 Over the years, Walzer’s theory has been applied to a dazzling variety of situations around the world. These include higher education in Malaysia,8 affordable housing in the us,9 and children’s rights in El 1 M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983). 2 McCrudden, in his recent reanalysis of equality, divides the concept into four categories. In the second category, he too envisages rights as distribuends, calling them ‘prized public goods:’ C McCrudden, ‘The new concept of equality’ (Paper prepared for the Academy of European Law conference, ‘Fight Against Discrimination:  The Race and Framework Employment Directives’, 2–​3 June 2003)  accessed 20 May 2018. 3 See section 3.3. below. 4 NL Rosenblum, ‘Moral Membership in a Postliberal State’ (1984) 36(4) World Politics 581. 5 ibid. 6 J Carens, ‘Complex Justice, Cultural Difference, and Political Community’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) 47. 7 All quotations from Walzer, Spheres of Justice (n 1) xiv. 8 H Stokke, ‘Reasonable Discrimination? Affirming Access to Higher Education in Malaysia’ [1999–​2000] Hum Rts Dev YB 189. 9 KD Adams, ‘Can Promise Enforcement Save Affordable Housing in the United States?’ (2004) 41 San Diego L Rev 643.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/9789004354265_003

Michael Walzer and Complex Equality

23

Salvador.10 Complex equality has been used in everything from the ­practice of paying research subjects for participating in clinical trials,11 to the communication strategies of police officers,12 and from the protection of personal data,13 to the creation of hierarchies of creditors following insolvency.14 Sectors in which the theory has been applied most recently include those of transportation,15 taxation16 and planning, specifically with regard to poverty alleviation in India.17 Could the equality strategy of the Court of Justice be added to this already impressive list? In this chapter, it is intended to sketch the basics of the theory (with due attention paid to the many who have written about it since 1983) and then, towards the end, to begin to consider whether it could be applied to the case-​law of the cjeu. 2.2

Complex Equality: A Thumb-​Nail Sketch

Michael Walzer holds that there is no one single distributive sphere; rather, there are different distributive spheres for different goods. The first major tenet of the theory of complex equality is that these spheres must be kept separate.

10 11 12 13 14 15 16 17

K Read, ‘When Is a Kid a Kid? Negotiating Children’s Rights in El Salvador’s Civil War’ (2002) 41(4) History of Religions 391. JA Anderson and C Weijer, ‘The Research Subject as Wage Earner’ (2002) 23 Theoretical Medicine 359. Phillip Chong Ho Shon, “Now You Got a Dead Baby on Your Hands”: Discursive Tyranny in “Cop Talk” (1998) xi (33) International Journal for the Semiotics of Law 275. J Van Den Hoven and PE Vermaas, ‘Nano-​Technology and Privacy:  On Continuous Surveillance Outside the Panopticon’ (2007) 32(3) J Med Philos 283. C Villiers, ‘Employees as creditors: a challenge for justice in insolvency law’ (1999) 20(7) Company Lawyer 222. K Martens, ‘Justice in transport as justice in accessibility: Applying Walzer’s “Spheres of Justice” to the transport sector’ (2012) 39(6) Transportation 1035. L Marriott, ‘Justice and the Justice System:  A Comparison of Tax Evasion and Welfare Fraud in Australia and New Zealand’ (2013) 22(2) Griffith Law Review 403. A Chettiparamb, ‘Articulating “public interest” through complexity theory’ (2016) 34(7) Environment and Planning C:  Government and Policy 1284. For a rare example of an application in the legal field, albeit on the procedural side, see the consideration of the (potential) selling of judicial authority in AJ Sebok, ‘Should the law preserve party control? Litigation investment, insurance law, and double standards’ (2015) 56(3) William and Mary law review 833.

24 Chapter 2 Within any one of these spheres, the good in question is distributed according to its shared social meaning, that is, according to an understanding of what the good means to the society in which the distribution is taking place, such understanding being shared by the members of this society (distributees and distributors alike). A  second tenet of the theory of complex equality is that these shared social meanings, or shared understandings, must be respected. Violation of these shared social meanings would occur if a person acquired a good, not in conformity with its shared social meaning, but instead by virtue of the prior possession of another good, belonging to a different sphere. This latter good is referred to in Walzer’s lexicon as a dominant good. If, having previously acquired the dominant good in its proper sphere, a person were to use this good to tyrannize (again, Walzer’s word) a neighbouring sphere, that is, to gain privileged or even exclusive access to the good or goods in this neighbouring sphere, then complex equality would be vitiated. Put simply, such a person may not use a dominant good to ‘jump the queue’ in a subsequent distribution of an entirely different good. In Walzerian terms, this distribution would be flawed. Only where the goods in each sphere are distributed according to their shared social meanings, and only according to their shared social meanings, is tyranny (or dominance) averted, allowing complex equality to prevail. The prohibition on tyranny, or dominance, is the third tenet of the theory of complex equality. The boundaries of the spheres must therefore be patrolled to avoid what Walzer calls boundary crossings, or blocked exchanges, wherein a successful recipient of a good in Sphere 1 is able to convert his success into automatic (and unwarranted) success in Sphere 2, thus by-​passing the distributive process intrinsic to Sphere 2. All such convertibility must be eliminated. A given sphere must be autonomous, that is, governed, internally, by the distributive rule germane thereto, and no other. The maintenance of the autonomy of spheres is vital to complex equality. It could be said that this is a fourth tenet of the theory of complex equality, but it is, to all intents and purposes, a restatement of the first tenet, described above, namely that spheres must be kept separate. As long as the boundaries of a sphere are secured against any contamination from neighbouring spheres, any distribution taking place within the sphere, grounded solely on the shared social meaning of the good in question, should be just.18 Of course there will be winners and losers. Not everyone will receive the same amount of the good, a result which Walzer calls simple equality, and 18

In this paragraph, the word ‘just’ is used in the basic sense that Walzer uses this word, that is, ‘not corrupt’.

Michael Walzer and Complex Equality

25

which he does not believe to be feasible. The situation in the sphere after the distribution will be at best what Walzer calls rough equality. However, it will not be the kind of gross inequality which is so characteristic of modern society. So long as the tenets set out above are adhered to, any inequalities which remain within the sphere will at least be just, that is, they will have come about for the right reasons, and not for the wrong ones; Walzer calls them ‘small inequalities.’ Moreover, with dominance having been eradicated, a winner in one sphere will be prevented from going on to claim immediate and unconditional victory in the other spheres. There is no dominant good which can ‘buy’ all of the others. This opens the door for a society of multiple winners, wherein each member can stake a claim for at least some of the successes on offer, in other words, those successes which they have properly deserved, or to which they are properly entitled. Each person will thus have some share of the benefits and burdens, privileges and chagrins, of the society in which they live. That is complex equality. A straightforward real-​world example would be the sphere of education. Education should be distributed by teachers, to students, in school. This is a fairly uncontroversial ‘shared understanding’ of the meaning of education, although of course other understandings are possible. However, the sphere of education must be protected from spillover from neighbouring spheres, for example, the sphere of money (students gaining advantage by virtue of wealth), the sphere of family (students gaining advantage via the influence of relatives), the sphere of politics (curricula being dictated by certain political factions) or the market economy (multinational corporations sponsoring schoolbooks as a means of ‘subliminal advertising’). As long as its borders are patrolled and it is kept autonomous, the situation obtaining in the sphere of education should be one of (rough) equality as between, say, student A and student B (although not, of course, as between the students and their teacher). It should also, momentarily anyway, be one of simple equality, but, almost as soon as the first lesson is completed, A may well have begun to overtake B academically.19 Indeed, A may leave school with top grades in all subjects, while B attains a bare pass. However, as long as A is prevented from monopolizing his success by converting it into success in all the other spheres,20 then B has at least the chance of succeeding elsewhere. The complex egalitarian rejects a 19

20

Walzer vehemently disagrees with Dworkin (and, to a lesser extent, Rawls) that individuals do not deserve their natural talents, and that all talents should therefore be artificially distributed among the entire population. Unless, of course, success in education forms part of the ‘shared understanding’ appropriate to the sphere in question.

26 Chapter 2 single big triumph by one person, in favour of many small triumphs by different people. With shared understandings respected, spheres constrained, and dominance precluded, complex equality will prevail.21 2.3

Negative Dominance

Negative dominance is first introduced by Walzer in his chapter on ‘Hard Work’ in Spheres of Justice. It builds on the traditional notion of dominance, that is, the power of a good to afford its possessor automatic access to other goods in other spheres. However, while possession of a positively dominant good might entitle its possessor to all, or at least some, of the good things in life, possession of a negatively dominant good brings about quite the opposite result. According to Walzer, such a good: commonly carries other negative goods in its train: poverty, insecurity, ill health, physical danger, dishonor, and degradation.22 He continues: poverty [is carried] into the sphere of money, degradation into the sphere of honor, weakness and resignation into the sphere of power.23 A possessor of a negatively dominant good, then, risks becoming ‘subordinate in all [spheres,] […] defeat in one area so often lead[ing] to defeat in another.’24 Rather than trying to be the King of every kingdom, to borrow Walzer’s early metaphor, the possessor of the negatively dominant good is forced to be the pauper of every kingdom; the negatively dominant good is not a good luck charm, but a bad luck charm. As Andre more philosophically puts it, People are complex emotional, social, cognitive wholes, who can find themselves in downward spirals:  one failure breeds others. We get discouraged, we lose our footing, we turn away from our friends.25 21 22 23 24 25

The sphere of education is discussed in more detail in C Sabbagh and N Resh, ‘Unfolding Justice Research in the Realm of Education’ (2016) 29(1) Social Justice Research 1. Walzer, Spheres of Justice (n 1) 165. ibid 183. J Andre, ‘Blocked Exchanges:  A Taxonomy’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (OUP 1995) 195. ibid.

Michael Walzer and Complex Equality

27

Later, Walzer himself was to take up the theme of failure: Failure pursues [excluded men and women] from sphere to sphere in the form of stereotyping, discrimination, and disregard, so that their condition is not in fact the product of a succession of autonomous decisions but of a single systemic decision or of an interconnected set. And for their children, exclusion is an inheritance; the qualities that supposedly produce it are now its products.26 Hervey also writes about exclusion, and, although she is not referring to negative dominance, or to complex equality at all, the ‘pattern of disadvantage’ which she describes is a similar concept: Members of racial, ethnic or religious (and indeed other) minorities are likely to be over-​represented in socially disadvantaged groups. This phenomenon is known as ‘social exclusion:’ a process by which access to income and other social benefits or services, such as health protection, housing and education, is restricted, and a pattern of disadvantage, manifesting itself in several interrelated ways, is perpetuated.27 Possession of a negatively dominant good can cause a boundary breach in exactly the same way as possession of a positively dominant good. The only difference, then, is in the likely reaction of the possessors. The declaration that a distribution is defective and must be rerun, this time with their good left ‘at the door,’ is likely to make the possessor of a negatively dominant good much happier than it might make the possessor of a positively dominant good.

26

27

M Walzer, ‘Response’ in Miller and Walzer (n 24) 291. Compare Walzer’s use of the word ‘stereotyping’ here, with Mark Bell’s reference to breaking down stereotypes quoted in section 4.4.2 below. An assumption made about somebody, or a class of people, can be one kind of negative dominant, following that person, or that class of people, from sphere to sphere, and jeopardizing their chances of success in each. T Hervey, ‘Putting Europe’s House in Order: Racism, Race Discrimination and Xenophobia after the Treaty of Amsterdam’ in D O’Keeffe and P Twomey (eds), Legal issues of the Amsterdam Treaty (Hart Publishing 1999) 333.

28 Chapter 2 2.4

The Concept of ‘Shared Understandings’ Investigated Further

The concept of ‘shared understandings’ is central to the theory of complex equality. All distributions need a fundamentum distributionis, or distributive principle, which makes it clear what each recipient should get, and, perhaps more importantly, why. For the complex egalitarian, this must be the shared social understanding of the meaning of the distributed good. Any distribution grounded on anything other than the shared social understanding of the meaning of the distributed good will be invalid. Thus it is absolutely essential to the theory that this understanding is determined correctly.28 Some critics have focused merely on the logistics of this determination. If there is to be a debate, who should participate? Where should it take place, and when? This type of criticism will be considered presently. Meanwhile, other critics have questioned Walzer’s entire premise that every social good indicates its own distributive principle, and that the shared understanding can be discerned from the good almost as easily as, say, a robber can be identified by his or her finger-​prints. To put it in Walzer’s own words, If we understand what it is, [and] what it means to those for whom it is a good, we understand how, by whom, and for what reasons it ought to be distributed.29 This implies that that there is one understanding to every one good. Naturally, many commentators have protested about this, claiming that there can be two, or even several, understandings to one good.30 Others go further and argue that this is not just another perspectivist point, in other words, that the good means different things to different people, but that even one and the same

28 ‘Correctly’ here referring to the determination process, not the understand ing. In complex equality, there is no single correct understanding; see discussion immediately below. 29 Walzer, Spheres of Justice (n 1) 9. 30 A Gutmann, ‘Justice across the Spheres’ in Miller and Walzer (n 24) (‘the social meanings of some goods are multiple’ on 99); Margo Trappenburg, ‘In Defence of Pure Pluralism:  Two Readings of Walzer’s Spheres of Justice’ (2000) 8(3) Journal of Political Philosophy 343 (‘sometimes different principles are used to distribute one particular social good’ on 345; ‘social goods have multiple or ambivalent meanings’ on 346).

Michael Walzer and Complex Equality

29

person might understand one good in two (or more) ways.31 Further, these understandings need not be mutually exclusive, but may well co-​exist. Still others restate the argument in terms not of understandings but of spheres, that is, that one good may find itself in two distributive spheres, depending on the identity of the potential recipients.32 It is perplexing that these critics indict such a differentiated theory on a charge of being too uniformitarian (‘one understanding to one good’). It is submitted that Walzer’s theory can quite easily accommodate multiple meanings, for multiple recipients. Walzer himself arguably entertained the idea of one good having two understandings in Spheres of Justice, where he described territory as being ‘a social good in a double sense,’33 representing two different things to two different sets of people. And he certainly entertained the idea twelve years later, when he gave the example of soup, which would mean one thing, and therefore be distributed in one way, to one group of people, in the grocery store, and would mean another thing, and therefore be distributed in another way, to another group of people, in a soup kitchen.34 He confirmed that the theory had no trouble, as a consequence, splitting soup into two different distributive spheres. Returning to the question of logistics, just how should the shared social understandings (of the meanings of distributed goods) be determined? What would such a determination look like? It is sometimes hard to picture exactly what Walzer has in mind here  –​some kind of meeting of villagers in a village hall, perhaps, or a gathering of citizens in the Roman Forum. Should a mass show of hands be used for the taking of decisions, as on the Pnyx hill in ancient Athens? Walzer himself makes no pretence of the fact that he is 31

32

33 34

M Rustin, ‘Equality in Post-​Modern Times’ in Miller and Walzer (n 24), giving health-​care and wealth as examples of goods with a double-​meaning; Macdonald gives love as his example: RA Macdonald, ‘Access to Justice and Law Reform’ (1990) 10 Windsor YB Access Just 287, 333. RB Thigpen and LA Downing, ‘Liberal and Communitarian Approaches to Justification’ (1989) 51(4) Review of Politics 533 (again giving medical care as an example). Mayer gives education and membership (inter alia) as his examples: R Mayer, ‘Michael Walzer, Industrial Democracy, and Complex Equality’ (2001) 29(2) Political Theory 237, 245–​246. Walzer, Spheres of Justice (n 1) 44. M Walzer, ‘Response’ in Miller and Walzer (n 24) 282. See also Walzer’s comment, made during an exchange of views which he conducted with Dworkin shortly after the publication of Spheres of Justice, that ‘[i]‌t is entirely possible, on my view, that for some goods we will have complex rather than unitary distributive principles’: M Walzer and R Dworkin, ‘Spheres of Justice: An Exchange’ (1983) 30(12) New York Review of Books 43, 44.

30 Chapter 2 more concerned with the theory than the practice: ‘I won’t try to describe how we might go about creating such a society. The description is hard enough’.35 In what follows, the word ‘debate’ will be used to describe the determination process. However, there is nothing significant about the choice of this particular word, and no special meanings should be read into it. The location of the debate will be given as ‘Forum’, but subject to the same caveat. Perhaps the most important thing to remember is that, since the shared understandings change from society to society,36 it makes sense that the precise style of the debate, as well as the nature of the Forum, will also change from society to society. But as well as questions of how and where, it is also necessary to consider questions of who. Who should be present at the debate? What characteristics do they need? How are they to be treated, both during and after the debate? Questions of when, or at least how often, will also crop up. If the intention is to design a procedure for ascertaining the shared social meaning of a good, and it is accepted, as discussed above, that goods have multiple meanings, then it follows that said shared social meaning will37 only be able to be ascertained via some form of deliberation, wherein the various 35

36

37

Walzer, Spheres of Justice (n 1)  xiv. And see B Barry, ‘Spherical Justice and Global Injustice’ in Miller and Walzer (n 24)  77:  ‘Walzer resolutely refuses to investigate the micro-​processes that go into the formation and sustenance of beliefs’. In the sense in which that word is used in the second paragraph of section 2.2, that is, the society in which the distribution is taking place. One could also call this the distributive community. The future and conditional tenses are used throughout this paragraph. For the purposes of this book, the creation of Walzer’s Forum, that is, a Forum capable of producing the shared social meanings necessary for complex equality, is, with one or two exceptions, something which has not happened yet. However, in reviewing Spheres of Justice, a number of critics seem to have read Walzer as meaning that the way in which a society distributes this or that good today is already, and without more, the correct distributive mechanism for that particular society for that particular good. By designing a theory dependent on local distributive mechanisms, they conclude, Walzer has made himself a hostage to the status quo. See in particular B Barry, ‘Intimations of Justice’ (1984) 84 Columbia Law Review 806 and J Cohen, Review of Spheres of Justice (1986) 83 Journal of Philosophy 457. Dworkin also seems to be of this view, with his oft-​quoted remark that if a society pursues Walzer’s approach based on shared traditions, ‘political theory will be only a mirror, uselessly reflecting a community’s consensus and division back upon itself’: Walzer and Dworkin (n 34) 46. Honneth seems to be of a similar opinion: ‘In [the case of Walzer] … criticism is grasped as nothing more than the reflexive form of a rationality that is itself anchored in the historical process’. See A Honneth, ‘Critical Theory’ in D Moran (ed), The Routledge companion to twentieth century philosophy (Routledge 2008) 784, 784.

Michael Walzer and Complex Equality

31

proposed meanings are put forward, and then weighed up as to their relative merits. The deliberation would end with the meaning considered best being selected. Since it is the social meaning that is required, that is, the ‘understanding of what the good means to the society in which the distribution is taking place,’38 it follows prima facie that all members of the society should be invited to the deliberation, and given a chance to propose meanings, and also prima facie that all members of the society should participate in the final selection. An examination of the many and varied types of democracy, not to mention the many and varied electoral systems, is of course beyond the scope of this book.39 But the society in question would obviously have to settle on a particular type, and on a particular system, in order to proceed. The deliberation has been given different names by different commentators. One of the most common is conflict. Also popular are dispute and argument. These words suggest that, in many people’s view, the process would not be altogether amicable. Settling, as discussed above, on the default word ‘debate’, the next question is:  who should be present at the debate? While all members of the society should, by rights, be invited (see previous paragraph), this may only bring about an impression of inclusion and fairness, not the real things. Okin (who discusses Walzer at some length) argues that women, for example, are ‘deprived of a voice,’40 male domination being so ingrained within modern society.41 Warren is undoubtedly right when he declares that, ‘inequalities are suspect when the voices of those affected are absent’.42 This is even reminiscent of the test for standing in Article 263 tfeu (formerly Article 230 ec). But, returning to Okin, how could there ever be ‘ “shared understandings” on abortion’43 where one of the major constituencies, namely the foetuses themselves, quite literally have no voice and, though very much affected, cannot

38 39 40 41

42 43

See supra section 2.2. The topic is touched upon, though, in Chapter 8 below. SM Okin, Justice, gender, and the family (Basic Books1989) 72. Walzer later admitted that he had underappreciated the extent of this ingraining –​the ‘totalizing character of patriarchalism’ –​and the danger it brought with it that meanings would be ‘pre-​interpreted to exclude’ women, though he continued to believe that good solutions, ‘transform[ations of] orthodoxies,’ were possible and indeed were taking place ‘from within:’ M Walzer, ‘Feminism and me’ (2013) 60(1) Dissent 50, 51–​52. ME Warren, ‘What Can Democratic Participation Mean Today?’ (2002) 30(5) Political Theory 677, 698. Okin (n 40) 66. Or rather on the right to life and the right to control one’s own body, which are the two rights which Okin mentions, since it is likely to be these which are being distributed, not abortions themselves.

32 Chapter 2 possibly participate in the debate? Armstrong sums up the problem very nicely when he talks of ‘the definition of the “us” ’44: Certainly one of the bones of contention that will remain for feminist theorists in particular is the question of who has the right to define this ‘we’, and whether it might not be a great deal more fragmented, and even unresolvable, than our rhetorical ‘we’s would indicate. In this sense defining the ‘we’ in advance of telling the story is wholly illegitimate, since defining the ‘we’ has to be a part of the story itself, open to negotiation and contention.45 If the presence of ‘those affected’ is compulsory at the debate, then two obvious groups in that category, the distributees and the distributors, would need to be present. And Armstrong is right that the identification of at least the first of these two groups is ‘every bit as controversial and political’46 as anything that might go on once the debate proper has commenced. The distributive community’s shared understandings cannot be ascertained until the community itself has been demarcated.47 Merkel also seems to be of the opinion that Walzer pays insufficient attention to this first step: It furthermore seems problematic that Walzer assumes a possible consensus on questions of distribution. That might be conceivable in his own idyllic upper-​class community of Princeton, but it would hardly work in the Bronx or in Harlem, which are barely an hour away.48 But if two of the major groups of ‘those affected’ are the potential distributees and the potential distributors, it is equally apparent that these two groups are likely to fundamentally disagree on the correct distributive principle to be employed. This problem is nicely dissected by Trappenburg, who in fact 44 45 46 47

48

C Armstrong, ‘Philosophical Interpretation in the Work of Michael Walzer’ (2000) 20(2) Politics 87, 91. ibid 90. ibid 91. On this analysis, there would need to be a debate before the debate. But this is surely right, because the right to decide on distributions (or decision-​making power) is a distribuend like any other, and thus its shared understanding would technically also need to be elucidated in a prior debate. But then the same problem would simply arise again in that debate, and so on, ad perpetuam. W Merkel, ‘Social justice and the three worlds of welfare capitalism’ (2002) 43(1) European Journal of Sociology 59, 65.

Michael Walzer and Complex Equality

33

identifies four orders of ‘actors:’ first-​order actors (for example, parliaments), second-​order actors (for example, local offices, boards, and so on –​that is, distributors), third-​order actors (potential recipients –​that is, distributees), and public opinion. Each order has a different motivation or driving force (except public opinion, which shares the same driving force as second-​order actors). According to Trappenburg,49 first-​order actors are motivated by overall efficiency, second-​order actors are motivated by (sphere-​specific) equity, and third-​order actors are driven by plain self-​interest. Thus, the question to be asked is not just who is present at the debate, but in what role do they find themselves on the day of distribution?50 And, again, the matter of the power-​balance within the Forum arises. Distributees can certainly be held to ransom by distributors, but occasionally it is distributors who find themselves powerless. An employee of a law firm, forced to take ‘false independent’ status51 by his employers for tax purposes (theirs not his or hers), will find himself or herself to be a distributor of services who is entirely subservient to his or her supposed distributees. In eu competition law, too, sellers can conceivably be held to ransom by (dominant) buyers.52 Will the two parties put their relative differences in power aside (or even, in Rawlsian terms, erect a ‘veil of ignorance’) during the debate, in order to ascertain, in a neutral fashion, the appropriate distributive principle? Or will, as Okin suggests, the ‘more powerful’ be able to silence, or render incoherent, those whom they dominate, so that the resultant ‘shared understanding’ is only shared in appearance?53 It is even possible that some of the ‘more powerful’ may do this unwittingly, entering the Forum with a distorted view of their community, so entrenched that 49 50 51 52

53

Trappenburg (n 30) 358–​9, in turn drawing on the work of Jon Elster, Local Justice: How Institutions Allocate Scarce Goods and Necessary Burdens (Russell Sage Foundation 1992). This links back to the point made above that one and the same person can easily understand one good in two or more ways. In other words, to become an independent service provider who then ‘invoices’ his former employer at the end of each month instead of receiving a salary. See Case 298/​83 cicce v Commission [1985] ecr 1105, [1986] 1 cmlr 486. In the uk, powerful supermarkets have a history of pressurizing dairy farmers to cut their farm-​gate price; for recent developments, see G Ruddick, ‘Tesco proposes deal with dairy farmers after milk price protests’ (Guardian, 4 November 2015)  accessed 6 May 2018. Walzer’s remark that ‘the consumer is not, and can never be, sovereign’ (Spheres of Justice (n 1) 113) is respectfully disagreed with. Okin (n 40) 112. Although one counter​argument might be that, in a properly realized Walzerian world, no party should hold significantly more power than any other.

34 Chapter 2 they ‘do not even consider it alterable,’54 such that the ‘shared understanding’ at which they think they have arrived is in fact ‘buil[t]‌on the prevailing ideologies.’55 The danger of an all-​pervasive ideology contaminating all debates is not lost on Walzer, who comments that, although entry for participants is free, ‘we may want to plead with some of them that they leave their conceptual baggage at the door’.56 And even if entry is free, implying that wealth makes no difference in the Forum, those with money may still wish to use it to influence the thinking of others, via education, communication, or some other means. This point has been made forcefully by Barry.57 According to him, any apparent consensus is actually the result of ‘power over communications’ (and so on), exerted by ‘the beneficiaries from the status quo.’58 This would give rise once again to the scenario wherein participants enter the Forum unwittingly brain-​washed, and then endorse an understanding of the good in question which is not theirs at all, but one planted by the powerful to perpetuate their own success (and perpetually to thwart the ambitions of the weak).59 Deliberate or inadvertent ill-​treatment of certain (weaker) factions during the debate is not the end of the story, however. How are participants, particularly those whose views did not survive into the final ‘consensus’, to be treated after the debate? Ball flags up this issue in the context of gay rights: [Gay men and lesbians] should be wary of a theory of political morality that does not contain the necessary protections against shifting majoritarian norms.60 Walzer himself sympathizes with those whose favoured ‘meaning’ was rejected in the Forum, and who now find themselves to be a minority:

54 55 56

57 58

59 60

ibid 113. ibid 72. As Rustin points out, if an injustice has ‘not already become the subject of contention within a society’, complex equality offers no grounds for intervening: Rustin (n 31) 31. Walzer, ‘Response’ (n 26) 295. He makes a similar point in Spheres of Justice: ‘All non-​ political goods have to be deposited outside: weapons and wallets, titles and degrees’ –​ Spheres of Justice (n 1) 304. Barry (n 35) 77–​8. Both quotations from ibid 78 (Barry’s italics). But, again, in a Walzerian world, would not the tyrannical use in the sphere of decision-​making (i.e. in the Forum) of success gained in the sphere of communications be prohibited as a ‘blocked exchange’? See L Tye, The Father of Spin: Edward L Bernays and The Birth of Public Relations (Crown Publications 1998) for some worrying examples of manufactured consent. Carlos A  Ball, ‘Communitarianism and Gay Rights’ (1999–​2000) 85 Cornell L Rev 443, 451.

Michael Walzer and Complex Equality

35

I don’t mean to dismiss the many problems that autonomy poses, especially with regard to non-​believers, who are forced to take on religion as a kind of ethnic affiliation. (I suspect that most religious communities have included significant numbers of non-​believers.) Their condition is not, however, entirely different from that of religious fundamentalists in a secular state. Another problem is more serious: it is very difficult to stop the religious majority from using state power across both communal boundaries and distributive spheres.61 With that last sentence, Walzer seems to be implying that, at least in a society in which complex equality is correctly practiced, the minority will be able to take comfort from the fact that the majority will be prevented from using their status as a dominant good; ‘membership of the majority’ will not be allowed to be a pass-​key to access all spheres. These issues are discussed in greater detail below, in Chapter 4, section 4.5.2(c), on the Tyranny of ‘Normalcy’. The last question to be asked is when, or at least how often, should the debate take place? Walzer himself is clear that ‘[s]‌ocial meanings are historical in character; and so distributions, and just and unjust distributions change over time’.62 And later: Boundaries […] are vulnerable to shifts in social meaning, and we have no choice but to live with the continual probes and incursions through which these shifts are worked out.63 It follows, then, that the debate itself must be continual. All the commentators seem to concur that this is the only solution. Danchin, for example, writes that ‘an unforced consensus must constantly be sought’:64 61

62 63

64

Walzer, ‘Response’ (n 26) 289. Scanlon also considers the problem of ‘those with divergent views’, post-​debate. He concludes that, in a private association, it would be acceptable to ‘deny [the relevant] goods to those who clearly lack [the required] beliefs.’ However, at the level of political society, goods such as civil and human rights must be extended even to people who ‘reject [society’s] most basic tenets’:  TM Scanlon, The difficulty of tolerance: Essays in political philosophy (cup 2003) 194–​5. Walzer, Spheres of Justice (n 1) 9. ibid 319. See also M Walzer, ‘Liberalism and the Art of Separation’ (1984) 12(3) Political Theory 315, 328, where he talks of the need for ‘continued revision’: ‘the arguing and the fighting have no visible end’. And four years later, he declares, ‘the interpretative enterprise goes on and on, never moving toward definitive closure’. See M Walzer, ‘Interpretation and Social Criticism’ in SM McMurrin (ed), The Tanner Lectures on Human Values, viii (cup 1988) at 24. Peter G Danchin, ‘Suspect Symbols: Value Pluralism as a Theory of Religious Freedom in International Law’ (2008) 33(1) Yale J Intl L 1, 53.

36 Chapter 2 For value pluralism not to lapse into the subjectivism of utopian universalism (religious or secular fundamentalism) on the one hand, or apologetic relativism (illiberal nationalism) on the other, it must strive for objectivity by continually seeking an overlapping consensus on the conflicting ends that divide cultures, groups, and individuals.65 Warren notes that ‘where individuals are empowered to serve as boundary patrols between spheres,’ justifications of inequalities (and therefore the distributive criteria which gave rise to them) would have to be ‘continually challenged and reworked.’66 Clark is heavily critical of Walzer, but certainly agrees that the distribution  of goods ‘is a matter of continuing argument rather than consensus.’67 Thigpen and Downing also talk about ‘continuing disagreements and unanswered questions,’68 observing, with regard to the presuppositions held by the participants of the debate, that ‘it is not possible to scrutinize every aspect of an intellectual inquiry at one time’.69 They conclude: [P]‌resuppositions may have been questioned yesterday, and they may be questioned again tomorrow. Those engaged in inquiry may reexamine their assumptions when those who question the continued usefulness of a concept or an approach gain the serious attention of their colleagues.70 2.5

An Objection: Walzer’s ‘Relativism,’ and The Need for An Override

A common objection to complex equality also takes as its point of departure the notion of ‘shared social meanings,’ that is, shared, to a greater or lesser extent,71 by the community in which the distribution is to take place. Allowing every society to be judged by its own yardstick, rather than insisting that all societies

65 66 67 68 69 70 71

ibid 50 (Danchin’s emphasis). Both quotations from Warren (n 42) 698. Samuel Clark, ‘Society against Societies: The possibility of transcultural criticism’ (2007) 13 Res Publica 107, 110. Thigpen and Downing (n 32) 546. ibid. ibid. See the previous section.

Michael Walzer and Complex Equality

37

be judged by a single, universal yardstick, has led to Walzer’s being labelled as a relativist.72 If the community is left to fix its own distributive criteria for every good, answerable only to ‘(internal) standards of justice,’73 what is to stop it settling on disagreeable, or even cruel, criteria? Rustin warns that Walzer’s theory may entail ‘practical indifference to barbarity.’74 Scanlon sums up the objection thus: Opponents of relativism thus commonly appeal to the possibility that the accepted norms of a society might license conduct which involves treating people in horrible ways.75 So what types of things might be classed as horrible? Nationalism is a popular example. Danchin confirms that cultural relativism, because of its penchant for ‘particularistic ideology’ and ‘anticosmopolitanism,’ may well bring in its train ‘variants of nationalism.’76 So if the Nuremburg Laws of 1935, which distributed (inter alia) the right to marry and even the right to engage in extra-​marital relations, had been found to be a manifestation of the ‘shared understanding’ of the German people at that time of those rights (and who should have them), does it follow, under complex equality, that those laws were just?77 Another example is slavery, where many commentators allege that Walzer himself is in violation of the rules. In a section discussing whether the 72

73 74 75 76 77

One of the earliest allegations of relativism came from Dworkin, leading to the exchange of views in the New York Review of Books mentioned above (supra n 34). Dworkin, in a review of Spheres of Justice in an earlier issue of the Review, later reprinted in his own A Matter of Principle, had commented, ‘We cannot leave justice to convention and anecdote’: R Dworkin, A Matter of Principle (Clarendon 1986) 220. Walzer, Spheres of Justice (n 1) 315. Rustin (n 31) 31. TM Scanlon, What we owe to each other (Bellknap Press of Harvard University Press 1998) 337. Danchin (n 64) 51. In her critique of Walzer, Shklar is particularly keen to press him on this point, which she sees as a great ‘difficulty’ of his work: ‘being governed by people just like oneself, representation as identity, has a history that Walzer does not want to face, but it is responsible for most of the horrors of the twentieth century’: JN Shklar, Political thought and political thinkers (S Hoffmann ed, University of Chicago Press 1998) 384. But note the interesting comment by one of the judges of the European Court of Human Rights that ‘the majority can always be wrong’: K Dzehtsiarou, European Consensus and the Legitimacy of the European Court of Human Rights (cup 2015) 201.

38 Chapter 2 Athenian treatment of metics was unjust (given the conception of citizenship that prevailed in Athens), Walzer leaves the question of slaves aside, because ‘the injustice of slavery is not disputed these days, at least not openly’.78 This judgment appears to be extrinsic to the community in question, and would seem to rely on extrinsic values. Carens describes the remark as ‘puzzling.’79 Is Walzer on the way to acknowledging the presence of an outer sphere (to borrow Carens’ model of concentric circles) containing ‘minimal standards of justice applicable to all contemporary states, regardless of their own particular histories, cultures, or political arrangements?’80 Opponents of abortion might also wish to apply such standards to states which allow the practice (presumably in accordance with shared social meanings).81 Those who take issue with female genital mutilation, living outside the countries in which this ritual is observed, might likewise need to call on them. One might describe these standards as transcultural, or universal. They are standards which override (or which those invoking them hope will override) the local norms at issue. However, given the nature of complex equality, the concept of and (arguable) need for an ‘override’ is not limited to inter-​ state disagreements, although these may provide the easiest or most sensational examples. All spheres, including and perhaps even especially those within the same state, are required to be, and to be kept, impregnable to outside influences. Within each distributive sphere, there is a specific social meaning which must be respected: ‘Dominance is ruled out only if social goods are distributed for distinct and “internal” reasons’.82 The structure provided by the social meanings is vital to the theory: ‘There are no external or universal principles that can replace it’.83 And yet, as several commentators have shown, principles that are certainly external to the distributive sphere concerned, and quite possibly universal as well, can be discerned at various points in Walzer’s argument. These principles, which (arguably) override the ban on boundary crossings, could be labelled as ‘interspherical,’84 or even ‘trans-​sphere.’85 The most common

78 79 80 81 82 83 84 85

Walzer, Spheres of Justice (n 1) 53. Carens (n 6) 58. ibid. But see supra text accompanying n 43. Walzer, Spheres of Justice (n 1) xv. ibid 314. This term used by Gutmann (n 30) 107. This term used by Walzer, ‘Response’ (n 26) 294.

Michael Walzer and Complex Equality

39

example is equal membership of the community, that is, equal citizenship. Walzer himself implies that this is a blanket principle, applicable across all spheres, in the third of his three ‘very general’ principles: ‘that [a]‌distribution must recognize and uphold the underlying equality of membership.’86 Seizing on this, critics have rushed to proclaim equal citizenship as a ‘referee criterion’ or ‘master principle,’87 a ‘criterion that seems to apply across [all] societal spheres,’88 and a ‘fundamental principle […] across other relational spheres.’89 According to Gutmann, equal citizenship ‘inform[s] distributive justice in many other spheres.’90 According to Miller, it ‘plays a regulative role’ but is not a fundamental principle.91 Den Hartogh sums up the issue in this way: It follows that the sphere of membership cannot be fenced off from the other spheres in the way required by the nondominance principle: the possession of membership determines or codetermines one’s share in the other spheres.92 Walzer later admitted that citizenship has a certain ‘centrality’ within the theory, and that it takes on ‘heightened instrumental and also symbolic value:’ ‘Indeed, this value may itself be critical, as both Miller and Swift suggest, in adjudicating the internal disputes.’93 Other concepts that critics have asserted are, or may be, overrides include responsibility,94 justice,95 morality,96 and recognition.97 A controversial addition to the list would be democracy itself. Armstrong puts the argument well: 86 87 88 89 90 91 92 93 94 95

96 97

Walzer, Spheres of Justice (n 1) 84. Both terms from Trappenburg (n 30) 347. ibid 346. James W Fox, Jr, ‘Relational Contract Theory and Democratic Citizenship’ (2003–​2004) 54 Case W Res L Rev 1, 37. Gutmann (n 30) 116. D Miller, ‘Introduction’ in Miller and Walzer (n 24) 3. G den Hartogh, ‘The Architectonic of Michael Walzer’s Theory of Justice’ (1999) 27(4) Political Theory 491, 494. Walzer, ‘Response’ (n 26) 87. Trappenburg (n 30) 346. Gutmann (n 30) 107 (and see also RJ van der Veen, ‘The Adjudicating Citizen: On Equal Membership in Walzer’s Theory of Justice’ (1999) 29(2) British Journal of Political Science 225, 228). ibid 112. ‘[S]‌omething that may be achieved in many spheres’:  R Keat, ‘Colonisation by the Market: Walzer on Recognition’ (1997) 5(1) Journal of Political Philosophy 93, 103.

40 Chapter 2 [T]‌he only way of ascertaining (choosing) our goals is the democratic way. This is the one a priori, transcendental and universal principle that Walzer will not drop: that we must not stop telling stories, and that only we can be the judges of which are the best.98 From this argument it is not a giant leap to conclude that the entire theory, since it necessarily has to apply across the board, is a universal principle and is thus in violation of itself. Ball articulates this nicely: Similarly, one can view the rule that spillover effects from one sphere to another always lead to injustice, as a form of metaprinciple from which no departure is possible without violating Walzer’s theory of justice.99 It would not be fair to finish this section, however, without briefly considering what Walzer himself has said on these topics over the years. It is submitted, for the purposes of this book, that Walzer all along entertained the idea of certain overarching norms. In 1983, in Spheres of Justice, he referred to ‘our common humanity,’ which he said gave rise to two rights: the right to life and the right to liberty.100 But anything else, he went on, would have to ‘follow from shared conceptions,’ and would be ‘local and particular in character.’101 In 1988, in Interpretation and Social Criticism,102 he modified this position slightly, accepting that there was what he called ‘a kind of minimal and universal moral code,’103 consisting of prohibitions on murder, deception, betrayal, and gross cruelty.104 In the same year’s Company of Critics,105 he insisted that it was not his wish to be an apologist ‘for this (or any other) society.’106 98 99 100 101

Armstrong (n 44) 90. Ball (n 60) 491. Walzer, Spheres of Justice (n 1) xv. ibid. However, prior to Spheres of Justice, it appears that he took a tougher line. In an article in 1981, for example, he wrote: ‘What is crucial, however, is that the redistributive pattern [the people] choose is not subject to authoritative correction’. See M Walzer, ‘Philosophy and Democracy’ (1981) 9(3) Political Theory 379, 385. 102 M Walzer, ‘Interpretation and Social Criticism’ in SM McMurrin (ed), The Tanner Lectures on Human Values, viii (cup 1988). 1 03 ibid 22. 104 Thus, according to Walzer, the Biblical prophets were able to employ the minimal code –​ ‘a kind of international law’ –​in order to criticize foreign nations with alien values (Jonah in Nineveh, for example): ‘don’t violate treaties, don’t kill innocent women and children, don’t transport whole nations into involuntary exile’. See ibid 76 and 78–​9. 1 05 M Walzer, The Company of Critics:  Social Criticism and Political Commitment in the Twentieth Century (2nd edn, Basic Books 2002). 106 ibid xix.

Michael Walzer and Complex Equality

41

In 1994, in Thick and Thin,107 he continued his discussion of the ‘moral minimum’ or ‘Minimal Morality.’108 Again, he considered that it ‘most likely’109 would consist of rules against murder, deceit, torture, oppression, and tyranny. But the point that he was keen to stress was that this ‘thin morality’ does not ‘leap from the philosopher’s mind like Athena from the head of Zeus’: it must be (or rather have been) ‘worked on.’110 Thus, even supposedly universal norms have been ‘work[ed] … out’111 by men and women who themselves are ‘creatures of history,’112 that is, particular people, from a particular time and place, with particular fears, beliefs, goals, and so on: ‘Minimalism … is less the product of persuasion than of mutual recognition among the protagonists of different fully developed moral cultures’.113 In 1995’s Pluralism, Justice, and Equality, he again conceded that ‘[m]‌urder, torture, and enslavement are wrongful features of any distributive process’:114 We need a theory of human rights (or its functional equivalent in other cultures) to set the basic parameters within which distributions take place.115 Finally, in 1996, in an essay entitled “Spheres of Affection,” he explicitly acknowledged the problem enunciated by Rustin and Scanlon at the beginning of this section, but offers his own warning in return: A particularism that excludes wider loyalties invites immoral conduct, but so does a cosmopolitanism that overrides narrower loyalties.116 107 108 109 110 111 112 113 114 115 116

M Walzer, Thick and Thin:  Moral Argument at Home and Abroad (University of Notre Dame Press 1994). ibid 9. ibid 10. All quotations from ibid 12. ibid 14. ibid 12. ibid 17, emphasis added. Walzer, ‘Response’ (n 26) 293, emphasis added. ibid. M Walzer, ‘Spheres of Affection’ in M Nussbaum and J Cohen (ed), For Love of Country, Debating the Limits of Patriotism (Beacon Press 1996) 125, 127. Shklar attempts to sum up (and maybe even rationalize) Walzer’s approach to human rights; he is ‘not really indifferent to human rights,’ she says, but a rights culture ‘is a threat to the clubby life of traditional groups’ for which she says he is ‘nostalgi[c]‌’: JN Shklar, Political thought and political thinkers (S Hoffmann ed, University of Chicago Press 1998) 384–​5.

42 Chapter 2 2.6 The cjeu and its Override: An Answer to the Objection In the context of European law in general, and the practice of the Court of Justice in particular, it is submitted that several overarching ‘codes’ are already in place that could, without the slightest difficulty, be pressed into service as ‘overrides’ to be used to ‘screen’ local, tradition-​sensitive distributive criteria, should the Court decide to adopt a Walzerian approach to its equality cases.117 Pre-​eminent amongst these, however, is the European Convention on Human Rights (echr), which in two important ways already has what might be called override status.118 Firstly, its entire raison d’être was to act as ‘a bulwark against any recrudescence of [Nazi or Fascist]

117 As well as the European Convention on Human Rights, which is dealt with in the text, there is the European Union’s own Charter of Fundamental Rights [2000] oj C364/​1, the United Nations Charter 1945, the Universal Declaration of Human Rights 1948, Convention No 111 of the International Labour Organisation 1958, the European Social Charter 1961, the International Covenant on Civil and Political Rights 1966, the International Covenant on Economic, Social and Cultural Rights 1966, the International Convention on the Elimination of All Forms of Racial Discrimination 1966 (‘cerd’), and the Convention on the Elimination of All Forms of Discrimination against Women 1979 (‘cedaw’). To these could even be added the Member States’ constitutions. 118 A rechristening of Walzer’s Minimal Morality from Thick and Thin (see previous section) as human rights was in fact already undertaken by Miller in 2012:  ‘We might equally well capture these universal principles of justice using the language of basic human rights: human rights identify the claims people can make on each other regardless of boundaries and memberships’ (D Miller, Justice for Earthlings: Essays in Political Philosophy (cup 2012)  145). In a footnote, he offers a caveat where rights requiring positive action on the part of the State are concerned (the example he gives is the right to subsistence). The European Convention on Human Rights contains no such positive rights, although the European Court of Human Rights has occasionally been prepared to  read a positive obligation into a negatively-​expressed right e.g. D v United Kingdom App no 30240/​96 (ECtHR, 2 May 1997), in which the Article 3 prohibition on torture was interpreted to mean that a national of St Kitts suffering from aids must be treated in the uk and not deported to a country with little or no capacity to look after him. Such interpretive frolics do increase the ‘override power’ of a right, so to speak, potentially tying the hands of the Walzerian judge by declaring certain distributional decisions to be illegal a priori. But clearly, whichever human rights document were chosen, adherence thereto brings with it conscription to the (overriding) interpretations of its constituent court. The capacity of human rights law to fix meanings, empowering a judge potentially to reclassify an instance of community-​mandated boundary defence as a boundary breach, will be returned to later.

Michael Walzer and Complex Equality

43

dictatorship;’119 by its very nature, therefore, it would override any law, however widely proclaimed or obeyed, that constituted a threat to the rights which it contained. Secondly, the Court of Justice itself has declared that European law enshrines the notion of human rights common to the eu’s Member States, and that: [I]‌international treaties for the protection of human rights on which the Member States have collaborated or of which they are signatories, can supply guidelines which should be followed within the framework of Community law.120 It has thus repeatedly used the echr as a yardstick by which to measure national, and eu, legislation.121 Furthermore, all eu Member States 119

AH Robertson and JG Merrills, Human Rights in the World: An introduction to the study of the international protection of human rights (4th edn, mup 1996) 120. 120 Case 44/​79 Liselotte Hauer v Land Rheinland-​Pfalz [1979] ecr 3727, para 15. At the end of the same paragraph the echr is referred to explicitly, as it is for much of the rest of the judgment. The declaration in Hauer was preceded by a number of declarations of a similar nature, albeit not referring directly to the echr. See Case 29/​69 Erich Stauder v City of Ulm [1969] ecr 419; Case 11/​70 Internationale Handelsgesellschaft mbH v Einfuhr-​ und Vorratsstelle für Getreide und Futtermittel [1970] ecr 1125; Case 4/​73 J Nold, Kohlen-​ und Baustoffgroßhandlung v Commission of the European Communities [1974] ecr 491. Eventually, the legislature caught up with the judiciary, and the obligation on the eu to respect the rights guaranteed by the echr was set out explicitly in Article 6 of the Treaty on European Union (the ‘Maastricht’ Treaty, signed in 1997), hereinafter ‘teu’. 121 Space forbids a full rehearsal of the cases in which the echr has played a prominent, or even decisive, role in proceedings. However, amongst the most well-​known of these cases (and with the right invoked in parentheses) are the following: Case 44/​79 Hauer (n 120)  (the right to property and the freedom to pursue a trade or profession), Case 63/​83 Regina v Kent Kirk [1984] ecr 2689 (no punishment without law), Case 222/​84 Marguerite Johnston v Chief Constable of the Royal Ulster Constabulary [1986] ecr 1651 (equal treatment between male and female workers), Joined cases 46/​87 and 227/​88 Hoechst AG v Commission of the European Communities [1989] ecr 2859 (the right to the inviolability of the home), Case C-​260/​89 Elliniki Radiophonia Tiléorassi ae and Panellinia Omospondia Syllogon Prossopikou v Dimotiki Etairia Pliroforissis and Sotirios Kouvelas and Nicolaos Avdellas and others [1991] ecr I-​2925 (the freedom of expression), Case C-​368/​ 95 Vereinigte Familiapress Zeitungsverlags-​und vertriebs GmbH v Heinrich Bauer Verlag [1997] ecr 3689 (the freedom of expression). Perhpas the rights case with the highest profile of recent years has been Joined Cases C-​402/​05 P and C-​415/​05 P Yassin Abdullah Kadi, Al Barakaat International Foundation v Council of the European Union, Commission of the European Communities, United Kingdom of Great Britain and Northern Ireland [2008] ecr I-​6351 (the right to a fair trial and the right to property).

44 Chapter 2 are signatories to the echr in their own right, which means that the rights enshrined therein may be invoked throughout the Union.122 To all intents and purposes, this means that all activity in the bloc –​including distributive activity –​ is already carried out under the gaze of the European Convention. The Convention sets universally-​applicable maximum and minimum parameters for behaviour, and behaviour which falls outside these parameters –​even if the result of consensus –​stands to be impugned. Although there are other candidates, then, it is proposed that the Court should use the echr, representing as it does both nation-​wide and continent-​wide consensus, to help it to scrutinize the ‘sphere-​specific’ distributive principles with which it is faced and to act, if necessary, as a kind of safety valve or circuit-​breaker. As will be argued at greater length elsewhere, the European Union is well-​suited to the theory of complex equality, and the ‘override question’ provides another illustration. The cjeu has both faced and solved the problems of if, when and how to override local, narrow rules where they clash with wider objectives; it is very much a past master when it comes to these particular conundra. Indeed, the entire European project could be seen as an ongoing exercise in finding and maintaining the right balance between diversity (many inner spheres; many sets of principles) and union (one outer sphere; one set of principles).123 As later examples from the case-​law will show, any instances of ‘barbarity,’ or ‘horrible’ treatment124 which may emerge from a given distributive community’s shared understandings will not be latent (as they could have been, and could have remained, under the Aristotelian approach), but patent:  quickly identifiable by both practitioners and judges, who of course are already extremely familiar with the echr. If a distributive criterion offended against the

122 While they may be invoked by any individual on European territory, regardless of nationality, they may only be invoked against the Signatory States themselves. However, via a form of drittwirkung (though not in the orthodox sense), the State may be held responsible where the breach is carried out by a second individual, for failing to have implemented the Convention sufficiently rigorously, and thus the Convention could be said to have horizontal as well as vertical effect (although again not in the orthodox sense). 123 The use of the adjectives ‘inner’ and ‘outer’ here draws (again) on Carens, who talks of an ‘innermost … circle’ and an ‘outermost circle’: Carens (n 6) 60 and 63. ‘United in diversity’ was at one time mooted as a possible motto for the eu. See the (later scrapped) Treaty establishing a Constitution for Europe [2004] oj C310/​1, Article I-​8. 124 The words in quotation marks taken from Rustin and Scanlon respectively, as cited at the beginning of section 2.5.

Michael Walzer and Complex Equality

45

‘Minimal Morality,’125 it could then be swiftly ‘rule[d]‌out’126 and any distribution based upon it declared flawed, leaving only the question of remedy outstanding.127

125 Walzer, Thick and Thin (n 107) 9. 126 Walzer, Spheres of Justice (n 1) 162. 127 This would hopefully provide reassurance to critics such as Benhabib, who, having noted that Walzer is ‘sceptical, or maybe better still, agnostic about universal human rights,’ wonders if reliance by a community on its own self-​understanding, unattended by such rights, might not lead to legal and moral compromise where refugees and asylees were concerned: S Benhabib, The rights of others (cup 2004) 118–​119. At the time of writing, a case is indeed pending before the Court of Justice wherein the European Commission accuses Hungary, Poland and the Czech Republic of failing to comply with a Council decision of September 2015, which, following the migrant crisis, had required them to accept a certain number of migrants each: -​-,​ ‘Relocation: Commission refers the Czech Republic, Hungary and Poland to the Court of Justice’ (European Commission Press Release, 7 December 2017)  accessed 7 May 2018. As the Charter of Fundamental Rights is mentioned in the latter decision, it seems likely that consideration of the Charter will form part of the Court’s work as it decides whether to find for the Commission or for the three Member States: Council Decision (eu) 2015/​1601 of 22 September 2015 establishing provisional measures in the area of international protection for the benefit of Italy and Greece [2015] oj L248/​80, Recital 45.

Chapter 3

The Principle of Equal Treatment of Persons Irrespective of Gender 3.1 Introduction In Chapter 1 some illustrations were given of how easily a comparison could be engineered to attain a certain result, for example, with regard to the choice of comparators.1 It was also seen how adjusting which properties were to be taken into account could transform or even reverse the result of a comparison.2 Thus it appears that what McColgan calls the ‘comparator-​driven approach’ to equality,3 and what Holtmaat and Tobler call the ‘symmetric approach’ to equality,4 is endlessly malleable. Nowhere is this more apparent (in the cjeu’s case-​law, at least) than in the field of gender discrimination. The Aristotelian ‘like cases alike’ (or ‘like for like’) test5 admits of two possible outcomes (‘discrimination’ and ‘not discrimination’), but each can be arrived at via two different routes, resulting in four conceivable permutations; see the table at Appendix I. As the result to be inserted in the second column is generally known in advance of performing the test, that is, it is generally already known whether Comparator 1 and Comparator 2 have been treated in like fashion or in unlike fashion, the column which usually (although not always) offers the best scope for ‘rigging’ (judicial or otherwise) is Column 1. It will be seen that a great deal hangs on the result of the enquiry as to whether the two comparators are like, or unlike. This in turn depends in large part upon which properties are to be taken into account for the purposes of the comparison, and (perhaps the same thing) upon which degree of ‘difference awareness’ is to be employed while doing the comparing. 1 See supra, section 1.2. 2 The agents of such engineering or adjusting may, in the first instance, be the Distributors themselves, but ultimately, and for the purposes of this book, most importantly, they may be the judges before whom the discrimination case is eventually brought. 3 Aileen McColgan, ‘Cracking the comparator problem: Discrimination, “Equal” Treatment and the Role of Comparisons’ (2006) 6 ehrlr 650, 650. 4 R Holtmaat and C Tobler, ‘Cedaw and the Eu’s Policy in the Field of Combating Gender Discrimination’ (2005) 12(4) mj 399, 417. 5 These two phrases are also borrowed from McColgan (n 3) 651 and 660.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/9789004354265_004

Equal treatment irrespective of gender

47

As Hervey and Shaw point out, ‘the archetypal question is that of whether a woman is “like” or “unlike” a man when she is pregnant’.6 This is a question the cjeu has now attempted to answer many times, while managing to avoid the discredited question, ‘Is a pregnant woman “like” or “unlike” a sick man?’ The question will be examined at section 3.4 below. However, before embarking on the case-​study proper, it is necessary to deal with two preliminary issues –​how the Walzerian analysis is to proceed (section 3.2) and, crucially, whether the theory can accommodate characteristics and voluntary choices (section 3.3). 3.2

Outline of Analysis and Taxonomy of Results

In analyzing a cjeu equality case from the Walzerian perspective, it is first necessary to identify the good being distributed, so as to be able to delineate the sphere at issue. The person or entity invoking the principle of equal treatment will either already be in this sphere, or else will be trying to gain access to it. Likewise, their (real or hypothetical) ‘comparator’7 will either already have entered the distributive sphere, or will be attempting to enter it. It will then be necessary to identify any boundary breaches, whether at the level of distributee, or at the level of distributor, or indeed anywhere else. All results except those arrived at by the distributive community itself or, in the event of a dispute, by the person universally accepted as arbiter,8 will be provisional only. Thus, for everyone else, a common answer to the question, ‘Has there been a breach of the principle of equality?’ is: ‘This cannot be known until the distributive community concerned has pronounced upon whether y is or is not part of its shared meaning of x.’9

6 TK Hervey and J Shaw, ‘Women, work and care: women’s dual role and double burden in Ec sex equality law’ (1998) 8(1) Journal of European Social Policy 43, 48. 7 Under the Aristotelian test. Complex equality does not actually entail comparison, and so can technically proceed without a comparator at all. This is discussed further below. 8 See section 7.2.2 below. 9 The same system of variables is used here as is used in Walzer’s own statement of the theory of complex equality in Spheres of Justice: ‘No social good x should be distributed to men and women who possess some other good y merely because they possess y and without regard to the meaning of x.’ See M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books1983) 20 (Walzer’s italics).

48 Chapter 3 For the purposes of this answer, x is the good being distributed and y is a good of which the potential distributee is already in possession. In what follows, this will be referred to as the Standard Contingent Reply. That said, there are times when a boundary breach is so blatant that it may simply be presumed. And even on other occasions, a good prediction can usually be made of what the result will be. Results, then, fall into two basic categories:  ‘breach’ or ‘no breach.’ An important point to remember is that there can still be a fair degree of inequality (of the sort that Walzer would call simple inequality) within the distributive sphere, even after the test has been applied and a finding of ‘no breach’ has been made. Such an inequality would be one of the ‘small inequalities’ to which Walzer referred in Spheres of Justice,10 a ‘just’ inequality in the sense that the reason for any distinction being made falls within the meaning of the good being distributed, and not outside it. Walzer never promised to ‘even […] up the balance,’ merely to ‘get […] the unevenness right.’11 The outcome of most social transactions involves a certain amount of inequality; one party rarely walks away from such a transaction in exactly the same position as the other party. And, where court cases are concerned, it is almost inconceivable that both parties would fare identically, unless they had come to some kind of (highly unusual) fifty-​fifty settlement. It has been clear since King Solomon’s day that there must be a winner and a loser; equality of outcome is neither possible nor even desirable.12 These are in most cases the benign inequalities of which life is necessarily made up. They do not offend against complex equality, because the ground upon which any distinction is founded is intrinsic to, rather than extrinsic to, the sphere at issue. In us terms, it is the non-​suspect nature of the ground which leads to the prima facie assumption that the distinction will turn out to be inoffensive. As Balkin puts it, such inequality is, the inevitable if occasionally unfortunate outcome of markets, individual private preferences, and judicial respect for legislation passed by democratically elected representatives of the people.13 10 11 12

13

ibid 17. See also more detailed examination in Chapter 2 above –​section 2.2. ibid 260. Referring to 1 Kings 3: 16–​28. Of course the outcome, while unequal on its face, may, looking at the bigger picture, lead to equality overall, for example, where a damages payment to one party simply restores the balance which existed before the contested event occurred. JM Balkin, ‘Plessy, Brown, and Grutter: A play in three acts’ (2004–​2005) 26 Cardozo L Rev 1689, 1717.

Equal treatment irrespective of gender

49

Tridimas explains it thus: [I]‌n the sphere of economic law, the principle of equality prohibits only measures imposing risks on traders beyond those which they can reasonably be expected to bear in the light of the underlying economic circumstances.14 What one might call ‘normal’ risks must just be run as usual; any negative outcome will not be actionable as a breach of the principle of equal treatment. In other words, the loser may not sue the winner. In what follows, then, many cases which do not disclose any boundary breach (‘no breach’ cases) may alternatively or in addition be referred to as ‘small inequalities’. 3.3

Characteristics and Voluntary Choices: Is the Theory Fit for Purpose?

A first issue that needs considering, especially in the context of suspect grounds, is the extent to which complex equality and boundary-​defence can work at all in the combating of discrimination. A typical boundary breach presupposes two distributions in two, neighbouring spheres.15 In the first distribution, X has distributed to him or her a good which will turn out to be dominant (or, perhaps more importantly for present purposes, negatively dominant); X then moves to the second sphere taking this good with them. In the second distribution, X has distributed to him or her, on the strength of his or her possession of said good, a second good (assuming that possession or otherwise of the first good actually has nothing to do with the second good, and does not form part of the second good’s shared meaning). A complex egalitarian would now consider that the boundary between the two spheres had been breached: X had used the first good tyrannically to gain the second, and so the second distribution was flawed. X should have left the first good ‘at the door’ (of the second sphere). This begs the somewhat difficult question: can a characteristic be 14

15

T Tridimas, ‘The Application of the Principle of Equality to Community Measures’ in A Dashwood and S O’Leary (eds), The Principle of Equal Treatment in EC Law (Sweet & Maxwell 1997) 228 –​emphasis added. The one described in this section is by no means the only sort. For example, the one described in this section is a boundary breach by a distributee. However, one could also have a boundary breach by a distributor. Nevertheless, the fact remains that all boundary breaches entail two distributions.

50 Chapter 3 regarded as the product of a distribution? Put another way, is a characteristic a distribuend, such that it can go on to be regarded as a dominant, or negatively dominant, good? It is worth starting with an easier version of this question, albeit one equally as important for the case-​law analysis which must be undertaken: can a voluntary choice be regarded as the product of a distribution? For example, when, in an age discrimination case called Bartsch, the claimant enters the ‘second sphere’ (where the contested distribuend is a survivor’s pension), she brings with her the fact that she chose to marry a man twenty one years her senior.16 She exercised a free preference. Except perhaps for classicists, there is no sphere run by Cupid, where he ‘distributes’ those with whom the ‘distributee’ must fall in love. However, that is to take too simplistic or literal a view of Walzer’s theory, which admittedly many critics do, especially when it comes to visualizing the network of spheres itself, and working out what constitutes a sphere in the first place. There is absolutely no reason why choice should not be a completely acceptable distributive criterion, as long as the relevant community is in agreement. Examples given by Walzer include the choosing of elected representatives (in the sphere of politics), and, indeed, the choosing of partners (in the sphere of romantic love). However, in both of the latter examples, it is the distributor who chooses. But a perfectly obvious example of a sphere where it is the distributee who chooses is the market itself.17 18 So, by analogy, it is not hard to envisage a ‘first sphere’ wherein Mrs Bartsch chose 16 17

18

Case C-​427/​06 Birgit Bartsch v Bosch und Siemens Hausgeräte (bsh) Altersfürsorge GmbH [2008] ecr I-​7245. Although no doubt Foucault would object that consumers are not making a free choice at all, but merely acting upon cues planted in them by ‘experts’ via the media. The individual’s acting, unquestioningly, on this expertise is one of what Foucault calls the ‘techniques of the self’, a kind of self-​regulation which in turn forms part of his theory of biopolitics: M Foucault, ‘The Political Technology of Individuals’ in LH Martin, H Gutman, PH Hutton (eds), Technologies of the self: a seminar with Michel Foucault (University of Massachusetts Press 1988). In Mayer’s critique of Walzer’s industrial democracy, he argues that an employee chooses to put himself under the authority of his manager; if that is correct, then the sphere of office would also become a sphere in which it is the distributee who chooses: R Mayer, ‘Michael Walzer, Industrial Democracy, and Complex Equality’ (2001) 29(2) Political Theory 237. Note post-​Brexit Britain’s apparent embrace of industrial democracy: ‘we will make sure … that the voices of workers are heard by the boards of publicly-​listed companies for the first time’ (T May, ‘Theresa May’s Brexit speech in full’ (17 January 2017)  accessed 7 May 2018).

Equal treatment irrespective of gender

51

the late Mr Bartsch as a marriage partner. (Again, to worry unduly about the identity of the distributor in this sphere is to subject a postmodern theory to seventeenth century-​style analysis: Walzer makes quite clear, in any event, that in matters of the heart potential lovers distribute themselves via what he calls ‘the gift of self.’)19 The conclusion, then, is that voluntary choices can indeed be regarded as the product of a distribution. Returning to the question of characteristics, then, it goes without saying that these are conceptually close, sometimes very close, to voluntary choices.20 Ball, for example, regards membership of the gay community as a voluntary choice,21 while others hold that homosexuality is more a matter of nature than nurture; for such people homosexuality would presumably more properly be described as a characteristic.22 Walzer seems happy on several occasions to at least entertain the idea of a boundary breach wherein the party in breach brings with them to the ‘second sphere’ a characteristic, the possession of which causes (or at least might cause) the second distribution to be flawed. The very first examples he gives draw on Pascal, and envisage the tyrannical importation into a foreign sphere of such matters as strength and ­handsomeness.23 Later, he speculates as to how charm might be used in the acquisition of love, 19 20

21

22

23

Walzer, Spheres of Justice (n 9) 238. ‘The sphere of private affairs is exactly like the market in commodities, except that these commodities own themselves’: ibid. Schiek’s three-​limbed schema for classifying the grounds of discrimination is instructive here (see also below, section 4.2.1). The second and third classifications are actual and unalterable biological differences (for the purposes of this section, ‘characteristics’), and differences which are the product of choice (for the purposes of this section, ‘voluntary choices’). The schema is explained in full in D Schiek, ‘A New Framework on Equal Treatment of Persons in Ec Law?’ (2002) 8(2) Eur LJ 290, particularly at 309–​310. The gay community is a ‘communit[y]‌of choice’, not a ‘[c]ommunit[y] into which ­people are born’:  Carlos A  Ball, ‘Communitarianism and Gay Rights’ (1999–​2000) 85 Cornell L Rev 443, 449–​450. Schiek, however, agrees with Ball that sexual orientation is a product of choice: Schiek (n 20) 310. Howard refers to it as a characteristic: E Howard, ‘The case for a considered hierarchy of discrimination grounds in eu law’ (2006) 13(4) mj 445, 454. This book takes no position on the matter. Wintemute also considers the proximity of characteristics and choices, this time in the context of gender discrimination: R Wintemute, ‘When is Pregnancy Discrimination Indirect Sex Discrimination?’ (1998) 27(1) Indus LJ 23. He wonders whether the fact that pregnancy is (usually) a result of the exercise of free will makes any difference to the way in which it should be viewed by discrimination law. As examples of choices of which only men are physically capable, he gives growing a beard, or donating sperm: ibid 27–​28. Walzer, Spheres of Justice (n 9) 18–​19. Another early example is eccentricity, which Walzer describes as ‘a social good like any other’: ibid 8.

52 Chapter 3 how being stimulating might be used in the acquisition of influence, or even how skill at bargaining might be used in the acquisition of pictures.24 In an important section on abstraction, he notes how many of a person’s ‘capacities,’25 and even the capacities ‘to make an effort or to endure [the] pain’26 needed to make use of the former capacities, are, at least in some philosophers’ eyes, merely ‘the arbitrary gift of nature.’27 Walzer objects to such abstraction, believing that a capacity such as (to take a random example) skill at playing the piano should enable one to acquire (say) public honour. Again, the ‘good’ which this person brings with them to the sphere of honour, and which, potentially, could be dominant (in this case, it is argued, not), is a characteristic. This implies that there was an earlier ‘distribution’, to all humans, of what Walzer calls the ‘goods in their minds.’28 One could hypothesize as to who the distributor in this original sphere might be –​Mother Nature, perhaps (echoing Walzer’s ‘gift of nature’), or some kind of Deity,29 or even, from the scientific point of view, one’s parents or other forebears.30 But again it is futile to become preoccupied with such matters in a postmodern theory. It is simply irrelevant; the use of the word ‘distribution’ in a case like this is really only a kind of shorthand. The real point is that all matters (including characteristics like piano-​playing ability) must stay within the sphere to which they are germane, unless their presence in a second matter’s sphere is sanctioned by the shared understanding of that second matter.

24

25 26 27 28 29

30

ibid 23–​24. Walzer also seems to regard what he calls ‘birthright’ (in other words, nobility) as a potentially dominant good; he even nominates the sphere in which it is distributed as ‘the sphere of birth and blood’: ibid 16. ibid 260. ibid 261. ibid. ibid. If Walzer’s tone is understood correctly, he actually regards this distribution as more important than a distribution of physical goods such as ‘hats and coats’: ibid. Walzer happily acknowledges God as a distributor when he talks about ‘divine grace’ in Chapter 10 of Spheres of Justice. Of course, God’s distributive criteria, both for Grace and (if applicable) characteristics, are unknown and unknowable. As Walzer laconically puts it, ‘we don’t know how [desert] figures’: Walzer, Spheres of Justice (n 9) 260. In the era of genetic engineering, it is not at all far-​fetched to talk about the distribution of genes, ordinarily by one’s parents, but occasionally by third parties via some manner of donation. Although it would vary from community to community, it seems likely that the correct distributive criterion for any such donation would be need. If it ever came about that wealth or status could ‘buy’ a person’s children superior genes, this would almost certainly be regarded as a boundary breach, or blocked transaction.

Equal treatment irrespective of gender

53

Critics have also happily embraced the idea of characteristics as goods capable of giving rise to boundary breaches. Trappenburg gives as one of her many examples the situation where, of two political candidates, one has clearly garnered more votes in an election, which is the only valid way of distributing parliamentary seats in most communities. However, she then asks whether the other candidate’s being trustworthy and decent, or being black, or being unattractive, might be a reason to cede the seat to him. Of course, if the seat were ceded to the second candidate on the grounds of one of these characteristics, there would be a very compelling argument for deeming this a boundary breach.31 The interim conclusion, then, must be that characteristics can indeed be regarded as the product of a distribution, even if it is a hypothetical one. Complex equality and boundary-​defence may thus be relied upon in the fight against discrimination, which is necessarily based on characteristics, or voluntary choices,32 possessed or made by the party discriminated against. One last word on characteristics and voluntary choices. A critique of over-​ systematization, even overdetermination, might properly be levelled against the foregoing paragraphs. It might even be thought that the whole project smacks of commodification of the entire business (pun intended) of being human, from the birth of an engineered baby to the death of an older lover. What about the transsexual experience, where the declaration by a midwife of what he or she perceives as a characteristic (gender) could in fact later be felt by the individual concerned as “non-​consensual gendering,” even “gendering violence”: Gender attribution is compulsory; it codes and deploys our bodies in ways that materially affect us, yet we choose neither our marks nor the meanings they carry.33 Here ‘characteristics’ are no such thing. Society, acting upon its own conception of culture and history, simply turns the newborn baby into whatever it considers most efficient to its purposes. A characteristic is something not of 31 32 33

See Margo Trappenburg, ‘In Defence of Pure Pluralism: Two Readings of Walzer’s Spheres of Justice’ (2000) 8(3) Journal of Political Philosophy 343, 350–​352. Or ascribed differences. This is the first of Schiek’s three categories –​see Schiek (n 20). All three quotations from S Stryker, ‘My words to Victor Frankenstein above the village of Chamounix: Performing transgender rage’ 1994(1) glq: A Journal of Lesbian and Gay Studies 237, 253–​4.

54 Chapter 3 the person, but at once imposed on them and, much more importantly, taken from them. And what is taken is the right to self-​identify, rendering laughable the concept of ‘voluntary choices’. If the distributor is a robber, then the judge risks becoming merely a ­robber’s accomplice. Possession can be corrupted and roles reversed. The ex-​possessor becomes the supplicant. However, from the Walzerian point of  view, the situation is now simply one in which it is the distribution in Sphere 1 which is contested, and the potential boundary breach is at the level of the Distributor. If the proper distributor of gender identity is oneself, then the midwife’s declaration is an act of trespass from his or her proper sphere, in which he or she distributes, say, epidurals, into an entirely separate distributive sphere beyond his or her office. 3.4 Pregnancy 3.4.1 The CJEU’s Case-​Law Under the Aristotelian test, as mentioned above, pregnant women have in the past be compared to sick men. This comparison has been described as ‘insulting,’34 ‘inappropriate,’35 ‘distasteful and unfortunate’36 and ‘inaccurate and politically damaging.’37 At best it could be called a somewhat troubling tribute laid at the feet of formal equality. But there is another, more ‘substantive’ course of action, and this is the one which the cjeu pursued across a series of landmark cases. This course of action was to do away with the test altogether, where discrimination on grounds of pregnancy was alleged, thus overcoming the absence of a male comparator. The Court introduced this new approach in a case called Dekker, where a woman who had applied for the post of instructor

34

J Dine and B Watt, ‘Introduction’ in J Dine and B Watt (eds), Discrimination Law: Concepts, Limitations and Justifications (Longman 1996) 4, referring to the uk cases of Turley v Allders Department Stores [1980] irlr 4 (woman dismissed on grounds of pregnancy not victim of sex discrimination –​sick man comparison used only by Ms P Smith in her dissent) and Hayes v Malleable Working Men’s Club and Institute [1985] irlr 367 (overturning Turley –​sick man comparison unanimously endorsed). 35 Nicola Lacey, ‘From Individual to Group?’ in B Hepple and EM Szyszcak (eds), Discrimination: The Limits of Law (Mansell 1992) 104. 36 H Fenwick and TK Hervey, ‘Sex equality in the single market:  New directions for the European Court of Justice’ (1995) 32 CML Rev 443, 450. 37 Clare McGlynn, ‘Social Policy: Equality, Maternity and Questions of Pay’ (1996) 21(4) EL Rev 327, 332.

Equal treatment irrespective of gender

55

at a training centre for young adults was not appointed on the grounds that she was three months’ pregnant.38 The cjeu stated: [O]‌nly women can be refused employment on grounds of pregnancy and such a refusal therefore constitutes direct discrimination on grounds of sex.39 In Webb, concerning a dismissal from employment rather than a refusal to employ, the Court stated the position in even more forceful terms: [T]‌here can be no question of comparing the situation of a woman who finds herself incapable, by reason of pregnancy discovered very shortly after the conclusion of the employment contract, of performing the task for which she was recruited with that of a man similarly incapable for medical or other reasons … As Mrs Webb rightly argues, pregnancy is not in any way comparable with a pathological condition.40 And, in the same case, Advocate General Tesauro confirmed that this approach was indeed one of ‘substantive equality,’41 which, in his view, was the outcome sought by those who had drafted the directive at issue, namely the Equal Treatment Directive.42 Habermann-​Beltermann,43 about the attempted termination of an employment contract, again on discovery by the employer that the employee was pregnant, completes this important trio of cases in which the cjeu totally rejects the need for comparison (as a prerequisite for a finding of discrimination).44 38 39 40 41 42

43 44

Case C-​177/​88 Elisabeth Johanna Pacifica Dekker v Stichting Vormingscentrum voor Jong Volwassenen (vjv-​Centrum) Plus [1990] ecr I-​3941. ibid para 12. Case C-​32/​93 Carole Louise Webb v EMO Air Cargo (UK) Ltd [1994] ECR I-​3567, paras 24–​25. ibid, Opinion of AG Tesauro, para 8. Council Directive 76/​207/​EEC of 9 February 1976 on the implementation of the principle of equal treatment for men and women as regards access to employment, vocational training and promotion, and working conditions [1976] oj L 39/​40 (‘Directive 76/​207’ or the ‘Equal Treatment Directive’). Case C-​421/​92 Gabriele Habermann-​Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb/​Opf eV [1994] ecr I-​1657. The Habermann-​Beltermann judgment has nevertheless attracted the criticism of some theorists for its endorsement of Germany’s law prohibiting pregnant women from undertaking night work, especially when this may be the only type of work which they can undertake, due to family responsibilities. See Fenwick and Hervey (n 36) 456.

56 Chapter 3 However, other entries in the cjeu’s canon of pregnancy cases are less clear-​cut, particularly where the topic at hand is pregnancy-​related ­illness. In the early case of Hertz v Aldi,45 the Court seemed to decide that a pregnancy-​related illness, manifesting itself after the end of maternity leave, could not be regarded as different from any other illness.46 Since male and female workers were equally exposed to illness, a woman suffering from such a complaint, at such a time, could not allege sexual discrimination if her absences resulted in dismissal (just as they would for a male colleague). But, as carefully explained by Advocate General Ruiz-​Jarabo Colomer in his Opinion in Brown v Rentokil,47 in the operative part of the Hertz judgment, the Court missed out the vital temporal caveat, so that it simply declared that the Equal Treatment Directive does not preclude dismissals which are the result of absences due to an illness attributable to pregnancy or confinement.48 Inevitably, this led the Court to refuse to make a finding of discrimination where a woman was fired as a result of absences caused by a pregnancy-​ related illness manifesting itself before her maternity leave:  Larsson.49 The situation was (at least partially) clarified in the abovementioned Brown case, which explicitly overruled Larsson. Here, the Court faced the fact that, since a pregnancy-​related illness is quite obviously part and parcel of pregnancy itself, to dismiss on the grounds of the former is to all intents and purposes to dismiss on the grounds of the latter, which would be contrary to Dekker, Habermann-​ Beltermann and Webb. Therefore, the Court held that dismissing a woman as a result of a pregnancy-​related illness arising before maternity leave, would be direct discrimination. However, the Court maintained the distinction based on the moment of first appearance of the illness, such that a pregnancy-​related illness arising after the end of maternity leave would be treated no differently from any other illness. 45

46 47 48 49

Case C-​179/​88 Handels-​og Kontorfunktionaerernes Forbund i Danmark [Acting for Birthe Vibeke Hertz] v Dansk Arbejdsgiverforening [Acting for Aldi Marked K/​S] [1990] ecr I-​3979. ibid para 16. Case C-​394/​96 Mary Brown v Rentokil Ltd. [1998] ecr I-​4185. Case C-​179/​88 Hertz (n 45) Operative part of Judgment. Case C-​400/​95 Handels-​og Kontorfunktionærernes Forbund i Danmark, acting on behalf of Helle Elisabeth Larsson v Dansk Handel & Service, acting on behalf of Føtex Supermarked A/​S [1997] ecr I- ​2757.

Equal treatment irrespective of gender

57

The saga of the pregnancy-​related illnesses continued some five months after Brown with Pedersen.50 Here, a Danish law resulted in the situation whereby, over a period of six months, a woman absent from work due to a pregnancy-​ related illness might receive, for the first three months, no pay, and, for the second three months, half pay (assuming, for the purposes of the example, the six months in question were the six months immediately preceding her confinement). Meanwhile, any other ill employee would receive full pay for the entire six-​month period. The cjeu followed Brown in acknowledging that the disorders and complications arising during pregnancy are to be regarded as a ‘specific feature of that condition.’51 Thus, the considerable drop in pay suffered by the woman with the pregnancy-​related illness represented discrimination.52 That said, a woman suffering from so-​called ‘routine … minor complaints,’53 similarly denied pay, was not to be regarded as having been discriminated against, that is, if she would not have qualified as ill under the normal test for paid leave. Two final examples are the 2005 case of McKenna,54 and the 2015 judgment in Rosselle.55 In McKenna, a pregnant employee of the North Western Health Board, who was absent for almost her whole term owing to a pregnancy-​related illness, was downgraded from full pay to half pay after 183 days pursuant to the Board’s sick-​leave scheme. At the moment that her pay was halved, she was still some two months away from taking maternity leave. Furthermore, the absence caused by the pregnancy-​related illness was ‘set against’ her total sick-​leave entitlement (365 days per four years). The Court considered that both matters fell within the scope of Article 141 ec56 and the Equal Pay Directive. However, neither the Article nor the Directive was breached. As regards the halving of the pay, the 50

51 52

53 54 55 56

Case C-​66/​96 Handels-​og Kontorfunktionærernes Forbund i Danmark, acting on behalf of Berit Høj Pedersen v Fællesforeningen for Danmarks Brugsforeninger and Dansk Tandlægeforening and Kristelig Funktionær-​Organisation v Dansk Handel & Service [1998] ecr I-​7327. ibid para 33, citing Case C-​394/​96 Brown (n 47) para 22. Please note that, as (with the exception of the final question) the issue raised in this case was pay rather than treatment, the Court dealt with it under Article 119 eec (subsequently Article 141 ec, and now Article 157 tfeu), and Council Directive 75/​117/​EEC of 10 February 1975 on the approximation of the laws of the Member States relating to the application of the principle of equal pay for men and women [1975] oj L 45/​19 (‘the Equal Pay Directive’). See, for example, Case C-​66/​96 Pedersen (n 50) para 42. Case C-​191/​03 North Western Health Board v Margaret McKenna [2005] ecr I-​7631. Judgment of 1 May 2015, Charlotte Rosselle v Institut national d’assurance maladie-​invalidité (inami) and Union nationale des mutualités libres, C-​65/​14, ECLI:EU:C:2015:339. As it then was –​now Article 157 tfeu.

58 Chapter 3 Court held that it was not necessary to pay women on maternity leave full pay, just as long as such pay as they did receive did not fall to such a level as to ‘undermine the purpose of maternity leave.’57 In reaching this conclusion the Court was inspired by the earlier case of Gillespie, in which it had first considered maternity leave as ‘pay’ within the meaning of Article 157 tfeu (as it is now), as well as the Directive on the Protection of Pregnant Workers, Directive 92/​85.58 As regards the offsetting of the absences against the total sick-​leave entitlement, the Court held that if it was possible for an employer to reduce pay during pregnancy, it must be possible for an employer to allow such offsetting too. In Rosselle, the claimant Ms Rosselle began working as a teacher in Belgium in September 2003, and in September 2008 she was appointed as an established public servant by the Flemish Community. Ms Rosselle obtained a non-​ active status from 1 September 2009 as a salaried employee, and she continued to work as a salaried employee until her maternity leave started on 11 January 2010. She applied to the unm (Union nationale des mutualités libres) for a maternity allowance as from 11 January 2010. This application was refused on the grounds that she had not completed the minimum contribution period, as a salaried employee, to qualify for the maternity allowance. According to the Belgian rules, had Ms Rosselle been dismissed as a public servant and then taken up work as a salaried employee, or had she resigned as a public servant and then done so, her former contributions would have been taken into account. However, her asking for and being assigned non-​active status, shortly before her maternity leave, did not have this effect and a person in Ms Rosselle’s position was required to begin contributing all over again. Ms Rosselle brought an action against the unm before the national court, which then made a reference to the Court of Justice asking whether these rules infringed the abovementioned Directive 92/​85. 57 58

McKenna (n 54) para 50. Case C-​342/​93 Joan Gillespie and others v Northern Health and Social Services Boards, Department of Health and Social Services, Eastern Health and Social Services Board and Southern Health and Social Services Board [1996] ecr I-​475; Council Directive 92/​85/​ EEC of 19 October 1992 on the introduction of measures to encourage improvements in the safety and health at work of pregnant workers and workers who have recently given birth or are breastfeeding (tenth individual Directive within the meaning of Article 16 (1) of Directive 89/​391/​EEC) [1992] oj L 348/​1), ‘Directive 92/​85’. A 2008 attempt to update this Directive has been fruitless, with the Commission finally withdrawing its proposal in 2015: Commission, “Proposal for a Directive of the European Parliament and of the Council amending Council Directive 92/​85/​e ec on the introduction of measures to encourage improvements in the safety and health at work of pregnant workers and workers who have recently given birth or are breastfeeding” com (2008) 637.

Equal treatment irrespective of gender

59

In its decision, the Court emphasized the established nature of a right to maternity leave, and the fact that an adequate maternity allowance is crucial in guaranteeing the health and safety of a pregnant worker. With reference to Directive 92/​85, which states that a Member State may not make an entitlement to maternity pay conditional upon periods of employment in excess of twelve months, the Court questioned whether this Directive precluded a Member State from imposing an additional six-​month contribution period when an established public servant is given a non-​active status in order to work as a salaried employee, even though that public servant has worked for more than twelve months immediately prior to the date of confinement. The Court considered that the wording ‘periods of previous employment’ in  Directive 92/​85 cannot be limited solely to the employment prior to the date of confinement, and a Member State therefore cannot impose an additional six-​month contribution period. To do so would ‘undermin[e]‌the minimum level of protection laid down in [the] Directive.’59 3.4.2 Preliminary Analysis The foregoing case-​law illustrates not only that the running of a ‘like for like’ test can entail problems, but that, in the case of pregnancy, these problems are so insuperable as to necessitate the relinquishing of the test altogether. The abandoning of comparison in the Dekker case is certainly an achievement in the quest for substantive equality,60 and many would regard it as only the first step in a much longer journey. But the cjeu seems reluctant to develop its doctrine of ‘noncomparative discrimination’ much beyond what might be called ‘pure pregnancy’ cases.61 Even in the adjacent field of pregnancy-​related illness, the cjeu has rushed to recover its fallen safety blanket. As Ellis has put it, referring to Hertz v Aldi:

59 60

61

Case C-​65/​14 Rosselle (n 55) 46. ‘By recognising that difference matters in itself and requires changes in the way it is dealt with, in order to eliminate the disadvantage historically resulting from difference, Community law is clearly paying a debt to the feminist legal theory that has disentangled the principle of equality from comparative evaluation’: M Barbera, ‘Not the same? The Judicial Role in the New Community Anti-​Discrimination Law Context’ (2002) 31 Indus LJ 82, 90–​1, footnotes omitted, extracted in A McColgan, Discrimination Law: Text, Cases and Materials (2nd edn Hart Publishing 2005) 26. Although it has expanded the doctrine to include cases of in vitro fertilisation, even prior to the transferring of the fertilised ova to the uterus: Case C-​506/​06 Sabine Mayr v Bäckerei und Konditorei Gerhard Flöckner ohg [2008] ecr I-​1017.

60 Chapter 3 At one level of abstraction, a ‘pregnancy-​related illness’ can be said to be something from which a person cannot suffer but for her female sex; at another level of abstraction, it can be said that ‘illness’ can be suffered alike by either sex.62 The Court in Hertz v Aldi chose to employ the second of these two ‘levels of abstraction,’ thus providing a nice example of what in cinematographical terms might be called ‘zooming out’, that is, the taking into account of fewer attributes rather than more. Ms Hertz and her hypothetical male comparator are both ‘ill’, rather than Ms Hertz being ‘ill and (formerly) pregnant’ and the comparator being ‘ill but not (formerly) pregnant’. In this way, Ms Hertz and the comparator can be regarded as ‘like’, and can receive the same treatment without violating the equality principle. Of course, as has been seen, the matter was a long way from resolved, with the final position being (roughly) that, in the case of pregnancy-​related illnesses arising prior to the taking of maternity leave, the noncomparative approach would be employed, and, in the case of pregnancy-​related illnesses arising after maternity leave, the comparative approach would be employed. Only a few months after the birth, then, a new mother is abstracted out of her specific context (new motherhood, including the risk that a pregnancy-​related illness may materialize). At first glance, the choice of the last day of maternity leave for this switch from ‘incomparable’ to ‘like’ seems a little arbitrary, the coldly legal rationale being, of course, that this day marks the end of the special ‘protection’ afforded to pregnant women.63 Doubts also linger as to how strictly this line-​in-​the-​sand is being policed. In McKenna, for example, it was held that absences resulting from a pregnancy-​related illness could be set against the employee’s total sick-​leave entitlement, without this amounting 62

63

Evelyn Ellis, ‘The definition of discrimination in European Community sex equality law’ (1994) 19(6) EL Rev 563, 567. A social scientist would refer to ‘level[s] of summation:’ J Ritchie and others, Qualitative Research Practice (Sage 2003) 281. This word borrowed from Article 2(3) of the Equal Treatment Directive, subsequently Article 2(7) of Directive 2002/​73/​E C of the European Parliament and of the Council of 23 September 2002 amending Council Directive 76/​207/​E EC on the implementation of the principle of equal treatment for men and women as regards access to employment, vocational training and promotion, and working conditions [2002] OJ L 269/15 (the ‘New Equal Treatment Directive’). Most recently the word can be found at Article 28(1) of Directive 2006/​54/​E C of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast) [2006] OJ L 204/23 (the ‘Recast Directive’).

Equal treatment irrespective of gender

61

to discrimination.64 Although Hertz and Brown are cited as authority, Ms McKenna’s absences were quite obviously prior to her maternity leave. For the employer to be allowed to include these in some future aggregate calculation, pursuant to the sick-​leave scheme, is surely tantamount to declaring that absences on the grounds of pregnancy-​related illness, prior to maternity leave, are no different from (in other words, are ‘like’) absences on the grounds of any other illness. This seems to be a return to the position in Larsson, which Brown had supposedly overruled. The Court’s approach to the distinction between pregnancy-​related illness and illness unrelated to pregnancy has been erratic in other ways. Brown appeared to state once and for all that a woman suffering from a pregnancy-​ related illness and a man suffering from some other illness were not alike, as regards vulnerability to dismissal (at least until the abovementioned line-​in-​ the-​sand was reached). But, since Pedersen, it has emerged that a woman suffering from a pregnancy-​related illness and a man suffering from some other illness are alike, as regards right to pay, and level of pay. It is worth considering the two cases in the light of the table at Appendix I. In Brown, Ms Brown (Comparator 1)  faced dismissal after twenty six weeks of absence, pursuant to a contractual clause. Any male employee (Comparator  2) would face exactly the same rule. This was like treatment, for the purposes of Column 2. So if the Court wanted to arrive at a finding of discrimination, it had to declare Comparator 1 and Comparator 2 to be unlike. The rule was then ‘applied in the same way to different situations,’65 thus constituting direct discrimination.66 However, in Pedersen, the factual set-​up was different. Here, Ms Pedersen (Comparator 1) faced three months on no pay and three months on half pay, in the six months leading up to her confinement, pursuant to Law No 516 of 23 July 1987. Meanwhile, a male employee who fell ill (Comparator 2) would enjoy six months on full pay, pursuant to the same law. This was unlike treatment. So, now, if the Court wished to arrive at a finding of discrimination, it was going to have to declare the two Comparators to be like. In fact the Court used noncomparative discrimination to get around this problem.67 However, the important point is that the cjeu could not repeat its remark from Brown about those suffering from pregnancy-​related illnesses and those suffering from other illnesses being in ‘different situations,’ because, given the factual

64 65 66 67

Please note that this finding was contrary to the Opinion of Advocate General Léger. Case C-​394/​96 Brown (n 47) para 31. ibid para 32. Case C-​66/​96 Pedersen (n 50) para 35.

62 Chapter 3 situation of the Pedersen case, this would amount to unlike being treated in unlike fashion, which could not, ipso facto, be discrimination.68 The distinction between a pregnant (female) worker and a non-​pregnant (male) worker has also been problematic. In Gillespie,69 for example, where seventeen plaintiffs who took maternity leave during 1988 complained that they did not enjoy the benefit of a (back-​dated) pay rise announced at the end of that year, the Court held that no provision of ec law ‘required that women should continue to receive full pay during maternity leave.’70 However, a woman on maternity leave should receive a pay rise awarded during that period, because ‘[t]‌o deny such an increase … would discriminate against her purely in her capacity as a worker since, had she not been pregnant, she would have received the pay rise.’71 Why only apply this reasoning to the second limb of the judgment (the pay rise), and not also to the first limb (the ‘normal’ pay)? In other words, in the same case, two comparisons are run. The same person (let her be A) is both ‘pregnant worker,’ and thus incomparable with B (a man), and ‘worker,’ and thus like B (such that, when treated in unlike fashion, as when B benefits from a pay rise but A does not, there is discrimination). Thus, the sort of adjustment referred to at the beginning of section 3.1 can be seen clearly here, carried out in the space of a single case. Turning finally to the doctrine of noncomparative discrimination itself, this has been discussed at length in other works.72 From a linguistic point of view, the doctrine seems at the very least paradoxical. ‘Discrimination’ comes from the Latin word discrimen, meaning ‘distinction.’73 A  thing cannot be 68

69 70

71 72

73

Indeed, by the time of the McKenna case, the Court had to acknowledge that there was one rule for dismissals, and another rule for pay:  see Case C-​191/​03 McKenna (n 54) para 58. The (implicit) equation of those suffering from pregnancy-​related illnesses and those suffering from other illnesses in Pedersen (as regards pay at least) led, in that case, to a levelling up of the situation in favour of those suffering from pregnancy-​related illnesses. But note how, by the time of McKenna, this same equation could be used to level down, that is, to subject women with pregnancy-​related illnesses to worse treatment than they might otherwise have expected. Case C-​342/​93 Gillespie (n 58). ibid para 20. The Court was clearly inspired here by the Directive on the Protection of Pregnant Workers (n58), which only calls for the payment of an ‘adequate allowance’ (Article 11). However, this Directive did not apply ratione temporis to the Gillespie case. ibid para 22. For example, T Macklem, Beyond Comparison:  Sex and Discrimination (cup 2003) 16–​19; Elisa Holmes, ‘Anti-​Discrimination Rights Without Equality’ (2005) 68(2) mlr 175, 186–​7. CT Onions (ed), The Shorter Oxford English Dictionary (3rd rev edn, Clarendon 1973) 564.

Equal treatment irrespective of gender

63

distinguished by itself, but only (it is submitted) in relation to something else.74 Put another way, anti-​discrimination law proceeds along the lines of what has been called ‘pair-​wise comparison’. ‘Singleton-​wise comparison’ would simply not work; a thing cannot be compared on its own.75 However, many words have evolved away from their original meanings, so perhaps one should not set too much store by linguistics. However, other problems remain. If a pregnant woman is incomparable, making any adverse treatment an instance of sex discrimination, what happens when an employer favours a non-​pregnant woman over a pregnant one? The latter has suffered sex discrimination even though both the advantaged party and the disadvantaged party are female. There may thus be a need to differentiate sex discrimination from pregnancy discrimination. Perhaps the last word should be given to Ellis, who eloquently expresses her opposition to the abandonment of comparison as follows: It is submitted that an element of comparability is important to the component of adverse impact; if direct discrimination is defined simply as ‘nasty treatment’ on the ground of sex, enormous discretion is placed in the hands of courts and tribunals, who remain overwhelmingly male in composition, to decide what is to the detriment or advantage of complainants, the majority of whom are female. For reasons of objectivity, it is preferable if the adversity of the treatment received by the complainant is measured by means of a comparison with the treatment received or receivable by a member of the opposite sex, placed in broadly the same circumstances as the complainant.76 3.4.3 A Walzerian Analysis A Walzerian perusal of the same case-​law reveals that complex equality can help to render the judgments more consistent and less confused, and can also offer a solution to the ‘noncomparative discrimination’ problem. The starting point for a Walzerian scrutiny of the Dekker case would be the ascertaining, by the relevant distributive community, of the shared meaning of 74

75 76

See, for a strong defence of this position, R Wintemute, ‘When is Pregnancy Discrimination Indirect Sex Discrimination?’ (1998) 27(1) Indus LJ 23: ‘Claims of discrimination without comparison are impossible’ (at 25). Unless the comparison is, for example, between the thing as it is and the thing as it might be: Macklem (n 72) 141. Ellis (n 62)  571–​2. For criticism of this statement, see, for example, GF Mancini and S O’Leary, ‘The New Frontiers of Sex Equality Law in the European Union’ (1999) 24(4) EL Rev 331, 338.

64 Chapter 3 the post of instructor. This would likely include some element of aptitude or competence (for example, in the training of young adults), but would almost certainly not include any consideration of whether or not the post-​holder was pregnant. The inclusion of such a consideration in the fundamentum distributionis, then, breaches Walzer’s rule that the distributive principle of any good must follow from its shared meaning. Complex equality having been contravened, the Walzerian analyst, like the Court, would find for Ms Dekker. The same result, mutatis mutandis, would be achieved in Webb and Habermann-​Beltermann. The holy trinity of the cjeu’s pregnancy case-​law remaining untouched, then, how would the Walzerian analyst deal with the other cases mentioned above? Here the results are not so clear-​cut, as they turn on shared meanings which are hard to forecast. In Brown, for example, the distribuend is continued employment with Rentokil.77 The company seems to be asking two questions of potential distributees, the first question being: do you have the ability to do the job? The second question, however, is: are you free from pregnancy-​related illnesses?78 The distributive community would probably see the meaning of the good at hand in terms of the first question only, and this would then determine the correct fundamentum distributionis: continued employment should go to those who are able to do the job. As in Dekker, inclusion of an alien element within the fundamentum distributionis would flaw the distribution and render it complexly inegalitarian; pregnancy would effectively become a negative dominant.79 The same two questions are asked of the employee by Aldi in Hertz v Aldi, the only difference being that in Brown the illness arose before maternity 77 78

79

Or, looked at the other way round, dismissal. A possible third question is, are you available to do the job? The distributive community might think it was fair that availability to do the job formed part of the shared meaning thereof, such that, in a case like Mahlburg, statutorily-​enforced non-​availability on grounds of pregnancy might be legitimately taken into account:  Case C-​207/​98 Silke-​ Karin Mahlburg v Land Mecklenburg-​Vorpommern [2000] ECR I-​549. However, it also seems fair that a short-​term absence should not be allowed to defeat a long-​term contract (this was the reasoning in Habermann-​Beltermann, followed in Mahlburg). This is a classic clash-​of-​meanings, then, as between the distributor and the distributee. It is true that, under orthodox complex equality, the final word would rest with the distributive community, possibly leaving pregnant women hostage to factions within a community which were hostile to working mothers. However, under the alternative proposal of mediated complexity, this balancing act would be the sole job of the Bench, which might even be able to operate a kind of ‘hardship rule’ to decide between conflicting meanings. Clearly, an entire lost career would outweigh a few months of lost revenue. See below section 8.3. Or a positive dominant if the distribuend was taken to be dismissal.

Equal treatment irrespective of gender

65

leave, while in Hertz it arose after maternity leave. Although the Court saw a difference, it is not clear that the Walzerian analyst would, meaning that the result for the latter would be identical to that reached in Brown, mutatis mutandis. Returning to Brown, it is just conceivable that some members of the community might argue that those with pregnancy-​related illnesses were too ill to do the job, ergo were not able to do the job; in other words, even without the second question, they would argue, the potential distributee’s state of health is part of the meaning of the good and, consequently, a part of the way it should be distributed. However, this definition of employment (continued or ab initio) is problematic, and overlaps with other debates which the same Forum would have had, or would have to have in the future, about the very meaning of a job, debates which would have to take in at least a hundred years of developments in social protection, employment law, and workers’ rights generally. In a nutshell, a job is not just an acquisition of labour-​power by the employer to achieve his or her (economic) ends, but is also a means of livelihood (usually the sole means of livelihood) for the employee and his or her family. It is unjust to simply cast him or her aside if he or she falls ill, not to mention dangerous, as ill employees struggle to conceal their medical problems and work on, for fear of losing their livelihoods. Paid sick leave is now mandatory for employers, and so dismissal on the grounds that someone has fallen ill is illegal.80 Going back to the Brown case, then, those who included the potential employee’s state of health in the meaning of continued employment would have to do so across the board: no job would be open to the ill, and all employees becoming ill during their employment could be sacked on the spot. This is very unlikely to be a viewpoint widely held within the distributive community as a whole. The most likely scenario, then, is that the distributive criterion adopted by Rentokil would be found to be at odds with the shared meaning of the distribuend, meaning that the Walzerian analyst, like the Court, would find for Ms Brown. Pregnancy-​related illnesses have nothing whatsoever to do with the performance of pest control services. This logic would seem to be applicable whether the pregnancy-​related illness manifested itself before or after confinement/​maternity leave, but it is crucial to know if pregnancy-​related illnesses are to be treated the same as other types of illnesses, or if they are to be accorded special protection. Could an employee’s contract be terminated on the strength of periods of absence caused by a pregnancy-​related illness, and, if so, would the requisite qualifying 80

That is, immediate dismissal; as time goes on, dismissal may become a legitimate option for the employer. See the paragraph-​after-​next.

66 Chapter 3 period begin to run from the beginning of the first leave of absence taken after the illness had manifested itself? Assuming that that was the case, the employee’s dismissal after, say, 26 weeks would be entirely legal, even though it would be, to all intents and purposes, dismissal on the grounds of pregnancy. It would be for the distributive community to work out these arrangements, but, as at the end of the paragraph-​before-​last, it seems highly unlikely that the entire community, made up as it will be of men and women a large number of whom will want to start their own families, would agree to meanings of employment or continued employment which stipulate that employees must have a clean bill of health, where the term ‘clean bill of health’ includes freedom from pregnancy-​related illnesses. While illnesses may eventually gain a legitimate status within the sphere of employment (much like intelligence within the sphere of education), such that any eventual inequalities based on ill-​health would themselves be legitimate (‘small inequalities’), any mention within the discussions of pregnancy should sound an urgent alarm. ‘This is a major boundary –​a boundary with the sphere of family and with the sphere of rights, to name but two’, such an alarm should say, ‘not just now but always.’ Pregnancy’s status within the sphere of employment should be illegitimate always.81 This would mean that pregnancy-​related illness could never feature in a fundamentum distributionis without causing a boundary breach and flawing the distribution. To answer the question posed at the beginning of this paragraph, this would hopefully give special protection to those suffering from pregnancy-​related illnesses not just before confinement/​maternity leave, but afterwards too. While it might be acceptable to fire an employee suffering from a ‘regular’ illness after, say, 26 weeks, it would not be acceptable to fire an employee suffering from a pregnancy-​related illness after that period, or (subject to the distributive community’s agreement) at all. This would separate those suffering from ‘regular’ illnesses and those suffering from pregnancy-​ related illnesses, and sick leave taken by the latter could not be used as grounds for termination of the employment contract. Considering Pedersen, here the very comparison being made was between a woman suffering from a pregnancy-​related illness, and any other (hypothetical) ill employee. Complex equality would require that the woman suffering from the pregnancy-​related illness receive the same salary as the woman (or man) suffering from a ‘regular’ illness, namely full pay. This conclusion mirrors that reached by the Court. 81

Except in the very rare situation that the job at hand required the post-​holder to be pregnant, in other words, that pregnancy was brought into the shared meaning of the distribuend.

Equal treatment irrespective of gender

67

In Gillespie, the distributive community’s shared understanding of a back-​ dated pay rise, and its shared understanding of pay itself, would determine how these things are to be distributed. These understandings may or may not make exceptions for those on maternity leave. If they do, then pregnancy’s presence within the distributive sphere becomes legitimate. But if they do not, then, as explained above, it would be illegitimate, representing a breach of boundaries and causing the distributions at issue to be flawed. What seems unlikely, though, is that one shared meaning would make special allowance for pregnant women, while the other would not. Thus use of a Walzerian approach to equality would avoid the sort of fractured result which the Court arrived at, finding discrimination in the case of the back-​dated pay rise but not discrimination in the case of pay itself. Turning to McKenna, it again seems probable that the distributive community as a whole would omit the reference to pregnancy, in ascertaining its shared meaning of pay and benefits, which North West Health Board includes in its distributive criterion. This difference of opinion would lead, as before, to an infringement of complex equality. It is worth noting how this verdict clashes with that of the Court, which held that inferior pay for those on maternity leave, and by extension for those absent by reason of a pregnancy-​related illness, was not discriminatory. Another very interesting point about McKenna is that it shows nicely how Walzer’s theory can be used non-​comparatively, that is, a complainant does not need a comparator in order for their situation to be assessed –​breaches of complex equality (or otherwise) can be deduced form observing the situation and circumstances of one person alone. Ms McKenna cannot point to any particular comparator. There are options, naturally, such as a woman on maternity leave, a non-​pregnant woman absent by reason of ‘traditional’ illness, a man absent by reason of ‘traditional’ illness, and so on. However, it is not obvious which of these the comparator should be. Walzer’s theory, meanwhile, allows her predicament to be exposed to the light of day without comparison being necessary, and the boundary breach, between the spheres of family and employment,82 is immediately evident.83 Her ‘nasty treatment,’ to use Ellis’ phrase,84 can be seen, and without undue discretion having to be afforded to the (possibly male) judges. 82 83

84

Or rights and employment. Hopefully, then, Walzer’s theory answers Patricia Cain’s call ‘to move beyond equality, which inevitably compares women to men, and to focus on women themselves’: PA Cain, ‘Feminism and the limits of equality’ (1989–​1990) 24 Ga L Rev 803, 806. See supra, text accompanying n 76.

68 Chapter 3 Finally, if one considers Rosselle through a Walzerian lens, it is maternity a­ llowance itself which would fall to have its meaning ascertained by the distributive community. It seems absurd that this community, if it found resignation from the public service and dismissal from the public service to be outwith the sphere of maternity allowance, would find a voluntary self-​assignment as ‘non-​active’ to be within it. Taking the Standard Contingent Reply as read, the meaning of maternity allowance might be a quantity of money to enable the expectant and/​ or new mother to maintain an adequate standard of living while on maternity leave so that she was not forced by financial concerns to either truncate this leave or not take it in the first place, thus jeopardizing her health and safety and that of the baby. Of course a requirement to have made contributions over a certain minimum period of employment could be imposed, and thus the ‘good’ of having made these contributions would be one which potential distributees would be required to bring into the sphere with them in order to qualify. This would not be a dominant in the Walzerian sense as it would be part of the meaning of the thing distributed. However, the fact that the relevant period of employment had or had not been brought to an end by resignation, by dismissal or by voluntary self-​reassignment to ‘non-​active’ status would seem unrelated to the grant of the allowance; in either event, the employee/​public servant could show that they had ‘paid in’ sufficiently to be able to lay claim to an award, and that –​the need to have first been a contributor before one could be a recipient –​is what would matter. Thus, Ms Rosselle’s six or so years of prior teaching, with their attendant contributions into the Belgian Sickness Insurance Fund, would be a perfectly adequate ‘entry token’ for her to be able to access the distribution of maternity allowance in this case, with or without their having been brought to an end by the obtaining at her own request of non-​active status after she fell pregnant. 3.5 Paternity Leave 3.5.1 The CJEU’s Case-​Law When the Recast Directive was adopted in 2006,85 with the intention of ­consolidating all of the existing secondary legislation on the topic of sex discrimination, the Framework Agreement on Parental Leave, Directive 96/​34,86 85 86

Directive 2006/​54 (n 63). Council Directive 96/​34/​EC of 3 June 1996 on the framework agreement on parental leave concluded by unice, ceep and the etuc [1996] oj L145/​4 (‘Directive 96/​34’). Since replaced by Council Directive 2010/​18/​EU of 8 March 2010 implementing the revised Framework Agreement on parental leave concluded by businesseurope, ueapme,

Equal treatment irrespective of gender

69

was left out. This was because it was felt that to include this directive within the consolidation exercise would require a different and more onerous Treaty base than Article 157(3) tfeu.87 However, the Recast Directive did protect employment rights during and following maternity leave, and, if provided, paternity leave, at Articles 2(2)(c), 15 and 16. What it did not do, though, was require the Member States to provide paternity leave. For a male claimant seeking a right to such leave, then, the choice was either to plead Directive 96/​34, or else to fall back on the equal treatment guarantee in the Recast Directive, or its predecessor, the Equal Treatment Directive.88 The leading case is Roca Álvarez,89 a judgment from 2011. Mr Roca Álvarez worked for Sesa Start, a Spanish company, and his wife was self-​ employed. When he applied to the company for paternity leave, his application was refused on the grounds of his wife’s not being an employee within the meaning of Article 37(4) of the Spanish Workers’ Statute; the mother’s being employed was an essential criterion of the entitlement to paternity leave in Spanish law. Relying on the Equal Treatment Directive,90 Mr Roca Álvarez argued that this was a breach of the principle of equal treatment and non-​discrimination on grounds of sex, and the national court made a reference. Advocate General Kokott recalled that, according to Article 5(1) of the Equal Treatment Directive, the application of the principle of equal treatment with regard to working conditions meant that men and women were to be granted the same conditions without discrimination on grounds of sex. She was satisfied that the Spanish law in question, which also permitted female employees to take one hour off to breastfeed a child under 9 months, was concerned with working conditions. She was also satisfied that the discrimination at issue was on the grounds of sex. The Spanish law only granted a reduction in working time to women; male workers at best had a derived right. She continues:

87 88 89 90

ceep and etuc and repealing Directive 96/​34/​EC [2010] oj L68/​13. The Commission has proposed to repeal and replace the latter Directive:  Commission, ‘Proposal for a Directive of the European Parliament and of the Council on work-​life balance for parents and carers and repealing Council Directive 2010/​18/​EU’ COM(2017)253 final. At the time of writing, this proposal is still undergoing Parliamentary scrutiny. As it is now. This piece of information taken from D Edward and R Lane, European Union Law (Elgar 2013) 871. Directive 76/​207 (n 42). Case C-​104/​09 Pedro Manuel Roca Álvarez v Sesa Start España ett sa [2010] ecr I-​8661. The Recast Directive did not apply ratione temporis to the facts of this case.

70 Chapter 3 In principle, men and women in the circumstances of this case too are in a comparable situation, as the principal purpose of the time off work at issue in this dispute is […] the taking care of a child. Since the provision in question confers the right expressly on female workers alone, a direct difference in treatment on grounds of sex consequently arises.91 She goes on to consider the exceptions contained within the Directive, but rejects them all. The Spanish legislation focuses on the child, and, as such, his or her ‘need to be taken care of can be met by allowing either the mother or the father time off work.’92 Rejecting a final argument based on positive action, the Advocate General concludes that the Spanish law offends against the principle of equal treatment as provided for in the Equal Treatment Directive. The cjeu readily agrees with Kokott AG’s finding that male and female working parents are comparable,93 with the national court having even admitted that the legislation at issue has become ‘detached from the biological fact of breastfeeding.’94 The Court is thus able to distinguish cases like Webb95 and Brown,96 where the Court had used the concept of noncomparative discrimination (substantive equality) to protect the special relationship between a woman and her child over the period which follows childbirth. The Spanish law at stake here, as interpreted by the Spanish courts, drew no distinction between breastfeeding and bottle feeding, and thus a pair-​wise comparison between working mothers and working fathers was appropriate on this occasion. The Court flatly rejected an argument from Spain that it was more difficult for a mother to enter into and remain in the world of work, therefore justifying differential treatment as between mothers and fathers. If anything, this law was liable to perpetuate a traditional distribution of the roles of men and women by keeping men in a role subsidiary to that of women in relation to the exercise of their parental duties.97

91 92 93

94 95 96 97

Case C-​104/​09 Roca Álvarez (n 89) Opinion of AG Kokott, para 30. ibid, Opinion of AG Kokott, para 40. Fredman notes that this lack of hesitation on the part of the Court ‘[w]‌ithout engaging in the contortions of earlier case-​law on the comparator’ is one of the ways in which this case ‘breaks new ground’: S Fredman, ‘Reversing roles: Bringing men into the frame’ (2014) 10(4) Int JLC 442, 453. Case C-​104/​09 Roca Álvarez (n 89) Opinion of AG Kokott, para 28. Case C-​32/​93 Webb (n 40). Case 394/​96 Brown (n 47). Case C-​104/​09 Roca Álvarez (n 89) para 36.

Equal treatment irrespective of gender

71

The law was thus precluded by the Equal Treatment Directive. A similar situation arose in the later case of Maïstrellis.98 Here, under Greek law, a pregnant judge is entitled to leave before and after childbirth, according to the provisions that apply to state civil servants. She may also request paid leave for up to nine months in order to raise her child. Male judges who become fathers could in principle also be entitled to parental leave, by analogy with the legislation applicable to women. However, Article 53(3), third paragraph, of the Civil Service Code, which, in the absence of specific provisions for judges, was applicable by analogy, included a restriction that, if the male judge’s wife did not work, he could not receive parental leave (unless she was ill). The applicant was a Greek judge. In December 2010, he applied for paid leave to care for his child born in October. On the grounds of Article 53(3) of the Civil Service Code, his request was rejected. The applicant challenged this decision by stating that the article was not compatible with Directive 96/​3499 or indeed with the Directive on the Protection of Pregnant Workers, Directive 92/​85.100 He commenced proceedings and the national court made a preliminary reference to the cjeu; in its reference, the national court also invoked Directive 2006/​54, the Recast Directive.101 Advocate General Kokott disagreed with the Greek government’s approach, which interpreted ‘working parents’ as meaning that both parents had to work in order to be entitled to parental leave. The Advocate General further emphasized that the right to parental leave is not a child’s right but the right of each parent, and therefore whether or not child support is guaranteed in the absence of parental leave is not the central question. The European legislation precluded parents from being denied this right altogether. The Greek article was therefore in breach of Directive 96/​34, and represented direct discrimination under Directive 2006/​54. The Court reached the same conclusion as Kokott AG, making the following interesting remark about context: [I]‌n interpreting a provision of eu law, it is necessary to consider not only its wording, but also the context in which it occurs and the objectives pursued by the rules of which it is part.102 98 Judgment of 16 July 2015, Konstantinos Maïstrellis v Ypourgos Dikaiosynis, Diafaneias kai Anthropinon Dikaiomaton, C-​222/​14, ECLI:EU:C:2015:473. 99 Directive 96/​34 (n 86). 100 Directive 92/​85 (n 58). 101 Directive 2006/​54 (n 63). 102 Case C-​222/​14 Maïstrellis (n 98) para 30. The Court took a similar approach to Directive 92/​85 in Case C-​65/​14 Rosselle (n 55).

72 Chapter 3 Investigating the context of Directive 96/​34, the Court agreed with the Advocate General that parental leave was an individual right, which each parent receives, and that it may not be denied altogether. Thus, in the case of a male judge, it may not be made conditional on his wife’s working. The Court ended its ruling with a consideration of Directive 2006/​54, and, like Advocate General Kokott, it was of the view that Greece’s rule here represented direct discrimination within the meaning of that directive, as a male judge with a (healthy) non-​working wife would not receive the benefit, while a female judge with a (healthy) non-​working husband would receive it. 3.5.2 A Walzerian Analysis It can be seen from Roca Álvarez that the Aristotelian approach works here to some extent as likes or comparables are being treated unlike from almost every angle of observation. However, one can also note the inconsistency with the earlier case-​law103 where the fact of being pregnant or of being a mother rendered the party concerned non-​comparable. Thus the malleability of the Aristotelian test, and the fear that it can be manipulated from case to case, remains problematic. On the other hand, some critics have noted the distinction as between this case and the earlier ones as a positive, not as a negative: It is worth noting the distinction between a benefit designed to offset the occupational disadvantages which arise for female workers as a result of being absent from work during the period following childbirth and a benefit designed to offset the occupational disadvantages which arise for female workers as a result of having brought up a child.104 The difficulty with Walzer here, though, is determining whether the parental leave on offer in the Spanish law is designed to help the baby, or to help the mother to stay in employment, or indeed something else. Although the government had said that the leave was designed to help the mother, the Court had said that it was designed to help the baby. As always, it is the distributive community which must resolve the dilemma as it is the distributive community which decides the social meaning of this kind of leave. Given the Standard Contingent Reply, one might speculate that the members of the community here would conclude that the meaning of the leave was to provide vital assistance to the baby in its first months of life. However, it cannot be ruled out 1 03 That discussed in section 3.4 above. 104 SA Flindall, ‘In the spirit of equality: breastfeeding leave for men’ (2010) 11 European Law Reporter 348, 350.

Equal treatment irrespective of gender

73

that, that decided, they may go on to specify that it would therefore not be available to a male worker whose female partner did not work and to a female worker whose male partner did not work, as these individuals would not be able to show that the baby for whom they were parentally responsible was in need of such assistance from them. This would produce the difference that, where under the Spanish law, pre-​Roca Álvarez, a female employee with a non-​ working male partner would have received the benefit but not vice versa, and post-​Roca Álvarez, a female employee with a non-​working male partner would have received the benefit and vice versa, in the Walzerian analysis, it is possible that neither a male nor a female employee with a non-​working partner would have received the leave on offer, as they would not hold the ‘entry token’ to the relevant distributive sphere, namely, responsibility for a baby in need of the assistance described. However, it must be stressed that the eventuality described in the previous paragraph would only arise where one or other partner was not working at all, not, as in the Roca Álvarez case itself, where a partner was self-​employed. A self-​employed person would have the same difficulty in providing the child with the requisite assistance as an employed person, and so their partner would still be able to show the entry token mentioned above, and would still be able to receive the leave sought, in casu, Mr Roca Álvarez would have received the benefit under a Walzerian regime105 because his wife was self-​employed. The foregoing analysis acts as a reminder that there can be more than one way to turn an asymmetry (the pre-​Roca Álvarez situation) into a symmetry (the post-​Roca Álvarez situation), with the Court of Justice recommending the granting of the leave to working parents with partners who were self-​employed and (presumably) working parents with partners who were not employed at all, but with –​perhaps –​the Walzerian judge only granting it to the first of those categories. The ramifications of a strict reading of the meaning of parental leave as a benefit to the baby are felt more harshly when one turns to Maïstrellis. Here it will be seen that, were that indeed to be the meaning arrived at by the distributive community, the outcome for Judge Maïstrellis would have been different under a Walzerian bench than it was under an Aristotelian one, with the judge still unable to claim parental leave as long as his wife was not working. The only way around this would be for the distributive community to come to the opinion that the meaning of parental leave also involved the sense of 105 Standard Contingent Reply taken as read, but a presumption having been made for the sake of argument that the distributive community reached the consensus that the meaning of a period of parental leave was to assist the baby in its first months of life.

74 Chapter 3 an opportunity for whichever parent was taking the leave to bond with their newborn child. Then their entry into the distributive sphere without being responsible for a baby in especial need of their care would not matter, as to merely be responsible for a baby, without more, would be to possess the relevant (non-​tyrannical) good granting the holder entitlement to the distribuend in question. It is noteworthy that in the first half of its judgment, the part pertaining to Directive 96/​34, the Court of Justice itself takes a contextual approach in order to discern who might or might not be ‘entitled’106 to the benefit at hand, not comparatively (‘A is entitled to the benefit therefore B must be entitled to it’), but non-​comparatively and contextually (‘A is entitled to the benefit because the context demands it’). The Court does not in fact deploy Aristotelian reasoning until the second half of the judgment concerning Directive 2006/​54. This shows a slight move away from simplicity and towards complexity, although it must still be borne in mind that the bench restricted itself to an interpretation of the legislature’s understanding of the context, not that of all families wherein one or more parent is working. One of the conclusions which can be drawn from this line of cases, then, is that the Walzerian emphasis on the meaning of the good being distributed, and its sole role in determining the correct distibutees, can occasionally spell a diminution in the number of potentially successful claimants as Walzer’s goal is not simply, even obsessively, to justify107 any anomalous distribution, as Aristotle’s seems to be, but, as quoted in Section 3.2 above, to ‘get […] the unevenness right.’108 His contribution is sense, not (necessarily) symmetry. 3.6 Pensions 3.6.1 The CJEU’s Case-​Law For many years, Member States finding themselves before the Court tried to rely on derogations vis-​à-​vis national pensionable ages, for example, at Article 7(1)(a) of the Social Security Directive,109 and at Article 9(1)(a) of the (at that 1 06 107 108 109

The Court’s word at Case C-​222/​14 Maïstrellis (n 98) para 41. In the mathematical or philosophical sense. M Walzer, Spheres of Justice (n 9) 260. Council Directive 79/​7/​EEC of 19 December 1978 on the progressive implementation of the principle of equal treatment for men and women in matters of social security [1979] oj L 6/​24 (‘Directive 79/​7’ or ‘the Social Security Directive’).

Equal treatment irrespective of gender

75

time) new Occupational Social Security Directive.110 The cjeu on the other hand largely ignored this and often seemed indirectly to be impugning discrimination in a State’s determination of pensionable age for the purposes of granting old-​age pensions. However, a crucial complication in the cjeu’s case-​ law on pensions emerges where the difference brought before the Court is not a difference in normal pensionable ages for men and women, but a second difference which subsumes the first. This complication can be seen in the early case of Burton v British Railways Board,111 in which the Board undertook an ‘internal reorganization,’ as a result of which it offered voluntary redundancy to male workers aged 60 and female workers aged 55. Mr Burton, aged 58, applied for voluntary redundancy but was rejected. Performing the ‘like for like’ test in order to ascertain whether or not there had been discrimination, the Court faced two ways of looking at the Board’s offer. First, for a male worker to be eligible for the payment of the voluntary redundancy benefit, he must have reached the age of 60, while for a female worker to be eligible for the payment of a voluntary redundancy benefit, she must have reached the age of 55 (the ‘first formulation’). On the first formulation, one could simply declare the offer to be a case of like workers, treated in unlike fashion. However, one could also conclude that both kinds of worker were eligible for the payment of the voluntary retirement benefit if they applied within 5 years preceding their normal minimum age of retirement. On this, second formulation, which the Court in Burton preferred, both kinds of worker were treated the same, that is, likes were treated alike. There was thus no discrimination. The second formulation could be called a ‘relative formulation’. While the first formulation is based on two, fixed numbers (60 and 55), the second formulation is based on one number (5), but this number is relative to two other, fixed numbers (65 and 60). The comparison, which appears to disclose no discrimination, is thus in reality relative to an existing distribution, which is discriminatory.112 It is easy to think of other examples. Suppose two men or women, A and B, both go to see a doctor for a vaccination prior to a foreign trip. The doctor 110 Council Directive 86/​378/​EEC of 24 July 1986 on the implementation of the principle of equal treatment for men and women in occupational social security schemes [1986] oj L 225/​40. This was later updated by Council Directive 96/​97/​EC of 20 December 1996 amending Directive 86/​378/​EEC on the implementation of the principle of equal treatment for men and women in occupational social security schemes [1997] oj L 46/​20. Both of these have now been overtaken by the Recast Directive. 111 Case 19/​81 Arthur Burton v British Railways Board [1982] ecr 554. 112 A relative formulation does not in fact have to contain any numbers as long as the distribution is based, for all comparators, on a single concept, the which concept on closer

76 Chapter 3 wishes to see the patients one year after the first injection for a booster. It would not make sense to make the second appointments of A and B on the same date, if the first appointments of A and B were on different dates. The original disparity here is functional. Even a one-​hundred-​metre race does not require all runners to finish at precisely the same spot, just one hundred metres from their respective starting points. Calling it a ‘one-​hundred-​metre race’ disguises the underlying disparities among the starting points; but, again, these disparities are functional. So the problem in Burton is not the taking into account of less or more attributes. There are two attributes per worker whichever of the two ‘ways of looking’ is employed: their gender, and then, either, their actual age of qualification for the voluntary redundancy benefit, or, the relative formulation of ‘five years from normal retirement age’. But, if the latter, this ‘attribute’ smuggles into the comparison a (disguised) difference. It is submitted that this difference is illegitimate,113 making the use of the disguise effectively a trick. In the doctor and race examples above, the hidden differences are functional and the use of the relative formulation is justified by a desire for clarity and ease. The (hidden) different numbers are actually irrelevant. It is the (given) same number which is relevant. The same cannot be said for Burton.114 Four years after Burton came Roberts v Tate & Lyle.115 Here, Ms Roberts found herself made redundant at the age of 53 under a mass redundancy; she was a member of Tate & Lyle’s ‘contracted out’ occupational pension scheme (that is, contracted out of the State retirement pension scheme). Under the final severance inspection varies significantly from one class of comparator to the next. For this reason, the use of relative formulations is a popular method of practicing indirect discrimination. 113 One of the supposed purposes of the pensionable age difference in the uk was apparently ‘to enable a married couple to retire at the same time (it being assumed that husbands were normally slightly older than their wives)’: Evelyn Ellis, eu Anti-​Discrimination Law (Oxford EC Law Library, OUP 2005) 380. Such an assumption does not function in a 21st century world where age differences between brides and grooms vary considerably. It also takes no account of same-​sex partnerships, or those workers who for whatever reason are single when they retire. That said, other purposes have been suggested for the difference. In the Pension Reform case of 1987, for example, the German Constitutional Court held that women should be allowed to retire earlier than men because of the ‘double burden’ which they endured from pregnancy and child care:  74 BVerfGE 163 (1987). See further DP Kommers, The Constitutional Jurisprudence of the Federal Republic of Germany (2nd edn, Duke up 1997) 295. 1 14 For further criticism of the Burton decision, see, for example, T Millett, ‘Sex equality: the influence of Community law in Great Britain’ (1986) 6 YB of Eur L 219, and ‘European Community law: sex equality and retirement age’ (1987) 36(3) iclq 616. 1 15 Case 151/​84 Joan Roberts v Tate & Lyle Industries Limited [1986] ecr 703.

Equal treatment irrespective of gender

77

terms, all employees made redundant were to be offered either a cash payment or an early pension, in other words, pensions were immediately payable to both men and women over the age of 55. Ms Roberts pleaded discrimination. Roberts v Tate & Lyle acts as a kind of obverse to Burton. Clearly, what Ms Roberts wanted was for the cjeu to view the severance terms on what in Burton had been the second formulation, that is, employing the technique of the ‘relative formulation’. It would then see that the relative formulation (seemingly) employed by Tate & Lyle had been: ‘Pensions should be immediately payable to male workers who are ten years away from the normal State pensionable age, and to female workers who are five years away from the normal State pensionable age’. Ms Roberts hoped that the Court would find this to be discriminatory, and call for its replacement with a non-​discriminatory relative formulation such as: ‘Pensions should be immediately payable to all workers who are ten years away from the national pensionable age’. This would produce the result that early pensions would be accorded to men aged 55, but also to women aged 50, thus bringing her within the scope for an immediate pension. But the cjeu, having been only too glad to endorse the relative formulation in Burton, now seemed to turn against the idea of relative formulations. Covert differences should be exposed to the light of day. On this reading, of course, Tate & Lyle’s severance terms seemed egalitarian (65 minus ten was 55; 60 minus five was 55). This could not be discrimination because it was ‘the grant of a pension to persons of the same age who are made redundant’ and amounted merely to ‘a collective measure adopted irrespective of the sex of those persons in order to guarantee them all the same rights.’116 It was Tate & Lyle, then, which had refused to ossify an original disparity here (namely the disparity in the State pensionable ages).117 But was the Court, by endorsing this, subtly showing its disapproval for the original disparity? And is it not merely a cosmetic remedy to equalize a secondary difference, while leaving the original difference intact? Does it even make sense?118 Another four years passed, and the Court was confronted with the case of Barber v Guardian Royal Exchange.119 Mr Barber had worked for Guardian 1 16 Both quotations from ibid para 36 (emphasis added). 117 In fact, Tate & Lyle had ossified the original disparity in the first version of the severance terms. However, its male workers, who would not then have received their early pensions until the age of 60, had complained. See ibid para 5. 118 Such a course of action is akin to the abovementioned doctor making the second appointments of both of his or her patients on the same date. 119 Case C-​262/​88 Douglas Harvey Barber v Guardian Royal Exchange Assurance Group [1990] ecr I-​1889.

78 Chapter 3 Royal Exchange (‘Guardian’), and its predecessor company, since 1948. A member of Guardian’s Pension Fund (another contracted-​out occupational pension scheme), he found himself dismissed by reason of redundancy at the age of 52. However, according to Guardian’s Guide to Severance Terms, those made redundant up to ten years preceding their normal pensionable age would receive an immediate pension. Those who had not yet attained this age (namely, 55 for men and 50 for women) would have to make do with a so-​called ‘deferred pension.’ The English Court of Appeal asked the cjeu whether it was discriminatory that where a man and a woman of the same age were made compulsorily redundant in the same circumstances, the woman received an immediate pension while the man received only a deferred pension. The Court held that the age condition in Guardian’s Severance Terms infringed Article 119 ec120 ‘even if the difference between the pensionable age for men and that for women is based on the one provided for by the national statutory scheme.’121 Relative formulations were thus driven out once and for all, or so it seemed. Three years after Barber came Birds Eye Walls v Roberts.122 Here, Ms Roberts, a packer for Birds Eye Walls, was affiliated to Unilever’s (contracted out) occupational pension scheme. Pursuant to this, those who were compelled on the grounds of ill health to take early retirement before reaching the statutory State pensionable age would receive from the employer an ex gratia payment called a bridging pension. This included an amount corresponding to a proportion of the State pension. However, from the age of 60 onwards, a female worker received a smaller bridging pension, on the grounds that she was now in receipt of the State pension itself. Ms Roberts alleged that this was a breach of Article 119 ec.123 Neither the Advocate General (Van Gerven AG) nor the Court agreed with Ms Roberts. Advocate General Van Gerven was swayed by an argument of the Commission’s that Birds Eye Walls is attempting to achieve substantive equality between the sexes by compensating for an inequality (difference in pensionable ages).124 In other words, although Birds Eye Walls’ behaviour in granting the relevant female workers less once they reached the age of 60 is an instance of inequality, 1 20 121 122 123 124

As it then was –​subsequently Article 141 ec, and now Article 157 tfeu. Case C-​262/​88 Barber (n 119) para 32 (emphasis added). Case C-​132/​92 Birds Eye Walls Ltd v Friedel M Roberts [1993] ecr I-​5579. As it then was –​subsequently Article 141 ec, and now Article 157 tfeu. Case C- 132/ 92 Birds Eye Walls (n 122) Section ii-​3 of the Report for the hearing.

Equal treatment irrespective of gender

79

it is designed to remove or prevent,125 or compensate for,126 another inequality, namely the fact that, without this reduction, men would receive less than their female counterparts between the ages of 60 and 65.127 The second inequality is, as the quotation from the Report for the hearing above shows, merely the difference in State pensionable age (in the uk), by another name. After quickly distinguishing Barber, the Advocate General concludes that Birds Eye Walls’ aim is actually to ensure equal pensions for men and women.128 The discrimination is thus justified, and Article 157 tfeu is not breached.129 The Court goes slightly further in that it refuses to make a finding of discrimination at all. Recalling the ‘like for like’ test, the cjeu declares that, the general principle of equal treatment laid down by Article 119 of the Treaty … presupposes that the men and women to whom it applies are in identical situations.130 But the financial position of a woman compelled on the grounds of ill health to take early retirement is not comparable to that of a man in the same situation, between the ages of 60 and 65:131 That difference as regards the objective premise, which necessarily entails that the amount of the bridging pension is not the same for men and women, cannot be considered discriminatory.132 Thus, Birds Eye Walls’ behaviour in reducing the amount of the bridging pension granted to the relevant female workers, after they reach the age of 60, does not breach Article 157 tfeu.133 The Court has thus lurched back to its earlier position of favouring relative formulations.134 But, going back to the analysis of Burton above, is the use of 125 Van Gerven AG’s choice of verbs: ibid, Opinion of AG Van Gerven, para 16. 126 The Commission’s choice of verb: ibid, Section ii-​3 of the Report for the hearing. 127 When the female counterparts would receive bridging pension plus State pension, while they (the men) would receive bridging pension only. 128 Case C-​132/​92 Birds Eye Walls (n 122) Opinion of AG Van Gerven, para 17. 129 ibid, Opinion of AG Van Gerven, para 18. 130 ibid para 17. Article 119 ec subsequently became Article 141 ec, and now Article 157 tfeu. 131 ibid para 20. 132 ibid. 133 ibid para 21. 134 Although, as the Advocate General nicely describes, the facts of this case lend themselves to both a non-​relative and a relative reading, the former leading to a conclusion

80 Chapter 3 this relative formulation a ‘trick’ to perpetuate a (non-​functional) disguised difference? Both the Advocate General and the Court would appear to reply to this question in the negative. Birds Eye Walls, they would argue, far from playing a trick to perpetuate the uk’s difference in pensionable ages, is actually trying to fashion the uneven landscape in which it finds itself, to generate evenness: fighting fire with fire, as it were. And, of course, applying a second inequality to an existing situation of inequality in order to bring about equality is not new: it is, as the Commission rightly stated,135 substantive equality. Nevertheless, this is a troubling result, because, as Ms Roberts correctly pointed out, Birds Eye Walls is ‘rely[ing] on differences in pensionable age for national schemes [which] is tantamount, in effect, to relying on discrimination based on sex.’136 And the Court had stated in Barber that Article 119 ec137 prohibited any discrimination with regard to pay as between men and women, whatever the system employed, even if said system is ‘based on … the national statutory scheme.’138 A second note of caution must also be sounded. The Court relies on the assumption that a female worker like Ms Roberts receives a State pension in the first place. But this is not necessarily the case. In the uk, married women who worked were given two options, the first being to pay their national insurance contributions at the full rate and then to receive a full State pension in their own right, and the second being to pay national insurance contributions at a reduced rate, but to receive a reduced State pension (or no State pension at all) when pensionable age was reached. In the precise circumstances of the case, Ms Roberts had taken the second option and therefore was not in receipt of a of direct discrimination, and the latter leading to a conclusion of indirect discrimination: ‘[I]‌t is possible to focus on the fact that the appellant pays to all female ex-​employees aged between 60 and 65 a lower bridging pension than to their male counterparts [non-​ relative formulation] … By the same token, however, emphasis can be laid on the fact that it calculates the bridging pension for all its ex-​employees in the same way, namely by deducting from their [gross retirement pension] such pension payments as they can claim from the State or from Unilever [relative formulation]’. See ibid, Opinion of AG Van Gerven, para 13. 135 ibid, Section ii-​3 of the Report for the hearing; see also ibid para 15. 136 ibid, Section ii-​2 of the Report for the hearing. 137 As it then was –​subsequently Article 141 ec, and now Article 157 tfeu. 1 38 Case C-​262/​88 Barber (n 119) para 32. The Court again appears to go back on Barber, thus upholding Birds Eye Walls, in the more recent case of Hlozek, where a difference as between the ages at which men and women were to be awarded bridging allowances, prima facie based on the respective Austrian pensionable ages, was endorsed: Case C-​19/​02 Viktor Hlozek v Roche Austria Gesellschaft mbH [2004] ecr I-​11491.

Equal treatment irrespective of gender

81

State pension, although, fortuitously, she was in receipt of a widow’s pension, which was equal in amount. The Court had been asked about the possibility of a claimant having no State pension to countervail, as it were, the reduction in the bridging pension, but took a most dismissive line on this, holding that the two options were ‘a matter in which married women … have freedom of choice.’139 The Court’s reasoning then becomes confused, with the judges first saying that they would not look to actual amounts when ruling on this kind of discrimination,140 but later –​apparently –​taking account of Ms Roberts’ receipt of a widow’s pension in this case.141 The worry is that another female claimant, identical to Ms Roberts but for the fact that her husband was still alive, would have found herself at 60, with no State pension or widow’s pension, but facing a massive drop in the bridging pension which was her only means of support. The Court’s ruling is only logical as long as the one thing replaces the other. As soon as there is no replacement, discrimination as between a female claimant and a male comparator, still in receipt of the full bridging pension, is obvious. 3.6.2 A Walzerian Analysis A Walzerian reading of the same pensions cases reveals two agreements with the Court, and two disagreements; it also suggests a greater degree of consistency when Walzer’s test is applied than would be the case under Aristotle. In Burton, the result would depend on the relevant distributive community’s shared understanding of a voluntary redundancy benefit. This community would have to decide if the most important thing was that all workers should receive it at the same biological age (say, 55  –​the ‘biological age understanding’), or if the most important thing was that all workers should receive it in the 5-​year period immediately preceding their national pensionable age. If the meaning is to provide a safety net for those workers who are selected for voluntary redundancy, then the availability of the benefit in the run-​up to the attainment of the national pensionable age, in other words, in the period during which redundancy is most likely to take effect, is critical (the ‘run-​up rationale’). This would suggest that the second of the above two options is the better, or at least the more logical, one, but of course it would be up to the community. However, one cannot help thinking that this might merely be a case of setting the rules within a well-​defined and autonomous distributive sphere, meaning that any inequality that emerged in the 139 140 141

Case C-​132/​92 Birds Eye Walls (n 122) para 27. ibid para 28. ibid para 31.

82 Chapter 3 final agreement would be a small inequality only, not a boundary breach.142 It appears, then, that the Walzerian analysis of the case would not produce a different outcome from the Aristotelian one. If it is presumed (subject to confirmation by the distributive community) that there is no boundary breach here, and therefore no violation of complex equality, it can be seen that this is an identical conclusion to the one reached by the Court, namely, that there was no discrimination in evidence in British Railways’ arrangements for voluntary redundancy. Roberts, on the other hand, is an example of a case where the Walzerian analyst would (or at least might) reach a different verdict to the one reached by his or her Aristotelian counterpart. The case, as has already been seen, is a model of the confusion produced by the ‘like for like’ test, since both of the interpretations of Tate & Lyle’s pension scheme (that of Tate & Lyle, and that of Ms Roberts) appear to satisfy it. Tate & Lyle sees an instance of likes treated alike (both a woman aged 53 and a man aged 53 would fail to acquire the early pension). Ms Roberts, meanwhile, sees an instance of likes treated unlike (a man seven years below his national pensionable age would acquire the pension, while a woman seven years below her national pensionable age would not). A Walzerian take on the case, on the other hand, would start with the distributive community, and its shared understanding of an early ­pension. What does ‘early’ mean? Does it mean 55, irrespective of gender (again, the ‘biological ​age understanding’)? Or does it mean five years prior to the employee’s national pensionable age? If the latter, could the ‘extra’ five years of eligibility enjoyed by men at the time of Ms Roberts’ action (namely, the period from age 55 to age 60) represent a sly advantage, making masculinity a positive dominant?143 It is at least possible, then, taking as read the Standard Contingent Reply, that Tate & Lyle’s pension scheme sanctions a distribution flawed by dint of a boundary breach. If this is the final conclusion, then it would obviously conflict with the Court’s actual decision, which was that Tate & Lyle had not violated the principal of equal treatment.144 142 But this would come down to whether the reason for the inequality was intrinsic to the sphere of voluntary redundancy benefits (as the run-​up rationale would be) or extrinsic thereto. A reason which turned on gender alone would imply an illegitimate crossing from the sphere of attributes to the sphere of voluntary redundancy benefits. See the discussion on Roberts, immediately below. 143 For a discussion of how an attribute can be a positive, or negative, dominant, see supra, section 3.3. 144 It is important to note, though, that, on this analysis, even after the flaw in the distribution was resolved (i.e. early pensions were granted to men at sixty, and women at fifty five), this would still not help Ms Roberts. An alternative approach would be to level down

Equal treatment irrespective of gender

83

Barber is almost a mirror image of Roberts, in that instead of the current system being based on the biological ​age understanding, and the complainant calling for the run-​up rationale, here the current system was based on the run-​ up rationale, but with Mr Barber calling for the biological a​ ge understanding to be used in its place. It would be up to the distributive community to decide what it understood by an immediate pension, and indeed what ‘immediate’ should mean in this context. If the meaning is to provide protection for those workers who are made redundant, then the availability of the benefit in the run-​up to the attainment of the national pensionable age, in other words, in the period during which redundancy is most likely to take effect, would seem, as in Burton, to be critical.145 Alternatively, the community might call for the immediate pension to be available to everyone at 55, regardless of gender, and come to the conclusion that the ‘extra’ five years of eligibility which the existing approach bestows on women (namely, the period from age 50 to age 55)146 was a secret benefit, turning the possession of female gender into a positive dominant. The Standard Contingent Reply prevails as usual, but the stronger of the two positions would seem to be the first one. If the distributive community assents, then the status quo would be vindicated and complex equality upheld. This would once again be at odds with the Court’s finding of discrimination on the part of Guardian Royal Exchange. It is worth noting how the Walzerian test produces consistency as between Barber and Burton (no breach of complex equality in either), where using the Aristotelian test in those cases led the Court to give two conflicting judgments (discrimination in Barber, no discrimination in Burton), even though the pension schemes utilized by the two employers were similar (although the one in Barber was contracted-​out while the one in Burton was not), and the complaints made by the two (male) plaintiffs were almost identical. The ‘problematic’ nature of the Burton judgment ‘in view of the more recent case law on pensions’ is thus overcome.147 rather than up, so that ten years prior to the employee’s national pensionable age was set as the threshold for an early pension. That way, women would enjoy the ‘extra’ five years too, and Ms Roberts would qualify for a pension at 53. As usual, this would be up to the distributive community. 145 Furthermore, it is not as though a male worker made redundant before the beginning of his run-​up period (like Mr Barber) would be left completely without help. They would still have access to the ‘deferred’ pension in the meantime. 1 46 These five years are not really ‘extra’ as the ones in Roberts were (see previous paragraph). They may occur earlier in the female worker’s life, but the aggregate number of qualifying years remains ten for both sexes. 147 Quotations from I Heide, ‘Sex equality and social security: selected rulings of the European Court of Justice’ (2004) 143(4) Intl Labour Rev 299, 327.

84 Chapter 3 Finally, there is Birds Eye. In this case, the starting point of a Walzerian examination would be for the distributive community to decide upon its shared understanding of a bridging pension. It would probably conclude that this was a means of providing relief to those workers who had taken early retirement on grounds of ill health, until the maturing of their State pension (at age 60 for a woman, and at age 65 for a man). It would therefore probably further conclude that the fact that the date on which the bridging pension began to be reduced corresponded to the date on which the State pension became payable was at the very least logical. A  female worker’s receiving a smaller bridging pension than a male worker between the ages of 60 and 65 would thus seem to be in keeping with the (likely) shared meaning of the distribuend in question. Taking as read the Standard Contingent Reply, it appears plausible that a Walzerian examination of this case would disclose no breach of complex equality; this would be the same result as that reached by the Court, which found no discrimination in the behaviour of Birds Eye Walls.148 However, it must also be remembered that the Court was happy to declare that women and men, that is, all women and all men, were not in ‘identical situations,’149 in other words, were not alike, ‘between the ages of 60 and 65.’150 As discussed above, this reasoning ignores the possibility that a female claimant, like in fact Ms Roberts, might not be in receipt of a State pension between those ages. The Aristotelian sledgehammer effect is thus seen once again. Aristotelian logic, as clearly invoked by the Court at paragraph 17 of its judgment, is unable to discern the subtle differences that may arise from one case to another; it is incapable of reaching a conclusion of ‘like in some cases, unlike in others’. A Walzerian approach, it is submitted, would have picked up on the difficulty here. As ventured in the previous paragraph, the distributive community might well decide that, to it, a bridging pension means ‘a means of providing relief to those workers who have taken early retirement on grounds of ill health, until the maturing of their State pension.’ But it follows that, if there is no State pension to mature, then the bridging pension must continue providing relief indefinitely. And if that 148 Likewise a Walzerian analysis of Case C-​19/​02 Hlozek (n 138), would most likely result in the distributive community’s declaring that the meaning of a bridging allowance was to provide a financial bridge between the termination of the worker’s employment and the receipt of the State pension (again, the run-​up rationale). Thus the difference, as between the sexes, of the age at which the worker was entitled to the bridging allowance made perfect sense, as long as the State pensionable ages themselves were different. Of course, for the latter to be altered, being as they are a separate distribution, would be a job for a separate, and presumably much larger, Forum. 149 Case C-​132/​92 Birds Eye Walls (n 122) para 17. 150 ibid para 20.

Equal treatment irrespective of gender

85

is what a bridging pension means, then distributing it in any other way violates complex equality. The fact of not being in possession of a matured State pension is the attribute which the complex egalitarian (assuming the meaning above is the correct one) would see as legitimate, that is, as legitimately able to swing the distribution. That is the attribute that forms part of the meaning, not the fact of being female or the fact of being male. If a future female claimant, then, without a matured State pension at 60, were to lay claim to her bridging pension after her sixtieth birthday, she would not be denied it, as she would still be able to show the requisite badge of entitlement. 3.7

The Question of Part-​Time and Full-​Time Work

3.7.1 The CJEU’s Case-​Law A good starting point for this section is the case of Bilka,151 which neatly straddles the dividing line between cases concerning pensions, on the one hand, and cases concerning part-​time work and full-​time work, on the other. Mrs Weber von Hertz had worked for Bilka, the German department store, for fifteen years, that is, from 1961 until 1976. Although for the first eleven years she had worked full-​time, from October 1972 onwards she worked part-​ time. Bilka declared that, in order to obtain a pension under its occupational pension scheme, an employee must have worked full-​time for at least fifteen years (over a total period of twenty years). Mrs Weber von Hertz alleged a breach of ec law on the grounds that women workers ‘were more likely than their male colleagues to take part-​time work so as to be able to care for their family and children.’152 The cjeu considers the main issue, which is of course discrimination. Although the Court does not use the term, it is indirect discrimination which is at stake here; Bilka is making receipt of the pension conditional upon a requirement (fifteen years’ full-​time work) applied equally to all, but with which a considerably smaller proportion of women than men (it is argued) can comply. It will be seen that the exact wording of the second limb of the test for indirect discrimination varies in the cjeu’s case-​law.153 In Bilka itself, the Court settles on the phrase ‘much lower proportion’: 151 Case 170/​84 Bilka –​Kaufhaus GmbH v Karin Weber von Hartz [1986] ecr 1607. 152 ibid para 6. 153 Since 1997, there has been a legislative definition at Article 2 of the Directive on Indirect Discrimination and Burden of Proof (Council Directive 97/​80/​EC of 15 December 1997 on the burden of proof in cases of discrimination based on sex [1998] oj L14/​6). This

86 Chapter 3 If, therefore, it should be found that a much lower proportion of women than of men work full time, the exclusion of part-​time workers from the occupational pension scheme would be contrary to Article 119 of the Treaty where […] that measure could not be explained by factors which exclude any discrimination on grounds of sex.154 As to this ‘explanation,’ known as an ‘objective justification,’ Bilka argues that the exclusion of the part-​time workers from the scheme ‘is intended solely to discourage part-​time work, since in general part-​time workers refuse to work in the late afternoon and on Saturdays;’155 the company was thus trying to make full-​time work ‘more attractive.’156 The cjeu gives little guidance on objective justifications, however, and their assessment remains a job for the national court: If the national court finds that the measures chosen by Bilka correspond to a real need on the part of the undertaking, are appropriate with a view to achieving the objectives pursued and are necessary to that end, the fact that the measures affect a far greater number of women than men is not sufficient to show that they constitute an infringement of Article 119.157 Finally, in an intrepid move, Mrs Weber von Hertz called for ‘periods during which workers have had to meet family responsibilities’ to be regarded as

154 155 156 157

reads: ‘[I]‌ndirect discrimination shall exist where an apparently neutral provision, criterion or practice disadvantages a substantially higher proportion of the members of one sex unless that provision, criterion or practice is appropriate and necessary and can be justified by objective factors unrelated to sex’. This definition was replaced in the New Equal Treatment Directive by a second one: ‘[I]ndirect discrimination [shall exist] where an apparently neutral provision, criterion or practice would put persons of one sex at a particular disadvantage compared with persons of the other sex, unless that provision, criterion or practice is objectively justified by a legitimate aim, and the means of achieving that aim are appropriate and necessary’. The New Equal Treatment Directive has since been overtaken by the Recast Directive, where the definition of indirect discrimination can be found at Article 2(1)(b); it is identical to the definition in the New Equal Treatment Directive. Case 170/​84 Bilka (n 151) para 29. Article 119 ec subsequently became Article 141 ec, and now Article 157 tfeu. ibid para 33. ibid. ibid para 36.

Equal treatment irrespective of gender

87

‘periods of full-​time work.’158 However, the Court dismissed this with the now-​ famous remark: Article 119 does not have the effect of requiring an employer to organize its occupational pension scheme in such a manner as to take into account the particular difficulties faced by persons with family responsibilities …159 Bilka built on the slightly earlier case of Jenkins,160 where part-​time employees (all but one of whom were female) were paid at a lower hourly rate than full-​ time employees, although both types of employee were doing the same work. The Court saw no problem with an employer’s full-​time rate being higher than its part-​time rate, as long as there was no distinction between men and women (that is, as long as male and female part-​timers were treated equally, and male and female full-​timers were treated equally). However, if it were established that a ‘considerably smaller percentage’161 of women than men performed ‘the minimum number of weekly working hours required in order to be able to claim the full-​time hourly rate,’162 then there would be a breach of (what is now) Article 157 tfeu, in the absence of an explanation not based on sex. It was for the national court to decide, regard being had, inter alia, to ‘the employer’s intention,’163 whether or not a difference based on weekly working hours was ‘in reality’164 discrimination based on the sex of the worker.165 In Stadt Lengerich v Helmig,166 six (part-​time) female employees sued their employers on the ground that part-​time employees were not receiving the 158 159 160 161 162 163 164 165

166

ibid para 39. ibid para 43. Case 96/​80 JP Jenkins v Kingsgate (Clothing Productions) Ltd [1981] ecr 911. ibid para 13. ibid. ibid para 14. ibid. Lester has called the Jenkins judgment ‘equivocal’ and ‘sibylline,’ particularly owing to this reference to the employer’s intention, which appeared to shield unintentional indirect discrimination from censure. He explains that the Court’s judgment was not ‘able to be used’ by the Employment Appeal Tribunal (when the case returned to the uk), so that ‘no-​one was any the wiser’: A Lester, ‘The Uncertain Trumpet –​References to the Court of Justice from the United Kingdom: Equal Pay and Equal Treatment without Sex Discrimination’ in HG Schemers (ed), Article 177 eec: Experiences and Problems (North Holland 1987) 164, 177–​183. Joined Cases C-​399/​92, C-​409/​92, C-​425/​92, C-​34/​93, C-​50/​93 and C-​78/​93 Stadt Lengerich v Angelika Helmig and Waltraud Schmidt v Deutsche Angestellten-​Krankenkasse

88 Chapter 3 same rate of pay for overtime as full-​time employees. In other words, a full-​ time employee, working in excess of the ‘ordinary working week of full-​time workers’167 (in this case, 38.5 hours), would receive payment at between 15 and 25% above the hourly rate. Meanwhile, a part-​time employee, working in excess of their contractual working hours (which varied), would receive payment at the normal rate. (In the event that the part-​time employee continued to work beyond the 38.5 hour threshold, then they would become entitled to the higher rate.) The plaintiffs alleged that this state of affairs breached Article 119 ec168 and the Equal Pay Directive. Advocate General Darmon dealt with the issue swiftly. Part-​time employees were not treated differently, per se, from full-​time employees. Both received overtime supplements when the weekly working time fixed by collective agreement for full-​time workers was exceeded. Far from remedying an inequality, to grant part-​time employees overtime supplements when their weekly working time was exceeded would actually ‘give rise to a real inequality,’169 as, for the same number of hours worked, some would get the overtime supplements while others would not. Ex Article 119 ec was, as things stood, not breached. The cjeu agreed with this position: In the circumstances considered in these proceedings, part-​time employees do receive the same overall pay as full-​time employees for the same number of hours worked.170 There was thus no discrimination in breach of ex Article 119 ec, and no violation of Article 1 of the Equal Pay Directive. In February 1996 came the case of Kuratorium für Dialyse und Nierentransplantation v Lewark.171 Ms Lewark was employed for 30.8 hours a week in the defendant’s ‘Care Unit,’ where she was one of eleven part-​time workers (ten women and one man). She was also the only part-​time worker on the Local Staff Council. In November 1990, with the defendant’s consent, and Elke Herzog v Arbeiter-​Samariter-​Bund Landverband Hamburg eV and Dagmar Lange v Bundesknappschaft Bochum and Angelika Kussfeld v Firma Detlef Bogdol GmbH and Ursula Ludewig v Kreis Segeberg [1994] ecr I-​5727. 167 Phrase used in ibid, Opinion of AG Darmon, para 3. 1 68 As it then was –​subsequently Article 141 ec, and now Article 157 tfeu. 1 69 Joined Cases C-​399/​92, C-​409/​92, C-​425/​92, C-​34/​93, C-​50/​93 and C-​78/​93, Helmig (n 166) Opinion of AG Darmon, para 29. 1 70 ibid para 27. 1 71 Case C-​457/​93 Kuratorium für Dialyse und Nierentransplantation eV v Johanna Lewark [1996] ecr I-​243.

Equal treatment irrespective of gender

89

she attended a full-​time training course in order to obtain the knowledge that was necessary for performing her functions on the Staff Council. The training course on 13 November 1990 lasted 7.5 hours. Ms Lewark would not normally have worked that day. The defendant paid her her normal salary for the week (30.8 hours), but failed to recompense her for the time spent on the training course. Since a full-​time worker would have been so recompensed, Ms Lewark considered that there had been a breach of Article 119 ec172 and the Equal Pay Directive. As regards whether there was a difference in treatment between those working part-​time and those working full-​time,173 the Court concluded that there was indeed such a difference in treatment; this was in fact ‘indisputable.’174 The Court then stated that, if it were the case that ‘a much lower proportion’175 of women than men worked full-​time, the exclusion of part-​time workers from the benefit in question (that is, compensation for the time spent on the training course) would be contrary to ex Article 119 ec, in the absence of an explanation not founded on sex. After examining the relevant statistics, the Court concluded that, in this particular case, there was (in principle) ‘indirect discrimination.’176 It of course remained open to Germany to show that the measure taken by the defendant reflected a legitimate aim of social policy, was appropriate to achieve that aim and was necessary in order to do so. It will be seen that this is the same three-​stage test as was proposed by the Court in Bilka.177 As usual, it was up to the national court to assess whether this test was satisfied. A month later, the Court reached an almost identical conclusion in the case of Freers.178 In the 2004 case of Wippel v Peek & Cloppenburg,179 Ms Wippel was employed by the Austrian company Peek & Cloppenburg pursuant to a system known as ‘work on demand.’ She had no fixed working hours, and the duration 1 72 As it then was –​subsequently Article 141 ec, and now Article 157 tfeu. 173 Note that the Court’s earlier, more lenient approach to such difference in treatment (see, for example, Case 96/​80 Jenkins (n 160) para 11) seemed to have now slightly hardened (see, for example, Joined Cases C-​399/​92, C-​409/​92, C-​425/​92, C-​34/​93, C-​50/​93 and C-​78/​93 Helmig (n 166) para 26). 174 Case C-​457/​93 Lewark (n 171) para 26. 175 ibid para 28. 176 ibid para 30. 177 Case 170/​84 Bilka (n 151) para 36, quoted above (see text accompanying n 157). 178 Case C-​278/​93 Edith Freers and Hannelore Speckmann v Deutsche Bundespost [1996] ecr I-​1165. 179 Case C-​313/​02 Nicole Wippel v Peek & Cloppenburg GmbH & Co kg [2004] ecr I-​9483.

90 Chapter 3 and ‘positioning’ of her working time was determined by agreement between the parties in each individual case. Specifically, the sales manager would ask her to work certain hours in a given week, and Ms Wippel could either accept or decline, without any need to give reasons. Importantly, she was not guaranteed any fixed income. Ms Wippel complained that the absence in her contract of an agreement as to working hours and the organizing of working time amounted to discrimination on the grounds of sex. The Austrian law on Working Time180 provided that the ‘positioning’ of normal working time was to be by agreement, as were any changes thereto. However, notwithstanding this, the ‘positioning’ of normal working time could be changed unilaterally by the employer, but only if, inter alia, there was a justification and two weeks’ notice was given to the worker. A later article of the same law applied these principles to part-​time work too. Advocate General Kokott started by considering the new Part Time Workers Directive.181 The latter, following the increasingly tough language in the case-​ law, had introduced a prohibition on discrimination against part-​time workers. However, the Advocate General concluded that there was no such discrimination in the legislation in this case (the Austrian law on Working Time treated full-​time workers and part-​time workers in the same way). Nor, according to the Advocate General, was there discrimination on the grounds of sex in the legislation in this case. Although the law made provision, in respect of full-​ time workers, for normal working time of 40 hours a week and 8 hours a day, no specific provision on maximum working time existed for part-​time workers. The absence of such a provision had a greater impact on women than men. However, in stark contrast to the referring court, which saw this as an instance of ‘like treated unlike’, the Advocate General decided that full-​timers and part-​ timers were not comparable ‘in this specific context.’182 Turning to the contract of employment, Kokott AG again found no discrimination against those working part-​time. If anything, a ‘work on demand’ contract which made no provision for (previously agreed) fixed working time would be ‘beneficial’ to ‘employees who are only able … to work irregular hours and in varying amounts.’183 With regards to sex discrimination, if all of those with fixed working hours (part-​time or full-​time) were compared with those 180 Arbeitszeitgesetz. 181 Council Directive 97/​81/​EC of 15 December 1997 concerning the Framework Agreement on part-​time work concluded by unice, ceep and the etuc –​Annex: Framework agreement on part-​time work [1998] oj L14/​9. 182 Case C-​313/​02 Wippel (n 179) Opinion of AG Kokott, para 93. 183 ibid, Opinion of AG Kokott, para 106.

Equal treatment irrespective of gender

91

employed on demand, the difference in percentage of female workers was so small as to be insignificant. There could be no discrimination if the numbers of women were the same on both sides of the comparison (those allotted a fixed working time, and those not).184 The Court agreed with the Advocate General in all of her conclusions. Of particular interest is that, as regards possible sex discrimination in the employment contract, the Court declared that since no full-​time worker at Peek & Cloppenburg had a contract like Ms Wippel’s, she effectively had no comparator. Without a comparison, there could not be a finding of less favourable treatment.185 The Court is here conspicuously moving away from the doctrine of noncomparative discrimination which it had championed in the pregnancy cases.186 Similarly, turning to the Equal Treatment Directive, the Court contemplated the same comparison as the Advocate General, namely, to compare all of those with fixed working hours (part-​time or full-​time) with those employed on demand. However, it went further than the Advocate General by abandoning the comparison altogether. Because the second group could accept or refuse work at will, while the first group could not, the two groups were simply ‘not analogous.’187 If they were unlike, they could of course be treated unlike. 3.7.2 Preliminary Analysis In Helmig, Advocate General Darmon mentions (and the Court does not disagree) that one rationale for the granting of overtime payments is the need for employers to be dissuaded from working their employees long hours. This in turn is because, according to the Advocate General, workers need leisure-​time and recuperation from fatigue. Therefore, as well as penalizing the employer, the overtime payment in some ways attempts to (partially) compensate the employee for their lost leisure and rest. By declaring that full-​time employees (mostly male) and part-​time employees (mostly female) were treated alike as regards overtime, the Court implied that a woman who (for the sake of example) works part-​time while raising a family loses neither leisure nor rest if she is 184 Of course, if all those working full-​time were compared with all those working part-​time, the difference in percentage of female workers was significant. However, neither the Advocate General nor the Court was minded to run this comparison. 185 Case C-​313/​02 Wippel (n 179) para 62. 186 To this it could conceivably be counterargued that the Court was merely constrained by the wording at Clause 4(1) of the Framework Agreement on part-​time work, annexed to the Part Time Workers Directive. 187 Case C-​313/​02 Wippel (n 179) para 64.

92 Chapter 3 required to work more than her usual shift (but less than a ‘full-​time’ shift). But this is erroneous if, as Mrs Weber von Hertz argued in Bilka, familial responsibility is regarded as work too. The point made in the previous paragraph raises a myriad issues, many if not all beyond the scope of this book. They cannot be done justice in a short space. Only a cursory examination of these issues will therefore be attempted here; the writers cited should be consulted for a full analysis of the arguments. At the heart of the matter is what the Court itself called the ‘double aim’ of Article 119 ec (later Article 141 ec, and now Article 157 tfeu),188 namely, its economic aim and its social aim. As Barnard excellently explains, the formal, like-​for-​like test lends itself perfectly to the economic role (‘creating a level playing field of competition’), but much less well to the social role.189 The economic sphere is centred on the market, and the market is historically predicated on a ‘male breadwinner’ model. Comparisons were thus traditionally to a ‘male norm,’190 and no allowance was made for women’s onerous caring duties (for example, arguably, by Stadt Lengerich and the other employers, endorsed by the Court, in Helmig).191 188

Case 43/​75 Gabrielle Defrenne v Société anonyme belge de navigation aérienne Sabena (no 2) [1976] ecr 455, paras 8 and 12. 189 Catherine Barnard, ‘The Economic Objectives of Article 119’ in TK Hervey and D O’Keeffe (eds), Sex equality law in the European Union (Wiley 1996) 321, 329. 190 The most celebrated writer on this particular aspect of the debate is Catherine MacKinnon:  ‘As male is the implicit reference for human, maleness will be the measure of equality in sex discrimination law … If male power is systemic, it is the regime.’ See C  MacKinnon, ‘Feminism, Marxism, Method, and the State:  Toward Feminist Jurisprudence’ (1983) 8 Signs 635, 644–​5 (MacKinnon’s emphasis). See also C MacKinnon, ‘Reflections on Sex Equality Under the Law’ (1990–​91) 100 Yale Law Journal 1281. 191 On the subject of women’s caring role, Everson has written, ‘it is the inability of [the European right of sexual equality] … to take caring obligations into account, which has failed European women as a whole’. See M Everson, ‘Women and Citizenship of the European Union’ in TK Hervey and D O’Keeffe (eds), Sex equality law in the European Union (Wiley 1996) 203, 209. The Court’s approach in Helmig has been described as ‘problematic’: Hervey and Shaw (n 6) 54. Ellis said that the judgment left the law in an ‘unsatisfactory state’: Evelyn Ellis, ‘Recent developments in European Community sex equality law’ (1998) 35 cml Rev 379, 383. Wentholt commented that it ‘expresse[d]‌a male point of view’: K Wentholt, ‘Formal and Substantive Equal Treatment: the Limitations and the Potential of the Legal Concept of Equality’ in T Loenen and PR Rodrigues (eds), Non-​ discrimination law:  comparative perspectives (Kluwer Law International 1999)  53, 63. Costello and Davies called it ‘dubious’: C Costello and G Davies, ‘The case law of the Court of Justice in the field of sex equality since 2000’ (2006) 43 cml Rev 1567, 1591.

Equal treatment irrespective of gender

93

That said, the Court’s record on the social dimension of Article 119 ec (later Article 141 ec, and now Article 157 tfeu) has not been entirely negative. Many of the successes can indeed be found in the case-​law on full-​time and part-​time work, and particularly the use of the doctrine of indirect discrimination. This doctrine, some argue, allows for consideration of groups, as opposed to individuals; a measure is impugned if it impacts negatively on a whole class of comparators, not just one.192 Others, however, refuse to see indirect discrimination as a victory for substantive equality. Hervey, for example, believes that the cjeu’s use of indirect discrimination is still essentially market-​driven, and warns that the doctrine has ‘significant limitations.’193 Finally, on the question of part-​time and full-​time work, there is something worrying about the Wippel judgment, in that the whole case turns on the ­comparison between workers on demand and workers with fixed hours. Can the very fact that workers on demand work on demand support, in and of itself, an a priori refusal to compare? This is reminiscent of Westen’s oft-​quoted complaint about the circularity of equality: ‘people who by a rule should be treated alike should by the rule be treated alike’.194 The comparison is over before it starts, because the fact that two things have already been treated differently (Column 2 at Appendix I), specifically, in this case, one worker has been given fixed hours and another has not, predetermines whether or not they are to be regarded as ‘like’ (Column 1 at Appendix I). On Westen’s analysis, while one may pretend that the distributing of goods takes place after the allocating of attributes, in actual fact the distribution has already taken place, when the attributes were allocated. The same ‘standard or rule’195 is applied to both events, and, having been applied the first time, it cannot help but be applied the second time; treatment thus folds into likeness.196 The consequence of 192 Mancini and O’Leary (n 76) 342–​3; Lisa Waddington, ‘The development of a new generation of sex equality directives’ (2004) 11(1) mj 3, 6. 193 TK Hervey, ‘Thirty Years of eu Sex Equality Law: Looking Backwards, Looking Forwards’ (2005) 12(4) mj 307, 311. The same author, writing with Helen Fenwick, further commented that, when faced with a conflict between equality for women and the interests of the market, the Court would ‘safeguard the market’: Fenwick and Hervey (n 36) 469. 194 P Westen, ‘The Empty Idea of Equality’ (1982) 95 Harv L Rev 537, 547. 195 ibid 545. 196 As Numhauser-​Henning puts it, referring to Wippel, ‘[the] problem consists of the fact that what is forbidden by the non-​discrimination provision –​differential treatment as regards employment conditions –​is at the same time part of what constitutes the groups that are to be compared. Different employment conditions pertaining to the mode of employment are a sine qua non’:  A Numhauser-​Henning, ‘eu sex equality law post-​ Amsterdam’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007)

94 Chapter 3 this from a judicial point of view is that, as McColgan puts it, the differences between comparators are ‘reducible to the very ground on the basis of which the claimant sought to challenge the discrimination.’197 3.7.3 A Walzerian Analysis Refracting the same cases through the prism of complex equality shows, firstly, how complex equality facilitates a more robust scrutiny of (to use the Aristotelian nomenclature) ‘objective justifications’. It also demonstrates how Walzer’s theory produces more predictable results (more often than not different from those reached by the Court), and allows for a more compassionate handling of cases like Helmig. In Bilka, for example, the fons et origo of the enquiry would be the meaning of a pension under an occupational pension scheme, as shared by the members of the relevant distributive community. Is such a pension to reward a commitment to full-​time work? Or is it to aid an ex-​employee in their old age? If the latter, then making the distribution conditional on the completion of 15 years’ full-​time work would seem inappropriate, or, put in Walzerian terms, the fundamentum distributionis would seem to be out of line with the meaning of the good being distributed.198 Those who had completed 15  years of full-​time work (of whom there would necessarily be more men than women) would now hold a badge of honour, garnered via an unconnected achievement in a different sphere, which would give them exclusive access to the sphere in issue. Bilka’s scheme thus provides good examples of both a flawed distribution and a boundary breach. The Court left open the possibility that Bilka’s behaviour in disenfranchising part-​time workers could be justified, the only justification put forward by Bilka being the desire to discourage part-​timers and to make full-​time 145, 155–​6. Another possible example is provided by Case C-​236/​09 Association Belge des Consommateurs Test-​Achats asbl and Others v Conseil des ministres [2011] ecr 773, discussed at section 3.9 below; see in particular the text accompanying n 264. 197 McColgan (n 3) 670, referring to the approach of Lord Hoffman in the English case of Carson: R (on the application of Carson) v Secretary of State for Work and Pensions [2005] ukhl 37; [2006] 1 ac 173. Costello and Davies comment that the Court’s handling of the comparability issue in Wippel was ‘regrettable’: Costello and Davies (n 191) 1594. 1 98 It would be different if the 15-​year stipulation was connected financially to the pension, that is, if 15 years of full-​time employment was somehow the minimum amount of salary-​ generating work needed to pay for it (via contributions deducted at source). However, it is clear that building up a sufficient quantity of contributions is not the issue here. Bilka stated before the Court that the stipulation was ‘solely to discourage part-​time work’: Case 170/​84 Bilka (n 151) para 33 (emphasis added).

Equal treatment irrespective of gender

95

work more appealing. The cjeu, then, while not expressly endorsing this justification (a job which it leaves to the national court), at least entertains it as a possibility. It will often be seen in this case-​study how the job of the judge (European or national) in assessing a proposed objective justification is similar to the job of the ‘gatekeepers’ of any given sphere in trying to decide whether a badge presented by a would-​be distributee may be legitimately brought in, or must be left ‘at the door’. The same sort of issues, whether justifying a discrimination or attempting to tie a badge to the business of a sphere, tend to crop up. However, it is not just the same process described in a different way, and Bilka is a good illustration of that. At the objective justification stage of the Aristotelian test, almost anything may be pleaded as a possible rationale for discrimination by a distributor;199 the justification put forward needs only to be remotely of interest or importance to him or her.200 The ‘gatekeeper’ stage of any proposed transaction under complex equality, meanwhile, is a much stricter affair; holders of badges without a proven connection to the ‘host’ sphere will be refused entry, unless they are prepared to put the badge aside. Thus, under a Walzerian analysis, Bilka’s sole suggested defence could be rejected straight away, without the need in this instance to wait for the national judge; it simply has nothing to do with the distribution of pensions. That is not to say that it may not be a valid or worthy goal, just that the wrong means were chosen to accomplish it. In Walzerian terms, Bilka pursued its objective in the wrong sphere. This is reminiscent of a famous line of ts Eliot’s: The last temptation is the greatest treason: to do the right deed for the wrong reason.201 In Bilka, of course, it was the ‘reason’ which was –​arguably –​right, and the ‘deed’ which was wrong.202 199 As long as it does not involve a suspect ground. Economic rationales, in certain types of internal market cases, are also excluded. 200 The proportionality test –​and particularly the appropriateness branch thereof –​is supposed to disqualify any that are too remote. 201 TS Eliot, Murder in the Cathedral (Faber and Faber 1935) 47. 202 The point being made here is not that the department store was correct in its policy of discouraging part-​time workers –​a policy which would obviously have had a disastrous impact on women employees –​merely that if the discouragement of part-​timers was the result that Bilka wanted to achieve, it should have opted for a more appropriate method by which to achieve it. To return to the education example, a schoolteacher may give one student an A grade and one student a C grade on the grounds of the two students’ relative

96 Chapter 3 Looking at the worst case scenario, and assuming that Bilka’s defence is accepted by the referring court, the overall result would be to exonerate the store of discrimination altogether. The Walzerian analysis, meanwhile, culminates in the finding of a clear boundary breach on the part of Bilka, that is, a violation of complex equality. This same pattern (Walzerian analysis revealing infraction of complex equality, while no infraction of equality principle found by cjeu at all) is repeated several times. Jenkins does not require a lengthy analysis here, as, from the legal point of view, it is very similar to Bilka. Kingsgate Ltd (the defendant company) may have had a legitimate wish to disincentivize part-​time workers, but it should not have used the hourly rate as a means by which to achieve this. To do so distorts the true meaning203 of the hourly rate. The distributive community204 would almost certainly hold that the hourly rate should represent recompense, or consideration, for one hour’s work. It is not a tool by which to penalize those who have chosen to work part-​time, thus turning their choice into a negative dominant.205 That fight had to be fought (if at all) in a different arena. In Helmig, Stadt Lengerich and the other employers denied their part-​time workers overtime supplements when they worked beyond their ordinary working week, but granted the same supplements to full-​time workers who had worked beyond their ordinary working week. This implied that a part-​time worker (almost certainly a woman) who worked beyond her ordinary working week had not lost any leisure-​time or recuperation from fatigue (a loss normally requiring compensation), and that the time between the end of her ordinary working week and the end of a full-​time worker’s ordinary working intelligence, but not on the grounds that the second was caught smoking while the first was not –​even though preventing minors from smoking is, in and of itself, a perfectly laudable  aim. The (potential) equation of Walzer to indirect discrimination/ objective justification is investigated in further detail below, at section 4.3.2(b). 203 Taking the Standard Contingent Reply as read. 2 04 There is no need to assume that this would only consist of the distributors (Kingsgate Ltd’s management) and the distributees (Kingsgate Ltd’s employees), and that it would therefore be intrinsically biased in favour of the full time (male) workers –​who would represent the majority –​and against the part time (female) workers. All interested parties are entitled to be present at a Forum, which might include Kingsgate’s shareholders, its customers, and other members of the local community affected by its decisions. Furthermore, if the theory of mediated complexity were preferred to that of complex equality stricto sensu, then the decision would fall to the Bench, which would considerably reduce the risk of bias. See Chapter 8 below. 205 For a discussion of how a choice can be a positive, or negative, dominant, see supra, section 3.3.

Equal treatment irrespective of gender

97

week was effectively ‘spare’ and at the disposal of the employer at no extra cost. There is little doubt that this represents a massive boundary breach. By belittling, if not abnegating, a part-​time female employee’s caring duties by effectively declaring that only after 38.5 hours in the workplace can anyone be regarded as fatigued, Stadt Lengerich and the other employers contribute to the perpetuation of abhorrent notions about the role of carers (for which read women) in today’s society. If they belong anywhere (and it is a big ‘if’ in the 21st century), notions of respective ‘roles’ of family members –​including emotional ‘roles’ –​belong within the sphere of kinship only. Their reiteration in the sphere of office (and others) is one of the worst boundary breaches that there is, and, as Armstrong points out, it cuts both ways  –​women are disadvantaged when those in other spheres assume female monopolization of caring work, and those assumptions (in a State’s economic policy, say, or its law and order policy) feed back into the domestic sphere, thus preserving the status quo: [T]‌he theory of complex equality provides us with a particularly clear framework for analysing the political nature of the personal.206 Boundary revision is urgently needed in this area, but the process is likely to be both lengthy and difficult. In Lewark there is arguably an even bigger affront to part-​time workers than there was in Helmig. The case turned once again on the temporal no-​man’s-​land between the end of a part-​time worker’s ordinary working week and the end of a full-​time worker’s ordinary working week. This time, however, the distribuend was recompense for a training day. The full-​time worker  who attended the training day over and above his ordinary working week207 would receive recompense. The part-​time worker, such as Ms Lewark, on the other hand, who attended the training day over and above her ordinary working week, that is, on a day on which she would otherwise not have been working, received none. At least Ms Helmig received her normal hourly rate for the extra work undertaken; Ms Lewark did not even receive that. Once again, it will fall to the distributive community to decide what it understands by a recompensed training day. It would be an odd community that did not answer this by simply saying that the recompense rewards the worker for 206 C Armstrong, ‘Complex equality: Beyond equality and difference’ (2002) 3(1) Feminist Theory 67, 80. 207 Or, presumably, as part of it. This detail is not made clear in the judgment.

98 Chapter 3 having attended the training day, and compensates them for the lost time and energy; this meaning would seem to apply to any attendee, whoever they are, however long their ordinary working week is, and wherever the training day comes within it. A fundamentum distributionis which made receipt of the recompense conditional on the length of working week/​position of training day would thus offend against this meaning, in breach of complex equality. A (presumably male) full-​time worker could use their full-​time status as a monopolizable badge of honour, which would come to act as a positive dominant. Since the Court this time found the employer’s behaviour to be discriminatory, Lewark is one case in this area where an Aristotelian analysis and a Walzerian analysis would appear to coincide. Finally, in Wippel v Peek & Cloppenburg, the task for the distributive community would be to arrive at a shared understanding of an agreement as to both working hours themselves, and the organization of working hours. Without putting words in the community’s mouth, it might say that such an agreement provides certainty for a worker, enabling them to plan their time better and thus giving them greater stability. As in the previous case, this shared meaning would seem to apply to any type of worker. Thus, a distributive principle which differentiated between those who worked on demand and those who had fixed working hours would appear to be at odds with this meaning, flawing the distribution and violating complex equality. The Walzerian outcome in Wippel diverges from the Court’s, since the Court found no discrimination in this case at all. It may be seen that the Walzerian approach is preferable here. From the moment Ms Wippel makes her complaint,208 the outcome is infinitely more straightforward and predictable than it would be using the circular Aristotelian test as described above. 208 The need for an actual complaint here is important, because, in the abstract, this case can seem perplexing. When the point is made in the text, for example, that the shared meaning considered would seem to apply to any type of worker, a counter-​argument might be that, while someone working on demand who wanted an agreement of this type would no doubt be grateful that they were included in the meaning of such an agreement, why should they want an agreement like this in the first place? Did they not choose work-​ on-​demand to give themselves flexibility in the organization of their time? Why would they want their working week so structured in advance? The answers to these questions are hard. However, once a real complainant comes forward, it is no longer particularly relevant why they might want the agreement. The fact is that they do want it, and they allege that their being denied it is discriminatory. The focus of the lawyer’s enquiry now, therefore, shifts from the reasons why a worker-​on-​demand would want their working hours organized, to whether the organization of their working hours, while no doubt contrary to custom or even inconvenient, is actually impossible (the non-​organization of their

Equal treatment irrespective of gender

99

3.8 Positive Action 3.8.1 The CJEU’s Case-​Law Much of the following case-​law turns on Article 2(4) of the (old) Equal Treatment Directive, and the question of whether or not this constitutes an exception to the general rule laid down in Article 2(1). Article 2(4) reads: This Directive shall be without prejudice to measures to promote equal opportunity for men and women, in particular by removing existing inequalities which affect women’s opportunities in the areas referred to in Article 1(1). In the 1995 case of Kalanke,209 Mr Kalanke and a woman called Mrs Glissmann both applied for the post of Section Manager of Bremen’s Parks Department. Since both candidates possessed the same qualifications, but since women were under-​represented in the sector, the female candidate was given priority on the basis of Paragraph 4 of the Law on Equal Treatment of Men and Women in the Public Service of the Land of Bremen.210 In proceedings brought by Mr Kalanke, the Court of Justice was asked if Article 2(4) of the Equal Treatment Directive covered Paragraph 4 of the German law, and, if not, whether Article 2(1) of said Directive required it to be disapplied. Advocate General Tesauro considered that what was meant by ‘equal opportunities’ in Article 2(4) was equality with respect to starting points, not points of arrival. However, Paragraph 4 of the German law was not designed to guarantee equality as regards starting points; it aimed to achieve equality as to result. Tesauro AG continued: This does not seem to me to fall within either the scope or the rationale of Article 2(4) of the directive.211 The objective of Article 2(4) was substantive equality. Differentiated treatment was permissible under this article, but only the sort of differentiated treatment

209 210 211

working hours being a vital component of the job itself), or merely possible, but considered unnecessary. If the latter, the question to be answered next is whether the inequality which this consideration causes is reasonable (in the sense of well-​reasoned). In other words, the reasons of the discriminator for discriminating are more important than the reasons of the victim of discrimination for wanting the discrimination to end. Case C-​450/​93 Eckhard Kalanke v Freie Hansestadt Bremen [1995] ecr I-​3051. Landesgleichstellungsgesetz. Case C-​450/​93 Kalanke (n 209) Opinion of AG Tesauro, para 13.

100 Chapter 3 needed to neutralize existing difference. Such treatment was discriminatory, but only in appearance; the existing difference legitimized the deviation from formal equality. The Advocate General went on: The rationale for the preferential treatment given to women lies in the general situation of disadvantage caused by past discrimination and the existing difficulties connected with playing a dual role.212 Quota systems, in his opinion, were ‘irrelevant’ to that end.213 What was needed were ‘measures relating to the organization of work’214 and ‘educational guidance and vocational training.’215 Any positive action had to not only derive from an existing obstacle, but also had to be temporary, that is, it would cease once the obstacle was removed. In this case, the German law sought to cure under-​representation. But: such a measure tends merely to rebalance the numbers of men and women, […] it will not remove the obstacles which brought about that situation.216 For these reasons, the Advocate General concluded that the German law was contrary to Article 2(1) of the Directive, and could not be excepted under Article 2(4). The cjeu agreed that the German law was contrary to Article 2(1) of the Directive. As regards the derogation in Article 2(4), the Court declared: National rules which guarantee women absolute and unconditional priority for appointment or promotion go beyond promoting equal opportunities and overstep the limits of the exception in Article 2(4) of the Directive.217 The German law was concerned, not with equality of opportunity, but with equality of result. It could not therefore be saved by Article 2(4).

2 12 213 214 215 216 217

ibid, Opinion of AG Tesauro, para 18. ibid. ibid. ibid, Opinion of AG Tesauro, para 19. ibid, Opinion of AG Tesauro, para 24. ibid para 22.

Equal treatment irrespective of gender

101

Two years later, the Court had a second opportunity to consider the scope of Article 2(4). In Marschall,218 Mr Marschall and a woman candidate both applied for a teaching position. Since both the candidates were equally suitable, and since fewer women than men were employed in the relevant pay bracket, the female candidate was appointed to the position on the basis of the Law on Officials of the Land of North Rhine-​Westphalia (‘the Law on Officials’).219 Mr Marschall brought an action, and the Administrative Court made a reference to the cjeu. Advocate General Jacobs noted that the Law on Officials contained a proviso; where there was under-​representation of women, women were to be given priority in the event of equal suitability ‘unless reasons specific to another candidate predominate.’ However, unlike a large number of the interveners, he did not see that this proviso was sufficient to merit diverging from the rule as laid down in Kalanke. The Law on Officials was discriminatory, and, notwithstanding the proviso, unlawful; it could not be saved by Article 2(4). The Court, meanwhile, held that Kalanke had only outlawed the granting of automatic priority to women. The proviso (or ‘saving clause’) altered the situation. Recalling Paragraph 22 of the Kalanke judgment,220 the Court stated that a guarantee of ‘absolute and unconditional priority’ went beyond the remit of Article 2(4). But a national rule like the Law on Officials, containing as it did a saving clause, did not exceed the limits of the article. That said, of course, if a ‘reason … specific to another [male] candidate’ were allowed to override a female candidate’s priority, that reason could not itself be discriminatory as against women. This important condition became the second limb of the Operative Part of the Judgment. In Badeck,221 forty six members of the Parliament of the Land of Hesse complained about the Hesse Equal Rights Law,222 which called for a large number of positive action initiatives to be put into practice in the public administration. The State Constitutional Court of Hesse made a reference to the cjeu. Advocate General Saggio found all of the initiatives to be non-​discriminatory, with the exception of a rule concerning appointment to collegiate bodies, wherein female candidates were to be granted appointments, irrespective of 218 219 220 221

Case C-​409/​95 Hellmut Marschall v Land Nordrhein-​Westfalen [1997] ecr I-​6363. Beamtengesetz für das Land Nordrhein-​Westfalen. Quoted above (see text accompanying n 217). Case C-​158/​97 Georg Badeck and Others, interveners:  Hessische Ministerpräsident and Landesanwalt beim Staatsgerichtshof des Landes Hessen [2000] ecr I-​1875. 222 Hessisches Gesetz über die Gleichberechtigung von Frauen und Männern und zum Abbau von Diskriminierungen von Frauen in der öffentlichen Verwaltung.

102 Chapter 3 suitability, up to a quota of fifty percent. The Advocate General thought that, since this rule ‘allow[ed] no exceptions,’223 it exceeded the limits of Article 2(4) and could not therefore be saved. The Court meanwhile found none of the initiatives to be discriminatory; the rule concerning collegiate bodies was (according to it) ‘non-​mandatory.’224 Positive action was also touched upon in the unusual case of Schnorbus.225 Here, a woman who had successfully passed the First State Examination in Law complained when she was passed over twice for admission to the practical legal training course. She was then forced to defer for some five months, pursuant to the Law on Legal Training,226 which permits the authorities to defer a person’s appointment for up to twelve months. She alleged that the combined effect of the Law on Legal Training and its implementing regulation227 was that those who had completed military service (who were necessarily all male) were able to take advantage of an exception to the deferral rule; they could plead ‘particular hardship’ and gain automatic acceptance. The Administrative Court sought a reference. Advocate General Jacobs’ conclusion was that this was indirect discrimination, but that it was objectively justified. Those who had undertaken military service had had the commencement of their legal studies delayed by approximately one year; the German measures compensated for this delay. Because the measures were justified, there was no need to consider them in the light of Article 2(4). However, interestingly, the Advocate General did not rule out the possibility of measures like those at issue (giving preferential treatment to men) coming within the scope of Article 2(4):  ‘the reference to existing inequalities which affect women’s opportunities [in Article 2(4)] is merely an example’.228 The Court agreed that this was indirect discrimination, and that it was justified by objective reasons. Finally, one might consider efta Surveillance Authority v Norway.229 Although this book is concerned with the Court of Justice of the European

223 224 225 226 227 228

229

Case C-​158/​97 Badeck (n 221) Opinion of AG Saggio, para 42. ibid para 65. Case C-​79/​99 Julia Schnorbus v Land Hessen [2000] ecr I-​10997. Juristenausbildungsgesetz. Juristische Ausbildungsordnung. Case C-​79/​99 Schnorbus (n 225) Opinion of AG Jacobs, para 53. Note that Article 141(4), a result of the revision of the EC Treaty carried out at Amsterdam, was sex neutral, as is its successor in the Treaty on the Functioning of the European Union, Article 157(4) tfeu. Case E-​1/​02 efta Surveillance Authority v The Kingdom of Norway [2003] efta Court Reports 1.

Equal treatment irrespective of gender

103

Union, rather than the efta Court, this case is instructive in that it contains discussion (and ultimately endorsement) of the way in which the cjeu construes Article 2(4) of the Equal Treatment Directive. In it, the efta Surveillance Authority brought an action against Norway claiming that, by applying its legislation so as to reserve a certain number of academic positions exclusively for women, Norway had failed to fulfill its obligations under the eea Agreement. The law in question was Article 30(3) of the Norwegian Act No. 22 relating to Universities and Colleges, the last sentence of which read: ‘The Board can decide that a post shall be advertised as only open to members of the underrepresented sex’.230 Before the efta Court, Norway requested that a different approach be taken to that of the cjeu, under whose interpretation of the Equal Treatment Directive, ‘affirmative action measures are legally defined as derogations from the prohibition on discrimination’.231 Rather, they should be viewed as ‘an intrinsic dimension’ of the prohibition.232 Norway admitted that its law permitted ‘automatic and unconditional preference’ for women,233 but argued that this was what substantive equality demanded. The efta Court, however, would not diverge from the cjeu’s approach. Article 2(4) of the Directive was a derogation, and absolute and unconditional priority (that is, the substituting of equality of representation for equality of opportunity) was forbidden. The last sentence of the Norwegian law at issue gave unconditional and automatic priority to women candidates, and such a measure could not fall within Article 2(4). Norway had therefore failed to fulfill its obligations. 3.8.2 Preliminary Analysis The debate on positive action is very much a continuation of the one considered above in the context of the question of part-​time and full-​time work. The principal similarity between the two debates is the shift which positive action entails between individual rights and group rights. This time, however, the Court is noticeably more cautious. Positive action may involve the needs of a group (for example, female managers, female teachers or male lawyers who have undertaken military service) overriding the needs of given individuals (for example, Mr Kalanke, Mr Marschall or Ms Schnorbus). If such overriding does indeed take place (as in the cases of Mr Marschall and Ms Schnorbus), 230 231 232 233

ibid para 2. ibid para 25. ibid. ibid para 26. And not just those women who were equally as qualified as, or better qualified than, men: see ibid para 45 (last sentence).

104 Chapter 3 the individuals in question must satisfy themselves with the thought that their misfortune served a greater good. But this raises questions, in particular as regards the Kantian ‘Categorical Imperative.’ Are such people now being treated simply as a means? The ‘individual versus group’ debate has generated a large volume of literature. In much of this, comparator-​driven, ‘individualistic,’ formal equality is condemned, and group-​oriented, substantive equality is often called for in its stead.234 As far as this particular group of case-​law is concerned, Marschall is regarded as the high-​water mark, described by Hervey as, ‘the closest the Court comes to recognising a substantive approach to equality.’235 Meanwhile, Kalanke has been described as ‘confused,’236 Schnorbus as ‘dismally confused,’237 and efta as ‘simplistic.’238 As with the debate on full-​time and part-​time work, full justice cannot be done to the debate on positive action in this book. There are compelling arguments on both sides, put forward in response to the complicated questions thrown up by the debate. These questions include whether there can ever be such a thing as a ‘debtor sex,’ owing ‘compensation’ to a ‘creditor sex’ for the historical wrongs done to the latter.239 If so, for how long must this compensation be paid?240 If indefinitely, is there not a risk that the 234 Lacey (n 35); Catherine Barnard, ‘The Principle of Equality in the Community Context: P, Grant, Kalanke and Marschall: Four Uneasy Bedfellows?’ (1998) 57 clj 352. For more on the weaknesses of ‘individualism’, see S Fredman, ‘Changing the Norm: Positive Duties In Equal Treatment Legislation’ (2005) 12 mj 369, 370ff. 235 Hervey (n 193) 316. 236 Hervey and Shaw (n 6) 59. 237 Costello and Davies (n 191) 1598. 238 Holtmaat and Tobler (n 4) 421. 239 The phrase ‘creditor or debtor … sexes’ can be found in Mancini and O’Leary (n 76) 345, referring to the concurring opinion of Scalia J in the us case of Adarand v Peña 515 US 200 (1995). For more discussion of the ‘compensation arguments,’ and the possible replacement of ‘backward-​looking’ strategies with ‘forward-​looking’ ones, see M Tomei, ‘Discrimination and equality at work: A review of the concepts’ (2003) 142(4) Intl Labour Rev 401. 240 It will be remembered that Tesauro AG in Kalanke insisted that positive action measures must be temporary. Also taking this position is Docksey: ‘The State would have to provide for monitoring the situation to ensure that single-​sex recruitment is brought to an end when a better balance is achieved’. See C Docksey, ‘The European Community and the promotion of equality’ in C McCrudden (ed), Women, Employment and European Equality Law (Eclipse 1987) 17. In Allen’s view, ‘time limiting affirmative action ensures that it is focused and more acceptable to those who cannot take the benefit of the action’: R Allen, ‘Article 13 ec, evolution and current contexts’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 50.

Equal treatment irrespective of gender

105

once over-​represented group will itself become under-​represented in due course?241 Furthermore, does such an arrangement only give rise to benefits for the creditor sex, or are there potential detriments, which should also be considered?242 Finally, returning to the question at the very beginning of this section, there is to this day phenomenal confusion over the issue of whether positive action is or is not an exception to the general rule against sex discrimination.243 Holtmaat and Tobler comment that it should in fact be regarded as ‘[sex equality’s] other, positive side.’244 The latter two writers, as well as Andrews245 and of course the Norwegian delegation in efta, draw attention to the (arguably) preferable wording to be found in the un Convention for the Elimination of 241 See Debra Franzese, ‘The Gender Curve:  An Analysis of Colleges’ Use of Affirmative Action Policies to Benefit Male Applicants’ (2007) 56 Am UL Rev 719. Franzese describes a significant under-​representation of men on us college campuses. It is not clear, though, whether this is the result of earlier positive action in favour of female students, or simply a natural increase in academic ambition on the part of women. 242 One detriment could be that women who have been appointed to posts on the basis of a quota will be viewed as in some way inferior. As Evelyn Ellis puts it, positive discrimination ‘would be patronizing to women and would undermine their achievements’: Ellis (n 191) 400. Docksey refers to the ‘possibility [for women] of being stigmatised as inferior candidates’: Docksey (n 240) 19. See also Tomei: ‘Preferential treatment … acts as a disincentive for members of the beneficiary groups to improve their skills [and] erodes their perceived competence in the eyes of society’ (supra n 239, at 413). 243 Wentholt accuses the Court of ‘undermin[ing] the effectiveness of affirmative action’ by treating it as a derogation: Wentholt (n 191) 60. Docksey also regards it as a derogation: supra n 240, at 17. Barnard calls for a ‘reconceptualis[ation]’ of discrimination such that positive action would no longer be regarded as an exception: Barnard (n 234) 371–​ 2. Masselot notes a conflict within the wording of the (later abandoned) Constitutional Treaty; Article ii-​83 presents positive action as an exception to gender equality, while Article iii-​214(4) presents positive action as part of gender equality: Annick Masselot, ‘The State of Gender Equality Law in the European Union’ (2007) 13(2) Eur LJ 152. Under the Lisbon Treaty, Article 141 ec remains –​with one technical exception –​untouched, although renumbered as Article 157; the new Article 157(4) is identical to (what would have been) Article iii-​214(4). Also, Article 6 teu now gives binding legal effect to the Charter of Fundamental Rights of the European Union, Article 23 of which is identical to (what would have been) Article ii-​83. So in fact the conflict identified by Masselot has not gone away. Numhauser-​Henning, too, observes that the Charter’s wording is ‘more narrow’ than that of the Treaty: Numhauser-​Henning (n 196) 153. 244 Holtmaat and Tobler (n 4) 414. 245 Jill Andrews, ‘National and International Sources of Women’s Right to Equal Employment Opportunities:  Equality in Law Versus Equality in Fact’ (1994) 14(2) Northwestern Journal of International Law and Business 413.

106 Chapter 3 All Forms of Discrimination Against Women (cedaw). However, as the efta Court (and Advocate General Jacobs in Marschall) pointed out, this Convention has for the moment no binding effect. 3.8.3 A Walzerian Analysis It is clear that, when it comes to positive action, it is better to take a broad approach to Walzer than a narrow one. Where such a complicated and pervasive social issue as gender inequality is at stake, it is necessary to contemplate the interplay of every sphere of justice with every other one, and the matter must be looked at in relation to the whole sphere and not just one particular element within it.246 In the context of the unequal treatment of women, Walzer entertains positive action (or affirmative action as it is known in the us), and furthermore he sees it as a component of, rather than an exception to, complex equality. He uses here the argument mentioned above in the context of part-time and full-time work,247 that sex discrimination may itself be a boundary breach, as notions of gender ‘roles’ within the family (part of what he calls the ‘structures of kinship’) are reiterated in other distributive spheres, unjustly locking women into stereotypes that do not have the slightest connection to these other spheres. As he puts it, The real domination of women has less to do with their familial place than with their exclusion from all other places … [W]‌hat is most important right now is that the market, as it actually functions and as we

246 If the narrow approach is insisted upon, then the results may be disappointing. This is because, on a more microscopic examination of the individual cases, disparities between (potential) shared understandings and the distributive criteria actually used do emerge. For example, there is no element of academic or managerial merit in Hesse’s distribution of college appointments in Badeck, while, without wishing to pre-​empt the particular distributive community at issue, a common understanding of such an appointment would be that the holder possessed some degree of aptitude either for teaching or for management (depending on the exact role in question). Thus recruitment strategies based entirely on positive action may, on the narrow reading, find themselves at odds with Walzer’s theory, and it is accepted that, in these circumstances, positive discrimination is an area to which complex equality has less to offer. However, the reader is reminded of the important comment made at section 1.3. above, that the theory is proffered here only as a potential new tool to be added to the judicial tool-​box, not as an all-​encompassing theory operating to the exclusion of all others. 247 See supra, text accompanying n 206.

Equal treatment irrespective of gender

107

understand its functioning, sets no internal bar to the participation of women. It is focused on the quality of goods and on the skill and energy of persons, not on kinship standing or sex.248 And as he nicely sums it up in a later work, patriarchy and male dominance are […] examples of distributive simplicity.249 Okin has probably gone the furthest of all commentators to develop this argument. While critical of Walzer’s use of ‘shared understandings’ as the basis for distribution,250 she praises his view that the various spheres of justice should be separated: Walzer’s theoretical framework –​separate spheres having to allow for different inequalities to exist side by side only insofar as a situation which he calls ‘dominance’ is not created –​has considerable force as a tool for social, and particularly for feminist, criticism.251 She continues that the implications of the separate spheres criterion for justice suggest what many feminists have been arguing […]: the unequal distribution of rights, benefits, responsibilities, and powers within the family is closely related to inequalities in the many other spheres of social and political life. There is a cyclical process at work, reinforcing the dominance of men over women, from home to work to what is conventionally referred to as the ‘political’ arena, and thence back home again.252 Women are still ‘designated and defined by their position within the ­family,’253 whereas: 248 Walzer, Spheres of Justice (n 9) 240–​1. 249 M Walzer, ‘Response’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) 290. 250 Walzer later confirmed that that was a principal point of divergence between them: M Walzer, ‘Feminism and me’ (2013) 60(1) Dissent 50. 251 SM Okin, Justice, gender, and the family (Basic Books 1989) 112. 252 ibid 113. 253 ibid.

108 Chapter 3 [s]‌ocial goods must be distributed in accordance with their own relevant reasons, not determined by women’s familial roles.254 If it were considered that, say, the Parks Department of Bremen, or even Bremen’s entire local administration, was characterized by an over-representation of male workers, then it might be concluded that recruitment procedures over the years had been biased in favour of men and had sought to reiterate structures of kinship (male ‘provider’/​breadwinner, female carer) within the sphere of office. As this would be a violation of complex equality, it would then be appropriate to bring the violation to an end by ensuring, henceforth, a fifty-​fifty gender divide within that Department/​administration. But if it were further considered that, given the engrained and institutional nature of the bias, preferential treatment should be accorded to women in the short term in order to bring about a fifty-​fifty gender divide in the long term, then, logically, that too should be sanctioned by complex equality. On that analysis, the Court reached the wrong decisions in Kalanke and efta, although the right ones in Marschall and Badeck. The Walzerian result in Schnorbus would depend on how the Court read the preferential treatment accorded to male Bar candidates in that case. If such preferential treatment, which was intended to make up for a year lost to military service, in fact perpetuated structures of kinship within the sphere of office, either by reiterating supposed gender ‘roles’255 in inappropriate fora, or simply by creating more male lawyers than female ones, then Ms Schnorbus was in the right in impugning it. If, on the other hand, it itself was a method of redressing an otherwise inevitable over-­ representation of women in the legal profession, then it could be condoned as a means of maintaining, not disrupting, the autonomy of the sphere in question. 3.9

Insurance Premiums and Benefits

3.9.1 The CJEU’s Case-​Law Following the adoption of a major new directive in 2004, the fight against sex discrimination moved for the first time outside of the area of employment and 254 ibid 114. The theory which Okin goes on to develop, the ‘abolition of gender’ (ibid 116), is by her own admission one which ‘would certainly not be agreed upon by all as desirable’ (ibid 180). Nevertheless, this is, in her opinion, the logical conclusion to Walzer’s theory, inasmuch as it applies to the gender issue. She thus goes further than the view expressed above that respective ‘roles’ of family members (if any) should be restricted to the sphere of kinship, by asserting that they should be abolished even from there. 255 The male as ‘fighter’, for example.

Equal treatment irrespective of gender

109

professional life, and into the area of the supply of, and access to, goods and services.256 This new legislation was bound to cause the Court of Justice to expend some new interpretative energy. However, so far, aside from infringement actions brought by the Commission, this energy has been entirely focused on one provision of the directive:  Article 5(2), permitting Member States to allow gender discrimination in the field of insurance premiums and benefits. This reads: Notwithstanding paragraph 1, Member States may decide before 21 December 2007 to permit proportionate differences in individuals’ premiums and benefits where the use of sex is a determining factor in the assessment of risk based on relevant and accurate actuarial and statistical data. The Member States concerned shall inform the Commission and ensure that accurate data relevant to the use of sex as a determining actuarial factor are compiled, published and regularly updated. These Member States shall review their decision five years after 21 December 2007, taking into account the Commission report referred to in Article 16, and shall forward the results of this review to the Commission. In the leading case of Test-​Achats,257 the Association Belge des Consommateurs Test-​Achats, a Belgian consumer association, along with two private individuals, sought the annulment of the Belgian legislation which transposed Article 5(2) of Directive 2004/​113 into Belgian law, claiming that this law was contrary to the principle of equality between men and women. The Belgian Constitutional Court made a reference to the Court of Justice. Delivering her Opinion in late 2010, Advocate General Kokott refused to accept that statistical differences between men and women reflected actual differences, at the level of ‘each individual.’258 To reach such a conclusion was, she said, to indulge in a ‘sweeping assumption:’ There is then a sweeping assumption that the different life expectancies of male and female insured persons, the difference in their propensity to take risks when driving and the difference in their inclination to utilise medical services –​which merely come to light statistically –​are 256 Council Directive 2004/​113/​EC of 13 December 2004 implementing the principle of equal treatment between men and women in the access to and supply of goods and services [2004] oj L373/​37 (‘Directive 2004/​113’). 257 Case C-​236/​09 Test-​Achats (n 196). 258 ibid, Opinion of AG Kokott, para 62.

110 Chapter 3 essentially due to their sex. In fact, however, as Test-​Achats has submitted without being contradicted, many other factors play an important role in the evaluation of the abovementioned insurance risks.259 The Advocate General thus held that Article 5(2) of the Directive was incompatible with the principle of equality and non-​discrimination, a decision in which she was followed, via slightly different arguments, by the Court in 2011. In the subsequent Finnish case of X,260 the male plaintiff, who had suffered a workplace accident, sought an extra sum of eur 278.89 in order that his compensation might attain the same level as that recoverable by a female worker in the same situation. Although the eu legislation at issue this time was the Social Security Directive,261 the Court was still satisfied that the Finnish accident insurance scheme represented a breach of the equality principle, and it advised the national court, in considering state liability, to bear in mind the Test-​Achats judgment with regard to whether or not Finland was liable for this breach. The case is thus an important reiteration of the Test-​Achats principle. 3.9.2 Preliminary Analysis Test-​Achats is a fascinating case which straddles the worlds of what is and what could be. As a preliminary point, the difference between the Advocate General’s Opinion and the Court’s judgment seem, for the purposes of this study, to be cosmetic only. The Advocate General had deemed Article 5(2) incompatible with eu human rights law altogether, while the Court implied that it had been compatible up to the date of the judgment, and would be for a short period thereafter, but held it to be incompatible from the date of 21 December 2012 onwards. The Advocate General’s reading of the situation is preferred to the traditionally elliptical pronouncement from the judges, and what is particularly interesting from the point of view of this study is the way in which she noted the complexity of the matter in hand –​the distribution of insurance premiums and benefits to individuals.262 In doing so she offered a glimpse of what 2 59 260 261 262

ibid, Opinion of AG Kokott, paras 61–​62. Judgment of 3 September 2014, Proceedings brought by X, C-​318/​13, ECLI:EU:C:2014:2133. Directive 79/​7 (n 109). ‘Admittedly, it is especially easy to implement distinctions on the basis of sex in respect of insurance products. The correct recording and evaluation of economic and social conditions and of the habits of insured persons is much more complicated and is also more difficult to verify, particularly since those factors may be subject to changes over time.’ See Case C- 236/ 09 Test- Achats (n 196) Opinion of AG Kokott, para 66 (emphasis added).

Equal treatment irrespective of gender

111

a Walzerian approach to premium-​calculation and benefit-​allocation might look like. Does the embracing of a proto-​Walzerian approach by a member of the Court spell a clean bill of health for the Court’s equality case law? The answer is no unfortunately, and this is due to the Court’s clinging to the Aristotelian test. Taking things step by step, the Court, at paragraph 28 of its judgment, recites the test. Then, at paragraph 29, it holds that: the comparability of situations must be assessed in the light of the subject-​matter and purpose of the eu measure which makes the distinction in question.263 Respectfully this seems to be the last nail in the coffin of Aristotle. If the difficulties of working out who and what was unlike, and who and what alike, was bad enough, now the reader is told that this can be carried out in the light of the avowed purpose of the measure, and not according to some extant, objective standpoint. Context is good, but focusing on the goal of the measure (and not the nature of the distribuend) is precisely the wrong context to pick, as a quick glance at the first line of the measure (in some jurisdictions even its name) will be enough for the judge to be able to engineer the comparison so that the like/​unlike decision is predetermined.264 To take a silly example, an act for the segregation of violinists and viola players, challenged on the grounds of breach of the principle of equality by one of these instrumentalists, would be condoned when the tribunal hearing the case (adopting the Court’s current approach) ruled that the parties were unlike, and being treated unlike, their lack of likeness forming the very ‘premiss’ (the Court’s word in Test-​Achats)265 of the act being contested. It would be easy to think of more serious examples than this. By close analysis of other provisions of the Directive, the Court is then able to satisfy itself that, the very premiss of the Directive is that ‘the respective situations of men and women with regard to insurance premiums and benefits contracted by them are comparable’.266 Article 5(2)’s advocacy of different 263 264

265 266

ibid para 29. See supra, section 1.2, for more examples of engineered comparisons. Brillat refers to the case as demonstrating ‘comparabilité prédeterminée:’ M Brillat, Le principe de non-​ discrimination à l’épreuve des rapports entre les droits européens (Fondation Varenne 2015) 151. Case C-​236/​09 Test-​Achats (n 196) para 30. ibid.

112 Chapter 3 treatment for comparable individuals thus breaches the Aristotelian principle, at least if left unchecked temporally; it must therefore be scrapped at the end of 2012.267 3.9.3 A Walzerian Analysis What would the Walzerian judge do? Fairly obviously, and as in fact already mentioned above, the distribuends are insurance premiums and benefits. Two hypothetical insured persons, one male and one female, enter the sphere in which this distribution is to take place, but must enter with their hands empty of everything save prior goods having to do with the meaning of insurance premiums and benefits. The usual disclaimer applying, life assurance (or pension insurance) is about contributing to a fund of money from which the insured person’s living expenses (or those of their survivors) are to be drawn in the future. Third party motor vehicle liability insurance is about contributing to a fund of money from which the innocent victim of a car accident caused by the insured person should be compensated. In the case of private health insurance, the fund of money to which contributions are made will be used to pay for any medical care needed by the insured person for so long as the policy is maintained. In all cases, the size of the contributions (in other words, the premium) must vary proportionately with the size of the risk that the fund will have to pay out; it seems self-​evident that those resorting to a fund more often (or for longer) should be greater contributors thereto. The characteristics possession of which should therefore be permissible in the sphere in question are, respectively, life expectancy, likelihood of crashing and likelihood of using the relevant services.268 The insured person’s gender is simply not on the list, and therefore cannot come into the sphere.269 As will be described in more detail later,270 the Walzerian approach is good at snarling assumptions; here that is seen in action. Assumptions perforate unsupervised boundary walls and pretty soon transform into foregone 267 Brillat also regards the Court here as having simply ‘reimposed’ the wishes of the legislator: Brillat (n 264) 152, translation author’s own. 268 This list drawn directly from that given by Kokott AG at paragraph 45 of her Opinion: Case C- 236/ 09 Test- Achats (n 196) Opinion of AG Kokott, para 45. 269 At any rate it is not on Advocate General Kokott’s list. Others may disagree, at least as regards some types of insurance contract. See for example P Watson, ‘Equality, fundamental rights and the limits of legislative discretion’ (2011) 36 el Rev 896 and E Caracciolo di Torella, ‘No sex please, we’re insurers’ (2013) 38(5) EL Rev 638. The final decision would be the distributive community’s; if it decided that for a certain type of contract gender was part of its meaning, then of course gender could legitimately be brought into the sphere in question. 270 See section 4.4.2 below.

Equal treatment irrespective of gender

113

conclusions:  ‘x happened to you yesterday, therefore y must happen to you today, and z will happen to you tomorrow’. On the other hand, boundary-​patrol of an insurance sphere will ensure that the distribution is performed ‘in a vacuum’ requiring only what the distribuend itself requires, and is not skewed by earlier choices, or gifts from Mother Nature or Fortuna. However, one problem arises. This is that, with the assumptions gone, the distribution becomes a much more ‘complicated’ business, to use Kokott AG’s word.271 The Advocate General is right to flag this up and also to note that the complication (or complexity) of a situation does not justify taking short cuts: Practical difficulties alone do not however justify the use, to an extent for reasons of convenience, of the insured person’s sex as a distinguishing criterion.272 And so readers of Test-​Achats catch the glimpse, mentioned above, of what a Walzerian approach to premium-​calculation and benefit-​allocation might look like. Decisions are taken case by case, with the myriad factors which really determine, say, life expectancy, investigated one by one –​the insured person’s caffeine consumption, their occupation and its attendant stress levels, the sports they may (or may not) practice273  –​a less elegant approach274 but a fairer one, a problem only to those whom simplicity would have favoured.275 Turning to X, it is clear that the Finnish government was concerned that, as, statistically, women live longer than men, female workers should receive a larger disability allowance in the event of a workplace injury than male workers, as the former may need this money to last longer. The Court dismissed this as a ‘generalization.’276 The Standard Contingent Reply taken as read, it seems likely that a Walzerian distributive community would see predicted length of life as a

271 272 273 274 275

276

C-​236/​09 Test-​Achats (n 196) Opinion of AG Kokott, para 66. ibid. List again inspired by Kokott AG’s: ibid, Opinion of AG Kokott, para 62. Paraphrasing Walzer’s assessment of his own theory. See Walzer, Spheres of Justice (n 9) 21. Caracciolo di Torella, though, sounds a note of warning that this more personalized investigation may be ‘more intrusive on [an] individual’s personal sphere’ and ‘more costly’: Caracciolo di Torella (n 269) 653. She is also unconvinced that gender is always an assumption-​based proxy in the field of insurance and holds that there may be some types of insurance contract where gender ‘could be a determining factor:’ ibid 652. Likewise Watson believes that the ‘non-​discriminatory nature of unisex premiums’ may be a ‘fundamentally flawed conclusion’: Watson (n 269) 902. Case C-​318/​13 X (n 260) para 38.

114 Chapter 3 permitted good within the sphere of this kind of accident insurance. However, as with Test-​Achats above, it is less likely that this community would see gender as a non-​tyrannical good, surely preferring instead the myriad lifestyle factors which genuinely form the basis of said prediction. 3.10

Concluding Thoughts

At the end of the first stop in the tour, then, the reader may be forgiven for having mixed emotions about the project thus far. On the one hand, the application of Walzerian analysis seemed capable of bringing about greater consistency,277 greater predictability and less confusion278 in the Court’s case-​law concerning gender equality, of offering a solution to the ‘noncomparative discrimination’ problem,279 of facilitating a more robust scrutiny of ‘objective justifications’280 and of allowing for a more compassionate handling of cases like Helmig.281 On the other hand, though, usage of a Walzerian test threw up difficult questions for a Forum,282 or the judges themselves,283 to resolve, and, perhaps above all, positive action –​a new way of thinking about sphere-​entry to rectify an old way of thinking about sphere-​entry –​seemed destined to fall foul of complex equality, stricto sensu, with its rigid boundary defence, unless the Forum itself could agree that such rectification formed part of the meaning(s) of the good(s) distributed. These problems will be taken in turn. Turning first to the Forum, it could be noted that, on the face of it, Walzer’s discursive approach to equality does have the merit of widening potential discussion to include under-represented groups. The public sphere has in the past (and depending how defined) helped to generate and disseminate alternative viewpoints, crucial to (among other movements) feminism.284 277 As between, for example, Case C-​342/​93 Gillespie (n 58), and the other pregnancy-​related illness cases, like Case C-​394/​96 Brown (n 47), and Case C-​66/​96 Pedersen (n 50). 278 In cases like Case C-​177/​88 Dekker (n 38), Case C-​32/​93 Webb (n 40), and Case C-​421/​92 Habermann-​Beltermann (n 43). 279 In a case like Case C-​191/​03 McKenna (n 54). 280 In a case like Case 170/​84 Bilka (n 151). 281 Joined Cases C-​399/​92, C-​409/​92, C-​425/​92, C-​34/​93, C-​50/​93 and C-​78/​93 Helmig (n 166). 282 Walzerian complexity. 283 Mediated complexity; see below, Chapter 8. 284 ‘[H]‌istorically the best feminist manifestos emerged from collective consciousness raising and discussion,’ comments bell hooks, in criticising the ‘lone queen’ approach to feminism of Sheryl Sandberg: bell hooks, ‘Dig Deep: Beyond Lean In’ (The Feminist Wire, 28 October 2013)  accessed

Equal treatment irrespective of gender

115

It is in the precise nature of the discussion, though, and the identity of the participants thereto, that the approach may become problematic. Today, society faces a serious decline in social democracy, indeed attacks in some countries on democracy itself, mirrored –​and in the context of a financial crisis and severe austerity politics –​by a serious rise in far right parties, outwardly racist and outwardly hostile to gender equality and lgbtqi equality too. Thus, merely replacing ‘top down’ with ‘bottom up’ cannot be regarded as a part of a solution to an equality deficit, where certain grassroots activists are themselves spreading such an illiberal message of hatred and intolerance. The alternative theory of mediated complexity, set out later in the book, may provide certain guarantees against the poisoning of shared meanings by such activists, although they may still bring their message to court, and, even if they do not, an illiberal bias on the part of the judge may leave litigants no better off. Turning now to the issue of the potential discord between positive action and complex equality, Amy Guttman, for one, has noted this weakness in Walzer’s theory, in relation to both race and gender, and has called for ‘[j]‌ustice across the spheres.’285 In relation to the meaning of goods (specifically jobs) which includes rectification of historical prejudice, she writes: This meaning is less popular … but it does not therefore have any less claim to govern the distribution of employment here and now.286 28 May 2018. In her celebrated response to Habermas, Nancy Fraser describes the ‘late twentieth-​century us feminist subaltern counterpublic, with its variegated array of journals, bookstores, publishing companies, film and video distribution networks, lecture series, research cent[re]s, academic program[me]s, conferences, conventions, festivals, and local meeting places.’ Crucially, however, she was not referring to ‘the public-​at-​large,’ but rather to a space for withdrawal and regrouping, and for training for ‘agitational activities directed toward [the] wider publics:’ N Fraser, ‘Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy’ (1990) 25/​26 Social Text 56, 67–​8. Meanwhile, a ‘single, comprehensive, overarching public’ would tend not to provide a venue for women’s participation, or for the exposure of domination, as it would be and remain ‘under the supervision of [the] dominant groups:’ ibid 66. However, other parts of Walzer’s theory, such as his focus on the social construction, and reconstruction, of meanings, could perhaps play a part in the incorporation of gender into stale, gendered debates. See further JA Tickner, Gendering world politics:  issues and approaches in the post-​Cold War era (Columbia University Press 2001) 27ff, referencing Whitworth. 285 A Gutmann, ‘Justice across the Spheres’ in Miller and Walzer (n 249) 99. 2 86 ibid 106.

116 Chapter 3 Of a distribution carried out in disregard of that meaning, she writes: This is unfair, and this unfairness is not adequately addressed by the claim of complex equality that careers, because they are careers, should be distributed according to talent … We need to add an important caveat to the claims of complex equality:  All goods should be distributed according to their social meanings if all other goods that serve overlapping social purposes are also distributed.287 Noting that non-​discrimination is an interspherical ideal, and warning that dominant social meanings will inevitably favour dominant groups, Guttman concludes: What complex equality characterizes as competing social meanings now looks more like a moral disagreement, and the moral disagreement enlists considerations that are not specific to the sphere of employment … In the context of an unjust society, justice is more interspherical than complex equality leads us to expect.288 On the face of things, Walzer’s theory has the potential to break sexual stereotyping, by fencing off irrelevant concerns, and unfencing relevant ones. However, there is a danger, to which Gutmann rightly alludes, that, as part of the fencing off, important moral considerations like fairness and the dignity of persons –​which crucially are not only not sphere-​specific, they are not good-​specific –​fall, or are pushed, by the wayside. This cannot be answered by filtering them back through the same machine, as the next things to be distributed, as ‘[t]‌hey are not … contingent cultural facts, which we just happen to hold and can readily change at our collective will’.289 From the necessary legal point of view of this book, that would seem to be an argument in favour of the Override, whereby non-​discrimination, a ‘shared ideal,’ certainly, but on this argument also a non-​contestable one, would ‘inform’ other shared understandings.290 Alternatively, while Gutmann stops short of declaring any of the foregoing as master principles, dominating all distributions without more, entrenched meanings, there seems no reason at all why a given society wishing to adapt 287 288 289 290

ibid 107. ibid 110–​11. ibid 119. ibid

Equal treatment irrespective of gender

117

complex equality for the purposes of the law, itself a challenge to Walzerian orthodoxy, should not adjust the tenets of the theory to waive sphere autonomy for any purpose it considered desirable; the overcoming of historical gender discrimination could be one such purpose. Indeed, as will be seen later, the entrenchment of certain meanings may be a necessary societal corollary of the adaptation itself.291 Opsahl wrote that affirmative action is justified by the fact that it advances class, or group, equality.292 With this in mind, one could contemplate an even bolder departure from Walzerian convention (at least on the face of it), by questioning why it is that only individuals move across sphere boundaries and not classes or groups. Were the latter to cross borders as one, there would seem to be no reason why they could not contest distributions  –​habitually, and tyrannically, skewed against them –​as one. Further, the effect of such a crossing would be to create, figuratively speaking, either a sphere within a sphere, or a trans-​sphere collective. Walzer’s theory favours both of these, regarding the latter as vital for fomenting cooperation and mitigating inequality,293 and, as regards the former, even requiring those running the macro sphere to extend advantages to the members of the micro sphere in order to ensure the equilibrium of co-​existence. This will be investigated further below.294 291 Below, ­chapter 9. 292 Torkel Opsahl, ‘Equality and non-​discrimination’ in T Opsahl, Law and equality: Selected articles on human rights (Ad Notam Gyldendal 1996) 194. 293 M Walzer, ‘Governing the globe: what is the best we can do?’ (2000) 47(4) Dissent 44, 47–52. 294 Below, section 9.5.

Chapter 4

The ‘Article 19’ Grounds: Racial or Ethnic Origin, Religion or Belief, Disability, Age, and Sexual Orientation 4.1 Introduction The insertion of a new Article 13 into the ec Treaty, via the Treaty of Amsterdam in 1997, and the subsequent adoption of two major Directives,1 gave the cjeu significant new tools with which to investigate discrimination based on five additional grounds: racial or ethnic origin, religion or belief, age, sexual orientation, and disability. Since the coming into force of the Treaty of Lisbon on 1 December 2009, Article 13 ec has been replaced by Article 19 tfeu, which –​ with a few technical exceptions –​is identical. It is of course wrong to suggest that these statuses were invented at Amsterdam; people were black and white, gay and straight, old and young, and so on, before 1997. It is just that Article 19 and its daughter Directives allowed differential treatment, grounded on such attributes, to be much more robustly objected to. Nevertheless, such treatment (as long as it fell within the ambit of the Treaty) could still have been objected to prior to the amendment by dint of the all-​encompassing ‘general principle’ of equality, the residual role of which, as will be seen, is a matter of some confusion.2

1 Council Directive 2000/​43/​e c of 29 June 2000 implementing the principle of equal treatment between persons irrespective of racial or ethnic origin [2000] oj L180/​22 (‘Directive 2000/​43’ or the ‘Race Equality Directive’) and Council Directive 2000/​78/​e c of 27 November 2000 establishing a general framework for equal treatment in employment and occupation [2000] oj L303/​16 (‘Directive 2000/​78’ or the ‘Framework Employment Directive’). 2 See Case C-​144/​04 Werner Mangold v Rüdiger Helm [2005] ecr I-​9981, and the comments on Mangold in Advocate General Sharpston’s Opinion in Case C-​427/​06 Birgit Bartsch v Bosch und Siemens Hausgeräte (bsh) Altersfürsorge GmbH [2008] ecr I-​7245. And even before the adoption of the Directives, the general principle was, it seemed, of limited use. For example, it could not be used to admonish Member States which refused to treat same-​ sex relationships (or even registered partnerships) as equivalent to marriage (or opposite-​ sex relationships outside marriage): Case C-​249/​96 Lisa Jacqueline Grant v South-​West Trains Ltd. [1998] ecr I-​621, Joined cases C-​122/​99 P and C-​125/​99 P D and Kingdom of Sweden v Council of the European Union [2001] ecr I-​4319.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/9789004354265_005

The ‘Article 19’ grounds

4.2

119

Racial/​ Ethnic Origin

4.2.1 The cjeu’s Case-​Law A good starting point for this section would be to clarify, at the outset, the differences between ‘racial or ethnic origin’ and ‘religion or belief,’ including how discrimination on these two grounds is not currently prevented to the same extent. It should not be thought that ‘racial or ethnic origin’ and ‘religion or belief’ are the same thing. Religion, if anything, overlaps with race. As Brown points out: In most cases, discrimination on grounds of belonging to a minority religion will disproportionately affect people belonging to an ethnic minority; in other words, such religious discrimination will constitute indirect race discrimination.3 Drawing a boundary between the two ideas can be ‘problematic.’4 Lester has made an impassioned plea (albeit in the uk context) that race and religion must be treated separately, pointing to a number of significant differences between the two concepts (for example, a person’s race may be visible, while his or her religion may be invisible).5 In the Jewish context, he goes on to explain, religious anti-​Semitism and racial anti-​Semitism are entirely distinct; they were prevalent at different times, and those affected by the second would not necessarily have been affected by the first.6 The separateness of the two notions has certainly not been lost on the Community institutions, which chose to leave religion to sit alongside the other grounds, while giving race its own Directive, entailing noticeably greater protection: the Race Equality Directive.7 The scope rationae materiae of the Race Equality Directive, which is set out in Chapter I  thereof, embraces not only employment, but also social protection (including social security and healthcare), social advantages, education, and access to and supply of goods and services which are available to the public (including housing).8 Discrimination 3 Christopher Brown, ‘The race directive:  towards equality for “all” the peoples of Europe?’ (2001–​2002) 21 yb of Eur L 195, 204–​5 (Brown’s italics). 4 M Bell, ‘eu anti-​racism policy: the leader of the pack?’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 187. 5 A Lester and P Uccellari, ‘Extending the equality duty to religion, conscience and belief: proceed with caution’ (2008) 5 ehrlr 567, 570. 6 ibid 572. 7 Directive 2000/​43. Full citation at n 1, supra. 8 ibid Article 3(1)(e)–​(h).

120 Chapter 4 on grounds of racial or ethnic origin occurring in any of these four additional arenas is thus outlawed, while discrimination affecting any of the other ‘new statuses’ would not be.9 Victims of race discrimination even receive greater protection than victims of gender discrimination, who admittedly are safeguarded in one of the additional arenas (goods and services), and part of another (social security), but not the remaining two.10 This has led to talk of a ‘hierarchy’ of grounds, with race at the top, followed by gender. Third place would be held jointly by religion, sexual orientation, disability and age,11 and last of all would be nationality.12 However, writers seem to vacillate between 9 10

11

12

Cf. the Framework Employment Directive, Article 3(1). The Equal Treatment Directive (Directive 76/​207, [1976] oj L39/​40) limited protection to employment (Articles 3, 4 and 5 of the original Directive, Article 3 only since amendment by Directive 2002/​73, [2002] oj L269/​15). The position is the same under the Recast Directive, Directive 2006/​54, [2006] oj L204/​23. Social security was dealt with in a separate instrument (Directive 79/​7, [1979] oj L6/​24 –​the Social Security Directive). In 2004, protection against sex discrimination was extended to cover goods and services, although, crucially, not education:  Council Directive 2004/​113/​ec of 13 December 2004 implementing the principle of equal treatment between men and women in the access to and supply of goods and services [2004] oj L373/​37 (‘Directive 2004/​113’) –​see Article 3(1) and Article 3(3). However, even within the arena of public goods and services, according to the legislator at least, protection against race discrimination was to be more extensive than protection against sex discrimination, because sex was able to be taken into account in setting insurance premiums, but race was not. See ibid Article 5(2); no such exception was to be found in the Race Equality Directive. See further Bell (n 4) 181–​4. However, as examined in the previous chapter, Article 5(2) was declared invalid (with effect from 21 December 2012) in the case of Test-​Achats, thus bringing Directive 2004/​113 into line with the Race Equality Directive:  Case C-​236/​09 Association Belge des Consommateurs Test-​Achats asbl and Others v Conseil des ministres [2011] ecr 773. Although it has been pointed out that age actually receives lesser treatment in comparison with the other three grounds in this list, mainly due to the extensive exceptions to be found in the Framework Employment Directive, Article 6.  See further H Meenan, ‘Age equality after the employment directive’ 10 mj 1 (2003) 9, 10, with the point also having been made by Gyulavári and Schiek: T Gyulavári, ‘Age discrimination: recent case law of the European Court of Justice’ (2013) 14 era Forum 377, 384 (‘age [is] at the bottom of the so-​called hierarchy of grounds’) and D Schiek, ‘Age discrimination before the ecj: Conceptual and theoretical issues’ (2011) 48(3) CML Rev 777, 796 (‘the ban on age discrimination in eu law is conceptually weaker than other bans on discrimination’). McGlynn has suggested that it is both age and disability which ‘occupy the lowest rung:’ C McGlynn, ‘ec legislation prohibiting age discrimination: “Towards a Europe for All Ages”?’ (2000) 3 C-​yels 279, 294. Nationality generally, that is. The protection in relation to the free movement of eu nationals is extensive. Thus, Martin puts nationality ahead of sex in his hierarchy-​of-​ treatment, as does Warnier: Martin D, Égalité et non-​discrimination dans la jurisprudence

The ‘Article 19’ grounds

121

lamenting the fact that the eu does not confer the same level of protection on all of the grounds,13 and accepting that the ‘specific characteristics’14 or ‘peculiarities’15 of the different grounds require different responses for each.16

13

14 15 16

communautaire:  étude critique à la lumière d’une approche comparatiste (Bruylant 2006) 590; N Warnier, ‘Les discriminations directes and indirectes dans la domaine de l’égalité homme-​femme et de l’égalité nationaux-​non-​nationaux’ (2006) 84(2) Revue de droit international et de droit comparé 225, 285. (eu) nationality discrimination will be dealt with in Chapter 5 below. For one description of the full hierarchy, see G Pitt, ‘Religion or belief:  aiming at the right target?’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 223–​4. Another can be found in E Howard, ‘The case for a considered hierarchy of discrimination grounds in eu law’ (2006) 13 (4) mj 445. Costello and Davies believe that the Court may make use of the general principle of equality to ‘flatten’ the hierarchy: C Costello and G Davies, ‘The case law of the Court of Justice in the field of sex equality since 2000’ (2006) 43 cml Rev 1567, 1574. ‘[N]‌o clear rationale has been given for the greater material and protective scope for sex and race … there is no obvious reason why sex and race should be favoured above the other grounds contained in Article 13’: H Meenan, ‘Conclusion’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 342–​3. ‘[I]t will remain difficult to explain to citizens why one form of discrimination enjoys a higher level of protection in law than another’: L Waddington and M Bell, ‘More equal than others: distinguishing European Union equality directives’ (2001) 38 cml Rev 587, 611. ‘[I]f Article 13 is to be a cornerstone provision of an emerging Union constitution, then the rights of certain citizens should not be more “fundamental” than those of others. By enacting [a specific Directive for race but] a Framework Directive for the other forms of discrimination, one is giving the impression that the right of someone from an ethnic minority to non-​ discrimination in, for example, the supply of goods and services is more important than that of a homosexual’: Brown (n 3) 223. M Bell and L Waddington, ‘Reflecting on inequalities in European equality law’ (2003) 28 ELRev 349, 368. H Meenan, ‘Age discrimination –​Of Cinderella and The Golden Bough’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 299. ‘As such these differences in treatment amount to a recognition of difference rather than the creation of a hierarchy’: Bell and Waddington (n 14) 368. ‘[Taking a differentiated approach to the new grounds] is not to endorse a hierarchy of inequality but rather to acknowledge the differences between them’: B Fitzpatrick, ‘The “mainstreaming” of sexual orientation into European equality law’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 313 (footnote omitted). Schiek in particular has attempted to explain the differences between the grounds, via her well-​known three-​pronged categorization in D Schiek, ‘A New Framework on Equal Treatment of Persons in ec Law?’ (2002) 8(2) Eur lj 290, particularly at 309–​310. ‘If there are different categories of characteristics, discrimination on grounds of which is forbidden, this may justify differences between different prohibitions to discriminate’: ibid 310. But for some criticisms of her approach, see Pitt (n 12) 224ff.

122 Chapter 4 With regards to race, the specific characteristics setting it apart from the other grounds include the fact that there is a political consensus on the need to combat racial discrimination (which there may not be in respect of the other grounds).17 With regards to gender, there is the fact that the ec Treaty itself appears to privilege this ground over the others.18 On 2 July 2008, the Commission made its proposal for a so-​called ‘horizontal’ Directive for the four grounds covered by the Framework Employment Directive, which would expand the prohibition on discrimination on those grounds (including religion or belief) to cover social protection (including social security and healthcare), social advantages, education, and access to and supply of goods and services which are commercially available to the public (including housing).19 Were this new directive ever to be adopted, it would of course have a huge impact on the ordering or even the existence of the ‘hierarchy of grounds’ discussed above. However, political wrangling seems to have left the proposal, until relatively recently, in legislative limbo. The sticking point seems to be obtaining unanimity in the Council, and in a Progress Report from 22 November 2016, the outgoing Slovak Presidency warned that there was ‘still a need for further work and political discussions before the required unanimity can be reached.’20

17

18 19

20

Suggested by Brown (n 3) 222. If the consensus argument is accepted, then the apparent favouritism shown to race is not really favouritism at all, but merely ‘realism’ (ibid 223) or ‘pragmatism’ (Meenan (n 15) 299, referring to E Barry, ‘Different Hierarchies –​ Enforcing Equality Law’ in C Costello and E Barry (eds), Equality in Diversity –​The New Equality Directives (Ashfield 2003) 414). As McGlynn points out, the Commission itself has admitted that its approach was steered by pragmatism: McGlynn (n 11) 288, referring to com(99)564 –​see [11] of the latter document. Further discussion of the pragmatism rationale can be found in Howard (n 12) 451ff. Suggested by Schiek (n 16) 300. See Commission (ec), ‘Proposal for a Council Directive on implementing the principle of equal treatment between persons irrespective of religion or belief, disability, age or sexual orientation’ com(2008)426 final, 2 July 2008. On 2 April 2009, the European Parliament adopted by 363 votes to 226, with 12 abstentions, a legislative resolution making significant amendments to the proposal. It has been debated in Council every year since, but at the time of writing, is still awaiting final decision. Progress Report 14284/​16, ‘Proposal for a Council Directive on implementing the principle of equal treatment between persons irrespective of religion or belief, disability, age or sexual orientation’ (Council of the European Union, 22 November 2016) 8. For a good description of the earlier discussions, see L Waddington, ‘Future prospects for eu equality law: lessons to be learnt from the proposed Equal Treatment Directive’ (2011) 36(2) ELRev 163.

The ‘Article 19’ grounds

123

With the differences between ‘racial or ethnic origin’ and ‘religion or belief’ having been established, one may turn to the cjeu’s case-​law on racial or ethnic origin, such as it is. To date only a small number of cases based directly on Directive 2000/​43 have come to judgment at the cjeu. In Firma Feryn,21 a Belgian company specializing in up-​and-​over doors refused to hire Moroccans to fit its products; one of the company’s directors publicly stated this policy both on television and in the press. On a reference from the Brussels Labour Court (an action having been brought by a Belgian public-​interest body), the cjeu started by reminding itself of the definition of direct discrimination in Article 2(2)(a) of Directive 2000/​43: [D]‌irect discrimination shall be taken to occur where one person is treated less favourably than another is, has been or would be treated in a comparable situation on grounds of racial or ethnic origin.22 It went on to hold that, regardless of the lack of an identifiable complainant in this particular case, [t]‌he fact that an employer declares publicly that it will not recruit employees of a certain ethnic or racial origin, something which is clearly likely to strongly dissuade certain candidates from submitting their candidature and, accordingly, to hinder their access to the labour market, constitutes direct discrimination in respect of recruitment within the meaning of Directive 2000/​43.23 Seven years later, on 16 July 2015, the Court delivered its judgment in chez  rb.24 The case concerned the allegedly discriminatory behaviour of the Bulgarian electricity company chez rb, which throughout Bulgaria had installed external electricity meters on telegraph poles at a height of 1.7m, but with certain exceptions. One of the exceptions was the Gizdova mahala district of Dupnitsa town, where the meters were installed at a height of six to seven metres. The company publicly declared that this was to protect the security of 21 22 23 24

Case C-​54/​07 Centrum voor gelijkheid van kansen en voor racismebestrijding v Firma Feryn nv [2008] ecr I-​5187. Directive 2000/​43 (n 1)  Article 2(2)(a), recalled in Case C-​54/​07 Firma Feryn (n 21) para 22. ibid para 25. Judgment of 16 July 2015, Chez Razpredelenie Bulgaria Ad v Komisia za zashtita ot diskriminatsia, C-​83/​14, ECLI:EU:C:2015:480.

124 Chapter 4 the electricity transmission network against electricity users making ‘unlawful connections.’25 However, the real, or underlying, reason was the prevention of tampering of the meters by Roma users of the service; Gizdova mahala was a majority Roma district, as were the other districts where the company behaved in this way. However, Ms Nikoleva, the Gizdova mahala resident who brought the case, was herself non-​Roma. Her complaint was therefore that she was suffering what Atrey has called ‘collateral disadvantage.’26 This included not having free access to her electricity meter readings, to which she was entitled as a consumer. On reference, the Court of Justice immediately dealt with the latter point by declaring that the concept of ‘discrimination on the grounds of ethnic origin’ did apply in circumstances where less favourable treatment was inflicted on persons of a certain origin, in a district in which the majority of inhabitants were of that origin, irrespective of whether it affected those of that origin or ‘those who, without possessing that origin, suffer, together with the former, the less favourable treatment.’27 Unusually, the Court went on to find that there was in this case both direct discrimination and, in the alternative, indirect discrimination. As regarded direct discrimination, the Court held that the requirement, in the abovementioned Article 2(2)(a),28 that one person be treated less favourably than another on grounds of ethnic origin could still be satisfied even where the practice which gave rise to the difference in treatment was also imposed on inhabitants not of that origin (like Ms Nikoleva) in the district concerned as long as the origin at issue was ‘shared by most of that district’s inhabitants.’29 With there now being a prima facie case of direct discrimination, the cjeu reiterated that it was for the electricity company to prove that they had not in fact discriminated directly on grounds of the majority Roma population’s ethnic origin when they installed the meters at the height of six to seven meters.30 The Court noted, though, that the company could not prove the alleged tampering, by the Roma, of the meters, and that this alleged tampering was merely a perception. As regarded the requirement in Article 2(2) (a) of less favourable treatment, the Court was satisfied that this was present, and as regarded the requirement of comparability as between those treated unfavourably and those treated favourably, the Court yet again stressed the 25 26 27 28 29 30

ibid para 31. S Atrey, ‘Redefining frontiers of eu discrimination law’ [2017] Apr pl 185, 188. Judgment in Case C-​83/​14 chez rb (n 24) para 60. Directive 2000/​43 (n 1) Article 2(2)(a), set out in text accompanying n 22. Judgment in Case C-​83/​14 chez rb (n 24) para 75. ibid para 77, referring to Directive 2000/​43 (n 1) Article 8(1).

The ‘Article 19’ grounds

125

unimportance of the collateralism point (impact on non-​Roma in the affected areas, lack of impact on Roma in the unaffected areas). Widening the field of reference somewhat, the cjeu was persuaded that ‘all final consumers of electricity who are supplied by the same distributor within an urban area’ were like,31 and therefore needed to be treated in like fashion. Next, the Court turned its attention to the wording of Article 2(2)(b) of the Race Equality Directive: [I]‌ndirect discrimination shall be taken to occur where an apparently neutral provision, criterion or practice would put persons of a racial or ethnic origin at a particular disadvantage compared with other persons, unless that provision, criterion or practice is objectively justified by a legitimate aim and the means of achieving that aim are appropriate and necessary.32 It will be seen that this definition of indirect discrimination is practically identical, mutatis mutandis, to that at Article 2(1)(b) of the Gender Recast Directive, encountered in the last chapter.33 On the wording, the Court was of the view that the phrase ‘apparently neutral’ meant ‘ostensibly’ neutral, or neutral ‘at first glance’,34 and that the phrase ‘a particular disadvantage’ did not mean that the disadvantage alleged needed to be serious, obvious or even particularly significant.35 Applying this to the case at hand in a bid to help the national court, the cjeu held that the practice of installing the meters at a height of six to seven meters to avoid the making of unlawful connections could be an ‘apparently neutral practice’ for the purposes of Article 2(2)(b) of the Race Equality Directive.36 Further, carrying out this practice in majority Roma districts like Gizdova mahala was likely to affect the Roma inhabitants of those districts in greater proportions than the non-​Roma ones, and thus to impose upon them a particular disadvantage.37 As regarded, finally,

31 32 33

34 35 36 37

Judgment in Case C-​83/​14 chez rb (n 24) para 90. Directive 2000/​43 (n 1) Article 2(2)(b). Directive 2006/​54/​e c of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast) [2006] oj L204/​23 (the ‘Recast Directive’). Judgment in Case C-​83/​14 chez rb (n 24) para 93. ibid para 99. ibid para 106. ibid para 107.

126 Chapter 4 the issue of objective justification, the Court seemed to be of the view that, while the securing of the electricity transmission network could be a legitimate aim within the meaning of Article 2(2)(b), the electricity company might struggle to show objectivity in the absence of specific evidence of tampering by these particular people at this particular time.38 Moreover, even it was a legitimate aim, the means utilised still had to be appropriate, which in theory they seemed to be,39 and necessary, which they were less likely to be, as there were other less restrictive, and less stigmatising, means of achieving the aim such as usage of alarmed meters.40 Concluding this section on the cjeu’s case law, then, one might worry that, with so few cases on the Race Equality Directive, victims of racial discrimination in the eu are purposely keeping silent. Academic research would seem to back this up.41 As Bell has pointed out in relation to sexual  orientation under the Framework Employment Directive, ‘absolutely low levels of litigation’ may be an ‘early warning sign’ that the Directive in question is not working.42 38 39 40

41

42

ibid paras 116–​117. ibid para 119. ibid paras 120–​122; the alternative meter is described in Advocate General Kokott’s Opinion at paras 128 to 130: Opinion of 12 March 2015, chez Razpredelenie Bulgaria ad v Komisia za zashtita ot diskriminatsia, C-​83/​14, ECLI:EU:C:2015:170. Cillo and Perocco describe how non-​Italian eu migrant workers in Italy have suffered discrimination in, for example, the manufacturing sector but the trade unions have felt powerless to act, one of the reasons mentioned being ‘the post-​crisis need to protect jobs:’ R Cillo and F Perocco, ‘Outsourced racism in Italy:  discrimination at work and trade union responses in three sectors’ (2015) 21(1) Transfer: European Review of Labour and Research 51, 61. In the same issue of Transfer, which was a special edition focused on ‘Challenging discrimination at work against ethnic minority and migrant workers,’ the examples given include Polish workers in the construction industry in Sweden, and Bulgarian Roma in the berry-​picking industry also in Sweden: C Thómqvist and S Bernhardsson, ‘Their own stories –​how Polish construction workers posted to Sweden experience their job situation, or resistance versus life projects’ (2015) 21(1) Transfer: European Review of Labour and Research 23 and N Mešić and C Woolfson, ‘Roma berry pickers in Sweden: economic crises and new contingents of the austeriat’ (2015) 21(1) Transfer: European Review of Labour and Research 37. With regard to the former, the point is made that some migrant workers offset their passivity in the face of poor treatment by focusing on their ‘life project’ (such as building a house in their home country). It therefore seems that such workers experience discrimination simply as a short-​term obstacle to be endured, not challenged. M Bell, ‘Publication review –​Sexual Orientation Discrimination in the European Union: National Laws and the Employment Equality Directive’ (2007) 4 ehrlr 481, 482.

The ‘Article 19’ grounds

127

4.2.2 A Walzerian Analysis (a) Explicit Bias: Firma Feryn In Firma Feryn, the post being distributed was that of a fitter of up-​and-​ over doors. The shared understanding, however, was not at all clear-​cut or predictable. A worrying aspect of the case was that, if Mr Feryn was to be believed, the exclusive hiring of non-​Moroccans was part of the shared understanding of this post within the relevant distributive community. He repeatedly said in both the press and on television that his customers ‘[did] not want Moroccans,’ and ‘[did not] want them coming into their homes.’ He continued: We must meet the customers’ requirements. […] I  want the firm to do well and I want us to achieve our turnover at the end of the year, and how can I  do that? I  must do it the way the customer wants it done!43 Presumably Mr Feryn would therefore argue that he was taking a Walzerian approach to the distribution of which he was (ultimately) in charge, and reflecting his community’s understanding of what it meant to be a fitter of up-​and-​over doors, by making sure that no such fitter was of Moroccan nationality.44 Of course, one man’s testimony cannot be determinative of a shared meaning. There may very well have been in Belgium at that time potential owners of up-​and-​over doors who did not hold the same racist views as Mr Feryn’s customers apparently did. But if, for the sake of argument, the shared meaning was indeed as Mr Feryn described it, how is a complex egalitarian to respond? Walzer himself entertains the idea of certain types of homogeneity, including ethnic or religious homogeneity, within certain workplaces: it may well happen that at a given time, in a given place, the most successful factory will be run largely by Italians, say, or by Mormons. I don’t see anything wrong with that.45 43 44

45

Case C-​54/​07 Firma Feryn (n 21) Opinion of AG Poiares Maduro, para 4. Fox describes a similar situation occurring in the us prior to the civil rights revolution: ‘employment was understood within its own sphere as properly discriminatory –​ a norm that applied to employers and employees alike (witness the strong support for racist employment policies among white unions)’: James W Fox, Jr, ‘Relational Contract Theory and Democratic Citizenship’ (2003–​2004) 54 Case W Res L Rev 1, 41–​42. Racism was quite simply ‘the custom of the relationship:’ ibid 42. M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) 162.

128 Chapter 4 However, he draws the line at hiring practices that use race as a criterion: In a society with a long history of racism, it would make sense to rule out racial criteria, hence to impose a minimal set of fair employment practices.46 The Firma Feryn case provides an example of the problematic situation discussed in section 2.5 above, where leaving a community to fix its own distributive criteria can lead to nationalism, or worse. As mentioned in that section, the solution is to build into the theory a kind of universal moral code against which the context-​dependent principles can be tested, and which, if necessary, can act as an override. It was submitted that Walzer had all along intended that the theory should incorporate such a ‘thin morality,’47 although he was at pains to point out that even this had a particularist dimension that could not be ignored. Later in the same chapter (section entitled, ‘The cjeu and its override: An answer to the objection’), it was discussed how the Court of Justice already had at its disposal an instrument which represented a kind of ‘European version’ of this thin morality.48 It was also discussed how simple it would be to identify any distributive criterion which offended against said morality, and to rule it out. In Firma Feryn, the apparently shared rule that no fitters of up-​and-​over doors may be of Moroccan nationality is just such a criterion. It offends against Article 14 and Protocol 12 of the echr. Relying on Europe’s Minimal Morality, then, as represented (at least in part) by this instrument, the judge at the cjeu would not have any difficulty in declaring

46 47 48

ibid. M Walzer, Thick and Thin:  Moral Argument at Home and Abroad (University of Notre Dame Press 1994) 12. It would, in fairness, be exceedingly thin, by Walzer’s standards. A  Council of Europe instrument such as the echr, for example, necessarily represents the combined ‘moralities’ of (in 2018) forty seven signatory states, to say nothing of the local ‘moralities’ of the smaller constituent communities within these states. But in many ways post-​war Europe was facing a ‘moment […] of crisis,’ and was ‘unite[d]‌[by] the sense of a common enemy’ (Nazism/​Fascism); this justified the construction of ‘an abstract version, a stick figure, a cartoon’ (all quotations from Walzer, Thick and Thin (n 47) 18). It could also be argued that in the half-​century or so since the War, the signatory states have grown closer and gone through more and more common experiences, lending a sort of communal thickness to the cartoon, and turning it into a statue.

The ‘Article 19’ grounds

129

that there had been a flawed distribution.49 He or she could then (subject to what is said at section 7.2.1 below) either rewrite the distributive criteria him or herself, or send the case back to the Labour Court in Brussels for it to rewrite them. (b) Rights without Rigmaroles: chez rb The Walzerian take on chez rb is summarized easily enough. The distributive sphere for free access to information with regard to one’s electricity usage only really admits of one entry token: that one is oneself a user of the service. Being of Roma origin, or living in an area predominated by people of Roma origin, cannot possibly form part of the meaning of this good, which, if it is not self-​evident, is as close to self-​evident as it is possible to be. Even if the electricity company genuinely feared unlawful connections being made, or indeed wished to keep consumers safe from the adverse health consequences of their own tampering, and assuming that the alarmed meter did not exist, or was prohibitively expensive for said company, then the only allowable (negative) dominant would surely be a conviction for having made unlawful connections, or for having tampered, in the past.50 Meanwhile the flaws of the Aristotelian test are all too evident in this case. The finding of an Article 2(2)(a) breach required two likes –​telephone users in Gizdova mahala and those in the rest of Dupnitsa town –​to be treated unlike. Although the necessary ground for the less favourable treatment (ethnic origin) was not explicit, it was ‘common ground’ and this was good enough.51 However, there were inevitably complications with the Court’s decision to reclassify the measure, which it had said was grounded on ethnic origin only a moment before, as not so grounded and therefore ‘apparently neutral’ for the purposes of Article 2(2)(b). This finding of an additional, if alternative, breach of the prohibition on indirect discrimination brought about a crisis in the like-​for-​like test as, as Atrey points out, the Roma in Gizdova mahala

49

50

51

Fox takes a slightly different route out of the problem (n 44), by using equal citizenship as his override. It will be recalled that equal citizenship was mooted as a possible universal or adjudicatory principle, already supplied by Walzer himself in Spheres of Justice:  see supra, section 2.5. Given that the right to free expression contains the right to receive information, and given that an electricity company, even a private one, would have enough connection to the State (guarantor of that right) as to be regarded as an emanation thereof for the purposes of eu Law, it is questionable whether even in the latter scenario it would be legal for the company to deny the convicted customer access to their usage data. Judgment in Case C-​83/​14 chez rb (n 24) para 31.

130 Chapter 4 could not be put at a ‘particular disadvantage’ where the non-​Roma in the district were treated the same,52 and widening the area of analysis could not help because the users of the telephone service in the other districts of the town, although better treated, were not in fact subject to the impugned measure!53 Indirect discrimination cannot obtain unless the ‘apparently neutral’ measure applies to all of the inhabitants of the area chosen. As things stood, in strict Aristotelian terms, unlikes (the users of the service in Gizdova mahala, and the users of the service in the other districts) were being treated unlike (the former subject to the measure, the latter not), meaning that there could be no indirect discrimination in this case at all. Needless to say critics have focussed on the Court’s alternative finding of indirect discrimination, Atrey calling it ‘superfluous,’54 ‘a poor fit’55 and ‘simplistic.’56 McCrudden warns that allowing Ms Nikoleva successfully to claim for what he calls ‘indirect indirect discrimination’ could eventually lead to a ‘marginalis[ing]’ of the primary victims of indirect discrimination.57 Important for this book, though, is the further evidence it provides of the malleability of the Aristolelian test to produce a desired result: [The Court’s] analysis under art 2(2)(b) shows the inconvenient interpretation it followed down the path of indirect discrimination just to reach a favourable finding.58 Atrey goes on to point out that one consequence of the Court’s action in chez rb is that the lines between the categories of direct discrimination and indirect discrimination have become blurred at eu level. Interestingly she calls for a ‘middle ground … to avoid rigmaroles.’59 One last point:  In the particular case in hand, it might be thought that, since only the Roma may carry the negative dominant of ethnic origin, only the Roma may successfully claim under a Walzerian system, and the admitted 52 53 54 55 56 57

58 59

Besides, treating the phone users of Gizdova mahala as an amorphous group one minute and then two antithetical groups the next is itself problematic. Atrey (n 26) 190–​1. ibid 189. ibid. ibid 193. C McCrudden, ‘The new architecture of eu Equality Law after chez: Did the Court of Justice reconceptualise direct and indirect discrimination?’ [2016] 1 European Equality Law Review 1, 8. Atrey (n 26) 193. ibid 195.

The ‘Article 19’ grounds

131

advance made by the Court in chez rb, in extending the protection from direct discrimination to those like Ms Nikolova suffering collateral disadvantage, would be reversed. While it is true that Ms Nikolova would only be able to show precisely the entry token that is in keeping with the distributed good’s shared meaning, it is not true that this would mean that she would fail in a claim against the electricity company. On the contrary, she would still be able to ask, and have the judge rule upon, why, if she had indeed shown the correct entry token, she was not receiving the good. The Walzerian system allows for the claims of those meeting the criteria inherent in a certain good who do not receive it as well as those not meeting the criteria who do receive it, whether or not these individuals additionally possess a negative dominant or more than one negative dominant, skewing their ‘otherwise’ valid claim. All one could say is that the claim of the Roma is a claim alleging a boundary breach (between the sphere of personal attributes and the sphere of access to electricity data), where the claim of Ms Nikolova would be a claim alleging a wrongful placement of the boundary in the first place (to exclude her when in fact she met the relevant distributive criteria). This section has briefly made a Walzerian analysis of chez rb and then noted at length the Court’s real-​life inability to escape Aristotelian ‘rigmaroles.’ Walzer, it is submitted, could provide the middle ground called for by Atrey, sidelining both limbs of the like-​for-​like test, corresponding to direct and indirect discrimination respectively, both vulnerable to judicial self-​contradiction or malleability, and instead focusing on the meanings of distributed goods and the boundaries between distributive spheres as the ready reckoners to resolve equality disputes irrespective of whether the claimant is the possessor of the protected characteristic, someone associated with them, or just someone suffering collaterally. 4.3

Religion and Belief

4.3.1 The cjeu’s Case-​Law Turning to ‘religion or belief,’ there have been two major judgments under the Framework Employment Directive (which will be discussed in detail in the following sections) in relation to this ground.60 Before dealing with those, however, it would be worthwhile to look at a case in which the European Court considered the ground some years prior to the passing of the Directive, in 1976 60

Judgment of 14 March 2017, Bougnaoui and Addh, C-​188/​15, ECLI:EU:C:2017:204; Judgment of 14 March 2017, Achbita v G4S, C-​157/​15, ECLI:EU:C:2017:203.

132 Chapter 4 to be precise. In the Prais case, Ms Prais requested an alternative date for the written test of a Council concours, because the date as it stood clashed with the Jewish festival of Shavuot (Pentecost).61 The request was refused, and a subsequent complaint was rejected. Ms Prais then sought the annulment of both the refusal and the rejection before the cjeu. She relied on Article 27(2) of the Staff Regulations, which stipulated that officials were to be selected without reference to race, creed or sex, as well as on the prohibition on religious discrimination (which formed part of the fundamental rights of the individual), and Article 9 of the echr. The Council counter​argued that these instruments were not to be understood as according to Ms Prais the rights which she was claiming; the Council would need an ‘elaborate administrative machinery’ if it had to take account of ‘all religions’ when fixing dates for its tests.62 The cjeu agreed. The principle of equal treatment in fact worked against Ms Prais, by requiring that the tests should take place on the same conditions for all candidates; this meant that all candidates would have to sit the tests on the same date. The interests of participants in avoiding certain unsuitable dates had to be balanced against this necessity. While the Court admitted that it would be ‘desirable’ that the Council informed itself ‘in a general way’ about dates of religious significance, the instruments cited did not impose a duty on it to avoid conflicts,63 especially when, firstly, it had not been told about Ms Prais’ difficulties in good time,64 and, secondly, other candidates had already been convoked.65 Post-​Directive, there have been two judgments concerning alleged discrimination on ground of religion and belief. In Achbita v G4S, Ms Achibta, who was a Muslim, made an allegation of such discrimination when her employer, G4S, dismissed her for wearing her headscarf at work.66 G4S had a written and Union-​endorsed code of conduct by which employees were ‘prohibited, 61 62 63

64

65 66

Case 130/​75 Vivien Prais v Council of the European Communities [1976] ecr 1589. Quotations from ibid para 11. Brillat says that the Court’s formula here is ‘more than ambiguous’ (translation author’s own), and certainly falls short of any kind of obligation on the institutions to monitor the festivals of minority religions: M Brillat, Le principe de non-​discrimination à l’épreuve des rapports entre les droits européens (Fondation Varenne 2015) 311. But, as de Dominicis has pointed out, while this enforced declaration of her religion might improve Ms Prais’ equality rights, it weakens her right to privacy: N de Dominicis, ‘La tension entre l’égalité de traitement et le droit à la vie privée: la protection des données à caractère personnel et l’emploi des personnes souffrant d’un handicap’ [2014] 2 Revue du droit de l’Union européenne 299, 313. Quotations from Case 130/​75 Prais (n 61) para 18. Judgment in Case C-​157/​15 Achbita (n 60).

The ‘Article 19’ grounds

133

in the workplace, from wearing any visible signs of their political, philosophical or religious beliefs and/​or from giving expression to any ritual arising from them.’67 Given that this seemed to be a neutral rule, uniformly applied, but one that put Muslim women wishing to wear a headscarf at a disadvantage compared with non-​Muslims, Advocate General Kokott arrived at the view that this was a case of indirect discrimination. However, after a painstaking analysis, the Advocate General concluded that G4S could plead an objective justification: [A]‌ban such as that imposed by G4S may … be regarded as being a genuine, determining and legitimate occupational requirement, within the meaning of Article 4(1) of Directive 2000/​78, which is in principle capable of justifying differences of treatment —​whether direct or indirect —​based on religion, provided that the principle of proportionality is respected.68 The Court took a similar line to its Advocate General, although its assertion that G4S’ prohibition on wearing an Islamic headscarf could be objectively justified by a legitimate aim, ‘such as the pursuit by the employer, in its relations with its customers, of a policy of political, philosophical and religious neutrality,’ has been criticised, by writers like Davies, as justifying prejudice with more prejudice and thus making a nonsense of the existence of legal protection against discrimination in the first place: Is it not incompatible with the idea of religious equality that an exclusionary rule be justified by a dislike of religion?69 Meanwhile, in Bougnaoui v Micropole, Ms Bougnaoui, who was also a Muslim, had been dismissed by her employer Micropole for the same reason –​wearing her headscarf to work. However, in this case, there was no company-​wide, religion-​neutral policy. In fact the dismissal seems to have arisen out of the complaint of one customer. Under these circumstances, Advocate General Sharpston, after a lengthy discussion of the meaning of the Muslim headscarf, and indeed of the meaning of French laïcité, came to the conclusion that this 67 68 69

ibid para 15. Opinion of 31 May 2016, Achbita v G4S, Case C-​157/​15, ECLI:EU:C:2016:382, para 128. G Davies, ‘Achbita v G4S:  Religious Equality Squeezed between Profit and Prejudice’ (European Law Blog, 6 April 2017)  accessed 14 May 2018; Court’s assertion is at Judgment in Case C- 157/ 15 Achbita (n 60) para 44.

134 Chapter 4 was a case of direct discrimination, and that this time the employer could not avail themselves of a derogation under Article 4(1) of the Directive, or indeed any other derogation. Were the Court to decide that this was a case of indirect discrimination, she was prepared to accept that ‘the interests of the employer’s business’ could in theory serve as the basis for an objective justification under Article 2(2)(b)(i) of the Directive, but that even then a ban on wearing the headscarf would need to be shown by the employer to have been a proportionate means of pursuing these interests.70 The Court, blaming the lack of clarity as between direct and indirect discrimination in this case on the ‘order for reference’71 went on to consider both possibilities, but left the final decision to the national court. Were the national court to find that Micropole’s rule/​policy was an instance of indirect discrimination, the cjeu said, then it would need to consider possible justification under Article 2(2)(b)(i) ‘such as the implementation, by Micropole, of a policy of neutrality vis-​à-​vis its customers, and if the means of achieving that aim [were] appropriate and necessary.’72 On the other hand, continued the Court of Justice, were the national court to find that the rule/​policy was an instance of direct discrimination, then it would need to look at whether such rule could be justified as a general occupational requirement under Article 4(1).73 However, after some analysis, the Court found that Micropole’s rule/​policy could not, in its opinion, qualify as such a requirement.74 4.3.2 A Walzerian Analysis (a) Implicit Bias: Prais v Council Cases concerning religion and race provide a nice example of what Walzer was describing at the beginning of Thick and Thin, discussed above, namely that moral codes do not just fall out of the sky, but that they are in fact the product of the particular history of the people, or peoples, laying claim to them. This means that, despite their claims to universality and neutrality, many such codes may have certain cultural ideas ‘embedded’ within them.75 70 71 72 73 74 75

Opinion of 13 July 2016, Bougnaoui and addh, Case C-​188/​15, ECLI:EU:C:2016:553, quotation at para 134. Judgment in Case C-​188/​15 Bougnaoui (n 60) para 31. ibid para 33. ibid para 34. ibid para 41. In the context of sexual orientation, Ball also talks about the role which ‘personal and intellectual attachments to particular western and liberal contemporary societies’ may play in the determination of supposedly universal principles: Carlos A Ball, ‘Communitarianism and Gay Rights’ (1999–​2000) 85 Cornell L Rev 443, 503.

The ‘Article 19’ grounds

135

Thus, in a case like Prais, the apparently neutral decision, first of the Council and then of the Court, to treat all candidates alike may have implicitly favoured candidates belonging to Europe’s dominant faith: Christianity. In other words, a hidden ‘norm’ (or ‘invisible baseline’)76 is endorsed, and with each endorsement further petrified, so that whatever sacrifices are to be made in the name of ‘like treatment’ are in fact exclusively made by the candidate of the non-​ dominant faith. This is not a meeting in the middle. Candidates of the non-​ dominant faith must be levelled, up or down, to ‘match’ those of the dominant faith. It is not as though, for example, a Christian candidate would ever be required to sit an examination on a Sunday (the Community institutions are closed on Saturdays and Sundays). As Danchin has put it (in the context of the ‘affaire du foulard’ in France), any exception to the general rule […] will be determined by the majority […] typically out of deference to the historical relationship between the nation and its dominant religion. […] [T]‌his exercise of national sovereignty will be neutral neither towards religion in general, nor to minority religions […] in particular.77 This is reminiscent of Schmitt’s famous comment, ‘The sovereign is:  He who decides on the state of exception.’78 While it is not the intention of this book to second-​guess shared understandings –​the whole point is that they must be arrived at by the distributive community itself –​nevertheless it is fairly certain that the shared understanding of the thing distributed in the Prais case, namely a post at the Council, would not contain any reference to the religion of the potential post-​holder. To perform work at the Council may require some experience of living in Europe, and no doubt a high level of verbal reasoning and other skills, but it is unlikely to require belief in any particular religion, or in any religion at all. It is true that an examination may have to be held to confirm the candidate’s possessing of the (pertinent) experience and skills. However, holding this examination on the feast-​day of any non-​dominant religion practiced within the Council’s jurisdiction allows those who can show a badge of membership of the dominant religion (according to whose calendar the 76 77 78

Peter G Danchin, ‘Suspect Symbols: Value Pluralism as a Theory of Religious Freedom in International Law’ (2008) 33(1) Yale J Intl L 1, 25. ibid (footnote omitted). ‘Souverän ist, wer über den Ausnahmezustand entscheidet’:  Carl Schmitt, Politische Theologie: Vier Kapitel zur Lehre von der Souveränität [Political Theology: Four Chapters towards a Theory of Sovereignty] (Duncker & Humblot 1990) 11.

136 Chapter 4 date has been set) privileged access to the distributive sphere, while effectively turning away, or at least discouraging, those whose badges say something else. Membership of the dominant religion is thus being used tyrannically, contrary to complex equality.79 The Court of Justice, if it found the shared understanding of posts at the Council to be as described, would therefore have had to find against the Council, and in favour of Ms Prais. As Danchin points out in a different context, such a finding would require ‘more than noninterference with the individual’s imagined sphere of liberty;’ it would require ‘public recognition of a plurality of different religious and cultural groups and ways of life.’80 Complex equality would even allow a policy of proportionality, such as Walzer describes in the context of schooling.81 The Council, as an institution which represents all Member States of the eu, might wish to have a work-​force representative of the eu’s population (if such an idea formed part of the shared meaning of ‘a job at the Council’) and ‘not dominated by the … reigning ideologies.’82 If proportionality were to be a component in the distributive mechanism employed by the Council, then holding the qualifying test on the same day as a Jewish festival would violate this mechanism by preventing an entire constituency within the European population (Jewish people) from having an opportunity to compete. Failure to respect a shared meaning would of course represent a further violation of complex equality. The Court’s job would simply be to identify that there had been a flawed distribution, and then, if possible, to right the wrong done. (b) Indirect Discrimination versus Walzer: Achbita and Bougnaoui By the time Achbita and Bougnaoui come to Court, of course, the Court has not only pioneered the concept of indirect discrimination, in cases like Kingsgate and Bilka discussed in the last chapter,83 but has inspired the eu legislature 79

80 81 82 83

As an alternative to viewing ‘membership of the dominant religion’ as the dominant good in this case, facilitating access of the majority to a good thing (the post at the Council, or at least the possibility thereof), one could also see Ms Prais as possessing a negative dominant (‘being Jewish’, ‘not being able to attend the test’), which grants access of the minority to a bad thing, or ‘negative good’ (exclusion from the test). Walzer discusses negative goods and negative dominance in Chapter 6 of Spheres of Justice, on ‘hard work’: Walzer, Spheres of Justice (n 45) 165–​183. See the discussion at section 2.3 above. Danchin (n 76) 11. Walzer, Spheres of Justice (n 45) 221–​224. ibid 226. Although Lenaerts notes that the first usage of the principle was in fact in an earlier case concerning the free movement of persons, namely, Case 152/​73 Giovanni Maria Sotgiu v Deutsche Bundespost [1974] ecr153: K Lenaerts, ‘L’égalité de traitement en droit communautaire: un principe unique aux apparences multiples’ [1991] cde 3, 13.

The ‘Article 19’ grounds

137

to enshrine this principle in various laws, not least Directive 2000/​78 itself. The Court now has a proper framework, it could be argued, to consider cases like Prais where a uniform, neutral requirement is applied to all employees (or potential ones), but puts persons having a particular religion or belief at a disadvantage. Indeed, the two new cases illustrate the principle nicely, as the vital element, the ‘apparently neutral provision, criterion or practice,’ as the Directive puts it,84 seems to be present in the case of Ms Achbita’s workplace, but absent in the case of Ms Bougnaoui’s, which explains the different conclusions of the two Advocates General. However, a finding of indirect discrimination brings with it greater opportunities of escape for the indicted employer, not only via Treaty-​or Directive-​ mandated derogations, but in fact via any legitimate aim shown to be proportionate. And the investigation of the purported legitimate aim can look, superficially, like a Walzerian analysis, as ascertaining the inclusion or not of the (usually negative) dominant held by the plaintiff within the meaning of the job distributed by the defendant, could be confused with compelling the defendant to defend their policy. Put another way, if something is a feature of the workplace able to be prayed in aid in an objective justification, then it is probably also part of the meaning of the job performed there. And indeed if it is the key attribute necessitating the apparently neutral policy, a sine qua non of that line of work, then it will almost certainly feature in the dna of the job when this falls to be analysed by a Walzerian. The questions, ‘Can the workplace live without the policy?’ and ‘Can the meaning live without the dominant?’ may thus amount to –​almost –​the same thing. Naturally the greater possibilities for exoneration tempt employers to push for a finding of indirect discrimination when in Court. Micropole seems to have done this in Ms Bougnaoui’s case, emphasising in its observations what it perceive[d]‌to be a fundamental contrast in this area of the law between the restriction of a right and the prohibition of discrimination.85 For the reasons stated above, it might be thought that what Advocate General Sharpston went on to call a conflict ‘between a restrictions-​based approach and one based on discrimination,’86 was simply an alternative way to express the difference between the Walzerian approach on the one hand, and the Aristotelian approach on the other. Sharpston AG’s comprehensive discussion 84 85 86

Article 2(2)(b). Opinion in Case 188/​15 Bougnaoui (n 70) para 58. ibid para 58ff.

138 Chapter 4 of said conflict, albeit prompted by the defendant and their rather strange framing of the debate, is refreshing, and she is right that, at least in the arena of eu law, and certainly where direct discrimination is concerned, the discrimination-​based approach has always dominated over the restrictions-​ based approach, which is reserved for cases of indirect discrimination.87 However, the Walzerian approach mooted in this book cannot be equated to the restrictions-​based approach described in the Opinion, not least because of its proposed application to indirect and direct discrimination cases and its focus on the rule qua rule, not qua restriction –​‘restrictions’ germane to the sphere in question would survive Walzerian scrutiny. The negative connotations of the loaded word ‘restrictions’ put the defendant on the backfoot from the get-​go, even if it is the approach they favour! Walzer avoids these negative connotations. He also avoids the ‘two wrongs make a right’ facility of the Aristotelian like-​for-​like test; it is implied within the indirect discrimination enquiry that, were the plaintiff and the rest of the workforce equally affected by the rule, there would be no further issue for Aristotle. This is sometimes known as the Equal Misery argument. However, complex equality may still be violated by a superfluous rule, even if it does apply to, and effect, everyone the same. Walzer acts to end the misery for everyone, and importantly ‘everyone’ may include some people who were put at a disadvantage but had chosen to suffer in silence. But were the rules superfluous on this occasion? On the face of it, being a practicing Muslim would appear to have nothing to do with the meaning of being a receptionist (Achbita), or indeed with the meaning of being a design engineer (Bougnaoui). However, in the case of the former, G4S’ regulation (although once an ‘unwritten rule’)88 that its staff should ‘dress neutrally’89 (an explicit policy of neutrality) complicates matters. And Micropole’s desire for ‘business … development’90 (i.e. to make a profit) by pleasing its customers, in turn by requiring Ms Bougnaoui to remove her veil, complicates them still further. Can this regulation, and this desire, be ‘written into’ the definitions of the two goods being distributed (that is, the two jobs), such that they will subsequently contaminate (for want of a better word) the meanings given to 87

88 89 90

Space forbids a longer discussion of the restrictions-​based approach, and the literature is copious, but a very good description can be found in G Davies, Nationality Discrimination in the European Internal Market (Kluwer 2003), especially 53ff, and this may serve as a nice starting point. Judgment in Case C-​157/​15 Achbita (n 60) para 11. ibid para 30. Opinion in Case 188/​15 Bougnaoui (n 70) para 23.

The ‘Article 19’ grounds

139

these two goods, and, ultimately, the way in which they are to be distributed? It is argued that the answer to this question should be in the negative. In juridifying Walzer’s system it is important not to lose sight of the tenets upon which the system is based. Put another way, the (figurative) citizens of a Walzerian world may not unstitch Walzer’s carefully crafted compartments from the inside, ‘defining’ their way back to caste, hierarchy and unfair advantage. A  Walzerian judge, acting in concert with a distributive community, or even acting alone, must still practice the Art of Separation. It is hierarchies –​ ‘the triumph of one good over all the others’91 –​which make for domination, and complex equality can only hope to counteract this domination where the goods are properly separated out. Thus, identifying the relevant good in Achbita as ‘a job at G4S, a company which pursues a policy of neutrality in banning its employees from wearing outward manifestations of their religion or belief,’ or the definition of the meaning of a job at G4S as including adherence to said policy, is impermissible per se, as it represents a breach of the boundary between the sphere of religion and the sphere of employment.92 One can tell that simply from the fact that the Court in Achbita felt it necessary, either to subjugate Ms Achbita’s right to her religious convictions under G4S’ freedom to run its business, or vice versa. This hierarchy-​creation is a red flag that the Art of Separation has not been practiced. It is submitted that there is a difference between the Aristotelian judge and the Walzerian judge in this case, and it is that, while the former can turn a breach of the equality principle into a non-​breach thereof if he or she assimilates the ‘unlikes’ in making the pursuit of a policy of neutrality a legitimate aim, the latter can never permit the boundary-​breaching meaning from being adopted as the correct one, because it is ipso facto impermissible.93 A community could decide to merge the sphere of religion and the sphere of employment, of course. It could then decide that religious concerns were employment ones and vice versa. G4S’ policy would then make sense and 91 92

93

Walzer, Thick and Thin (n 47). Such an identification, or definition, of the good at issue in Bougnaoui (i.e. the job at Micropole) would be similarly impermissible, the fact that the policy of neutrality was explicit in the one case and implicit in the other making no difference. One could say that Walzer rescues Aristotle from the more extreme manifestations of his obsession with symmetry. May’s reworking of Rancière’s theory of equal intelligence into a more general one of ‘equal worthiness,’ first encountered in Chapter 1, produces a similarly asymmetric effect: immigrants protesting their lot as against citizens would be protected by the theory, while heterosexuals protesting against equal rights for gays and lesbians would not be. See T May, Nonviolent resistance: a philosophical introduction (Polity 2015) 146. See supra, section 3.7.3, for a previous encounter with the (potential) equation of Walzer with indirect discrimination/ objective justification.

140 Chapter 4 the meaning of a G4S job could justly include references to religion. A priest in such a community might justly refuse to give the Last Rites to an unemployed parishioner. Such a society would not see a difference between the two matters: the occupational would be the clerical and the clerical would be the occupational. However, as long as the society in question denominated Employment and Religion as two separate spheres, the Art of Separation would demand that the two matters be kept apart to prevent hierarchisation between them. A  macro merger can be allowed because it represents the elimination of a boundary. However, a micro merger at the level of a single good (‘religion to be left at the door if you come and work for G4S’) can never be anything other than a boundary breach. Thus while the Aristotelian judge can abolish the separation in the substance of his or her judgment, the Walzerian judge can only safeguard the separation as it forms part of his or her way of judging.94 4.4 Age 4.4.1 The cjeu’s Case-​Law (a) Introduction: Mangold and Palacios Age has been described as ‘particularly difficult,’95 and even ‘the hardest of the four grounds’96 covered by the Framework Employment Directive.97 As discussed above, though,98 it is less a case of hardness or simplicity, or better or worse, than it is different courses for different horses. The ‘differences’ of age, as a ground for discrimination, are manifold, including the 94

95 96

97 98

A further interesting feature of the Bougnaoui and Achbita cases that will not be explored here, as it does not transfer across to the Walzerian context, is the way in which the Court ignored, and thus to all intents and purposes quashed, any claims which the two employees may have had in relation to intersectional discrimination, that is, the combined effect on them of both sex discrimination and religious discrimination. The Court had only a few months earlier taken a similarly dismissive line on intersectionality in its Judgment of 24 November 2016, Parris, C-​443/​15, ECLI:EU:C:2016:897, where the combination at issue was that of age discrimination and discrimination on grounds of sexual orientation. Meenan (n 15) 284. T Osbourne, ‘Will the European Union Directive on equal treatment fulfil its purpose of combating age discrimination in employment?’ (2004) 38(3) The International Lawyer 867, 880. Directive 2000/​78 (n 1). Text accompanying n 13, et seq.

The ‘Article 19’ grounds

141

obvious fact that noone’s age remains static.99 In the case of Age Concern (­discussed  below), Advocate General Mazák acknowledged the ‘genuine difference between age and the other grounds,’ including age’s ‘fluid’ nature.100 Those discriminating on grounds of age have less to fear from Directive 2000/​78 than those discriminating on one of the other grounds. For a start, national provisions laying down retirement ages are outside the purview of the Directive.101 In addition, payments made under state social security schemes, or social protection schemes, are exempted.102 As well as these two compulsory exemptions, the Directive provides for a number of optional ones (optional on the part of the Member States, that is). The prohibition on age discrimination, for example, need not be applied in respect of the armed forces.103 The fixing of ages in occupational pension schemes may also contravene the ban, if a Member State so decides.104 It is Article 6(1) of the Framework Employment Directive, though, which affords Member States the biggest scope to discriminate on grounds of age, or to permit such discrimination, without penalty.105 The first paragraph reads:

99

Of course, one’s religion, or one’s sexual orientation, or one’s being disabled or not, can change over time. But these statuses are very unlikely to change as often, or as inevitably, as age. Advocate General Jacobs has commented that, while a ground like sex involves a ‘binary criterion’, age is ‘a point on a scale’:  Case C-​227/​04 P Maria-​ Luise Lindorfer v Council of the European Union [2007] ecr I-​6767, First Opinion of AG Jacobs, para 84. That said, sex itself is now starting to be considered as a continuum: U  Lembke, ‘Tackling sex discrimination to achieve gender equality? Conceptions of sex and gender in eu non-​discrimination law and policies’ [2016] 2 European Equality Law Review 46, 47, and reference therein. Meanwhile, the mutability or immutability of race continues to be hotly contested, as the Rachel Dolezal controversy has demonstrated: R Blakely, ‘Race activist continues to insist white is black’ The Times (London, 17 June 2015). 100 Case C-​388/​07 Incorporated Trustees of the National Council on Ageing (Age Concern England) v Secretary of State for Business, Enterprise and Regulatory Reform [2009] ecr I-​1569, Opinion of AG Mazák, para 74. 101 Recital 14. 102 Article 3(3). 103 Article 3(4). 104 Article 6(2). Skidmore has commented that this derogation ‘runs the risk that it may result in other forms of discrimination:’ P Skidmore, ‘Ec Framework Directive on Equal Treatment in Employment: Towards a Comprehensive Community Anti-​Discrimination Policy?’ (2001) 30 ilj 126, 130. 1 05 Schiek has described Article 6 as being ‘loaden with exceptions’: Schiek (n 16) 301.

142 Chapter 4 Member States may provide that differences of treatment on grounds of age shall not constitute discrimination, if […] they are objectively and reasonably justified by a legitimate aim, […] and if the means of achieving that aim are appropriate and necessary. Thus, as long as a legitimate aim can be shown, which passes a standard objective justification test106 and a standard proportionality test, any instance of age discrimination  –​including direct age discrimination  –​ can, theoretically, evade the Directive. As Meenan has rightly pointed out, Article 6(1) is ‘infinitely elastic.’107 Waddington and Bell have called it ‘open-​ended.’108 Others have gone even further, accusing Article 6(1) of ‘legalis[ing] age discrimination’109 and warning that the exception ‘could end up swallowing’ the rule.110 Age was in fact the subject of the very first preliminary reference made to the Court of Justice specifically on Directive 2000/​78: Mangold v Helm.111 The case concerned the use of fixed-​term employment contracts. In Germany, the law placed two curbs on the use of fixed-​term employment contracts (in order to prevent their misuse), requiring an objective reason justifying the fixed term, or, alternatively, imposing limits on the number of renewals (three) or on duration of use (two years). However, German law permitted fixed-​term contracts, even without the above restrictions, if the employee was aged 60 or over. When Germany came to transpose Directive 2000/​78, by means of the ‘Law on part-​time working and fixed term contracts’112 (the ‘ftc Law’), this threshold was lowered to 58. In 2002, the threshold was again lowered, by the so-​called ‘Hertz Law,’ to 52.

106 It was argued in the Age Concern reference (Case C-​388/​07 Age Concern (n 100)), by Age Concern, that the additional words ‘and reasonably’ added an extra hurdle to the objective justification test in Article 6(1). However, this argument was rejected by both the Advocate General (see para 79 of the Opinion of AG Mazák) and the Court (see para 65 of the Judgment). 107 Meenan (n 15) 297. 108 Waddington and Bell (n 13) 599. 109 Eurolink Age cited in L Waddington, ‘Article 13 Ec: Setting Priorities in the Proposal for a Horizontal Employment Directive’ (2000) 29(2) ilj 176, 179, discussing the Directive when it was still at the proposal stage. 110 Osbourne (n 96) 874. 111 Case C-​144/​04 Mangold (n 2). 112 Gesetz über Teilzeitarbeit und befristete Arbeitsverträge und zur Änderung und Aufhebung arbeitsrechtlicher Bestimmungen.

The ‘Article 19’ grounds

143

In 2003, at the age of 56, Mr Mangold was hired by Mr Helm (a lawyer) on a fixed-​term employment contract; the contract contained no objective justification, or stipulation of a maximum number of renewals, or stipulation of a maximum duration. In proceedings brought by Mr Mangold, the Court of Justice was asked whether there was an incompatibility between ec law (in particular, Directive 2000/​78) and the German laws which deprived employees over 58 (and later 52) of legal protection against abuse, by employers, of this kind of contract. Advocate General Tizzano recalls the wording of Article 6(1) of the Directive, and then comments that the difference in treatment on grounds of age (as between those who can enter into fixed-​term contracts without restrictions, and those who cannot) ‘is… self-​evident.’113 He goes on to opine that this difference in treatment is objectively justified, albeit implicitly, by the aim of ‘enhancing the employability of unemployed older workers who… have particular trouble finding new employment.’114 In other words, the lack of a requirement to provide protection to the potential employee boosts his or her prospects of employment, as the employer will find such a candidate more attractive than one for whom protection must be provided. However, according to the Advocate General, this objective justification fails the proportionality test. The denial of stability within the employee’s employment relationship, which on the Advocate General’s reading of the rules could start as early as the age of 50, and would last until retirement, is therefore permanent. Such permanent ‘exclusion from… safeguards’115 is, in Tizzano’s view, disproportionate. The Opinion ends with consideration of a small point which has gone on to have a big significance. The Advocate General suggests that, in declaring the German laws to be incompatible with ec law, the Court should use, not Article 6 of Directive 2000/​78, but the general principle of equality itself. After all, the analysis (difference in treatment, objective justification, proportionality) would be identical. The benefit would be that general principles have horizontal direct effect, while Directives do not (and, in any case, the deadline for transposition of the Directive had not, at the material time, passed). The Court largely follows its Advocate General. The difference in treatment on grounds of age, brought about by the ftc Law and the Hertz Law, would be discrimination unless the conditions listed in Article 6(1) were met. 113 Case C-​144/​04 Mangold (n 2) Opinion of AG Tizzano, para 88. 114 ibid, Opinion of AG Tizzano, para 90. 115 ibid, Opinion of AG Tizzano, para 96.

144 Chapter 4 As regards objective justification, the Court, like Advocate General Tizzano, was satisfied that the German laws served the purpose of ‘promot[ing] the vocational integration of unemployed older workers, in so far as they encounter considerable difficulties in finding work.’116 Turning to proportionality, the Court commented that Member States had a ‘broad discretion in their choice of measures [to achieve employment policy objectives].’117 However, the ftc Law, combined with the Hertz Law, deprived all workers over the age of 52 of protection against abuse by employers, via indefinitely renewable fixed-​term contracts, and therefore of ‘stable employment.’118 It made no difference ‘whether or not they were unemployed before the contract was concluded,’119 or what the duration of any period of unemployment had been. This failure by the German authorities to take into consideration ‘the personal situation of the person concerned,’120 or indeed the structure of the labour market as a whole, meant that the German law went ‘beyond what [was] appropriate and necessary.’121 The imposition of what might be called a ‘blanket’ age-​threshold meant that Germany had failed the proportionality test. Finally, the Court picked up Advocate General Tizzano’s point about the preferability of using the general principle of equality in deciding this case, rather than the Directive. However, in a controversial move, the Court went further, by claiming that the principle of non-​discrimination on grounds of age was itself a general principle of Community law.122 The Court’s baptism of a general principle of non-​discrimination on grounds of age has provoked ire among some critics,123 who are unconvinced as to the sources124 of the new principle (which are, according to the Court, various international instruments, including the echr and the iccpr, and ‘the constitutional traditions common to the Member States’).125 It is well known that at the time of the drafting of the Amsterdam Treaty, the Member States, while happy to countenance the new Article 19 (then Article 13), were adamant that 116 117 118 119 120 121 122 123 124 125

ibid para 59. ibid para 63. ibid para 64. ibid. ibid para 65. ibid. ibid para 75. Although Meenan describes it as ‘enormously helpful:’ Meenan (n 15) 306. Or ‘alleged sources’ as Arnull puts it: A Arnull, ‘Editorial: Out with the old …’ (2006) 31 EL Rev 1, 2. Case C-​144/​04 Mangold (n 2) para 75.

The ‘Article 19’ grounds

145

it should not create rights of general application.126 Furthermore, many writers speculate as to whether a precedent has now been set, allowing all of the other grounds covered by Article 19 to be ‘upgraded’ (to use Muir’s word) to general principles.127 If so, this in turn raises questions about how certain other cases were decided.128 Almost two years after Mangold, the cjeu gave judgment in the case of Palacios de la Villa v Cortefiel,129 which concerned compulsory retirement in Spain. A Spanish law of 2005 legalized compulsory retirement clauses within collective agreements; workers over a given age could be compulsorily retired as long as this was ‘consistent with employment policy’ (the ‘employment policy aim’), and as long as they had completed the minimum contribution period for a pension. The 2005 law was given retrospective effect by means of the so-​ called ‘Single Transitional Provision’ (the ‘stp’), although the stp mentioned only the second of those two conditions, not the first. In proceedings brought by Mr Palacios de la Villa against his employer Cortefiel, on being compulsorily retired at 65 pursuant to the relevant collective agreement, the Court of Justice was asked whether the stp was precluded by the principle of equal treatment. Advocate General Mazák was of the opinion that the stp was not covered by the Framework Employment Directive, because retirement ages were exempted from the remit of the Directive by Recital 14.130 The Court of Justice disagreed. The matter at issue was the prevention of future participation in the labour force. This fell within Article 3(1)(c) of the Directive, which provided that the Directive would apply to ‘employment and working conditions, including dismissals and pay.’ Turning to Article 6(1), the Court acknowledged that no employment policy aim was mentioned explicitly in the stp, but nevertheless held that such an aim (in this case, regulating the national labour market for the purposes of checking unemployment) could be inferred from the context. As regards the proportionality test, the Court felt that the taking of this particular measure in order to facilitate access to the labour market (in other words, removing older workers in order to free up posts for younger 126 This can be confirmed by contrasting the wording of Article 19 tfeu (then Article 13 ec) with Article 18 tfeu (then Article 12 ec) prohibiting discrimination on grounds of nationality, as discussed by E Muir, ‘Enhancing the effects of Community law on national employment policies: the Mangold case’ (2006) 31(6) EL Rev 879, 889. 127 ibid 889–​890; Meenan (n 15) 306; Arnull (n 124) 2. 128 See below, text accompanying n 203. 129 Case C-​411/​05 Félix Palacios de la Villa v Cortefiel Servicios Sa [2007] ecr I-​8531. 130 In the event that the Court came to the opposite view, the Advocate General’s alternative submission was that the stp was justified and proportionate.

146 Chapter 4 workers) was ‘not unreasonable.’131 Also counting in favour of Spain’s conduct were the fact that a safeguard condition had been included to ensure that no older worker was retired before they had earned a full pension, and the fact that the ‘compulsory retirement mechanism’132 could only be used in the context of a collective agreement, which ensured flexibility, and consideration of any industry-​specific factors. In the light of all of this, the Court’s conclusion was that Spain’s action in passing the stp had been ‘appropriate and necessary.’ No mention was made of the general principle of non-​discrimination on grounds of age from Mangold. O’Cinneide has commented that, in Palacios, the objective justification test was applied ‘with considerable rigour,’ with the justification itself being subjected to ‘close scrutiny.’133 Connolly meanwhile appears to take the opposite view, criticizing the Court for its ‘generous’ approach (or ‘light touch’ as he puts it) in accepting an implicit purpose for the stp, discerned from ‘evidence,’ as opposed to an explicit one. He regards Palacios as a ‘departure from the stricter approach’ of cases like Mangold, with the Court apparently prepared to approve ‘vague’ goals as justifications; according to him, Spain’s attempt to redistribute jobs from the old to the young is little more than a ‘ruse,’ leaving ‘those most in need of protection’ without it.134 Schiek agrees that the Court in Palacios engages in a ‘loose standard of scrutiny’ of Spain’s justifications, compared to the ‘strict standard’ employed in Mangold.135 (b) Compulsory Retirement and Other Age-​Dependent Restrictions The issue of compulsory retirement has returned to the Court of Justice in the cases (among others) of Age Concern,136 Petersen137 and Fuchs and Köhler.138 In the Age Concern reference, the English charity Age Concern (among ­others) had sought a judicial review of certain provisions of the Employment

131 Case C-​411/​05 Palacios (n 129) para 72. 132 ibid para 74. 133 Both phrases from C O’Cinneide, ‘Age Discrimination and Mandatory Retirement’ (2008) 6/​7 Eur Anti-​discrimination L Rev 13, 15. 134 Quotations from M Connolly, ‘The ecj signals a light touch towards age discrimination and compulsory retirement’ (2007) 81 Emp lb 1, 3 (emphasis added). 135 Schiek (n 11) 777, 788 and 785. 136 Case C-​388/​07 Age Concern (n 100). 137 Case C-​341/​08 Domnica Petersen v Berufungsausschuss für Zahnärzte für den Bezirk Westfalen-​Lippe [2010] ecr I-​47. 138 Judgment of 21 July 2011, Gerhard Fuchs and Peter Köhler v Land Hessen, Joined Cases C-​159/​10 and C-​160/​10, ECLI:EU:C:2011:508.

The ‘Article 19’ grounds

147

Equality (Age) Regulations 2006, which was one of the measures adopted by the uk for the purpose of transposing Directive 2000/​78. Age Concern argued that the transposition was flawed. According to Age Concern, Article 6(1) did not permit Member States to introduce a general defence of justification for direct age discrimination, but that was just what the Regulations did (in stipulating that employers may dismiss employees at the age of 65 without having to show any objective justification whatsoever). According to Age Concern, for the Directive to be correctly transposed, the discriminator (that is, the employer) would still need to show a legitimate aim, and proportionality. The uk disputed this. The High Court made a reference to Luxembourg. In September 2008, Advocate General Mazák delivered his Opinion. He found without difficulty, following Palacios, that the Regulations fell within the scope of the Directive. Turning to the substance, Age Concern had argued that the Member States had to specify which differences in treatment were potentially capable of being justified, and had to specify what aims were capable of justifying such differences. The national court in its questions had referred to the possibility of the Member State drawing up a ‘list’ to define the justifiable differences. The uk rejected Age Concern’s argument, and stated that a list would be unrealistic and inappropriate. The Advocate General agreed; implementing legislation had to be specific and sufficiently clear (which, in his opinion, the Regulations were), but a list would be ‘impossible.’139 Thus, the Regulations were not incompatible with Article 6(1). Member States (but not, importantly, individual employers) needed to proffer an identifiable (although not necessarily express) ‘legitimate aim,’ and the means employed to achieve it needed to be “appropriate and necessary.’ However, the Advocate General left it to the national court to decide whether, in this particular case, these two obligations had been discharged. In March 2009, the Court handed down its judgment. Having confirmed that the Regulations did fall within the scope of the Directive, the Court went on to explain that, when transposing a directive, Member States had to make sure that the implementing law was effective, but that they retained a broad discretion as to the methods to be used. Thus, the uk could not be impugned for having chosen not to draw up a list of justifiable differences. Neither did it matter if legitimate aims were not stated explicitly, as long as they could still be identified. However, both the ascertaining of the legitimate aim, and the proportionality test, were jobs for the national court. The only guidance which the Court would give was that, when weighing up whether a specific measure was ‘appropriate and necessary,’ the national court should reject

139

Case C-​388/​07 Age Concern (n 100) Opinion of AG Mazák, para 54.

148 Chapter 4 ‘[m]‌ere generalizations:’140 a generalization could not constitute evidence of the suitability of a measure for its purpose.141 In Petersen,142 a German dentist was informed that her authorization to provide ‘panel’ dental care (that is, dental care to those insured under the statutory health insurance scheme) was to expire when she reached the age of 68. Ms Petersen having brought an action, a reference was made to the cjeu asking the Court whether setting a maximum age of 68 for practice as a panel dentist was an objective and reasonable measure to protect the health of patients insured under the statutory scheme. The Court in fact looked at three possible justifications for the German rule, which it said had been ‘mentioned’ by the referring court, although only the first had been ‘relied on.’143 With regard to this justification, the Court held that, given that there was an exception to the rule to the effect that dentists outside the panel system could work on after 68, the proffered objective of protection of public health could not be regarded as legitimate. However, the Court said that the rule could potentially be justified on both of the other grounds, namely, the protection of the financial balance of the German health system and the need to distribute employment opportunities among the generations. In Fuchs and Köhler,144 two State Prosecutors from Land Hessen, Germany, brought an action after they were compulsorily retired at 65, pursuant to some legislation passed by the Land under the federal law implementing Directive 2000/​78. The referring court set out a number of justifications potentially to be relied on by Land Hessen, reproduced numerically below for ease of reference: 1. To ensure an appropriate age structure in order to encourage the recruitment and promotion of young people; 2. To improve personnel management;

140 ibid para 51. 141 It is worth noting that the uk’s Regulations did not contain a safeguard condition, like the one in Palacios, requiring that employees not be compulsorily retired if they had not yet acquired sufficient pension entitlements. As the safeguard condition was such a key plank in the Court’s finding (in Palacios) that the stp was proportionate, surely the absence of such a condition can only lead to a finding of lack of proportionality? This point is made by both O’Cinneide (n 133) 19 and Connolly (n 134) 4. For an interesting analysis of the cjeu judgment, see M Connolly, ‘The Heyday case: much ado about very little’ (2009) 90 Emp LB 2. 142 Case C-​341/​08 Petersen (n 137). 143 ibid para 38. 144 Judgment in Joined Cases C-​159/​10 and C-​160/​10 Fuchs and Köhler (n 138).

The ‘Article 19’ grounds

149

3. To preclude individual legal disputes with older employees over their continued fitness for service; and 4. To make budgetary savings. The Court had no doubt that the compulsory retirement of the Prosecutors at 65 represented a difference in treatment (with those below 65) on grounds of age. Turning to the matter of objective justification, the Judges were on the face of it happy to accept all four of the above justifications (though pausing to check that safeguarding of an appropriate age structure was, where justice was concerned, a public interest and not just an interest of the employers). When it came to the proportionality enquiry, the test of suitability was passed (if the Prosecutors were not sacked, they would not go), and, in relation to all four of the justifications, the cjeu was satisfied that this action did not go beyond what was necessary to achieve the aims/​outcomes in question. The Court did however note in relation to Justification 1 that it was a matter for Land Hessen to find the right balance between maintaining an appropriate age structure and the protection of older workers, and in relation to Justification 4, that this justification could only be used (as here) in conjunction with others, and not on its own. (c) Maximum Age Limit for Commencing a Career Sometimes an individual is told that they are too old, not at the end of a period of employment, but when they apply for the job in the first place. This potentially puts the employer at odds with the ban on age discrimination in Directive 2000/​78. At the opposite end of the scale are those who, if not exactly told they are too young for the job, nevertheless suffer a detriment in comparison to older colleagues. The first of these behaviours will be considered in this section, the second in the section which follows. In the case of Wolf,145 Mr Wolf was refused entrance to the German fire service at the age of 31 on the grounds that, pursuant to German law, there was a maximum recruitment age of 30. The German Government claimed that this breach of the ban on age discrimination was justified by the following aims: 1. To ensure long careers for officials; 2. To limit the amount of social benefits paid; 3. To set up a balanced age structure within an occupation; and 4. To ensure a minimum period of service before retirement. 145

Case C-​229/​08 Colin Wolf v Stadt Frankfurt am Main [2010] ecr I-​1.

150 Chapter 4 Furthermore, the Government was satisfied that the setting of a maximum recruitment age for intermediate career posts was appropriate and necessary in relation to the aims pursued. Mr Wolf’s action having been referred to the Court of Justice, Advocate General Bot asked himself:  What constituted the essential aim of the maximum recruitment age? He decided that it was the third of the aims mentioned above, namely, the need for operational capacity to ensure the proper functioning of the fire service.146 Indeed, he pointed out, the legitimacy of this aim was confirmed by the wording of Recital 18 to the Directive concerning the preservation of the operational capacity of, inter alia, the emergency services.147 The Advocate General continued that holders of intermediate career posts in the fire service were working in the field, placing a high physical demand on their bodies over a long period. If a thirty year old underwent the two years of training, he would then only be able to provide the fire service with eighteen, or perhaps even only thirteen, years of service (the exact age at which the high physical demand mentioned could no longer be met not being precisely ascertained).148 Besides, the proper functioning of the fire service required that the young should in time replace the old, which meant that there was a need for young fire fighters to enter the service.149 The difference in treatment was therefore, in Bot AG’s opinion, justified, either under Article 6(1) of Directive 2000/​78, or, in the alternative, under Article 4(1) of the Directive, which permits exceptions to the prohibition on age discrimination where there exists a ‘genuine occupational requirement’ of the job in question.150 On the question of proportionality, the Advocate General was satisfied that the maximum recruitment age did not go beyond what was necessary, and concluded his Opinion by observing that, in the event that the Court chose to decide the case under Article 6(1), it should note that Member States had a broad discretion in relation to social and employment policy.151 The Court agreed with the Advocate General’s assessment of the essential aim of the measure in question, and that it did result in a difference in treatment on grounds of age. The Court’s judgment drew attention for its heavy emphasis on Article 4(1) of the Directive as the appropriate provision on the basis of which to decide the case. The Court decided from the 1 46 147 148 149 150 151

ibid, Opinion of AG Bot, para 37. ibid, Opinion of AG Bot, para 38. ibid, Opinion of AG Bot, para 31. ibid, Opinion of AG Bot, paras 32–​34. ibid, Opinion of AG Bot, para 35ff. ibid, Opinion of AG Bot, para 44.

The ‘Article 19’ grounds

151

wording of Article 4(1) that, for a plea under this article to succeed, the genuine occupational requirement in issue had to be a requirement in relation to another characteristic related to age, which in the case of Mr Wolf was, according to the Court, physical fitness: possession of full physical capacity was related to age because those over forty five or fifty years of age could not carry out high physical tasks in the same way that the young fire fighters could. Meanwhile, in respect of objective justification, the Court was satisfied that the concern to preserve occupational capacity was a legitimate justification, again citing Recital 18 in evidence, and that the setting of a maximum recruitment age did not go beyond what was necessary to achieve this.152 (d) Discounting of Seniority Accrued Before a Minimum Age In Hütter,153 Mr Hütter challenged an Austrian law by which, where a benefit was to be given to a worker in recognition of their professional experience, such experience attained before the age of majority could not be taken into account. In his own case, Mr Hütter had undertaken an apprenticeship at the same time as another individual twenty two months his senior, with the result that this individual gained more accreditation for their professional experience then he did. On reference to the Court of Justice, the Austrian Government informed the Court that the law in question was intended to fulfil three aims. Firstly, it was intended to bring about equal treatment as between apprentices and those young people who have stayed in school. Secondly, it was intended to reduce costs. Thirdly, it was intended to integrate young people into the labour market. Intervening in the case, the Danish Government argued that this law should not be precluded by the Framework Employment Directive, as there was a legitimate vocational training objective being pursued, and the Member States had a wide discretion in these matters according to the Court’s own jurisprudence. Meanwhile, the Commission argued that, although all three of the stated aims were acceptable on the face of it, the law in question was not proportionate and necessary for attaining them. The Court was inclined to agree with the Commission. Having found a difference in treatment as between those who completed professional experience before the age of majority and those who completed it afterwards, the Court went on to hold that although the three aims put forward by the Austrian Government were legitimate vocational training objectives, which the Court had held to come 152 153

ibid, paras 35–​44. Case C-​88/​08 David Hütter v Technische Universität Graz [2009] ecr I-​5325.

152 Chapter 4 within the scope of the Article 6(1) exception in Age Concern,154 the means which Austria had chosen to achieve those aims (or at least the first and third) were not ‘appropriate or necessary’ to their achievement, within the meaning of the same provision. With regard to the third one, for example, it could be seen that the law in question affected workers throughout their careers, and therefore could not be said to only help young people.155 Kücükdeveci156 concerned a German law under which periods of employment completed by an employee before reaching the age of twenty five could not be taken into account in calculating the notice period for dismissal. Ms Kücükdeveci had been employed by Swedex between the ages of eighteen and twenty eight, but in 2006, on dismissing her, Swedex only gave her one month’s notice corresponding to the period of three years after her twenty fifth birthday; had she been employed for ten years following her twenty fifth birthday, she would have received three months’ notice. Ms Kücükdeveci challenged the German law, alleging age discrimination. In his Opinion, Advocate General Bot, stressing that he preferred to use the Directive rather than the general principle inaugurated in the Mangold decision,157 found there to have been a difference of treatment in Ms Kücükdeveci’s case, that is, between those who joined the company before their twenty fifth birthday and those who joined it afterwards. With this decided, the next issue was objective justification under Article 6 of the Directive, which allowed for age discrimination in pursuit of social policy objectives. Germany pleaded three main justifications for this legislation (passed in 1926): afford employers greater flexibility in personnel management by • Toalleviating the burden on them in respect of the dismissal of young

• •

workers, on the grounds that It is reasonable to expect a greater degree of personal or occupational mobility from younger workers, and To facilitate the recruitment of young workers.

Bot AG was not convinced that flexibility of personnel management was of a ‘public interest nature,’ as this phrase had been used by the Court in 1 54 155 156 157

ibid, para 41, referring to Case C-​388/​07 Age Concern (n 100) para 46. Case C-​88/​08 Hütter (n 153) paras 44–​50. Case C-​555/​07 Seda Kücükdeveci v Swedex GmbH & Co kg [2010] ecr I-​365. According to Bot AG, to confine himself to interpreting the general principle would ‘deprive Directive 2000/​78 of all useful effect:’ ibid, Opinion of AG Bot, para 34.

The ‘Article 19’ grounds

153

Age Concern,158 and dismissed the argument about the facilitation of the recruitment of young workers as ‘theoretical.’159 As for the justification ­concerning the supposed ease with which younger workers could react to losing their job, the Advocate General stated that, while this may have been true in 1926, it was not true today, and in any event the law in question went beyond what was appropriate and necessary to achieve such objectives, as all workers who initiate an employment relationship before the age of 25 and who are dismissed … shortly after reaching that age are excluded, in a general way, regardless of their personal and family situation or their level of training, from an important part of the protection afforded to workers in case of dismissal.160 The Court was a little ambiguous on the question of whether to use the general principle or the Directive in deciding the case, holding that it is the general principle of European Union law prohibiting all discrimination on grounds of age, as given expression in Directive 2000/​78, which must be the basis of the examination.161 Going on to consider whether there was a difference in treatment in Ms Kücükdeveci’s case, the Court answered this in the affirmative, and then moved on to justification under Article 6(1). The judges seemed satisfied that the three purported justifications mentioned above were, at least on the face of it, allowable under Article 6. However, like the Advocate General, they had doubts as to the appropriateness of the German law. With regard to the justification having to do with the supposed greater degree of mobility possessed by young workers, for example, this was not consonant with the fact that the law in question would penalize an individual who joined the company before their twenty fifth birthday ‘whatever their age at the time of dismissal.’162 The German law was therefore illegal. 158 Case C-​388/​07 Age Concern (n 100) para 46. 159 Case C-​555/​07 Kücükdeveci (n 156) Opinion of AG Bot, para 45. 160 ibid, Opinion of AG Bot, para 52 (emphasis added). It will be remembered that the Court in Age Concern had ruled out the use of ‘mere generalizations’ in justifying derogations from the prohibition of age discrimination: Case C-​388/​07 Age Concern (n 100) para 51. 161 Case C-​555/​07 Kücükdeveci (n 156) para 27 (emphasis added). 162 ibid para 40.

154 Chapter 4 4.4.2 A Walzerian Analysis Complex equality is very well suited to the task of helping judges to decide discrimination cases owing to its novel focus on distribution, which facilitates the exposure of one of the mainsprings of discrimination: assumptions. That assumptions are one of the major sources of discrimination, across all grounds, has been well explained by Bell (in the gender and sexual orientation discrimination contexts), McGlynn (in the age discrimination context) and Quinn (in the disability discrimination context). Bell writes: Discrimination against women is largely based on … gender-​based assumptions relating to [women’s] social role. The physical sex of the individual may be the marker through which discrimination is visualized, but it is not the physical characteristics of women per se which is at the root of the discrimination against them. In other words, a woman is not denied promotion to management level because the employer believes her physical body is inappropriate, but because of arbitrary and prejudicial assumptions about the capabilities of women as a social group, and their proper place in society. The underlying objective of anti-​ discrimination legislation in this perspective is to breakdown [sic] such gendered stereotypes.163 But sometimes the anti-​discrimination legislation not only fails to break down the stereotypes, it reinforces them. McGlynn’s concern, for example, is not so much the way assumptions inform the thinking of society’s discriminators (such as employers), but the way in which they pervade even the legislative arena, so that ‘objective’ justifications for discrimination suggested by an instrument like the Framework Employment Directive are little more than these same assumptions, repackaged: [T]‌he suggested legitimate justifications for differential treatment on grounds of age in [the Framework Employment Directive] may be based on assumptions regarding national retirement ages […]. Equally, justifications for age discrimination based on requirements of professional training, for example, may also be based on ageist assumptions regarding the nature of the labour market and the desirable age of employees. […] These assumptions appear to be out of step with the moves towards a more flexible labour market, ‘portfolio’ careers and the removal of 163 M Bell, ‘Shifting Conceptions of Sexual Discrimination at the Court of Justice: from P v S to Grant v swt’ (1999) 5(1) Eur LJ 63, 65 (Bell’s emphasis, footnote omitted).

The ‘Article 19’ grounds

155

assumptions about the nature of jobs and the nature of the employees required to fill them.164 Tyranny, as has been explained before, involves having a ‘badge’ from another sphere, which actually only has Meaning X, but which is allocated Meaning Y by the gate-​keepers of the neighbouring sphere. This in turn involves making an assumption; the gate-​keepers of the neighbouring sphere do not need to examine each newcomer separately, when they can simply rely on a long-​held rule of thumb (‘If something has Meaning X, it must also have Meaning Y’). Quinn describes this very well: [D]‌iscrimination may be motivated … by the use of proxies or stereotypes concerning the assumed characteristics of group members. […] Disability is commonly –​and mistakenly –​taken as a proxy for inability to perform the routine tasks of life. […] [I]t is … impermissible to use [the proxies] to cloud rational judgments about individual ability since it is always possible that individuals will not conform to the stereotype. It is fundamentally unfair not to afford everyone an equal chance of proving themselves. […] [T]he proxies are highly inaccurate and rest on encrusted layers of unexamined presuppositions that have piled up over the centuries.165 As might be predicted, assumptions make for very porous boundaries between spheres, and are thus prime targets for the complex egalitarian. 164 McGlynn (n 11) 290–​1. Even the judiciary is not immune to assumptions. In the religious discrimination case of Şahin v Turkey App no 44774/​98 (ECtHR, 10 November 2005), the European Court of Human Rights justified the inequality caused when Ms Şahin was forced to obey a secular dress code, by reasoning that the wearing of the Islamic headscarf was ‘perceived as a compulsory religious duty’, imposed by men, sometimes against the wishes of the wearer, thus violating the ‘equality before the law of men and women’ (paras 115–​116). In his dissenting judgment, Judge Tulkens strongly admonished the majority for relying on mere perceptions about this particular religious practice and its application (see the Dissenting Opinion of Judge Tulkens, para 12), which Danchin went as far as to call ‘an essentialized and unexamined set of assumptions’: Danchin (n 76) 27. 165 G Quinn, ‘Disability discrimination law in the European Union’ in Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 244–​5. Schiek describes the process of stereotyping as ‘assimilationist pressure:’ Schiek (n 135)  781. Miller regards income as one of the most important proxies. If a person has a high income, it is often, if wrongly, assumed that they must have the skills to succeed in all of the other spheres: D Miller, ‘Complex Equality’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup1995) 213.

156 Chapter 4 In a case like Age Concern, what is really being distributed (from a Walzerian point of view) is continued employment, with youth, at least potentially, acting as a dominant good.166 When a worker under the age of 65 flashes their ‘badge of youth,’ youth is apparently converted into ‘ability to do the job,’167 and access to the sphere of employment is granted (or extended). As Walzer himself puts it, ‘[t]‌he conversion process violates the common understandings of the goods at stake’.168 A judge would hear evidence from both sides. Counsel for the Government might well claim that the common understanding of work in the uk was that only younger people could do it, meaning that, in fact, there had been no boundary breach at all. But this would obviously have to be convincingly proven. If it was not (as seems likely), the judge could conclude without difficulty that the distribution of work in the uk was flawed, and could then speedily move on to the arguably more important business of putting this right (or of sending the matter back to the uk for rectification there).169 No protracted, not to mention distracting, comparison between those under 65 and those over that age would be remotely necessary. It can thus be seen how use of Walzer’s theory would enable a judge to pinpoint, with ease, the true core of the case, that is, the false assumption being made and perpetuated on the boundary between the two spheres at issue.170 Exactly the same boundary breach, predicated on the same assumption, can be observed in Mangold and Palacios (although the exact distribuends vary). In both cases, it is a natural or legal person who has allegedly discriminated on grounds of age, not a Member State.171 But for the purposes of this section it makes no difference to the outcome. Counsel for Mr Helm would have to convince the judge that being young formed part of the common understanding of a restricted fixed term employment contract (and the employment certainty that went with it) in Germany. And counsel for 166 It would also, of course, be possible to view the case the other way around –​the distribuend would then be compulsory retirement, and the (negative) dominant would be old age. 167 For the purposes of this section, it is taken for granted that ‘ability to do the job’ is the correct (and only) distributive criterion within the sphere of employment. Of course a distributive community might settle on a different ‘shared meaning’ of employment. 168 Walzer, Spheres of Justice (n 45) 12. 169 Of course, in the event, the Court of Justice’s contribution to this case was entirely technical, thus shunting the whole of the principal issue (age discrimination) back to the national court. 170 The Sphere of Employment and the ‘Sphere of Age.’ It is wholeheartedly admitted that the latter ‘sphere’ is difficult to picture, unless age is (continuously) distributed to man by Old Father Time. However, a non-​literal reading of Walzer’s theory is urged. See section 3.3 for more on whether characteristics are (or even need to be) distribuends. 171 Although Member State legislation is very much at stake.

The ‘Article 19’ grounds

157

Cortefiel would have to convince the judge that being young formed part of the common understanding of continued employment (the same task as the uk’s in Age Concern) in Spain. Another case where Walzer’s theory would facilitate the pinpointing of the true core of the case would be Petersen. Here, those able to show the ‘badge of youth’ were accorded authorizations to provide panel dental care, while those over 68 came away empty-​handed. The Court reasoned that the fact that the age restriction did not apply vis-​à-​vis non-​panel work fatally undermined the German Government’s claim that its purpose was to protect public health. This reasoning is sound, but relies on comparison between the two sectors to show up a contradiction in the German Government’s approach; the danger is that if the age restriction had been applied across the board, the Court would have found it to be justified. This kind of approach fails to answer the question which the case is really about:  do dentists over 68 make more mistakes than any other dentists? A complex egalitarian would ask whether the fact of being below 68 formed part of the meaning of dental care, and the provision thereof. While in this case both approaches might well have led to the same result (with the age restriction failing the justification test –​at least on this ground), it is submitted that the Walzerian approach reaches the result via the more appropriate route. Perusal of the raft of age discrimination cases since 2010, most from Germany, reveals similar themes. Particularly prevalent are the rule of thumb issue mentioned above –​see, for example, how age is used as a proxy for physical fitness in the Wolf case172 –​and the stereotype-​reinforcement point –​see, for example, the cjeu’s endorsement of this equivalence (age-​ fitness) in its judgment. The Walzerian judge meanwhile would enquire as to the meaning of being a firefighter in the community concerned. For the sake of argument, this may be supposed to be the status of being employed by the fire brigade for the purposes of assisting in the putting out of, and rescuing of citizens from, fires, which, necessitating as it does the ability among other things to perform a fireman’s lift, requires physical fitness on the part of said employee. Now, while the badge of fitness (part of the meaning) may be legitimately flashed by candidates wishing to enter this sphere, production of the badge of age, or more specifically, the status of being over or under thirty (not part of the meaning), may not be required of such entrants. Walzerian boundary patrol compels the exclusion of this status as a dominant (negative or positive, depending), and the prohibition of the job-​for-​age exchange as one blocked under a regime of Complex 172

Case C-​229/​08 Wolf (n 145).

158 Chapter 4 Equality. Those who are fit over thirty, and interestingly those unfit under thirty, would be admitted and rejected respectively. It is conceded, though, that use of the Walzerian model would demand a casuistic or empirical approach; that is the inevitable result of dispensing with rules of thumb, generalisations, and other ways employers may try to take a short cut from (among other things) knowing who the candidate is, to knowing whether or not they are suitable. Thus it will be the result anyway in those cases where the Court has rejected an assumption-​born objective justification. As Connolly puts it, [E]‌mployers should assess each candidate according to his or her physical strength.173 Schiek also comments, An approach based on principles of non-​discrimination would suggest that employers must make individual assessments rather than using age-​ related stereotypes.174 Thus, the increased costs of individual testing (what one might call a move to subjective justification) are on their way, even if the model of equality utilised by the Court remains unchanged. Returning to the stereotype-​reinforcement, or ‘rubber stamp’ point, a number of the post-​2010 age discrimination cases seem to exemplify the contagion of assumptions from employers to legislature, and from legislature to judiciary, with each simply endorsing the last, like the mutual-​verification process of the ship’s captain and the watchmaker.175 It has already been seen how well-​suited the Aristotelian test is to this phenomenon, with comparability/​objective justification a matter for employers or (especially where age is concerned) States, and proportionality a matter for judges, and each confirming the others’ prejudices in an endless feedback loop. As Rodgers has noted, 173 M Connolly, ‘The coalition government and age discrimination’ [2012] 2 Journal of Business Law 144, 158. 174 Schiek (n 135) 793. 175 Mirowski nicely summarizes the story behind this phenomenon, sometimes known as the Zanzibar Effect, as follows: ‘[A]‌ship’s captain […] fired a gun every day at noon according to a clock that had been set by comparison with the timepieces in a shop of a watchmaker in Zanzibar, who set his clocks by the report of the gun at noon.’ See P Mirowski, The effortless economy of science? (Duke University Press 2004) 148.

The ‘Article 19’ grounds

159

instead of the proportionality test serving as a significant restriction on member states seeking to justify discrimination on the grounds of age, in some cases, it has hardly represented a barrier at all.176 This is clearly on display in Wolf,177 where, according to Schiek, the Court accepted reference to stereotyping, to the effect that older people will always become incapable of performing physically demanding tasks.178 The Court’s temporary return to an outlawing of generalisations in Kücükdeveci179 bucked the trend a little, and showed the judges interpolating their own analysis at the proportionality stage, and testing the robustness of the State’s logic, for example, vis-​à-​vis the increased mobility of the young or its supposed positive impact on employer flexibility, rather than just nodding it through. As Bribosia and Bombois comment, the exclusion of such blanket principles is the least the Court can do, failing which the proportionality test will become, ‘a simple blind validation of the wishes of the relevant rule’s author.’180 The Walzerian judge avoids this danger by eschewing the simple dichotomy of comparability/​objective justification on the one hand, and proportionality on the other, which lends itself perfectly to the mutual-​verification process mentioned above, in favour of one single decision, or rather decision-​making process, albeit one which continues ad infinitum. Certainly the Walzerian judge is ‘reviewing’ the Distributor’s decision, but pursuant to the same, single and independent criterion –​the meaning of the distribuend. If the distribution did not match the meaning, then it is flawed and declared invalid. If the meaning has shifted, then the distribution may have to shift with it.181 Thus, 176 L Rodgers, ‘Labour law and the “public interest:” discrimination and beyond’ (2011) 2(4) European Labour Law Journal 302, 315. 177 Case C-​229/​08 Wolf (n 145). 178 Schiek (n 135) 793 (footnote omitted). 179 Case C-​555/​07 Kücükdeveci (n 156). 180 ‘une simple validation aveugle des volontés de l’auteur de la norme examinée’: E Bribosia and T Bombois, ‘Interdiction de la discrimination en raison de l’âge:  du principe, de ses exceptions at de quelque hésitations  –​Réflexions sur des arrêts Wolf, Petersen et Kücükdeveci de la Cour de justice de l’Union européenne’ (2011) 47(1) Revue trimestrielle du droit européen 41, 58. English translation author’s own. 181 Joly makes the excellent point, for example, in relation to Wolf, that ways of fire-​fighting could change over time: L Joly, L’emploi des personnes handicapées entre discrimination et

160 Chapter 4 any initial wrong is corrected, rather than simply partnered with a second wrong in the hope that the two will make a right. The Walzerian judge must investigate the meaning of the good de novo to fulfil his or her task and may not simply ‘trust’ the Distributor’s definition, in the way that the Aristotelian judge may, and often does, simply trust the employer or State’s comparability analysis and/​or objective justification, or worse, as seen in a number of these cases, trust the employer’s analysis/​justification because it is based on the State’s. This regrettable result is especially likely in the field of age discrimination where the directive which the judge is charged with interpreting itself nominates the States as the proper author-​designers of ban-​defying rationales and mechanisms. Under Aristotle, the State’s rationale and/​or mechanism are sanctioned because it is the State’s rationale and/​or mechanism which count. Under Walzer, the situation is quite the opposite: the Distributor’s definition is queried because it is the community’s definition which counts. 4.5

Sexual Orientation

4.5.1 The cjeu’s Case-​Law This section will begin with an appraisal of the Court of Justice’s approach to the question of transsexuality in the landmark case of P v S.182 It is stressed that this is not to imply that a person’s transsexuality has anything to do with their sexual orientation. It does not, and indeed the Court acknowledged this in P v S, where the Luxembourg judges rightly recognized that transsexuality, and gender reassignment, are a matter of gender.183 However, although the judges égalité (Dalloz 2013) 219. Opsahl makes the same point: ‘What is held to be justified is not static, but changes with time; what was held to be a differentiation yesterday may be tantamount to discrimination tomorrow’: Torkel Opsahl, ‘Equality and non-​discrimination’ in T Opsahl, Law and equality: Selected articles on human rights (Ad Notam Gyldendal 1996) 175. 182 Case C-​13/​94 P v S and Cornwall County Council [1996] ecr I-​2143. 183 As Waaldijk and Bonini-​Baraldi explain, the concept of ‘sexual orientation’ is sometimes given a wider meaning to include phenomena that are related to (what they call) ‘sex-​ as-​gender’, such as transsexuality or transvestism. And indeed in some Member States’ laws, ‘sex-​as-​gender’ phenomena have been included within the concept of sexual orientation, for example, transvestism in Denmark. However, in the light of P v S, Waaldijk and Bonini-​Baraldi are of the view that the appropriate course is to treat transsexuality as a matter of sex, not sexual orientation. See further K Waaldijk and M Bonini-​Baraldi, Sexual orientation discrimination in the European Union:  national laws and the Employment Equality Directive (tmc Asser Press 2006) 96–​97. Others regret the submerging or ‘covert

The ‘Article 19’ grounds

161

were prepared to stretch their traditional understanding of the concept of gender to aid a transsexual employee, in the later case of Grant,184 they famously drew the line at aiding a homosexual one, at least until there was a stand-​alone prohibition for discrimination on grounds of orientation.185 In order to understand Grant and its successors it is thus helpful to keep in mind the transsexuality cases, starting with P v S. Transsexuality usually involves a medical condition called gender disphoria, wherein the patient’s biological sex (sometimes called ‘chromosomal’ sex) does not correspond to their sexual identity; in most cases, the condition can be eased by medical treatment culminating (if desired) in surgery for gender reassignment. P was a manager at an educational establishment, of which S was the principal and chief executive. P was a transsexual whose biological sex was male, but whose sexual identity was female. (In the uk at the material time, it was impossible to have the sex originally attributed to a person altered in the register of births, marriages and deaths.) P was dismissed after she186 underwent gender reassignment surgery, and brought an action claiming sex discrimination. The Industrial Tribunal at Truro made a reference. Advocate General Tesauro was not convinced by the argument (made by the uk) that the correct comparator for a male-​to-​female transsexual was a female-​to-​male transsexual (who of course would have been treated equally badly, meaning that there could not have been any gender discrimination –​ the so-​called ‘equal misery’ argument). Rather, he imagined what things would have been like if P had remained a man. Obviously, in that case, she would not have been dismissed. He concluded: ‘where unfavourable treatment of a transsexual is related to […] a change of sex, there is discrimination […] on grounds of sex’.187 To argue that this was not discrimination ‘between the two sexes’188 inclusion’ of ‘discrimination by reason of gender reassignment’ within the general category of sex, and would prefer to see gender identity as an autonomous ground. See M Bell, Anti-​Discrimination Law and the European Union (Oxford Studies in European Law, oup 2002) 110, with reference at footnote 123 to a failed initiative by ilga-​Europe. 184 Case C-​249/​96 Grant (n 2). 185 ibid para 48. 186 It is proposed to follow Advocate General Tesauro’s lead in referring to P as a female, that is, taking into account her sexual identity as opposed to her biological sex: Case C-​13/​94 P v S (n 182) Opinion of AG Tesauro, para 4. 1 87 ibid, Opinion of AG Tesauro, para 18. 188 ibid, Opinion of AG Tesauro, para 20. Indeed, as Flynn points out, since at all times P remained legally a man, technically the comparison offered by Tesauro, and later the Court, is between two men. L Flynn, ‘Annotation P v S and Cornwall County Council’ (1997) 34 cml Rev 367, 377.

162 Chapter 4 would be, he said, ‘quibbling.’189 The Advocate General called for a ‘broader perspective,’190 and warned the Court against making what he saw as a ‘moral condemnation’191 of transsexuals. Because there was ‘no precise provision’ on point, he explained in closing, what was needed was equality by inference.192 The Court answered the call made by its Advocate General, and, in a very short decision, similarly rejected the ‘equal misery’ argument. According to the Court, the Equal Treatment Directive193 included within its scope discrimination arising from gender reassignment. It continued: Such discrimination is based, essentially if not exclusively, on the sex of the person concerned. Where a person is dismissed on the ground that he or she intends to undergo, or has undergone, gender reassignment, he or she is treated unfavourably by comparison with persons of the sex to which he or she was deemed to belong before undergoing gender reassignment.194 S’s behaviour in dismissing P was therefore precluded by the Directive. Almost two years after the judgment in P v S came the case of Grant.195 Ms Grant was engaged as a clerical officer by (the company which became) South-​West Trains in 1993. Mr Potter, her predecessor, had obtained travel concessions for his ‘female cohabitee,’ pursuant to the employment contract in conjunction with the Staff Travel Facilities Privilege Ticket Regulations (‘the Regulations’). However, when Ms Grant applied for travel concessions for her female cohabitee, pursuant to the same two documents, in 1995, she was refused. The reason given was that, according to the Regulations, travel concessions were only granted to common law spouses (with whom a ‘meaningful relationship’ had been shared for two years or more) of the opposite sex, not of the same sex. Ms Grant brought an action for sex discrimination, and the Industrial Tribunal at Southampton made a reference. Advocate General Elmer, first holding that travel concessions were equivalent to pay, proceeded to consider the case under Article 119 ec196 alone. 189 190 191 192 193 194 195 196

Case C-​13/​94 P v S (n 182) Opinion of AG Tesauro, para 20. ibid, Opinion of AG Tesauro, para 23. ibid, Opinion of AG Tesauro, para 24. ibid, Opinion of AG Tesauro. Directive 76/207 (n 10). Case C-​13/​94 P v S (n 182) para 21 (emphasis added). Case C-​249/​96 Grant (n 2). Subsequently Article 141 ec, now Article 157 tfeu.

The ‘Article 19’ grounds

163

Considering the same quotation as that given above,197 he noted that the Court in P v S had found for P, not because she had been discriminated against on grounds of transsexuality (although she had been), but because she had been discriminated against on grounds of sex. He further noted that the Court had rejected the ‘equal misery’ argument (male transsexuals and female transsexuals treated alike), and had thus taken ‘a decisive step away from an interpretation of the principle of equal treatment based on the traditional comparison between a female and a male employee.’198 Posing the question whether gender was the causative factor (of the unfavourable treatment) in this case, the Advocate General observed that the Regulations made the travel concessions conditional on the cohabitee’s being of the ‘opposite sex’ to the employee. Thus, gender was in fact ‘the only decisive criterion.’199 South-​West Trains’ behaviour was therefore, in Elmer’s view, gender discrimination within the meaning of Article 119. Furthermore, as it was direct discrimination, it could not, in his view, be objectively justified. The Court, however, took a different view. The first question it asked itself was whether the Regulations constituted discrimination based directly on the sex of the worker. Ms Grant had been refused the travel concessions simply because she did not satisfy the conditions, conditions which applied regardless of the sex of the worker concerned. Using the ‘equal misery’ argument, the Court went on: Thus travel concessions are refused to a male worker if he is living with a person of the same sex, just as they are to a female worker if she is living with a person of the same sex.200 197 Text accompanying n 194. 198 Case C-​249/​96 Grant (n 2) Opinion of AG Elmer, para 15. 199 ibid, Opinion of AG Elmer, para 23 (emphasis added). However, as he goes on to discuss, that meant not just the gender of the employee, but also the gender of the cohabitee. He concludes, though, that use of an ‘abstract criterion’, like ‘opposite sex’, could ‘make no difference’ (para 25). It is interesting to note how useful the doctrine of associative discrimination would have been to Ms Grant here. But of course the Court did not recognize this doctrine until the case of Coleman v Attridge Law in 2008 (see below n 280, and accompanying text). Alternatively, one could hold, as Barnard does, that the complainant remains the only object of scrutiny, but that she is scrutinized with regard to two attributes, not just one: ‘gender plus sexuality.’ As she continues: ‘Yet, for a valid sex discrimination analysis, the comparison must change only the sex of the complaining individual and must hold all other circumstances constant’: C Barnard, ‘Some are more equal than others: the decision of the Court of Justice in Grant v South-​West Trains’ (1999) 1 C-​YELS 147, 153. 200 Case C-​249/​96 Grant (n 2) para 27.

164 Chapter 4 South-​West Trains’ behaviour could not therefore be regarded as discrimination on grounds of sex. The second question which the Court asked itself concerned the following three categories of relationship: 1 stable relationships between two persons of the same sex; 2 (a) stable relationships outside marriage between two persons of opposite sex; (b) marriage between two persons of opposite sex. The Court needed to ascertain whether Community law required that Category 1 should be regarded by all employers as equivalent to Category 2(a) or Category 2(b). The Court held that, although Member States were increasingly treating Category 1 as equivalent to Category 2(a) or Category 2(b), the Community had not yet adopted rules providing for such equivalence. Thus, in the Court’s opinion, the correct way to proceed was to continue to regard Category 1 as not equivalent to Categories 2(a) or (b), an approach which had been endorsed by the European Court of Human Rights (or at least its Commission).201 Thus, South-​West Trains was not required to treat the situation of Ms Grant and her partner as equivalent to that of Mr Potter and his partner. The final question which the Court asked itself was whether discrimination on grounds of sexual orientation constituted discrimination on grounds of sex. Having held that P v S was to be confined to its facts (that is, was relevant only to cases involving gender reassignment), the Court answered this question in the negative; discrimination based on sexual orientation was not covered by Article 119.202 2 01 ibid para 33, and the several case references therein. 202 As Bell has written, ‘Grant was crucial in creating a separation in Community law between “sex” and “sexual orientation” ’: Bell (n 183) 110. But not all writers agree with this ‘separation.’ Wintemute, for example, believes that discrimination on the grounds of sexual orientation is sex discrimination:  R Wintemute, ‘Recognising New Kinds of Direct Sex Discrimination:  Transsexualism, Sexual Orientation and Dress Codes’ (1997) 60 mlr 334, cited and discussed in Barnard (n 199) 153–​4. Barnard refers to Wintemute’s argument as ‘fancy footwork’ –​ibid 154 –​and then proceeds to offer her own version –​ibid 155–​8. Mulder comments that ‘sexuality and gender discrimination can hardly be considered separately, despite separate legal consideration’: J Mulder, ‘Some more equal than others? Matrimonial benefits and the cjeu’s case law on discrimination on the grounds of sexual orientation’ (2012) 19 mj 505, 514. The conflation of sex discrimination and sexual orientation discrimination is also the favoured approach of the Human Rights Committee when interpreting the iccpr: Toonen v Australia, Communication No 488/​ 1992, UN Doc CCPR/​C/​50/​D/​488/​1992 (1994).

The ‘Article 19’ grounds

165

Grant was of course decided before the Framework Employment Directive, so that Ms Grant’s main problem was to try to shoehorn a sexual orientation-​ discrimination point into the sex-​discrimination regime. With sexual orientation’s now having its own, dedicated prohibition, it is highly unlikely that she would have lost had the case been brought today. However, the age case of Mangold203 now poses an odd riddle. If all of the Article 19 grounds are (and therefore always were) general principles of Community law (as age was declared to be in that case), then why could the (supposed) general principle of non-​discrimination on grounds of sexual orientation not be invoked by the Court in Grant? Critics have made much of the inconsistencies between P v S and Grant, the most obvious difference being the Court’s rejection of the ‘equal misery’ comparison in the former, and acceptance of the same in the latter. Canor calls this ‘interpretative acrobatics.’204 She cites Flynn, who in 1997 (a year before Grant) wrote that the correct comparator for a female employee who is sexually attracted to or has sexual relations with women is a male employee who is sexually attracted to or has sexual relations with women.205 Had the Court in Grant taken this view, it would effectively have been comparing Ms Grant with Mr Potter, which was what she wanted all along. The comparison proposed by Flynn is also regarded as the correct one for Grant by Koppelmann206 (who makes a scathing attack on the ‘equal misery’ comparison actually chosen), and, it would seem, by Mancini and O’Leary.207 Others question the need for comparison at all, and wonder whether the establishment of detriment to Ms Grant could have been sufficient on its own, by analogy with Dekker,208 a pregnancy case.209 These include Bell,210 who furthermore regards P v S 203 Case C-​144/​04 Mangold (n 2). 204 I Canor, ‘Equality for Lesbians and Gay Men in the European Community Legal Order –​ “they shall be male and female”?’ (2000) 7 mj 273, 276. 205 Flynn (n 188) 382, cited in ibid 278. 206 A Koppelman, ‘The Miscegenation Analogy in Europe, or, Lisa Grant meets Adolf Hitler’ in R Wintemute and M Andenæs (eds), Legal recognition of same-​sex partnerships: A study of national, European and international law (Hart 2001). 207 GF Mancini and S O’Leary, ‘The New Frontiers of Sex Equality Law in the European Union’ (1999) 24(4) EL Rev 331, 350. 208 Case C-​177/​88 Elisabeth Johanna Pacifica Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-​Centrum) Plus [1990] ecr I-​3941. See supra, Chapter 3, text accompanying n 38, et seq. 209 For a nice discussion of the pros and cons of this approach, albeit written a year before Grant, see Flynn (n 188) 376ff. 210 M Bell, ‘Shifting Conceptions of Sexual Discrimination at the Court of Justice: from P v S to Grant v swt’ (1999) 5(1) Eur LJ 63, 67 and 74.

166 Chapter 4 and Grant as ‘impossible to reconcile,’211 and Carey, who also mentions the non-​comparative approach.212 Carey refers to the Court’s ‘illogical application of tests’ and ‘incoherence.’213 He puts the apparent u-​turn down to the relative sizes of the two communities at issue, with the Court perhaps being encouraged by the small number of transsexuals in the ec, only to then be discouraged by the large number of homosexuals (and the ‘potentially significant financial consequences’ of a positive ruling).214 But as he rightly states, ‘equality should apply, by definition, to everyone and not just a group occupying a given minimum percentage of the population’.215 As well as this economic motive for the Court’s behaviour,216 others have also detected a political, or even a moral, one.217 The Court had another opportunity to consider the equivalence or otherwise of same-​sex partnerships and marriage in the 2001 case of D and Kingdom of Sweden v Council,218 this time with the added complication that the Member State concerned (Sweden) permitted the legal registration of same-​sex partnerships, and that under Swedish law registered partners were treated as equivalent to married persons. D was a Swedish official of the Council, who was in a registered partnership with another Swedish national. D requested that the Council should treat his status as registered partner as equivalent to marriage in order for him to be entitled to the so-​called ‘household allowance.’ However, the Council refused on the grounds that the Staff Regulations did not allow a registered partnership to be treated as equivalent to marriage. On an application for annulment of this refusal, the Court of First Instance found for the Council. On appeal to the Court of Justice, D, supported by Sweden, claimed inter alia that the Council was bound by the provisions of Swedish law. However, Advocate General Mischo, referring to the famous free movement case of Reed,219 stated that any interpretation of a legal term on the basis of social 211 ibid 74. 212 N Carey, ‘From obloquy to equality: in the shadow of abnormal situations’ (2001) 20 yb of Eur L 79, 100. 213 ibid 99. 214 Quotation from Bell/​ ILGA-​Europe, Equality for lesbians and gay men: a relevant issue in the civil and social dialogue (ilga-​Europe 1998) 12. 215 Carey (n 212) 104. 216 Also mentioned by Barnard:  C Barnard, ‘The Principle of Equality in the Community Context:  P, Grant, Kalanke and Marschall:  Four Uneasy Bedfellows?’ (1998) 57 clj 352, 357. 217 Carey (n 212); Bell (n 210). 218 Joined cases C-​122/​99 P and C-​125/​99 P D (n 2). 219 Case 59/​85 State of the Netherlands v Ann Florence Reed [1986] ecr 1283.

The ‘Article 19’ grounds

167

developments must take into account the situation in the whole Community, not merely one Member State, and, at the material time, there existed in only three out of fifteen Member States the legal category of registered partnership, assimilated to marriage. Moreover, even Swedish law acknowledged a distinction between the two categories. The two categories did not have the same name, and there were a number of legal differences between them, for example pertaining to the adoption of children. In the Advocate General’s phrase, Sweden ‘did not wish to give two persons of the same sex unqualified access to the legal category of marriage.’220 Finally, since the Court in Grant had found no equivalence between a stable relationship between two persons of the same sex and a stable relationship between two persons of opposite sex, it followed that there was no equivalence, in Community law, between a registered partnership and a marriage.221 The Court agreed that the concept of a marriage and the concept of a registered partnership were distinct, and dismissed the appeal.222 D of course pre-​dates the Framework Employment Directive, and would undoubtedly be decided differently today. But the same riddle, deriving from Mangold,223 may be asked (mutatis mutandis) in relation to D, as was previously asked in relation to Grant.224 A counterpoint to D is the transsexuality case of kb.225 Here, a female employee of the uk’s National Health Service, kb, was concerned that her long-​ time partner, R, would not be provided for after her death. This was because R was a female-​to-​male transsexual, and the nhs’ pension scheme only provided for a survivor’s pension to be payable to a member’s surviving spouse, where spouse meant solely a person to whom the member had been married. 2 20 Joined cases C-​122/​99 P and C-​125/​99 P D (n 2) Opinion of AG Mischo, para 77. 221 As Carey points out, though, the reference to Grant here is ‘not wholly obvious:’ ‘Contrary to Grant who was legally single and without specific obligations towards her partner, D is, under Swedish law, obliged to fulfil duties towards his partner akin to those expected of one spouse towards the other.’ See Carey (n 212) 105. 222 Contrast the position in Canada, where the Supreme Court of Ontario held that Ontario must extend the definition of ‘spouse’ to include same-​sex partners: Case of M as discussed in Mancini and O’Leary (n 207) 347. And see now Maruko below (text accompanying n 234, et seq). For a very interesting discussion of this issue, especially in the cross-​border context, see K Boele-​Woelki, ‘The Legal Recognition of Same-​Sex Relationships within the European Union’ (2007–​8) 82 Tul L Rev 1949. 223 Case C-​144/​04 Mangold (n 2). 224 See supra, text accompanying n 203. 225 Case C-​117/​01 KB  v National Health Service Pensions Agency and Secretary of State for Health [2004] ecr I-​541.

168 Chapter 4 Even after his gender reassignment surgery, R was prevented from marrying kb by uk law (as it stood at the material time). During proceedings brought by kb, the Court was asked whether the exclusion of a transsexual partner from the scheme constituted sex discrimination. Advocate General Ruiz-​Jarabo Colomer, reiterating that a survivor’s pension constituted pay, recalled that in the 2002 ecthr case of Goodwin v uk,226 the European Court of Human Rights had held that the impossibility of a transsexual to marry (that is, to marry a person of the opposite sex to their post-​operative, or ‘acquired,’ sex) was a breach of the right to marry. Returning to kb’s case, the uk argued that a male unmarried person and a female unmarried person would fare exactly the same under the pension scheme in question, regardless of transsexuality or sexual orientation. However, the Advocate General took the view that the national rule which outlawed marriage between a transsexual individual, on the one hand, and a person of the opposite sex to their post-​operative sex, on the other hand (‘transsexual marriage’) was contrary to Community law, given that thirteen out of fifteen Member States did recognize such marriages, and especially given the judgment in Goodwin. Furthermore, this rule denied transsexuals access to a widow’s or widower’s pension. He concluded: The Court of Justice must ensure that the exercise of rights protected by the Treaty remains free of any prohibited discrimination and also that those rights are not made conditional on requirements which are contrary to European public policy.227 The impediment to marriage (and therefore the denial of access to the pension) was based on the gender reassignment of the person concerned. This was covered by Article 141 ec228 following P v S. Therefore, it followed that Article 141 ec precluded the national rule which outlawed transsexual marriage. The Court agreed wholeheartedly with the Advocate General: Legislation, such as that at issue in the main proceedings, which, in breach of the echr, prevents a couple such as kb and R from fulfilling the marriage requirement which must be met for one of them to be able to benefit from part of the pay of the other must be regarded as being, in principle, incompatible with the requirements of Article 141 ec.229 226 227 228 229

Christine Goodwin v the United Kingdom App no 28957/​95 (ECtHR, 11 July 2002). Case C-​117/​01 kb (n 225) Opinion of AG Ruiz-​Jarabo Colomer, para 75. Now Article 157 tfeu. Case C-​117/​01 kb (n 225) para 34.

The ‘Article 19’ grounds

169

Thus, it can be seen how Community law before Directive 2000/​78 was ­prepared to remove the national rule impeding kb, a transsexual, but was not prepared to remove the national rule impeding D, a homosexual.230 The case of Goodwin was also central to a later transsexuality decision: Richards.231 The uk, in response to the censure from the Strasbourg court, had adopted the Gender Recognition Act 2004, by which a post-​operative transsexual could have their ‘acquired’ gender officially recognized and taken into account in administrative and legal matters. Unfortunately it did not come into force until 2005, which was too late for Ms Richards. She was a post-​ operative male-​to-​female transsexual. Wishing to retire at the age of 60, she applied to the Department of Work and Pensions for a retirement pension. However, because her chromosomal sex was male, and because (pending the coming into force of the new act) there was no way for this to be changed, she was told that she could not have a retirement pension until the age of 65 (the pensionable age for men). She appealed. At second instance, a reference was made to the cjeu. Advocate General Jacobs noted that Ms Richards was denied her pension in circumstances where she would have been entitled to it had she been registered as a woman at birth, that is, had she been a person whose identity was not the result of gender reassignment surgery. This posed a slight problem for the choice of comparator, however. In P v S, which would seem to be the obvious precedent for a gender reassignment case, a post-​operative male-​to-​female transsexual was compared to a man whose identity was not the result of gender reassignment surgery.232 If the same comparison was run in Ms Richards’ case, Ms Richards would not be found to have been discriminated against. Unlike P, she did not want to be treated the same as a chromosomal man, because, if she were, she would indeed have to wait until the age of 65 for her pension! However, the Advocate General found greater inspiration from the case of kb. There, a post-​operative female-​to-​male transsexual was compared to a man whose identity was not the result of gender reassignment surgery. Applying this comparison, mutatis mutandis, to Richards’ case, she (a post-​operative male-​ to-​female transsexual) would also be compared to someone who possessed her 230 That is, the rule or framework of rules by which Sweden only allowed persons of the same sex to have registered partners, but denied them access to the legal category of marriage itself. See supra Mischo AG’s comment in text accompanying n 220. 231 Case C-​423/​04 Sarah Margaret Richards v Secretary of State for Work and Pensions [2006] ecr I-​3585. 232 To borrow a phrase from Skidmore, P was ‘compared … with her former self:’ P Skidmore, ‘Sex, Gender and Comparators in Employment Discrimination’ (1997) 26 Indus lj 51, 59.

170 Chapter 4 acquired gender, but who possessed it not as the result of gender reassignment surgery (that is, in Richards’ case, a woman whose chromosomal sex was also female). From such a comparison it would be plainly visible that, while Ms Richards received nothing at 60, the chromosomal woman received her full pension  –​clearly a case of like treated unlike, and therefore discrimination contrary to ec law. The Court used almost exactly the same reasoning to achieve exactly the same result. As a consequence of Ms Richards’ inability to have her ‘new’233 gender recognized, she had been the victim of unequal treatment amounting to discrimination. As in kb, a national rule which stood in the way of a Member State worker’s fulfilling the necessary requirements to attain a right protected by Community law, was incompatible with ec law and had to be removed. Returning to sexual orientation, a number of cases have been decided by the cjeu since the coming into force of the Framework Employment Directive, the first of these being Maruko.234 Mr Maruko, the surviving same-​sex partner of a costume designer, was denied a widower’s pension by the body which administered the ‘Pay scheme for Germany’s theatres’ (‘the scheme’), to which Mr Maruko’s late partner had been affiliated. During proceedings brought by Mr Maruko, a reference was made to Luxembourg. Advocate General Ruiz-​Jarabo Colomer first had to decide whether the scheme (and the pension which Mr Maruko was claiming under it) fell within the material scope of the Directive at all. For example, was the pension a matter of social security (in which case it fell outside the Directive’s scope by virtue of Article 3(3)), or did it instead constitute ‘pay’ (in which case it fell within the Directive’s scope by virtue of Article 3(1)(c))? The test for discerning whether a pension under a given scheme is social security or ‘pay’ is basically whether it is determined more by considerations of social policy (social security), or more by the employment relationship (‘pay’). Having considered a number of factors, the Advocate General’s conclusion was that this pension was derived from the employment relationship (of Mr Maruko’s partner), and therefore that it constituted ‘pay’ within the meaning of Article 141 and Directive 2000/​78.235 Turning to the question of marital status, Ruiz-​Jarabo Colomer first considered Recital 22 of the Directive, which reads: ‘This Directive is without prejudice to national laws on marital status and the benefits dependent thereon’. However, he concluded that Recitals have no binding force, and that, although it is true 2 33 The Court’s word. Case C-​423/​04 Richards (n 231) para 28. 234 Case C-​267/​06 Tadao Maruko v Versorgungsanstalt der deutschen Bühnen [2008] ecr I-​1757. 235 ibid, Opinion of AG Ruiz-​Jarabo Colomer, paras 65–​72.

The ‘Article 19’ grounds

171

that the Community has no powers with regard to marital status, Member States had to exercise their competence in a manner which did not infringe Community law. Having reviewed Grant, D and kb, the Advocate General held that the refusal to grant Mr Maruko a pension was not based directly on sexual orientation, but rather on the fact that he was not married to his partner. The refusal did however constitute indirect discrimination, because, under German law, only opposite-​sex partners could get married, not same-​sex partners. In a break with D, the Advocate General went on to hold that the legal situation of persons in a registered legal partnership (as Mr Maruko and his partner had been), and the legal situation of spouses, were compatible. Thus, like having been treated unlike, the refusal to grant the pension was (indirect) discrimination, on grounds of sexual orientation, contrary to Directive 2000/​ 78, and furthermore was incapable of objective justification.236 The Court concurred that the pension was ‘pay’ and therefore that it fell within the scope of the Directive. It also entirely agreed with the Advocate General’s dismissal of Recital 22. On the substance, the Court noted that Germany had created for persons of the same sex ‘a separate regime,’ the conditions of which had been gradually made ‘equivalent’ to marriage.237 Having provisionally established comparability,238 the Court had no difficulty in deciding that surviving same-​sex partners were being treated less favourably than surviving spouses. Thus, the German legislation concerned was discriminatory on grounds of sexual orientation.239 The Court had another opportunity to consider the matter of civil partnerships in Germany, this time in relation to supplementary pension payments, in the case of Römer.240 Here, Mr Römer, who had formed a civil partnership under German law with Mr U, sued Hamburg for paying him his supplementary retirement pension in accordance with a less favourable tax category than would have been used in the case of a married pensioner. This was pursuant to Paragraph 10(6) of the First Law on Retirement Pensions (‘First rgg’).241 2 36 ibid, Opinion of AG Ruiz-​Jarabo Colomer, paras 73–​104. 237 Quotations from ibid para 67. 238 Although note how the cjeu leaves the final word on this to the national court: ibid para 72. See also the last sentence of para 73. 239 The Court makes a slight change to Ruiz-​Jarabo Colomer AG’s Opinion here. In its view, the discrimination was direct, not indirect. Tobler and Waaldijk make some interesting suggestions as to why this might have been at C Tobler and K Waaldijk, ‘Annotation Maruko’ (2009) 46 CML Rev 723. 240 Judgment of 10 May 2011, Jürgen Römer v Freie und Hansestadt Hamburg, Case C-​147/​08, ECLI:EU:C:2011:286. 241 Erstes Ruhegeldgesetz der Freien und Hansestadt Hamburg.

172 Chapter 4 On reference from the first instance court, the Court of Justice was asked, as in Maruko, to first of all discern whether supplementary pension payments constituted social security or ‘pay’ in order to determine whether they fell outside or within Directive 2000/​78, and, assuming within, whether the First rgg was a ‘law on marital status’ and so outwith the Directive’s scope courtesy of Recital 22. Should these hurdles be overcome, the Court was then asked whether Paragraph 10(6) represented direct or indirect discrimination under the Directive. The Court was happy to dismiss the national court’s concerns about the scope of the Directive rationae materiae.242 It went on to consider discrimination, starting, in the Aristotelian tradition, with comparability. In this case, as in Maruko, the comparability paradigm was more akin to that in D than that in Grant, meaning that pensioners of the same sex who had entered into a civil partnership were being compared to pensioners of the opposite sex who had entered into marriage, the latter institution not being available to same sex partners, and, exactly as in Maruko, the Court was convinced of the gradual equivalence, and hence comparability, of the regime of civil partnerships with the regime of marriage in Germany.243 With the usual deference to the national court’s right to make the final decision, the Court went on to assert that there had been a difference in treatment, as the contributions made by married pensioners and those made by pensioners in civil partnerships were identical, although the resultant pension payments were (once taxation was factored in) different: likes had been treated unlike. Since this discrimination was, in the Court’s view, direct, the question of justification did not arise. Despite Mr Römer’s victory, though, the judgment did end on a rather negative note, as it was ruled that he would only be able to claim compensatory back payments from the date of transposition of the Directive (in 2003) and not from the date of his entering the civil partnership with Mr U (in 2001), the Court being apparently unwilling to extend its Mangold doctrine to cover the two years in between. One 242 The supplementary retirement pension at issue did constitute ‘pay’: Case C-​147/​08 Römer (n 240) para 33. In Dittrich, Klinke and Müller, another sexual orientation case, the Court was again called upon by a German court to ascertain whether a benefit, this time the Beihilfe, or assistance to public servants in case of illness, was or was not ‘pay’, and again it was. The national court did however seem to have taken Maruko and Römer to heart in respect of comparability, and so that matter did not form part of the reference: Judgment of 6 December 2012, Bundesrepublik Deutschland v Karen Dittrich and Robert Klinke and Jörg-​Detlef Müller v Bundesrepublik Deutschland, Joined Cases C-​124/​11, C-​125/​11 and C-​143/​11, ECLI:EU:C:2012:771. 243 Case C-​147/​08 Römer (n 240) paras 44–​48 inclusive.

The ‘Article 19’ grounds

173

reason for this was, said the Court, that Paragraph 10(6) of the First rgg was not ‘a measure implementing Directive 2000/​78.’244 Finally, one might mention briefly one other post-​Framework Directive case on the subject of sexual orientation. In Asociaţia accept,245 which concerned the Romanian football club FC Steaua, a Mr Becali, variously described as the club’s ‘banker’ and ‘patron,’ made some public remarks to the effect that he would rather hire a junior player than a homosexual, whereupon an action was brought by accept, a non-​governmental organisation set up to protect lgbt rights, claiming that these remarks amounted to direct discrimination in matters of recruitment. The Consiliul Naţional pentru Combaterea Discriminării (National Council for Combatting Discrimination) delivered its judgment in October 2010, stating firstly that these remarks were not made in the context of an employment relationship, secondly that Mr Becali was not an employer, and thirdly that, although it found that the remarks were discrimination in the form of harassment, it would only issue a warning as the six month limitation period for the imposition of a fine (under national law) had elapsed. accept then challenged this decision before the national court, which referred four questions to the Court of Justice. The Court held that, firstly, in order to establish ‘facts from which it may be presumed that there has been direct or indirect discrimination’ under Article 10 of the Directive, the maker of a statement does not need to have legal capacity to define recruitment policy for, or to bind, the entity on whose behalf the statement is made; the Court noted that FC Steaua had not distanced itself from Mr Becali’s comments. Secondly, the fact that no actual negotiations were underway for the hiring of any particular footballer did not make a difference. Thirdly, and contrary to what the national court feared, it would not breach any player’s right to privacy if FC Steaua were required to demonstrate that there had been no breach of the 244 ibid para 63. This somewhat presupposed that the (amended) law on fixed term contracts in Mangold was such a measure. On the substance, Advocate General Jääskinen had reached the same conclusions as the Court did (allowing also for an alternative finding that the Hamburg law was unjustified indirect discrimination), but unlike the Court the Advocate General had been prepared to entertain the possibility (admittedly among others) that Mr Römer had a redressable wrong dating back to the start of his civil partnership, not just to the date of transposition of the Directive: see Opinion of 15 July 2010, Jürgen Römer v Freie und Hansestadt Hamburg, Case C-​147/​08, ECLI:EU:C:2010:425, paras 145–​148 inclusive. 245 Judgment of 25 April 2013, Asociaţia accept v Consiliul Naţional pentru Combaterea Discriminării, Case C-​81/​12, ECLI:EU:C:2013:275.

174 Chapter 4 principle of equal treatment;246 the sexual orientation of existing players did not need to be shown, merely (for example) the existence of express provisions of the club’s recruitment policy demonstrating compliance with the Directive. Finally, with regard to the rule in Article 17 of the Directive that sanctions for breach must be effective, proportionate and dissuasive, and bearing in mind that it was for the national court to decide the case, the cjeu expressed the view that a warning might not be commensurate with the seriousness of the matter under discussion. 4.5.2 A Walzerian Analysis (a) Sexual Orientation, Quintessential Boundary Conflict: Grant The ground of sexual orientation really comes down to the debate between those who think that marriage is about the gender of the parties, and those who think that it is more about free choice. Ball sees these two interpretations as ‘mutually exclusive,’ and calls the situation a ‘log jam.’247 The issue of same-​ sex unions thus provides a classic example of a dispute, within a community, as to the meaning of a distributed good (a marriage certificate). And it could be argued that, whatever meaning is decided upon, and whatever law is passed, any law on this subject would represent a boundary breach: the sphere of power invading the sphere of kinship, which Walzer calls ‘the deepest understanding of tyranny.’248 Attempting to present the shared understanding of sovereignty in the form of a list, he notes at point 2: ‘[The state’s] officials cannot control the marriages of their subjects’.249 Unsurprisingly, then, complex equality has been largely embraced by those seeking an end to discrimination on grounds of sexual orientation. Indeed, Ball rightly notes another, crucial, boundary breach: Even if one were to conclude that gay men and lesbians are not entitled to marry because their relationships do not fit within the definition of marriage as determined by our shared traditions, Walzer’s theory of justice would prohibit that view from spilling over into other spheres such as that of employment.250 246 Under the Directive, where the facts referred to in Article 10 have been shown by a complainant, the burden of proof is reversed and it is then for the defendant to prove that there has not been direct or indirect discrimination: Directive 2000/​78 (n 1) Article 17. 247 Ball (n 75) 502 and 504. 248 Walzer, Spheres of Justice (n 45) 228. Referring to the us of 1983, he notes that homosexual marriage is ‘legally unrecognized and politically controversial:’ ibid. 249 ibid 283. 250 Ball (n 75) 505.

The ‘Article 19’ grounds

175

Most of the cases discussed at section 4.5.1 directly concern the employment sphere (and one thinks particularly of Grant and D), and the others concern it indirectly in any event. Why is society’s view of gay marriage being allowed to invade this sphere? For example, Grant seems to entail the ultimate violation of a shared meaning. It is hard to conceive of any community wherein its understanding of train tickets would include the sexual orientation of the potential passenger, and that of the potential passenger’s partner. It seems appropriate, then, to turn to Grant first. In section 3.3 above, it was explained how, if Walzer’s theory was not read literally, a free choice or even a characteristic could be considered as a distribuend in a fictional ‘first’ sphere. It is submitted that this is how Ms Grant’s homosexuality must be viewed.251 However, for the purposes of this case, Ms Grant’s homosexuality is not just a good; it is a negatively dominant good. When Ms Grant arrived in the sphere of office in possession of this negatively dominant good, she was excluded from the distribution and not allocated any travel concessions.252 The dominant good, being irrelevant to the shared meaning of the travel concessions (see above),253 was present in the sphere of office tyrannically, rendering the distribution flawed. Either the European Court or the national court would then likely order the rerunning of the distribution, with Ms Grant’s homosexuality taken out of the picture (left, as it were, ‘at the door’ of the sphere). It can thus be seen how a Walzerian analysis of the Grant case produces a much better result for Ms Grant, and avoids any need for tortuous comparisons between different types of couple, and different types of partner. (b) Negative Dominance, Non-​Stop Debate: P v S, kb, Richards As mentioned in section 2.3, the concept of negative dominance is especially important in the case of a suspect ground. This is because, although a litigant could be complaining of a boundary breach by the holders of positively dominant goods (for example, ‘the able-​bodied’ in disability discrimination 251 It does not matter for the purposes of this book whether Ms Grant’s homosexuality was chosen by her, or whether it was an innate characteristic. The book does not take a position on the question. 252 Please note that, just as a positively dominant good can procure for its owner both good things and the absence of bad things (hard work, for example), so a negatively dominant good can procure for its owner both bad things and the absence of good things (travel concessions, for example). 253 One is reluctant to assume a shared meaning, but, as mentioned earlier, it is difficult to imagine that any community would include sexual orientation in the definition of travel concessions. It would be open to counsel for South West Trains to argue the contrary, but, for the purposes of this section, it is presumed that any such argument would fail.

176 Chapter 4 cases), it is more common that the complainant sees their own minority (or under-represented) status as a negatively dominant good. Good examples from section 4.5.1 include P v S and kb. Assuming that the distribuend is ‘continued employment,’ it seems foolish to refer to the dominant in P v S as ‘not being transsexual;’254 it is surely more truthful and realistic to refer to the negative dominant (possessed by P) of ‘being transsexual.’ Likewise, in kb, where the distribuend is a widow(er)’s pension, it makes more sense to say that the negative dominant is, again, ‘being transsexual’ (although in this case referring to the claimant’s partner, rather than the claimant themselves). These two cases were resolved happily by the cjeu, and a Walzerian approach would produce the same result, the Court being highly unlikely to be convinced that transsexuality forms any part at all of the shared meanings of employment or pensions. The situation becomes slightly more rarified when one turns to Richards. Here, it was not transsexuality alone which was costing Ms Richards her pension at 60; it was the fact of being a woman whose identity was the result of gender reassignment surgery (as opposed to a woman whose identity was not the result of gender reassignment surgery). But if this is made the (negative) dominant, then the distribution can easily be declared flawed, as long as it is held that (in the eyes of the community in question)255 it makes no difference, for the receipt of a pension at 60, whether the pensioner is chromosomally female or female as a result of surgery. Richards raises another interesting point. It is highly unlikely that the uk legislature intended to prejudice transsexuals when it declared that women should receive their pensions at 60. It is just that gender reassignment surgery is a newer phenomenon. But this merely underlines the fact that the debate (at which shared meanings are determined) must be ongoing,256 and must be accessible by all, including those with ‘external’257 ideas or ideas ‘from outside the … mainstream.’258 As Rustin points out, it is only in fundamentalist societies ‘that arguments are only admissible in debate if they are already elements of accepted doctrine.’259 If a contingency that was simply not thought of, or else deemed too remote, at the time of the setting of the original distributive criteria later materializes, or if a circumstance that was not anticipated later occurs, then there must be 2 54 And see below, section (c). 255 Or the Bench, if the alternative theory of mediated complexity is preferred; see below Chapter 8. 256 See supra, section 2.4 and in particular the quotations given at the end of that section. 257 M Rustin, ‘Equality in Post-​Modern Times’ in Miller and Walzer (n 165) 36. 258 ibid. 259 ibid.

The ‘Article 19’ grounds

177

a facility for reopening the debate and modifying the criteria as appropriate. This may indeed mean ‘carry[ing] out new surveys every time somebody [comes] up with an objection that might affect […] attitudes in the population.’260 (c) The Tyranny of ‘Normalcy’ In P v S, kb and Richards, then, it is seen how in claims involving suspect grounds, it is often preferable to regard the case as one of negative dominance. The claimant has taken their characteristic or choice from the sphere to which it was germane to one in which it was, quite literally, not welcome. It is this action which flaws the distribution, and causes it to have to be rerun. The alternative is less palatable and could be called: The Tyranny of ‘Normalcy.’ There is little doubt that a teacher who was not transsexual would have been kept on by S and Cornwall County Council. Likewise, kb would have had no difficulty in acquiring a widower’s pension for a partner who was not transsexual. And it goes without saying that a British woman whose identity was not the result of gender reassignment surgery would receive her pension at 60. Those without minority or under-represented status, what one might call the ‘normal,’ appear to be able to take their ‘normalcy’ from sphere to sphere, gathering all the available goods as they go. ‘Normalcy’ here is used somewhat ironically and within quotation marks. It means being standard or being average, where ‘standard’ and ‘average’ are words used to describe someone all of whose characteristics and choices are the characteristics and choices generally possessed or made by the majority in the society concerned. The Norm could be seen as setting itself, shifting in accordance with the shifts in the majority’s traits and preferences. Alternatively, as Foucault suggested with his ‘bio-​power’ thesis, the State (borrowing tactics of regulation from what he called ‘the Disciplines’) may be exerting disciplinary, but also normalizing, power on the members of society, encouraging them to classify and sort themselves, so as to move ever closer to the Norm. In such a programme of mass homogenization, (State-​sanctioned) ‘normalcy’ would take on a much more significant role; it would be very much the goal of the members of the community, with the law complicit as a mere legitimation of the other powers.261 2 60 J Elster, ‘The Empirical Study of Justice’ in Miller and Walzer (n 165) 92. 261 ‘The norm has a very different epistemological structure [to law]:  it is supposed to be experimental; it keeps to the facts whose differences, divergencies and hierarchies it privileges. It multiplies inequalities and brings them out wherever it takes hold: inequalities in face of death, in face of sickness and knowledge etc. It … ‘unequalizes’, because it is set up on the basis of its capacity to locate the most minimal disparities’: F Ewald, ‘Justice, Equality, Judgement: On “Social Justice” ’ in G Teubner (ed), Juridification of Social Spheres (de Gruyter 1987) 107.

178 Chapter 4 But whether self-​created or imposed (or perhaps a bit of both), ‘normalcy’ cannot be ruled out as a dominant good. It could even be regarded as having its own sphere. However, it is a sprawling sphere, very much ‘out of control’ and in need of robust boundary revision. ‘Normalcy’ is constantly looking to colonize the spheres around it, cajoling and bullying, tempting other spheres to give up their autonomy and join it,262 or else simply grabbing them outright. The ‘normal’ also very much favour the Aristotelian test for equality, since they are the Guardians of ‘Likeness:’ they decide what (or who) is like, but also, more importantly, what (or who) is unlike.263 However, ‘normalcy’ moves about insidiously and has a chameleon-​like nature; it can, by definition, fit in anywhere, so that its secret dominance goes practically unnoticed. In any given distribution, it would be difficult to define exactly what it was at that precise moment, and therefore to know which, if any, distributee possessed it. Likewise it would be difficult to know which characteristics he or she would have to leave ‘at the door’ in order to make the distribution more just. Thus the use of negative dominance would still seem to be the best approach (from the Walzerian point of view) for cases involving minority statuses or under-representation. There is no knock-​out blow, then, or p ­ erhaps that should be iron boundary, to defeat the tyranny of a ubiquitous but ever-​mutable ‘normalcy.’ Instead it must be fought indirectly, case by case. Nevertheless, the Tyranny of ‘Normalcy’ remains an interesting topic. It would be naïve to deny that county councils, national health services and secretaries of state (in other words, distributors) all have some kind of ‘Norm’ in their minds when they distribute.264 And Foucault may well be right that normalizing powers 262 Foucault’s ‘bio-​power’ operates (reflexively) in the same way. No one wants to be the one who does not fit in, and this much-​feared consequence of resistance pushes everyone on to conform even more. 263 ‘[A]‌society which perceives heterosexuality as the commonly prevailing normal sexual behaviour […] which others should be compared to and should aspire to assimilate into, might sanction whoever fails to do so by approving of discrimination against that individual’: Canor (n 204) 277. ‘[D]oes the equal misery approach ensure that those outside the club of normality are to remain there but, while outside, are to be treated equally inter se? Following this approach equality fast becomes a means of protecting normality. Normality, in turn, becomes a vicious circle wherein only those who define its parameters are included [telling those who are excluded to …] fit in as we do or else do not bother complaining that it is cold outside’: Carey (n 212) 103 (Carey’s italics). 264 That is why it is essential that all members of the community, distributors and distributees, are present at the debate at which the distributive criteria are ascertained.

The ‘Article 19’ grounds

179

(and therefore the Norm itself) are ‘institutionalized… by legal instruments’265 (the Births and Deaths Registration Act 1953 in Richards, for example). To take one of the other Article 19 grounds, disability, it is often the case that Health and Safety laws hamper the progress of the disabled, thus reinforcing the dominance of ‘Normalcy.’ The Tyranny of ‘Normalcy’ will be returned to later. (d) Distinctly Identical: D versus Maruko and Römer Finally, D, Maruko and Römer provide a neat example of the absurdity of the Aristotelian approach to equality cases, and the relative wisdom of Walzer’s. In D, a registered (same-​sex) partnership was treated as ‘unlike’ a marriage, but seven years later in Maruko, a registered (same-​sex) partnership was treated, at least pending the national court’s confirmation, as ‘like’ a marriage, with the same equivalence being found, three years after that, in Römer.266 If a 265 A Barron, ‘Foucault and Law’ in J Penner, D Schiff and R Nobles (eds), Introduction to Legal Theory and Jurisprudence (Butterworths 2002) 994. 266 One might commend the Court’s newfound contextual approach in both Maruko and Römer. In Römer, for example, Advocate General Jääskinen talked of not taking into account matters which are ‘irrelevant to the specific situation’ (Opinion in Case C-​147/​08 Römer (n 244)  para 93)  and the Court itself talked of carrying out its comparability analysis in ‘a specific and concrete manner in the light of the benefit concerned,’ not in ‘a global and abstract manner’ (n 240, para 42). In both of these examples, the Court’s approach in the earlier Maruko was cited as the template. However, this is not the same as, and should not be confused with, the contextual approach being put forward in this book, as the questions being asked by the Court are not the same. ‘Are the married couple and the couple having registered a civil partnership comparable as regards survivor’s/​retirement pensions?’ is the question being asked by the cjeu in Maruko and Römer, and it is still not quite the right question. ‘Is having married or having registered a civil partnership part of the shared meaning of a survivor’s/​retirement pension?’ is the right question. It might be argued that they come to almost the same thing. Indeed, in Maruko/​ Römer, the two questions seem to reach the same result (that Mr Maruko and Mr Römer deserve equal treatment with men having married), just via a different route –​the pension context discounting the relevance or importance of marital/​civil status in the first, and marital/​civil status being discounted from the sphere of pensions in the second. From this it might be concluded that therefore the Court’s current, more nuanced approach does well enough. However, it is submitted that this would be a dangerous conclusion, because, although the two questions led to a similar result here, in another case they might not. Indeed in another case, the current or ongoing difference in distribution might by itself lead to a negative answer to the first question (that is, a finding of non-​comparability), and in turn, discarding rare recourse to indirect discrimination, a continuation in unequal treatment for the two men; after all, unlikes being treated unlike spells no breach of Aristotle’s formula. It would be in

180 Chapter 4 traditional comparative approach is insisted upon, it is as a minimum to be hoped that the same comparison could be run in the same way twice, ceteris paribus, so that the Aristotelian rule could at least be seen to obey itself. D and Maruko/​Römer provide proof positive, if proof were needed, that the outcome of equality cases is as likely to be influenced by politics and opportunism as it is by logic or law. The fact that the Framework Employment Directive was not implemented in 2001, and that D could thus only rely on the general principle of equal treatment, and that therefore it fell to the Court alone, without Member State backing, to declare the identity of same-​sex partnerships and marriages, undoubtedly explains the Council’s victory in D: the political stakes were simply too high for the judges at the cjeu, and their nerve failed.267 However, the fact that the Framework Employment Directive was implemented in 2008, and that Mr Maruko could rely on it without recourse to the general principle, and that the Court could thus find for him while leaving the legislature to bear any responsibility, similarly explains the claimant’s victory in Maruko, and the victory of a subsequent litigant like Mr Römer. It is not the political u-​turn in this story which is sad; political u-​turns happen every day, and the second result is indisputably better than the first. What is sad is the way the Aristotelian test can be so obviously moulded to produce

such a case that the second question would show its superiority, as its answer would not change. Aristotle’s rubber stamping or inequality-​reinforcement tendencies are stamped out because the good will always expel the dominant, while the context may only sometimes expel the (relevance of the) status, other times confirming it. The judge’s decision to be a slave to or break with the past, to reify or rewrite the context, to maintain or change, would seem to be entirely a matter for his or her whim that day, and unlike a refusal to change the meaning of a good, made following a thorough public airing of the evidence of said meaning, a refusal to change a context draws far less attention to itself and is indeed hidden from outside scrutiny. The prejudiced judge not only does not need smoke and mirrors to distract attention from a telltale mechanism, he or she can exploit the mechanism to tell exactly the tale he or she wants, the crime its own watertight alibi. 2 67 Mulder, who considers whether D might have been a matter of civil status discrimination, not sexuality discrimination, would disagree that the Court was ever that keen to begin with. She states that in D, the Court ‘chooses to uphold and strengthen the normative effect of marriage’ as part of a wider campaign of ‘imposing heteronormativity.’ She holds that even the later cases, like Maruko and Römer, only came about after same sex relationships had been, in her word, ‘normalise[d]‌;’ she urges the Court to take a ‘less assimilationist’ approach. Interestingly, she refers to the challenge to the courts ‘to fundamentally reconsider their understanding of gender, sex and sexuality:’ Mulder (n 202) 507, 514, 516, 518, 521 (emphasis added).

The ‘Article 19’ grounds

181

the needed result. Either two things are alike or they are not; the mere happenstance of which remedy may or may not be available on the day of the comparison cannot possibly affect this.268 Looking at the three cases from the Walzerian point of view, the three distribuends at issue are obviously a household allowance (in D), a survivor’s pension (in Maruko) and a retirement pension (in Römer). In all three cases the dominant would appear to be the same: the status of being married, that is, that the tie between any claimant and their partner (or late partner) was one of marriage. Thus there was a boundary breach in all three cases, unless counsel for the Council, or for the German Theatre Pension Institution, or for Hamburg, could convince the Court that the status of being married (in the sense of a union between two heterosexuals) was part of the shared understandings of household allowances or survivor’s/​retirement pensions in the communities concerned: Germany in Maruko/​ Römer, and the whole of the eu (or perhaps just the workforce of the Council) in D. Debates held in these two communities would reveal the shared understandings. Alternatively,269 a summary of the relevant debate could easily be presented by counsel, calling witnesses if necessary, always with the possibility for opposing counsel to cross-​examine. Opposing counsel could also make their own presentation and call their own witnesses, and it would fall to the Bench to make the final decision. In either alternative, all that would then be left to do would be to declare whether the distribution under consideration was flawed or not. The results in the three cases cannot be predicted. Nor can it be predicted whether the results in Maruko, Römer or D would be the same or different as between themselves. If, as the Court hinted, there was indeed some kind of change of view in Europe, with regard to gay marriage, between 2001 and 2008, this would be reflected in the result. If not, this would similarly be reflected in the result. D’s own community, and Mr Maruko’s and Mr Römer’s own community, would thus bear the true responsibility for deciding whether, in their 268 Opponents of judicial activism might argue that the Court was right to wait for the Framework Employment Directive to be implemented before pronouncing on gay marriages, and that therefore there was a substantive difference between 2001 and 2008: agreement between the Member States as to the undesirability of discrimination on the grounds of sexual orientation. But this would be a nonsense. The Framework Employment Directive still existed in 2001 (even if not on the date of D’s application). Any ‘agreement’ relied upon in 2008 could have been just as easily relied upon (if tacitly) in 2001. And what exactly is the point of the general principle if not to allow the Court to correct any inequalities which it may encounter in the course of hearing a case? 269 According to the theory of mediated complexity presented in Chapter 8 below.

182 Chapter 4 respective cases, a registered partnership was or was not the same as a (heterosexual) marriage.270 The Court would not have to fear making an unpopular decision, because the decision would, ultimately, be that of the populace itself. 4.6 Disability 4.6.1 The cjeu’s Case-​Law Disability is perhaps the most complex of the Article 19 grounds, not least because of the many definitional complications which the word ‘disability’ presents. On one view, disability discrimination is entirely ‘a function of the social reception’ of the disabled person, and nothing whatsoever to do with ‘the functional state of [his or her] body or mind.’271 In other words, it is an ascribed difference (to use the first of Schiek’s three categories),272 based only on the reactions and opinions of others. The person to whom the ascription is applied may not consider themselves ‘disabled’, or different, at all. In a study into happiness, published in The Lancet, children with cerebral palsy were found to be as happy as those not suffering from this condition. Professor Allan Colver, who led the research, commented: for the person with cerebral palsy –​that’s who they are, and as they grow up and develop their sense of self, that disability is indistinguishable from their identity as human beings.273 Bell has summed up the problem which this poses in the legal context: The meaning of ‘disability’ is contested and gaps emerge between self-​ perception and external categorizations. The definition of disability for 270 And of course it would be open to all counsel to make provision, in their pleadings, for an adverse shared meaning (pure Walzerian complexity) or for the possibility of their side losing on the question of the shared meaning (mediated complexity); they could request that, in that eventuality, the Bench apply the ‘override’ (discussed above, for example, at section 2.6) in order to mitigate any resultant injustice. The Court would no doubt want this request, like any other request, to be fully reasoned. And it would almost certainly demand that counsel produce very strong and compelling grounds, before it would consider overriding the duly ascertained wishes of the distributive community. 271 Quotations from JE Bickenbach, ‘Disability and Equality’ (2003) 2 jl & Equal 7, 8. 272 See Schiek (n 16). 273 D Lawson, ‘From Pentecost island to modern Britain, the futility of trying to measure happiness’ The Independent (London 6 July 2007) 33.

The ‘Article 19’ grounds

183

the purposes of discrimination law may not correspond to an individual’s assessment of whether they have a disability.274 The Framework Employment Directive has tried to skirt round this problem by eschewing a definition of ‘disability’ altogether. However, while the absence of a definition should have permitted judges to adopt the ‘greatest range’ of possible meanings when construing the word,275 in fact the cjeu opted for a very narrow understanding of ‘disability’ in the first preliminary reference on the subject. In Chacón Navas,276 the issue was whether mere ‘sickness’ was covered by the prohibition of discrimination on grounds of disability in Article 1 of the Directive. The Court answered this in the ­negative. Quinn has described the decision as ‘disappointing to say the least’277 and ‘an aberration.’278 Meenan has commented that ‘[i]‌t would be a pity if Chacón Navas was to remain the final word … on the issue of definition’.279 The second preliminary reference tackled by the Court on disability was the case of Coleman v Attridge Law.280 This again concerned the scope of the Directive, this time whether it covered what is sometimes called ‘associative discrimination,’ also known as ‘transferred discrimination.’281 This is the phenomenon whereby the complainant has suffered discrimination not as a result of their own disability, but as a result of a third party’s. Ms Coleman was working for a firm of solicitors when, in 2002, she gave birth to a son who was disabled. She alleged that, inter alia, when she sought to take time off to care for her son she was called ‘lazy,’ and she was also denied the same flexibility as regards her working arrangement as was granted to her colleagues with non-​ disabled children. She sued for constructive dismissal, and the South London Employment Tribunal made a reference. Advocate General Poiares Maduro was clear that the Directive could indeed be applied to discrimination by association:

2 74 Bell (n 4) 197. 275 The quotation is from DL Hosking, ‘Great Expectations: Protection from discrimination because of disability in Community law’ (2006) 31(5) el Rev 667, 681. 276 Case C-​13/​05 Sonia Chacón Navas v Eurest Colectividades sa [2006] ecr I-​6467. 277 Quinn (n 165) 274. 278 ibid 277. 279 Meenan (n 13) 348. 280 Case C-​303/​06 S. Coleman v Attridge Law and Steve Law [2008] ecr I-​5603 281 See for example M Pilgerstorfer and S Forshaw, ‘Transferred discrimination in European law’ (2008) 37(4) Indus lj 384.

184 Chapter 4 [T]‌he Directive does not allow the hostility an employer may have against people belonging to the enumerated suspect classifications to function as the basis for any kind of less favourable treatment in the context of employment and occupation … The Directive does not come into play only when the claimant is disabled herself but every time there is an instance of less favourable treatment because of disability.282 The Court followed this argument: Although, in a situation such as that in the present case, the person who is subject to direct discrimination on grounds of disability is not herself disabled, the fact remains that it is the disability which […] is the ground for the less favourable treatment […]. […] Directive 2000/​78, which seeks to combat all forms of discrimination on grounds of disability in the field of employment and occupation, applies not to a particular category of person but by reference to the grounds mentioned in Article 1.283 If the Directive applied only to those having the characteristic classified as suspect, and not to those associated with them, then, according to the Court, it would be ‘deprive[d]‌… of an important element of its effectiveness.’284 Unlike in Chacón Navas, then, the Court in Coleman opted for a wide interpretation of the scope of the Directive (albeit the scope rationae personae, not rationae materiae).285 The next case to come before the Court concerning disability was Odar.286 Here, Mr Odar’s employer, Baxter Deutschland GmbH, offered its workers who were made redundant on operational grounds certain social plan benefits. In calculating these, for those under 54, their length of service was taken into account (‘the standard method’). For those over 54, however, length of service was not taken into account, and the calculation was instead made on the basis of the earliest date on which their pension would begin (‘the special formula’). Mr Odar, aged 59, claimed that this rule prejudiced him on the grounds of his 282

Case C-​303/​06 Coleman (n 280) Opinion of AG Poiares Maduro, paras 22–​23 (emphasis added). 283 ibid para 50. 284 ibid para 51. 285 For an interesting note on the case, see L Waddington, ‘Annotation Coleman’ (2009) 46 cml Rev 665. Joly has called it ‘pioneering’: Joly (n 181). 2 86 Judgment of 6 December 2012, Johann Odar v Baxter Deutschland GmbH, Case C-​152/​11, ECLI:EU:C:2012:772.

The ‘Article 19’ grounds

185

disabled status, which brought with it entitlement to an early pension and, consequently, less entitlement to the benefits offered by Baxter. The German court made a reference, asking for the cjeu’s opinion as to whether the rule in question and/​or the method of calculation derived therefrom breached Directive 2000/​78, either on grounds of age or on grounds of disability. The Court of Justice was unconcerned with the rule itself, but it did wish to scrutinize the method of calculation. It found no breach of the prohibition of discrimination on grounds of age, because the difference in treatment as between those under 54 and those over 54 was justifiable in that it pursued the objectives of: granting compensation for the future, protecting younger workers and facilitating their reintegration into employment, whilst taking account of the need to achieve a fair distribution of limited financial resources in a social plan.287 It was also proportionate, as (inter alia) there was a safeguard ensuring that even those who would receive nil under the special formula would still receive 50% of the compensation which would have been due to them under the standard method. However, turning to the prohibition of discrimination on grounds of disability, the Court was prepared to find a breach here, as this time the difference in treatment was as between disabled workers and non-​disabled workers in the same age bracket who were approaching retirement, with the disabled workers’ early entitlement to their pension causing them to receive less under the special formula where pensionable age was a decisive criterion. Although the method of calculation, with its inherent differences in treatment, was still objectively justifiable for the reasons stated above, this time it failed the test of proportionality, as the severely disabled ‘face greater difficulties in finding new employment’288 and have ‘financial requirements arising from their disability which cannot be adjusted.’289 Thus, according to the Court, the 287 ibid para 42. 288 ibid para 69. 289 ibid. It is worth noting how Germany’s argument in relation to proportionality was that ‘the lower amount of compensation on termination received by Dr Odar is justified in the case of severely disabled workers by the advantage they have consisting in entitlement to claim a retirement pension as from three years earlier than non-​disabled workers’ (ibid para 66), but that this argument was rejected by the Court because ‘there is discrimination based on disability when the disputed measure is not justified by objective factors unrelated to such discrimination’ (ibid para 67, emphasis added). This is, firstly, because

186 Chapter 4 difference in treatment as between those eligible for their pensions at 60 (the disabled) and those eligible for their pensions at 63 (the non-​disabled) went beyond what was necessary to achieve the social policy objectives pursued by Germany.290 This does however leave the reader with the question as to how this situation was to be made good. Were disabled workers to be treated as though they would not receive their pensions for another three years, or should the standard method be applied to them as it was for those workers under 54? The Court returned to the Directive’s scope rationae materiae in hk Danmark,291 where employees of two different Danish firms were fired after they had taken sick leave, in the one case, owing to lumbar pain and, in the other case, following a car accident. In both cases, their employers made use of the Danish Law on the Legal Relationship between Employers and Salaried Employees (‘fl’),292 Paragraph 5(2) of which provided for a shortened notice period (one month) once an employee had been on paid sick leave for 120 days over a period of twelve consecutive months. The trade union hk Danmark brought an action, on behalf of both employees, before the Maritime and Commercial Court. It stated that they should have been offered reduced working hours rather than being fired, and that Paragraph 5(2) was only for employees who were ill, not the disabled. The Maritime and Commercial Court made a reference to the cjeu. In its questions, it focussed on the definition of ‘disability,’ whether reduced working hours could be ‘reasonable accommodation’ within the meaning of Article 5 of the Court’s interesting decision to treat this as part of the proportionality analysis, rather than an objective justification for the difference in treatment concerned. The justifications which had been accepted earlier in relation to age discrimination in fact referred to a different instance of inequality, which arguably went to the rationale of the rule itself, not the rationale for a disparity of treatment as between two groups affected by one aspect of it. However, a second and more important reason why this small part of the decision is worth noting is because of the Court’s assertion that an objective justification (and again this implies that the three year advantage was viewed by the Court as such a justification, and not a matter of proportionality) could not be related to the subject matter of the discrimination (in other words, the protected ground) itself. Cf Judgment of 18 May 2015, Dupré v eeas, Case F-​11/​14, ecli:eu:F:2015:47, considered at section 6.3.3 below, in which the absence of comparability was to all intents and purposes used as the justification. 290 Case C-152/ 11 Odar (n 286) para 70. 2 91 Judgment of 11 April 2013, hk Danmark, acting on behalf of Jette Ring v Dansk almennyttigt Boligselskab and hk Danmark, acting on behalf of Lone Skouboe Werge v Dansk Arbejdsgiverforening acting on behalf of Pro Display A/​S, Joined cases C-​335/​11 and C-​337/​11, ECLI:EU:C:2013:222. 292 Lov om retsforholdet mellem arbejdsgivere og funktionærer.

The ‘Article 19’ grounds

187

of the Directive, and the extent to which Paragraph 5(2) of the fl could be utilized by employers in the case of disability. The Court of Justice, recognizing that the United Nations Convention on the Rights of Persons with Disabilities had been approved on behalf of the European Community293 since the handing down of the Chacón Navas judgment, now amended its definition of disability from the latter judgment to reflect that contained in the un Convention, namely: a limitation which results in particular from physical, mental or psychological impairments which in interaction with various barriers may hinder the full and effective participation of the person concerned in professional life on an equal basis with other workers.294 The Court added that the impairments concerned must be ‘long-​term.’295 The Court stressed that it did not matter if the origin of the disability was illness, although an illness not entailing a limitation as described would not be covered (Chacón Navas). Nor did it matter, according to the Court, if the disabled person retained some ability to work, or whether or not they required special equipment to do so. With regard to reduced working hours, the Court was satisfied that the definition of accommodation measures under Article 5 of the Directive was wide enough to cover them (as long as their provision did not entail a disproportionate burden for the employer). Furthermore, if the employee’s absence, triggering the operation of Paragraph 5(2) of the fl, was in fact due to the employer’s failure to accommodate and not due to the worker’s disability itself, then such operation would be illegal. Finally, the Court turned its attention to the essence of Paragraph 5(2) of the fl, namely, whether or not dismissal, of a disabled worker, unaccompanied by the usual safeguards, on grounds of having taken sick leave, was precluded by the Directive. Shackled by the Aristotelian test, the Court could only answer this question in comparison with a non-​disabled worker. It held that there was indirect discrimination of the disabled worker, as he or she was more ‘exposed to the risk of application of the shortened notice period’296 than the non-​disabled worker. It also held that this discrimination could be objectively justified on grounds of the encouragement of recruitment and the promotion 2 93 294 295 296

By Council Decision 2010/​48/​EC of 26 November 2009 [2010] oj L23/​35. Judgment in Joined cases C-​335/​11 and C-​337/​11 hk Danmark (n 291) para 38. ibid para 39. ibid para 76.

188 Chapter 4 of labour market flexibility, but left the question of proportionality up to the national court. In Z,297 Ms Z, who was a teacher, suffered from a condition whereby she was born without a uterus and could not conceive children; she did have functioning ovaries. She and her husband opted for surrogacy and, following in vitro fertilization in Ireland, the fertilized egg was transferred to the surrogate mother in California (where the law provides for detailed regulation of surrogate pregnancies and births). Ms Z travelled with her husband to California for a number of weeks in April and May 2010 in order to be present at the birth, on 28 April 2010, and then returned with the baby to Ireland. In February 2010, Ms Z had applied to the Government Department with responsibility for the school where she worked, requesting leave equivalent to adoptive leave (which did exist in Ireland at the time under the Adoptive Leave Act 1995). The Department refused and Ms Z brought an action in the Equality Tribunal, claiming discrimination on grounds of gender, family status and disability. On reference to the Court of Justice, the latter court seemed to take a narrow view with regard to all three of these alleged grounds of discrimination. There could be no gender discrimination, direct or indirect, it held, because a man who had used surrogacy would receive no leave either (the equal misery argument). There could not be discrimination on grounds of pregnancy, it continued, because Ms Z had never been pregnant. Finally, reasoned the Court, there could not be discrimination on grounds of disability because, although the condition from which Ms Z suffered was a ‘limitation which result[ed] in particular from [a]‌physical, mental or psychological impairment […],’ within the meaning of the first part of the definition of disability given in hk Danmark,298 and indeed was long term, it did not satisfy the second part of the definition in that it did not hinder Ms Z’s full and effective participation in professional life on an equal basis with other workers. One last case to consider is Kaltoft,299 in which Mr Kaltoft, following his dismissal as a childminder, sought compensation for discrimination on grounds of obesity. The District Court of Kolding, Denmark, asked the Court of Justice inter alia whether it was against eu law for a public sector employer 297 Judgment of 18 March 2014, Z v A Government department and The Board of management of a community school, C-​363/​12, ECLI:EU:C:2014:159. 298 See supra, text accompanying n 294. 299 Judgment of 12 December 2014, Fag og Arbejde (foa), acting on behalf of Karsten Kaltoft v Kommunernes Landsforening (kl), acting on behalf of the Municipality of Billund, C-​354/​ 13, ECLI:EU:C:2014:2463.

The ‘Article 19’ grounds

189

to discriminate on grounds of obesity, whether obesity was a disability under Directive 2000/​78, and if it was a disability, which criteria would be decisive. The Court replied that there was no general principle of non-​discrimination on grounds of obesity in eu law. Turning to the issue of Directive 2000/​78, though, the Court repeated its definition of disability from hk Danmark,300 and stated that this could cover an obese worker, if their full and effective participation in professional life (on an equal basis with other workers) was hindered.301 4.6.2 A Walzerian Analysis (a) Introduction: Chacón Navas and Coleman In Chacón Navas, it can be seen that the Court decided against allowing sickness either to form part of one of the original grounds in the Framework Employment Directive (specifically disability), or to be a ground in its own right. Thus, the comparison (between a sick employee and a healthy one) was never run.302 Since complex equality is only being mooted in this book as a complement to traditional comparison within court cases at the cjeu,303 there is perhaps no need to consider Ms Chacón Navas’ case from the Walzerian angle either. However, this decision will be revisited later on. Coleman, meanwhile, offers a nice chance to review some of the features of the Walzerian approach already encountered in the earlier analyses. While a number of things were being distributed by Mr Law to Ms Coleman, they could perhaps all be summed up in a single idea: flexibility. Ms Coleman lost 3 00 Judgment in Joined cases C-​335/​11 and C-​337/​11 hk Danmark (n 291). 301 The Court’s denial that Ms Z was disabled in Z, and ambiguity as to whether Mr Kaltoft was disabled in Kaltoft, and its consequent refusal to perform the like-​for-​like test in either case, can be contrasted with Judgment of 22 May 2014, Wolfgang Glatzel v Freistaat Bayern, C-​356/​12, ECLI:EU:C:2014:350, a case of alleged discrimination on grounds of disability falling outside the parameters of the Framework Directive. Here, despite the Court’s uncertainty as to whether or not Mr Glatzel was disabled (ibid para 47), it still went on to declare that there was a difference in treatment and to consider objective justification. 302 It is not exactly clear why the Court only mentioned the general principle (Case C-13/05 Chacón Navas (n 276) para 56), but did not actually make use of it; so far as can be seen the case falls within the scope of Community law. One explanation is that this course of action was not alluded to by the national court in its questions. However, that should not have stopped the Court from raising the matter proprio motu. 303 Not as some kind of all-​embracing programme for societal reform. Of course a trickle-​ down effect, from the Court to the wider world, cannot be ruled out, and is even to be expected, where distribution is concerned.

190 Chapter 4 out in the distribution of flexibility owing to a characteristic (albeit of her child) which acted as a negative dominant, potentially rendering the distribution defective.304 First, though, the Court would have to decide whether having a disabled child formed any part of the shared meaning of flexibility in Ms Coleman’s community. Assuming that it decided that it did not, the distribution would be declared defective, and a rerun ordered.305 This time Ms Coleman would receive her share of flexibility (of working arrangements, hours, and so on) on the same terms as workers with non-​disabled children. But a painstaking dissection of the ways in which they might be different from her, and she from them, would be unnecessary. (b) Double Distribution: Odar and hk Danmark A certain complication arises though when attempting to analyse both Odar and hk Danmark from the Walzerian perspective, and that is that they are what might be called ‘double distribution’ cases, in other words, that the litigant in question will be a distributee twice, and thus in two different spheres. Taking Odar first, here Mr Odar, and of course all of the other employees of Baxter, are first recipients of social plan benefits, and later of pensions. Age could be part of the meaning of social plan benefits, with the younger workers being given less help and the older more, but that meaning would then lend itself to a distribution based either on the worker’s length of service, or on how long they had left until they qualified for their pension, in all cases, not the current system with its seemingly arbitrary move from the first method to the second at the age of 54. Thus a Walzerian judge would give a different answer 304 Note how, once again, it would be possible to view this case the other way around, so that workers with non-​disabled children would be the ones responsible for the boundary conflict, via their (tyrannical) possession of a positively dominant good. In this alternative scenario one can detect, again, the operation of the Tyranny of ‘Normalcy’. 305 In the Forum, Ms Coleman would presumably argue that flexibility, as applied to an employee, meant the ability to vary one’s working arrangements to take account –​within reasonable limits  –​of certain other circumstances, including personal circumstances. Mr Law on the other hand might argue that it meant the same thing, but with the caveat, ‘such ability only being enjoyable by parents of non-​disabled children.’ It is hard to see how he could make this argument at all, let  alone in a convincing enough manner to attract a majority of votes (or whatever else the chosen system for electing the winning meaning required). However, even if he had succeeded in this argument in the Forum, counsel for Ms Coleman could invoke the Override before the Court, asserting that this meaning infringed Article 14 and Protocol No 12 of the European Convention (at the very least). Alternatively, the choice of meanings itself could be left to the Bench via the alternative theory of mediated complexity, for which see below Chapter 8.

The ‘Article 19’ grounds

191

to the third question than did the cjeu, finding the current arrangement to be discrimination on grounds of age. Turning to the fourth question, disability could also be part of the meaning of social plan benefits, but would that result, as presently, in the disabled worker receiving less? If the distributive forum had considered it fair, when distributing pensions, to give disabled workers their pensions early, which seems possible, and if the distributive criterion chosen for social plan benefits was how long they had left until they qualified for their pension, then, in order to stay true to the meanings of both goods, the community would it seems have to award fewer benefits to the disabled worker, as now. Thus once again the Walzerian would reach the opposite result to the Aristotelian, finding the current arrangement not to be discriminatory. If, though, length of service was elected as the proper entry token to the sphere of social plan benefits, then the disabled worker would not lose out in comparison to his or her non-​disabled colleague, and a litigant like Mr Odar would have no cause of action at all, where his disability was concerned. Turning to hk Danmark, here the two litigants, Ms Ring and Ms Skouboe Werge, as well as the other employees of the companies concerned, were first to be recipients of a job, and later of dismissal/​a notice period. The meaning of a (normal) notice period given by employer to employee on sacking him or her, which would of course be up to the community concerned, might be held to be a measure to help a worker, faced with imminent unemployment, to find a new job and otherwise reorganise their lives, without loss of income in the meantime, and therefore keeping the instability of losing their job to a minimum. A measure like Paragraph 5(2) of the fl, then, providing for a short notice period, would seem to offend against this meaning, and thus against the principle of equality as between employees who had taken sick leave and those who had not. A disabled worker, who would be likely to have taken more sick leave than his or her non-​disabled colleagues, would need the notice period as much as they would. However, what of the argument, made by the Danish government, that this measure helped to ‘encourage employers to recruit and maintain in their employment workers who are particularly likely to have repeated absences because of illness’?306 And, if accepted by the members of the Walzerian Forum, and particularly the disabled members, would this incentive for employers, bringing with it a reduction in the employee’s notice period, form part of the meaning of ‘a (normal) notice period’, or part of the meaning of ‘a job’, or even just part of the good of a job itself?

306 Judgment in Joined cases C-​335/​11 and C-​337/​11 hk Danmark (n 291) para 78.

192 Chapter 4 Starting with the assumption that it formed part of the meaning of ‘a (normal) notice period,’ it would be seen that such meaning would now be twofold, first, the provision of assistance to the worker to find a new job, but also second, an incentive to the employer to hire a worker who may take a lot of sick days, thus (arguably) costing him or her more than to hire a worker without a propensity to take sick days. It would be the Forum’s job to decide (perhaps) on a sliding scale: as the worker’s likelihood of taking sick days increased by x amount, their notice period would decrease by y amount. Thus, the Forum members would have to debate, negotiate and eventually strike a balance –​ at the stage of distributing the job –​between the workers’ interests and the employers’ interests. The model which they ended up with may look like the one currently set out in Paragraph 5(2) of the fl, or it may not. An employee’s x status in the distributive sphere of jobs would then, by agreement, be able to sway their award in the distributive sphere of notice periods by y amount, that link or bridge between the two spheres, while technically a failure to properly ring fence these two spheres, would not breach complex equality as the members of the community concerned would have decided that likelihood of taking sick days would be part of the meaning of (length of) notice periods. However, what of the extra complication of litigants, like Ms Ring and Ms Skouboe Werge, who only became ill/​disabled after beginning their job with the employer concerned? Such employer would have hired them on the basis of a three month notice period and now, it is submitted, would have to give them the three month notice period as previously agreed; to be true to Walzer, the meaning on the day of the distribution is surely the only one that counts. Being ill/​disabled may entail, again by agreement, a specialised outcome as discussed above, and the ill or disabled Forum member knows what is in his or her interests and will (hopefully) argue for it. But becoming ill or disabled should not entail this specialised outcome, and the employee in question should be entitled to the outcome they expected all along.307 This last point illustrates an interesting difference between existing laws and the Walzerian Forum. Existing laws may assist in the production of the sort of ‘chain reaction’ of failure described in section 2.3 above, what one might call exponential negativity, because they are set in stone and an individual whose circumstances change faces immediate reclassification across the statute book. Of course that will entail an amplification effect as the legal consequences of 307 The only possible way around this would be, again, for the employers in the Forum deciding on the meaning of a job, to convince the others that the inclusion into the employment contract of a clause altering the notice period in the case of subsequent illness/​ disability was or should be part of that meaning.

The ‘Article 19’ grounds

193

the original misfortune become known one after the next. Complex egalitarian fora may help to alleviate this as they operate more on a day-​to-​day basis. A good bargain yesterday need not become a bad bargain today by blind operation of pre-​existing law, as meanings are malleable enough to take the change of circumstances into account, or, perhaps more crucially, exclude the change of circumstances as irrelevant. One could go further and say that yesterday’s bargain may simply not be reopened; if possession in the good has already changed hands then that closes that particular ‘deal.’ Changes in future, on either side, can only have prospective effect (in other words, the next time that that particular distributor and that particular distributee do business), not retrospective effect vis-​à-​vis deals already done. Why should the employer ‘save money,’ via a reduced notice period, in the case of Ms Ring and Ms Skouboe Werge, where the quid pro quo for that saving (the provision of better job prospects for the ill or the disabled) has not been/​could not be performed in the case of these two women? Why should there be an alleviation of the burden on the employer without a concomitant experience of benefit on the part of the employee(s)? The employer should take Ms Ring and Ms Skouboe Werge as he or she found them on the day of the distribution, and they are entitled to have the good, and any previously debated follow-​on goods, with its or their meaning or meanings –​to them, on that day –​intact. Their circumstances changed, not their good. Both Walzerian law and current law take your new circumstances into account tomorrow. But only current law will reopen today’s business, confiscating or reclassifying like some invisible loss adjustor. Law has small print enabling changes in circumstance to be anticipated and provided for, not always to the changed party’s benefit. A meaning is a simpler animal. It can change, and this may affect a future (unrelated) distribution, but the old meaning had no small print, no contingency plan for this change, and so past distributions based on it are unaffected. Benefits and burdens of the change thus lie where they fall; benefits cannot be preordered, burdens cannot be preexcluded. If the notice period is part of the meaning of a job, in its three month ­version, as protector of stability for a sacked worker, and in its one month version, as enhancer of attractiveness for a disabled worker, then being disabled or not may be legitimately present in the sphere of job distribution, but only as a determinant of the terms of the job once offered, not as determinant of whether or not an offer should be made (the distributive criterion there being merit). However, the inclusion of disability within the sphere must originate from the wishes of the disabled workers in the Forum, viewed by them not as a negative attribute robbing them of two months’ notice, but as a positive attribute facilitating their recruitment (albeit on inferior terms) in the

194 Chapter 4 first place. Should the disabled constituency within the Forum abandon this pragmatic approach in favour of a more political one, insisting that employers should need no ‘consolation prize’ for hiring their members, then the bridge referred to earlier would be dismantled, and the status of being disabled would become simply a negative dominant, barred from entry to the sphere, leaving disabled candidates free to present themselves on the sole basis of their merit. It is submitted that there would be some danger in allowing considerations based on a suspect ground into the meaning of a good, possibly even leading to Nuremburg-​style laws, and so perhaps this is an argument in favour of leaving the definition of the meaning to a judge or judges. (c) Incomparable Walzer: Z and Kaltoft (and Chacón Navas again) The more recent disability decisions again confirm how convoluted comparisons under the Aristotelian test can be mercifully side-​stepped using a Walzerian one. Nowhere is this better illustrated than in the case of Z, where, at different points of the argument, comparisons are mooted as between: couples wherein the mother gives birth to the child herself (henceforth denoted by the preceding adjective ‘biological’) and couples who make use of surrogacy; as between parents who adopt and parents using surrogacy; as between biological mothers and mothers making use of surrogacy, and as between biological mothers who breast feed and those making use of surrogacy who breast feed. Meanwhile, from the Walzerian point of view, the case is relatively straightforward, and once the meaning of maternity/​adoptive leave is ascertained (for the sake of argument, let this be a period of paid leave allowed by an employer to enable the parent –​chromosomal or adoptive –​of a newly born child to bond with said child, whether biologically or emotionally or both), then it would not seem at all controversial to declare Ms Z’s condition as outwith that meaning and therefore not something to be considered within the distributive sphere at all (that is, not something relevant to the distribution of leave). As it stands, discrimination law requires categories. Were the rules of ­complex equality to be codified into law, the resultant Walzerian law would supplant and extend discrimination law, as any attribute on the basis of which a distribution is sought to be based (in the real world, post hoc) may be held to be a positive or negative dominant, stripped of its dominating power by being blocked from the sphere in which it (currently) commands undue reverence, without need of its prior inclusion into a suspect category, a procedure which can have the effect of leaving the wrongfully distinguished without remedy under the present arrangement, as in Z, the distinction in question perhaps not having even been anticipated. Put another way, under Walzerian equality one

The ‘Article 19’ grounds

195

is suspicious of everything. This is not to denigrate or trivialize the existing suspect grounds, but merely to say that privilege is irrelevantly invoked too often, and sometimes too subtly, to be left to –​or to be left solely to –​a predetermined list.308 As seen in Z, under Aristotle, a category error can work massively in favour of a discriminator. Under Walzer, a ‘sphere error’ will, or should, always catch them out. Discrimination law needs to be about not only the indictment of discriminators, but about their identification in the first place. Another and perhaps more sinister aspect of the current test is revealed in the recent case law. Although equal treatment is at issue in the Framework Directive,309 and this presupposes that those who pray the Directive in aid will have the opportunity to have their situation considered in contradistinction to the situation of some other or others, this will only be the case if they can bring themselves within one of the categories covered. However, the new definition of ‘disability,’ handed down by the Court in hk Danmark, with its qualifying clause, ‘on an equal basis with other workers,’310 contains within it a hidden comparative test, enabling a judge to run their own, unseen a priori comparison of the plaintiff with those around him or her and then hold there to be no inequality present, locking the plaintiff out of the definition and thus denying them a proper running of the test, framed in the terms which they seek, that is, whether or not they –​a disabled person –​have been treated like or unlike a non-​disabled person. This is what happens in Z, where the issue of Ms Z’s comparative disadvantage compared to her co-​workers is dealt with in two paragraphs by the judges, these two paragraphs in turn referring to three paragraphs in the Advocate General’s Opinion, but none of these apparently considering the psychological effect which the absence of maternity/​adoptive leave might have on a person with Ms Z’s condition. It cannot be ruled out that she would not have gone to work at all if she had known that she would not receive the desired leave. This would it is submitted be enough to establish a hindering of the full and effective participation of Ms Z in professional life on an equal basis with other workers. This is turn would have brought with it disability status under the hk Danmark definition, and protection against the unequal treatment of Ms Z compared to those whose children were born by the worker themselves or else adopted. It is not argued that this definitional comparison, run surreptitiously before the main one, and, in the case of Z, precluding the main one altogether, will always produce an unfairness, just that it 308 ‘[T]‌he privilege of irrelevance’ is MacKinnon’s phrase: C MacKinnon, Feminism Unmodified: Discourses on Life and Law (Harvard up 1987) 4. 309 Supra n 1, Articles 1, 2(1), 5, 7(1), 7(2), 8, 9, 10(1), 13(1), 14 and 16. 310 See supra text accompanying n 294.

196 Chapter 4 seems to carry with it a danger of unfairness if abused, a danger which the non-​ comparative approach of Walzer avoids as the precise categorisation of Ms Z as either disabled or non-​disabled is not to the fore (as discussed above). Can it be ruled out, for example, in the later case of Kaltoft, that the Municipality of Billund would use the fact that Mr Kaltoft was able to work as a childminder for fifteen years to find his participation in professional life to be equal to anyone else’s, thus similarly denying him first disability status, and then statutory protection? This course of action is even hinted at approvingly by the Court in its judgment.311 However, a doctor, if asked, might well say that Mr Kaltoft’s equal participation in professional life was jeopardised by his condition, as it ruled him out of many roles open to others. Presumably he would then be regarded as disabled in respect of those roles.312 Consequently, if childminding was in fact the only job he could do, it seems a little unfair to deny him disability status just when he needs it most. Again, a Walzerian approach would avoid this potential sleight of hand. In the light of these thoughts prompted by the Z and Kaltoft judgments, one might revisit the decision not to consider Ms Chacón Navas’ case from the Walzerian angle. Ms Chacón Navas could rely on a Walzerian system to investigate whether she had been discriminated against, even if ‘only’ sick and not disabled, where, as she discovered, the current comparative regime, with its need for preestablished grounds of comparison, leaves her without a remedy, and even the new definition of disability, post-​h k Danmark, with its latent a priori comparison, seems destined to perennially refute the ‘unlikeness’ of any worker whose very work obliterates their claim to unequal participation in professional life. 4.7

Concluding Thoughts

And so it can be tentatively concluded that the application of a Walzerian test to equality alongside the Aristotelian one would fare the cjeu well in dealing with the Article 19 grounds, quickly exposing harmful assumptions (which 311 Judgment in Case C-​354/​13 Kaltoft (n 299) para 62. 312 In an excellent note on Z, Boujeka also points out the novel situational-​functional approach to disability adopted by the Court in that case, rendering an individual disabled in one context and non-​disabled in another, effectively stigmatizing the situation, and reducing the concept of a ‘disabled person’ to little more than a mechanism for figuring out where to draw the line between the protected and the unprotected: A Boujeka, ‘La gestation pour autrui et le handicap confrontés au principe de non-​discrimination en droit de l’Union’ (2014) 31(7612) Le Dalloz 1811, 1815.

The ‘Article 19’ grounds

197

otherwise either would only be exposed slowly and painfully, or would not be exposed at all), and avoiding tortuous hair-​splitting comparisons. Decisions in cases like Prais,313 Palacios,314 Grant315 and Z316 might well be transformed, and even where the Aristotelian result would not necessarily be different from the Walzerian one, it can be seen, in cases like Age Concern,317 Petersen318 and Coleman,319 how the route for arriving at this decision might well be easier and/​or more appropriate under a Walzerian regime. At the same time, this is not to ignore the difficulties which judicial use of Walzer might entail, particularly where the ascertaining of shared meanings is concerned (Firma Feryn,320 D,321 Maruko,322 Römer);323 these difficulties are fully discussed, and a solution proposed, in Chapter 8 below. 313 314 315 316 317 318 319 320 321 322 323

Case 130/​75 Prais (n 61). Case C-​411/​05 Palacios (n 129). Case C-​249/​96 Grant (n 2). Judgment in Case C-​363/​12 Z (n 297). C-​388/​07 Age Concern (n 100). Case C-​341/​08 Petersen (n 137). Case C-​303/​06 Coleman (n 280). Case C-​54/​07 Firma Feryn (n 21). Joined cases C-​122/​99 P and C-​125/​99 P D (n 2). Case C-​267/​06 Maruko (n 234). Judgment in Case C-​147/​08 Römer (n 240).

Chapter 5

Nationality Discrimination 5.1 Introduction Nationality has been described as the legal bond between a person and a State.1 The operative word here is ‘legal.’ Unlike gender, race, or any of the other grounds of discrimination so far considered, a person’s nationality comes into existence by operation of law, and can sometimes even cease to exist by operation of law.2 This legal dimension greatly affects how society views inequality grounded on nationality. While discrimination in matters relating to nationality should certainly be, if at all possible, ‘avoid[ed],’3 and while discriminating against others on grounds of nationality (or supposed nationality) has certainly led to unfairness and even great hardship on occasion,4 the fact remains that where a legal relationship is concerned, it makes no sense, from a purely legal point of view, and in the normal course of events, to treat those between whom the relationship pertains in exactly the same way as those between whom the relationship does not pertain, at least vis-​à-​vis the matters with which the relationship is concerned. To act in any other way would be to render the relationship redundant. Entering into a relationship creates rights among the parties in which those outside the relationship simply cannot share; the creation of exclusive rights is the whole point of the exercise. One could draw an analogy (although not a perfect one) with the doctrine (in English law) of privity of contract. An individual who is not a party to a contract cannot derive rights thereunder. He is a stranger to the contract. The exclusion of the stranger is not only not unlawful, it is (in contract law) mandatory. Similarly, the foreigner5 should not be surprised to find him or herself excluded from a number of the bargains struck 1 European Convention on Nationality, Article 2(a). There are of course other, non-​legal senses of the term. 2 In the European Union, possession or otherwise of nationality is a matter for national law, not eu law. See the Declaration on Nationality of a Member State, appended to the Treaty of Maastricht: [1992] oj C191/​98. 3 European Convention on Nationality, 5th recital to Preamble. 4 For example, the expulsion of Germans and Hungarians from Czechoslovakia after World War II; the detainment and internment of large numbers of Japanese Americans in the wake of Japan’s attack on Pearl Harbour. 5 Étranger in French, extranjero in Spanish.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/​9 789004354265_​0 06

Nationality Discrimination

199

between the nationals of the State he or she is visiting, and the State itself. He or she is simply not privy to them. However, the creation of the European Union puts a different complexion on things. It is an entity sui generis, establishing, as the cjeu famously declared in the Van Gend & Loos case, a ‘new legal order,’6 bringing with it new legal relationships not supplanting, but supplementing existing ones. In Walzerian terms, this can in some cases mean the co-​existence of two, mutually inclusive, distributive communities –​a micro sphere within a macro sphere. The stranger may now only be a stranger at the micro level; at the macro level he or she is a friend, a fellow distributee of certain rights and obligations, a co-​contractor. His or her claim to equal treatment is not just stronger now, it is insuperable. But this dual (legal) personality is often stifled by Member States, and the eu’s split-​level decision-​making, and double context, is misunderstood or ignored, as will be seen.7 5.2

Nationality Discrimination in eu Law

Discrimination on the grounds of nationality is prohibited by Article 12 of the Treaty of Rome, now Article 18 tfeu, although the prohibition is reiterated (sometimes only implicitly) on several occasions thereafter, perhaps most noticeably in the provisions setting out the free movement of persons,8 establishment9 and services,10 and in a number of pieces of secondary legislation.11 6 7

8 9 10 11

Case 26/​62 nv Algemene Transport-​en Expeditie Onderneming van Gend & Loos v Netherlands Inland Revenue Administration [1963] ecr 1, at Part IIB of the grounds. Walzer is perfectly comfortable with the idea of individuals living in spheres-​within-​ spheres, and thus owing, as it were, a double allegiance. See his description of what he calls multinational empires in M Walzer, On Toleration (Yale University Press 1997) 14–​19. The mere fact that the inner sphere’s inhabitants were  –​in some respects, anyway  –​ governed from without rather than from within does not breach complex equality. See Article 39(2) ec –​now, since the coming into force of the Treaty of Lisbon, Article 45(2) tfeu. See Article 43 ec, second paragraph –​now, since the coming into force of the Treaty of Lisbon, Article 49 tfeu, second paragraph. See Article 50 ec  –​now, since the coming into force of the Treaty of Lisbon, Article 57 tfeu. This means nationality of an ec Member State, though. Third country nationals, subject to one or two exceptions, are excluded. The cjeu has made it clear that the principle of non-​discrimination on grounds of nationality may not be invoked where non-​member countries are at issue, for example, Case 52/​81 Offene Handelsgesellschaft in Firma Werner Faust v Commission of the European Communities [1982] ecr 3745.

200 Chapter 5 This is because the outlawing of discrimination on grounds of nationality lies at the heart of the internal market project. As Van der Mei has put it, Nationality requirements are by their very nature at odds with the objectives of the Community. The common market could never have been realised, and could never operate effectively, if Member States were free to exclude nationals of other Member States from their territory and labour markets.12 In the field of free movement, the Court of Justice has heard countless cases involving nationality requirements, actual or constructive, most of which boil down to a dissonance between the treatment of a person or entity national to the so-​called ‘host’ State and the treatment of a person or entity entering the host State from another Member State (the ‘home’ State), or indeed from a third country. Where discrimination on the grounds of nationality is alleged, the Court uses Aristotelian reasoning in making its finding, with the same unsatisfactory results seen earlier in relation to the other grounds. In Angonese,13 for example, the Court faced its perennial problem of identifying the right comparators. In this well-​known case, an Italian man was banned from applying for a job at a bank in the Bolzano region of Italy, on the grounds that he did not hold a certain certificate, awarded on the successful completion of an exam and known as the patentino, attesting his fluency in German; the fact that he was fully bilingual, having learnt German by other means, made no difference. The patentino being a local qualification, many other Italians, in other parts of the country, would have been similarly turned down, making the case, according to some, one of residency discrimination and not nationality discrimination; the correct comparators would therefore have been a resident of Bolzano and a non-​resident of Bolzano. However, the Court held the bank’s condition to be discriminatory on grounds of nationality, preferring to compare residents of Bolzano with non-​Italians. Davies regards Angonese as further proof of the Court’s ‘shaky grasp of who should be compared with who,’14 his principal example being a case about free entry for local children and oap’s into the 12 13 14

A Pieter van der Mei, Free Movement of Persons within the European Community (Hart 2003) 70. Case C-​281/​98 Roman Angonese v Cassa di Risparmio di Bolzano SpA [2000] ecr I-​4139. G Davies, ‘ “Any Place I Hang My Hat?” or: Residence is the New Nationality’ (2005) 11(1) Eur LJ 43, 46. A Walzerian inquiry would reveal, almost definitely, that part of the meaning of a job at a bank in the German-​speaking region of Italy –​or at least the opportunity thereof –​ was the need to be bilingual, not ‘possession of the patentino’ (although of course the latter

Nationality Discrimination

201

museums of Florence: Commission v Italian Republic.15 Again, in holding that there had been nationality discrimination, the Court ran a comparison between the resident and the foreigner, rather than between the resident and the non-​ resident; Davies calls this ‘logically regrettable.’16 However, even after the comparators are selected, the Court still struggles with the question of whether two people or things are like or unlike, and whether their treatment was like or unlike. In another case concerning a regional requirement, this time in the context of the free movement of goods, Walloon Waste,17 it was necessary to compare waste produced in Wallonia with imported waste (as the Walloon Regional Executive wished to treat these two things in an unlike fashion). As Wilsher has explained, the outcome of the comparison depends entirely on whether the observer used a market equivalence method, or a regulatory equivalence method.18 From the point of view of market equivalence, the two types of waste are of course alike. But from the point of view of regulatory equivalence, waste produced at home and imported waste are different; the environmental policy pursued by the Executive –​in keeping with the principle that waste should be disposed of as close as possible to its source –​is satisfied by the first but not by the second, making them (in the eyes of a regulator) unlike.19

15 16 17 18 19

is one of the ways of satisfying the former). As things stood, the patentino was an archetypal dominant –​those possessing it gained automatic and exclusive access to other goods, while those without it were turned away. Although this Walzerian analysis reaches the same result as the Aristotelian one (a breach of the equality principle), it can be seen that the first is much easier than the second. It avoids both the problem of choosing comparators (job-​ applicants from two different regions of Italy, or job-​applicants from two different Member States), and the problem of one of the comparators’ needing to be hypothetical. Case C-​388/​01 Commission v Italy [2001] ecr I-​721. Davies (n 14) 49. Case C-​2/​90 European Commission v Belgium [1992] ecr I-​4431. Daniel Wilsher, ‘Does Keck discrimination make any sense? An assessment of the non-​ discrimination principle within the European single market’ (2008) 33(1) EL Rev 3, 7. In the event, the Court mixed the two approaches by finding discrimination, only to justify it, controversially, on grounds of imperative requirements of environmental protection. A Walzerian approach would have quickly distilled the case down to its essential question: in the business of waste-​treatment, is the distance that waste has travelled a pertinent factor (in which case the breach of the border between the sphere of nationality and the sphere of waste-​management is legitimate), or an extraneous factor (in which case it is illegitimate)? Guessing, for the sake of argument, that it is a pertinent factor (in accordance with the principle mentioned in the main text), then Belgium should have won the case outright. Awkward –​even absurd –​questions about whether two lots of waste are like or unlike are avoided, as, for the record, are obscure and impenetrable legal riddles such as whether direct discrimination can or cannot be objectively justified.

202 Chapter 5 The tax case of Gilly20 also presented difficulties at the stage of the like-​ for-​like analysis. Here, a woman who lived on the Franco-​German border, on the French side, who taught in a German state school, and who was of dual (French and German) nationality, found herself taxed in Germany, pursuant to a Franco-​German Double Taxation Convention. Although married, her non-​residence in Germany meant that she could not benefit from a so-​called ‘splitting system,’ in accordance with German federal income tax laws; this meant that she was treated as though she were single, with the result that she had to pay more tax in Germany than she would have had to had she been taxed in France, or than she would have had to had she been resident in Germany. The Court examined the rationale for the German federal income tax laws. This was that non-​residents usually have some other source of income, where they actually live, and the income received in the State where they are not resident is just a part of their total income, while, where the State is taxing a resident, they take the view that this is all of or the main, concentrated part of the taxpayer’s total income.21 Accepting this rationale, the Court refused to find discrimination as between a person living in Germany and paying tax in Germany, and a person living in France and paying tax in Germany.22 However, as Vanistendael23 has noted, Mrs Gilly’s being treated as a single person is highly discriminatory, and the rationale was flawed: clearly her teacher’s salary, in Germany, was the main part of (if not all of) her total income. As usual, Aristotle can be used to condone, perpetuate and, in some ways, further legitimize an existing (unfair) distinction. The German federal income tax laws said (wrongly) that the non-​resident and the resident were unlike, and the cjeu simply grafted Mrs Gilly’s case on top of this pre-​existent, and unsound, analysis –​ treating the unlike in unlike fashion could not be discrimination. Rather, the resident and the non-​resident (or the taxpayer with German nationality and the taxpayer with dual Franco-​German nationality) should have been treated as like, making their unlike treatment unjustified discrimination. The confusion is made worse by the fact that, just over three years before Gilly, the Court had unveiled its so-​called ‘Schumacker doctrine,’24 by which the resident and the non-​resident had been treated as like, but only in situations where the 20 21 22 23 24

Case C-​336/​96 Mr and Mrs Robert Gilly v Directeur des services fiscaux du Bas-​Rhin [1998] ecr I-​2793. ibid para 49. ibid para 50. F Vanistendael, ‘Annotation Gilly’ (2000) 37 CMLRev 167. Case C-​279/​93 Finanzamt Köln-​Altstadt v Roland Schumacker [1995] ecr I-​225.

Nationality Discrimination

203

non-​resident received the major part of his or her income in the Member State of employment, a condition for comparability which, one would have thought, Mrs Gilly herself satisfied.25 These examples hopefully illustrate that the cjeu’s reliance on the Aristotelian test for equality has been as problematic in the field of nationality discrimination as it has been in the other fields so far considered. Using a sample of cases, this chapter attempts to ascertain if and to what extent a Walzerian analysis would improve matters. 5.3

A Specific Example: The Free Movement of Persons

As was mentioned in Chapter 2, Walzer holds equal citizenship to be an essential prerequisite for the establishment of complex equality.26 It is crucial that every member of the distributive community should have a fair and equal 25

26

The determinative difference between the two cases, according to the Court, was that Mrs Gilly also paid tax in her State of residence (France), while Mr Schumacker did not pay any tax in his (Belgium). Vanistendael holds this distinction to be ‘irrelevant:’ Vanistendael (n 23) 178. For further discussion of the two cases, see N Bammens, The principle of non-​discrimination in international and European tax law (ibfd 2012), particularly Chapter 13, and B Hagen, ‘Schumacker Comparability –​the Conditions of the Schumacker Doctrine’ in K Dziurdz and C Marchgraber (eds), Non-​Discrimination in European and Tax Treaty Law: Open Issues and Recent Challenges (Linde Verlag 2015) 19. A  Walzerian approach would have allowed the Court to face the real question, which was whether non-​residency and/​or non-​nationality were pertinent factors for Germany to take into account when distributing the more lenient tax demands. If the distributive community agreed with the Court that the correct badge to be shown at the entrance to this particular distributive sphere was receipt of all or nearly all one’s income in Germany, then that was the only legitimate dominant. Any other badges (residency in Germany, German nationality) would have to be left at the door. However, if the community felt that the meaning of these more lenient demands had to do with their role in providing fairness across a household where a taxpayer was married, then it would be the status of being married (or an equivalent arrangement resulting in financial codependence) which should determine their distribution; this would seem to be Vanistendael’s position. In either event, by focusing on the good and its meaning, Walzer cuts through the assumptions being made that (say) non-​residence in a state somehow equates to a smaller contribution being made to that state’s exchequer. For another example of the Court’s difficulties with the like-​for-​like stage, see the analysis and discussion of Vigier below, section 5.6: Case 70/​80 Tamara Vigier v Bundesversicherungsanstalt für Angestellte [1981] ecr 229. Supra section 2.5.

204 Chapter 5 chance to participate in the Forum, which means that even the most recent newcomer must be granted full political rights:  ‘every immigrant and every resident is a citizen, too’.27 It is obvious that the existence of a second-​class citizenry, such as the metics in ancient Greece, as described by Walzer in his chapter on ‘Membership,’28 would pose a serious problem for the complex egalitarian. In the us, there exists such an underclass, whose members are known as ‘resident aliens.’ Linda Bosniak, for one, has applied Walzer’s theory to the problem of resident aliens, with some degree of success.29 She draws on cases where the us courts have attempted to ascertain which, if any, rights, enjoyed by nationals, may be lawfully withheld from aliens. She then uses the concept of sphere-​separability to explain why some denials of rights by States were held to be (or should have been held to be) impermissible. For example, in the Graham case, the Supreme Court found that the denial of certain State benefits to resident aliens of foreign origin was akin to denying them a right to reside in the first place, in other words, it was an invasion, by the sphere of welfare benefits (a State concern), of the sphere of membership (a Federal concern).30 Such decisions, and the tensions between State competence and Federal competence which they sometimes reveal, are very helpful when one comes to consider the cjeu’s case-​law on the free movement of persons. Over the many years that it has been hearing cases on the subject, the cjeu has faced a plethora of distribuends: educational grants,31 childbirth loans,32 rail fare reduction cards,33 places on training courses,34 and so on. The question 27

28 29 30

31 32 33 34

M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) 52. See also Walzer’s later remark, ‘Participants in economy and law, [guests] ought to be able to regard themselves as potential or future participants in politics as well’: ibid 60. See the discussion at ibid 53–​5. See L Bosniak, The citizen and the alien: dilemmas of contemporary membership (Princeton University Press 2006). John O Graham, Commissioner, Department of Public Welfare, State of Arizona, Appellant, v Carmen Richardson, Etc.; William P Sailer et al Appellants, v Elsie Mary Jane Leger and Beryl Jervis, 403 us 365, 91 S Ct 1848. Case 76/​72 Michel S v Fonds national de reclassement social des handicapés [1973] ecr 457. Case 65/​81 Francesco Reina and Letizia Reina v Landeskreditbank Baden-​Württemberg [1982] ecr 33. Case 32/​75 Anita Cristini v Société nationale des chemins de fer français [1975] ecr 1085. Case 152/​82 Sandro Forcheri and his wife Marisa Forcheri, née Marino, v Belgian State and asbl Institut Supérieur de Sciences Humaines Appliquées -​Ecole Ouvrière Supérieure [1983] ecr 2323.

Nationality Discrimination

205

was, could a Member State vary the way it distributed these things according to whether it was confronted with one of its own nationals, or with a national of another Member State exercising their right of free movement (or even a member of the latter’s family)? Or did it have to treat both categories of person in the same way? The Court declared in the majority of cases that equal treatment of national and migrant was required; what was good for the host State goose was good for the home State gander. This idea can be plainly seen in a number of the early judgments, for example: If the widow and infant children of a national of the Member State in question are entitled to such cards provided that the request had been made by the father before his death, the same must apply where the deceased father was a migrant worker and a national of another Member State.35 And: [T]‌o require of a national of another Member State lawfully established in the first Member State an enrolment fee which is not required of its own nationals […] constitutes discrimination by reason of nationality.36 These sweeping, Aristotelian statements, insisting that Member States could not act towards nationals from other Member States differently to the way they acted towards their own nationals, in other words, that likes must be treated alike, conveyed in a way that did not draw too much attention to itself the fundamental axiom: that migrants and nationals were like. The Aristotelian test’s knack for cloaking a hidden message, and the ease with which it could be used as a rubber stamp for a foregone conclusion,37 were very much on display in these cases. Even where the Court did wish to find for the Member State, in most cases38 it did this by declaring whatever the migrant or his family member were seeking to be outside the scope of Community (primary or secondary) law.39 This reasoning also, at least partially, cloaked a difficult issue: the division of competences between the eec (as it then was) and the Member States. But the message was clear enough. Migrants and nationals could be regarded 35 36 37 38 39

Case 32/​75 Cristini (n 33) para 15. Case 152/​82 Forcheri (n 34) para 18. See section 9.1 below. But not all. See the discussion of Case 336/​96 Gilly (text accompanying n 20). Or occasionally even national law e.g. Case 70/​80 Vigier (n 25).

206 Chapter 5 as unlike when the matter under consideration was Member State competence, but as soon as the matter under consideration became Community competence, the two groups were to be regarded, almost without fail, as indistinguishable.40 The Aristotelian test is a blunt instrument at the best of times, but using it in this fettered way renders the results even less convincing. Above all, the sneaking suspicion that the Court’s decisions are entirely political, and that Aristotle is merely being used as so much whitewash, is raised throughout this case-​law. 5.4

The Early Case-​Law

A seeming agreement between the judge who finds the differential treatment of national and migrant to be discriminatory, and the complex egalitarian who discerns the exertion of dominance41 in the non-​distribution of the benefit concerned to the migrant, occurs often in the early case-​law, as will be seen in this section and the next. In the 1970s and 1980s, in case after case, the cjeu allowed the immutable logic of free movement –​turning yesterday’s ‘unlikes’ into today’s ‘likes’ –​to trump any other consideration. It was thus discriminatory for the French to deny citizens of other Member States an allowance for handicapped adults,42 a rail fare reduction card,43 or an allowance for women with children.44 Similarly 40

41 42 43 44

If a Member State did want to plead some difference (as between its own citizens and newcomers), it could only do this, it seemed, at the stage of objective justification, and even then with little chance of success, e.g. Case 237/​78 Caisse régionale d’assurance maladie de Lille (cram) v Diamante Palermo, née Toia [1979] ecr 2645; Case C-​147/​03 Commission v Republic of Austria [2005] ecr I-​5969. Somek has called this phenomenon ‘the proportionalisation of difference:’ A Somek, ‘Solidarity Decomposed: Being and Time in European Citizenship’ (2007) 32(6) EL Rev 787, 816. Or the imposition of negative dominance. Case 63/​76 Vito Inzirillo v Caisse d’allocations familiales de l’arrondissement de Lyon [1976] ecr 2057. Case 32/​75 Cristini (n 33). Case 237/​78 Palermo née Toia (n 40). The allowance was granted when the woman concerned reached the age of 65. The distributive community would therefore (probably) find that the ‘meaning’ of the allowance was assistance for the elderly, in which case the presence of nationality within the distributive sphere (the nationality of the recipient’s children in this case) was tyrannical –​an undue invasion by the sphere of membership. However, another of the conditions for the granting of the allowance was that the woman

Nationality Discrimination

207

it was discriminatory for Belgium to deny eu migrants an old age pension45 or a special unemployment benefit for young workers.46 The protection of the Treaty also extended to those nationals of other Member States who had resided in the host State over a period of time. They could lay claim, on a like-​ for-​like basis, to any and all entitlements to which nationals could lay claim, irrespective of national rules purporting to set a minimum residence threshold. Thus, in Scrivner,47 Belgium’s insistence that uk nationals should have resided in the country for five years before being entitled to receive the minimum subsistence allowance (‘minimex’) was found to be discriminatory by the Court. Likewise, in Frascogna,48 France’s setting of a fifteen-​year residence requirement as a condition precedent for the receipt of a special old-​age allowance was held to be nationality discrimination by the cjeu. 5.5

A Walzerian Analysis

To take the example of the Frilli49 case from the previous section, where Belgium sought to deny Ms Frilli, an Italian national, its guaranteed income for old people, a complex egalitarian would wonder whether Ms Frilli’s ‘being Italian,’ or ‘alienage’ to use Bosniak’s word, constituted a negative dominant disqualifying her from the distribution of this particular benefit, or indeed whether (elderly) Belgian citizens’ ‘citizen status’ acted as a positive dominant for them. While, as usual, no definitive answer could be given until the distributive community concerned had stipulated what its shared understanding of a ‘guaranteed income for old people’ was, it is possible to state, provisionally, that Ms Frilli should receive the benefit, unless ‘being Belgian’ was in some way part of said understanding. A guaranteed income for old

45 46 47 48 49

concerned should have had at least five children. If the distributive community found, then, that the meaning of the whole scheme was increasing France’s birth rate, it might well be justified in saying that the nationality of the children born did form part of this meaning; presumably children born with non-​French nationality would be excluded from the calculation of France’s birth rate, thus precluding their mother from receiving the allowance. Case 1/​72 Rita Frilli v Belgian State [1972] ecr 457; Case 261/​83 Carmela Castelli v Office National des Pensions pour Travailleurs Salariés (onpts) [1984] ecr 3199. Case 94/​84 Office national de l’emploi v Joszef Deak [1985] ecr 1873. Case 122/​84 Kenneth Scrivner and Carol Cole v Centre public d’aide sociale de Chastre [1985] ecr 1027. Case 256/​86 Maria Frascogna v Caisse des dépôts et consignations [1987] ecr 3431. Case 1/​72 Frilli (n 45).

208 Chapter 5 people, at least on the face of things, is not about being Belgian, but about being old.50 Likewise, the unlikelihood of ‘being French’ forming part of the shared meaning of a disability allowance, a travel card or a maternity benefit, or of ‘being Belgian’ forming part of the shared meaning of a pension for those at the end of their working life, or social assistance for those at the beginning, means that the host States’ inclusion of these attributes in their distributive criteria represents a boundary breach, causing the resultant distributions to be flawed. The same applies to the various forms of residence requirement. Such rules would, almost certainly, be inimical to a regime of complex equality, as it is unlikely that long-​term residence in Belgium would form part of the distributive community’s shared understanding of, say, a benefit intended to alleviate the hardships of unemployment. A Walzerian view of the Frascogna case51 would similarly result in the condemnation of the Member State, as long as the relevant distributive community did not consider long-​term residence in France to be a part of its shared understanding of a benefit designed to alleviate the hardships of old age.52 5.6

Cases Deriving from the Second World War

In more complicated cases, however, the Aristotelian avenue and the Walzerian avenue lead, or at least might lead, in different directions. In Fossi,53 the German Government was of the opinion that an Italian national, habitually resident outside the Federal Republic of Germany, should be denied a total disablement pension claimed under a social security law dating back to the Third Reich (he had worked in a mine in Sudetenland during the War). The cjeu, taking the view that this benefit fell outside the scope of the pertinent Community laws, agreed with this opinion.

50

51 52

53

Of course, the distributive community could respond that only those old people who had contributed to this benefit by paying their Belgian taxes should receive it. However, this would be not so much a nationality requirement as a residence requirement. This touches on the difficult issue of solidarity, discussed further below at 5.15. Case 256/​86 Frascogna (n 48). But, again, would the communities not be within their rights to argue that pay-​outs should be denied to those who had never paid anything in, such as new-​arrivals? See the section on solidarity below at 5.15. Case 79/​76 Carlo Fossi v Bundesknappschaft [1977] ecr 667.

Nationality Discrimination

209

Tinelli54 was a very similar case, with a claimant similarly disappointed. However, there the purpose of the law under consideration was given as: to facilitate the re-​integration, following events connected with the National Socialist regime and the Second World War, of exiles and refugees who contribute by their work to reconstruction in the Federal Republic of Germany.55 In the case of Even,56 the good in question was an early retirement pension, granted by Belgium, without any reduction, to Belgian veterans of the Second World War. Mr Even was a French national in receipt of a war service pension under French legislation; when he moved to Belgium and applied for the Belgian veterans’ pension, his application was refused. The cjeu provided no relief, declaring that the benefit fell outside the scope of the relevant Community Regulation. The purpose of the veterans’ pension was: to offer to Belgian workers who fought in the Allied forces between 10 May 1940 and 8 May 1945 and suffer incapacity for work attributable to an act of war a testimony of national recognition for the hardships suffered during that period and to grant them, by increasing the rate of the early retirement pension, a benefit by reason of the services thus rendered to [Belgium].57 Finally, among the other cases deriving from the Second World War, Vigier58 is also worth considering. Ms Vigier, despite being born in Germany (in 1923), was of French nationality and resided in France. As a ‘victim of persecution’ within the meaning of Germany’s Federal Compensation Law, and indeed having already received some compensation from Germany for loss of educational opportunities, Ms Vigier applied in 1975 for authorization to make retroactive payment of contributions with regard to invalidity and old age insurance. The Federal Insurance Office rejected this application; according to the relevant law, Ms Vigier had to have already paid at least one contribution 54 55 56 57 58

Case 144/​78 Renzo Tinelli v Berufsgenossenschaft der Chemischen Industrie [1979] ecr 757. ibid para 7. Case 207/​78 Criminal proceedings against Gilbert Even and Office national des pensions pour travailleurs salariés (onpts) [1979] ecr 2019. ibid para 12. Case 70/​80 Vigier (n 25).

210 Chapter 5 to the competent (German) institution in order to be regarded as an ‘insured person.’ Ms Vigier argued that contributions which she had made in France ought to have been taken into account, as if they had been made in Germany. The cjeu disagreed with this; such equivalence was not required within the circumstances of this particular case. 5.7

A Walzerian Analysis

A Walzerian analysis of the cases described in the previous section will now be attempted. With regard to Fossi,59 Walzer might wonder what the relevant distributive community would say was the meaning of the benefit at issue. The Federal Republic of Germany had assumed certain of the obligations incumbent upon the mineworkers’ social security institutions in existence before 1945, and had amended the ‘Imperial Law’ concerned.60 The purpose of this amendment, as stated by the Government before the Court, was ‘to alleviate certain situations which arose out of events connected with the National Socialist regime and the Second World War.’61 Now if the distributive community was of the same view as regards the meaning of the law, and the benefits awarded thereunder, it might take the view that an Italian permanently resident in Italy was entitled to receive the pension. The alleviation of hardships caused by the Nazis, if that was indeed the project, would not appear to come to an end on reaching the national border: the Nazis caused hardship in many different countries. If Mr Fossi had a hardship caused by the Nazis which needed alleviating, it would surely need alleviating wherever he was. The presence of elements, like nationality and residence, proper to the sphere of membership, in the sphere of distribution (of the type of pension sought by Mr Fossi) would therefore be tyrannical. Once confined to their own sphere, however, the distribution could take place in the correct manner, and in perfect accordance with the meaning of the goods being distributed. Turning to Tinelli,62 assuming that the ‘meaning’ of the relevant law and the pensions available thereunder, as set out in the German law and quoted above, was also the one endorsed by the distributive community, then Mr Tinelli’s being resident outside the Federal Republic of Germany might indeed 59 60 61 62

Case 79/​76 Fossi (n 53). ibid para 2. ibid para 7. Case 144/​78 Tinelli (n 54).

Nationality Discrimination

211

have a bearing on whether or not he should receive the pension. If the pension rewarded those who had contributed by their work to reconstruction in the Federal Republic, then those who had spent their working lives in Italy (as Mr Tinelli had done) would be unlikely to qualify. In other words, residence in this case formed part of the shared meaning of the good at issue; its presence in the distributive sphere was thus appropriate, not tyrannical. A Walzerian enquiry, so it seems, would exonerate Belgium exactly as the cjeu had done, but arguably on better grounds than simply dodging the issue as an exclusive national competence, despite the clear impact of these national laws on eu migrant workers. In relation to Even,63 while again not pre-​empting the distributive community, it seems likely that ‘being Belgian’ would form part of its shared understanding of the pension here. On a strict reading of the relevant law, as quoted above, nationality of Belgium and, perhaps more interestingly, loyalty to Belgium would seem to be absolutely central to this particular good. Mr Even’s having fought for France would take him outside the range of potential distributees, and, from a Walzerian point of view quite legitimately, outside the distributive sphere.64 Falling to be analysed last of all is Vigier.65 For a complex egalitarian, the first line of enquiry would be the meaning of the distribuend, as shared by the members of the distributive community. It is hard to anticipate what might happen here. If the distributive community includes all of the victims of Nazi persecution, it seems odd in the extreme that they should decide that those victims who now live outside Germany (and who therefore find it hard to make the requisite payment) must be regarded as having foregone their compensatory rights within Germany. Victims of persecution very often live outside the country where that persecution took place, the persecution itself having driven them away. On this analysis, merely possessing the status of victim should be enough to guarantee someone in the position of Ms Vigier the right to participate in the distribution, even from France. On the other hand, if this good is less an act of compensation for an earlier wrong, and more a 63 64

65

Case 207/​78 Even (n 56). But contrast Case 9/​78 Directeur régional de la Sécurité sociale de Nancy v Paulin Gillard and Caisse régionale d’assurance maladie du Nord-​Est, Nancy [1978] ecr 1661. Here, a Belgian ex-​POW was refused a special old age pension by the French authorities. Without usurping the role of the distributive community, it seems likely that a Walzerian analysis would produce a positive result for Mr Gillard, as the French law in question talked about a claimant’s wartime service ‘in the French or Allied Forces’ (ibid para 5, emphasis added). Case 70/​80 Vigier (n 25).

212 Chapter 5 simple return on an earlier investment, then it is the payment of the contribution, rather than the status of victim, which should be determinative; those outside the country and therefore less able to contribute can legitimately be excluded. It can be seen that the meaning of the distribuend in Vigier is very ambiguous. Walzer’s method, which allocates the ‘casting vote,’ as it were, to the distributive community itself, provides a straight path to a resolution of the problem.66 Whatever the distributive community says goes. An Aristotelian analysis of the case, meanwhile, simply exacerbates the earlier ambiguity. It seems logical that someone who manages to pay the stipulated contribution (henceforth a ‘payer’) is more likely to be a German resident, and therefore more likely to be a German national, than a non-​payer, who is more likely to be a resident of another Member State, and therefore more likely to be of a non-​German nationality. It is submitted, then, that the rule set out in Germany’s law concerning victims of persecution is indirectly discriminatory against non-​Germans (with no obvious justification), breaching if nothing else Article 18 tfeu (then Article 12 ec). In other words, it is a case of unlikes being treated alike: Germans and non-​Germans differ as to the ease with which they can find out about, get information on, and finally make the necessary contribution, but both are made subject to the same rule –​a classic example of the second limb of the Aristotelian test in action, one might have thought. However, the cjeu nicely illustrates how unforeseeable the test is by declaring there to have been no discrimination at all. The Court relied on an earlier case as authority for the proposition that a condition of ‘prior affiliation’ was acceptable, and that this did not mean that the Member State making the condition was compelled ‘to treat as equivalent insurance periods completed in another Member State.’67 The two groups of potential payers remain un-​alike, and their treatment remains like, but neither the general principle of equality, nor any of its specific enunciations within eu legislation, would seem to have been breached.68

66 67 68

This is the orthodox position, anyway. For an alternative version of the theory, mediated complexity, in which the casting vote is given to the Judge, see below Chapter 8. Case 70/​80 Vigier (n 25) para 19. Contrast Case 237/​78 Palermo née Toia (n 40) where mothers of French nationality and mothers of non-​French nationality faced the same rule, even though the mothers of French nationality were always going to have an easier time fulfilling it (that is, bearing French children) than their non-​French counterparts. Here, the Court had no difficulty in finding the rule to be indirectly discriminatory.

Nationality Discrimination

5.8

213

The Citizenship Case-​Law: Early Days

In Chapter 2, it was explained how even Walzer himself understood the need for an Override in cases where complex equality inadvertently sanctioned, via its particularism, disagreeable attitudes or practices. He called it the Minimal Morality. Later in the same chapter it was suggested that, for eu purposes, the echr could fulfill this role, especially seeing as it represented a consensus among all the Member States with regard to human rights and fundamental freedoms. In the cjeu’s more recent case-​law on the subject of the free movement of persons, one can detect the emergence of a not dissimilar mechanism in eu law, even in its current, simple egalitarian form, as will be discussed in this and the next four sections. A good example comes from Advocate General Jacobs in the case of Konstantinidis.69 Here, a Greek national who established himself as a masseur in Germany complained when he was forced to accept an ‘insulting [and] unpronounceable’70 transliteration of his name on official documents; the German court made a reference. Jacobs was of the opinion not only that this state of affairs constituted discrimination on grounds of nationality,71 but also that it could be impugned, if need be, as an indistinctly applicable restriction on the freedom of establishment.72 Reaching this last conclusion, he stated: [A Community national invoking the free movement provisions] is […] entitled to assume that, wherever he goes to earn his living in the European Community, he will be treated in accordance with a common code of fundamental values, in particular those laid down in the European Convention on Human Rights. In other words, he is entitled to say ‘civis europeus sum’ and to invoke that status in order to oppose any violation of his fundamental rights.73 69 70 71

72

73

Case C-​168/​91 Christos Konstantinidis v Stadt Altensteig [1993] ecr I-​1191. ibid, Opinion of AG Jacobs, para 12. See ibid, Opinion of AG Jacobs, para 26: ‘The entries made in official registers […] are of such obvious importance that the migrant worker should be entitled to demand that he, like any citizen of the host country, be properly identified in those documents and have his name written in a manner that is not insulting and offensive to him’. See further para 27: ‘[A]‌s regards entries in official registers he is entitled to the same treatment as German nationals’. Following ‘Cassis de Dijon’ [Case 120/​78 Rewe-​Zentral ag v Bundesmonopolverwaltung für Branntwein [1979] ecr  649] in relation to goods and Case 76/​90 Manfred Säger v Dennemeyer & Co Ltd. [1991] ecr I-​4221 in relation to services. Case C-​168/​91 Konstantinidis (n 69) Opinion of AG Jacobs, para 46.

214 Chapter 5 A person exercising his right to freedom of establishment ‘should in general have to comply with the local legislation;’74 this is reminiscent of Walzer’s insistence that all activity within a given sphere should be regulated by the rules intrinsic to that sphere. It is nicely captured by the saying, ‘When in Rome, do as the Romans do’. But the Advocate General questions ‘whether … a disproportionate restriction or one entirely devoid of justification could be applied against a national of another Member State.’75 In other words, regardless of whether or not it is the custom of the host State and regardless of whether or not all the nationals of the host State have to do it as well, if a practice beaches the ‘common code of fundamental values,’ the migrant may lawfully object to it. In the case in hand, of course, it was only Mr Konstantinidis who was being denied the right to be identified properly. By declaring this to be a restriction on free movement, the Advocate General, and indeed the Court,76 brought the treatment of Mr Konstantinidis into line with that of ordinary German nationals whose names were spelt using the Roman alphabet and thus not amended. Community law granted him the equality which national law had withheld. This case suggests that a move to the Walzerian system of equality would not be as big a shock for the Court as might have been thought. The logic followed by Jacobs here (local meaning in the first instance, with Override in reserve) is almost identical to that which a complex egalitarian might have used. Some years later, Advocate General Jacobs adopted a similar technique in the case of Bickel and Franz.77 Here, two defendants in criminal proceedings in Bolzano (Italy), both from German-​speaking countries, requested that their trials be carried out in German. Owing to the presence of the German-​speaking minority in the Trentino-​Alto Adige Region, there were laws in place allowing judicial proceedings to be conducted in German in courts in that region. Therefore, any locals who had made this request would have been accommodated. The Bolzano court asked the cjeu whether it could refuse the two men’s requests, or whether such a refusal would constitute discrimination on grounds of nationality. While considering whether the subject-​matters of the two men’s trials were within the scope of the Treaty, Advocate General Jacobs made use of the 74 75 76

77

ibid, Opinion of AG Jacobs, para 48. ibid. In a short judgment, the Court held that Germany’s behaviour could be a violation of the right of establishment, provided that its effect (on Mr Konstantinidis) was felt when he was pursuing his occupation, for example, if potential clients were confusing him with other persons. Case C-​274/​96 Criminal proceedings against Horst Otto Bickel and Ulrich Franz [1998] ecr I-​7637.

Nationality Discrimination

215

(at that time) new citizenship provision at Article 18 (ex 8a) ec, now Article 21 tfeu. Of this provision, he commented: The notion of citizenship of the Union implies a commonality of rights and obligations uniting Union citizens by a common bond transcending Member State nationality. The introduction of that notion was largely inspired by the concern to bring the Union closer to its citizens and to give expression to its character as more than a purely economic union.78 This transcendent ‘commonality of rights,’ like the earlier ‘common code of fundamental values,’ is not dissimilar to the Override, and Jacobs seems to be using it in a similar way, that is, as a backstop, to be deployed if all else fails. But not everyone has been quite so optimistic about the advent of eu citizenship. More, for example, wrote that ‘it lacks connection to a set of common or shared values and, moreover, to a proper democratic process’.79 It was indeed hard to see what the ‘common bond’ was, especially in cases like Sala from the same year,80 where the good at issue was not the right to use one’s own surname or one’s own language in the host State, but to spend the host State’s taxpayers’ money in the form of a child-​raising allowance. Jacobs was probably on safer ground linking his ‘Minimal Morality’ to the echr, as he did in Konstantinidis. In the event, the Advocate General found that there was nationality discrimination in Bickel and Franz, and did not accept any of Italy’s proposed justifications for treating the two foreigners differently to the local population. The Court agreed with this conclusion. 5.9

A Walzerian Analysis

With regard to Konstantinidis, it is submitted that Advocate General Jacobs ­utilized the European Convention on Human Rights as a sort of Override providing eu migrants, who would otherwise be compliant with the local norms of the country to which they had moved, with recourse to a common code of values,81 78 79 80 81

ibid, Opinion of AG Jacobs, para 23. G More, ‘The Principle of Equal Treatment: From Market Unifier to Fundamental Right?’ in P Craig and G de Búrca (eds), The Evolution of eu Law (oup 1999) 539. Case C-​85/​96 María Martínez Sala v Freistaat Bayern [1998] ecr I-​2691. Although clearly not entirely common if the migrant is finding one of the values in the code to be underprotected, or not protected at all, in the Host State.

216 Chapter 5 protected by eu Law and thus supreme over those local norms, where they were subjected to treatment in contravention of one of the values stated therein. However, the case is interesting as it is not clear whether Mr Konstantinidis was alone outside the relevant sphere here, or whether he was or at least could have been joined by other Germans who, for example, wanted to change their name to one which utilized Cyrillic characters. If that was the case, then the outcome would depend not on whether or not German society considered that registration of one’s civil status included the necessity to be German, but on whether or not German society considered that such registration included the right to be named, spelling and all, as one would wish. It could be anticipated that either it would turn out, following judicial inquiry, that it did, or that, if it did not, the Walzerian judge would invoke the Override to impose full observance of the right to a name on the Germans, like it or not, exactly as Jacobs AG had done. The Court’s more Treaty-​bound response to the facts of the case puts Mr Konstantinidis’ migrant worker status, with its right not to be obstructed whatever the treatment of locals, front and centre. Like in cases concerning the free movement of goods, then, the Court clearly felt itself able to move ‘beyond ­discrimination’ into an entrave or obstacle-​based approach, whereby violations of the Treaty’s free movement provisions could be found even where treatment of the migrant and the Member State nationals was, at least in law, identical.82 It is further submitted that the question posed by the Advocate General (namely, ‘is the right to a name violated where an individual is forced to have their name transcribed in Roman characters against their will?’) is superior to that posed by the Court (namely, ‘is a Greek masseur put at a disadvantage compared to German masseurs if he is forced to have his name transcribed in Roman characters against his will?’) The case is thus a nice example of how the Walzerian approach, exemplified here by the Advocate General’s Opinion, cuts to the heart of a problem, unlike the Aristotelian approach which would in this case simply perpetuate the German rule,83 violation or not (the ‘Equal Misery’ 82

83

As in the previous chapter, G Davies, Nationality Discrimination in the European Internal Market (Kluwer 2003), especially 53ff, is recommended as a good starting point for further research into the obstacle-​or restriction-​based approach. Oddly, in this case, the transliteration rule vis-​à-​vis registration of civil status was technically an international one, not a national one, and in the judgment it is stated that both Germany and Greece were parties to the Convention on the Representation of Names and Surnames in Registers of Civil Status of 13 September 1973: Case C-​168/​91 Konstantinidis (n 69) para 5. However, this does not mean that Mr Konstantinidis did not move from one sphere to another. He did. He simply moved from the sphere where that was interpreted

Nationality Discrimination

217

argument again), or be abandoned altogether in favour of an obstacle-​based approach,84 as in fact happened. In Bickel and Franz, the boundary breach is stark. Mr Franz’s ‘being German’ and Mr Bickel’s ‘being Austrian’ are almost certainly negative dominants, subject to whatever the members of the distributive community decide is the shared meaning of the right to ask for German to be used in judicial proceedings in Bolzano. To justify denying the right to foreigners, they would have to resolve that the meaning of the right was the protection of their minority (the German-​speakers of Bolzano) and their minority only. But this would be perverse and would surely require application of the Override. A right to ask for German to be used in judicial proceedings is to help German-​speaking defendants, to save money (on translation and interpretation), and generally to ensure the smooth-​running of the trial and thus to aid justice. Any other consideration, such as local loyalty, would without doubt be regarded by a complex egalitarian as extraneous.85 5.10

The Citizenship Case-​Law Evolves

The Court’s ‘new’ technique for widening the scope rationae personae of Community law, namely via an expansive reading of Article 21 (then Article 18), was set to continue and indeed to become its pre-​eminent method for dealing with free movement of persons cases. An example is Grzelczyk.86 The case concerned a French national who went to study at a university in Belgium. For the first three years he supported himself by working, but in the fourth year he

84 85 86

to mean that Roman characters should be transposed into Greek ones, into the sphere where that was interpreted to mean that Greek characters should be transposed into Roman ones, and his custom of spelling his name in Greek, or at least transliterating it the ‘Greek way,’ went from a positive dominant to a negative dominant as he did so. Quaere whether the representatives of the States Party to the Convention considered the impact of the agreed-​upon rule on their nationals when the latter travelled, or only the expediency that the rule allowed for their domestic civil services. A Walzerian Forum, constituted at international level and made up of all of the people affected by the distribution, would of course consider this. The obstacle-​based approach, it is admitted, also shares some genetic material with Walzer’s test. Note how Bickel and Franz is the reverse of a case like Case 207/​78 Even (n 56), where local loyalties were intrinsic to the distributive sphere. Case C-​184/​99 Rudy Grzelczyk v Centre public d’aide sociale d’Ottignies-​Louvain-​la-​Neuve [2001] ecr I-​6193.

218 Chapter 5 applied for Belgium’s minimum subsistence allowance (the ‘minimex’). This application was rejected on the grounds that he was a student, and therefore did not fall within the scope of the appropriate Regulation (Regulation 1612/​ 68);87 it was also established Community law that, while migrant students who fulfilled the requisite conditions could sometimes receive assistance from the host State with regard to their tutorial fees, they could not receive any kind of maintenance grant.88 Following Bickel and Franz, the Court gave a wide interpretation to the citizenship provision (Article 21, then Article 18), so that it encompassed any Member State national who was ‘exercis[ing] [their] right to move and reside freely in another Member State.’89 This of course covered Mr Grzelczyk. Article 18 (ex 12) could thus be utilized to guarantee him equality of treatment vis-​à-​vis students who were native to Belgium (and who received the minimex). The Court acknowledged that this was a departure from the earlier student case-​law, especially Brown,90 but justified this on the grounds that, since that case was decided, a number of developments had occurred in eu law, including the signing of the Maastricht Treaty and the creation of Union citizenship. Nevertheless, the Court still insisted on fulfillment of the two conditions from the Students’ Directive (Directive 93/​96),91 namely, firstly, that the student must have sufficient resources to avoid becoming a burden on the public finances of the host Member State, and, secondly, that the student must be covered by sickness insurance.92 In the slightly later case of D’Hoop, however, the Court accepted that there may have to be a ‘real link’ between migrant and State in order for an allowance to be granted.93 A  ‘real link’ condition was also accepted as legitimate 87

88 89 90 91 92

93

Regulation (eec) No 1612/​68 of the Council of 15 October 1968 on freedom of movement for workers within the Community [1968] oj English special edition:  Series I Chapter 1968(II)/​475. See, for example, Case 293/​83 Françoise Gravier v City of Liège [1985] ecr 593; Case 197/​ 86 Steven Malcolm Brown v The Secretary of State for Scotland [1988] ecr 3205. Case C-​184/​99 Grzelczyk (n 86) para 33. Case 197/​86 Brown (n 88). Council Directive 93/​96/​EEC of 29 October 1993 on the right of residence for students [1993] oj L317/​59. And the host Member State was not allowed to automatically interpret the mere fact of asking for the allowance as proof of non-​fulfilment of the first condition. See Case C-​184/​ 99 Grzelczyk (n 86) paras 42 and 43. Case C-​224/​98 Marie-​Nathalie D’Hoop v Office national de l’emploi [2002] ecr I-​6191. Specifically, between the migrant and the State’s employment market. However, in the Court’s view, the Royal Decree at issue in this case went too far. Note Advocate General Geelhoed’s Walzerian language here; he focuses on the distribution of tideover allowances

Nationality Discrimination

219

in Collins.94 In Bidar,95 the Court endorsed the making of university funding conditional upon the existence of a ‘genuine link’ between the student and the host Member State. 5.11

A Walzerian Analysis

From a Walzerian perspective, the facts of Grzelczyk96 present a real danger of boundary breach, occurring at the level, not of the distributees this time, but of the distributors themselves. Since withholding the minimex from Mr Grzelczyk would to all intents and purposes force him to return to France, one could allege that the Distributor in the sphere of minimum subsistence allowances was attempting to usurp the role of the immigration services, by analogy with the us case of Graham.97 This would of course be an illegitimate crossing of the boundary between the two spheres. Meanwhile, the decisions in D’Hoop, Collins and Bidar, from the Walzerian point of view, seem to represent a recognition that what is being distributed is not just subsistence, but the plaintiff’s future in the host State. Taking the Standard Contingent Reply as read, need and the nature of the plaintiff’s past relationship with the host State would seem to be much more appropriate distributive principles, than simply a combination of nationality and student-​status. 5.12

Interim Conclusion

Looking at the citizenship case-​law from 1993 to 2005, then, one does not necessarily have to use counterfactual analysis to evaluate the Court’s performance, as this performance can be viewed from both an Aristotelian

94 95 96 97

and notes that there is a conflict between the distributive criterion and the good being distributed: ‘[T]‌he refusal to allow nationals access to the programmes purely because they completed their education in another Member State appears problematic … [This refusal] does not to my mind square with the declared objective’. See ibid, Opinion of AG Geelhoed, paras 45–​56. Case C-​138/​02 Brian Francis Collins v Secretary of State for Work and Pensions [2004] ecr I-​2703. Case C-​209/​03 The Queen, on the application of Dany Bidar v London Borough of Ealing and Secretary of State for Education and Skills [2005] ecr I-​2119. Case C-​184/​99 Grzelczyk (n 86). See supra, text accompanying n 30.

220 Chapter 5 perspective and, putatively, the beginnings of a Walzerian perspective. From the Aristotelian perspective, then, it can be seen that, at least up until 2005, the Court was happy, subject only to the establishment of a real or genuine link, to equate the non-​working eu migrant, qua eu citizen, with the non-​ working national. Meanwhile, the suggested Walzerian perspective reveals two facts: Firstly, that the Court seemed to have discovered the universalising effect of the citizenship provisions, which at least one Advocate General tied to the equally universalising effect of the echr, to override the hostility often shown by local Distributors to eu migrants. Secondly, that even when the Court moved to restrict its earlier discovery, it arguably did so in a way which respected the tenets of complex equality, meaning that distributions might now be deemed just without use, or exclusive use, of the citizenship ‘Override’. 5.13

The Citizenship Case-​Law Since 2005

Martin is right to comment that, in time, the new concept of the free movement of citizens ‘swallowed up’ the old concept of the free movement of workers, but that as a result it became increasingly difficult to make any sense of the latter.98 One example he gives is the 2009 case of Vatsouras,99 where a certain German social assistance benefit was granted to illegal immigrants (from non-​eu States) but not to eu migrant workers, such as Mr Vatsouras, who had in the meantime lost their jobs. Challenged before the Court of Justice, the Court refused to apply the prohibition on nationality discrimination to a situation where an eu citizen was being compared to a third country national, even though it had applied the prohibition in that way on at least one previous ­occasion.100 If Union citizenship was indeed destined to be the ‘fundamental status of nationals of the Member States,’ as the Court had asserted in

98 D Martin, ‘La libre circulation des personnes: au-​delà de l’évolution et des révolutions, la perpétuelle quête de sens’ [2012] 1 rae 85, 88, translation author’s own. 99 Joined cases C-​22/​08 and C-​23/​08 Athanasios Vatsouras and Josif Koupatantze v Arbeitsgemeinschaft (Arge) Nürnberg 900 [2009] ecr I-​4585. 100 Case C-​176/​96 Jyri Lehtonen and Castors Canada Dry Namur-​Braine asbl v Fédération royale belge des sociétés de basket-​ball asbl (frbsb) [2000] ecr I-​2681 (basketball players from other Member States compared with those from non-​e u countries). For more on this point, see D Martin, ‘L’interdiction de discriminer en raison de la nationalité: quelques arrêts troublants’ [2009–​2010] 2 rae 237.

Nationality Discrimination

221

Grzelczyk,101 how could their being treated less favourably than third country immigrants possibly be explained? The citizenship case-​law reached a major milestone in 2011 with the handing down of the Zambrano judgment.102 In this judgment, the third country national father of a Belgian child was granted both a right of residence in Belgium and a Belgian work permit, solely on the basis of the child’s citizenship rights under the Treaty. According to the Court, to expel Mr Zambrano would be a ‘depriv[ation] … of the genuine enjoyment of the substance of [the child’s] rights,’103 and this was despite the fact that the child had neither moved to a different Member State, nor (obviously) undertaken any work. But where the earlier case of Sala104 had been an instance of ‘radical equality,’105 the Court equating an economically inactive eu migrant with a Member State national worker, in Zambrano no equation was being carried out at all. That was because there was only one party before the Court: Mr Zambrano’s child, who was an economically inactive Member State national. The right to non-​discrimination on grounds of nationality, whether in its Treaty or secondary legislative manifestations, or as an unwritten constituent right of eu citizenship, is necessarily predicated on the idea that the right-​holder would have travelled to a different Member State, wherein the discrimination might arise in the first place.106 This strange turn in the Court’s logic towards the Purely Internal Situation results in the abandonment of the equality principle altogether. From these two cases, it might be suggested that the equation of eu citizen to (non-​working) national from the end of the last section was under threat. In the first one, far from being level with the non-​working national, the eu citizen found himself lower down the hierarchy even than a third country migrant, and an illegal one at that. In the second one, the plaintiff, this time himself a third country national, fared better than the eu migrant plaintiff had done in the first one, but this was due to his being the father of an eu citizen who had neither moved, nor worked, and whom the Court had equated with noone at all. 101 102

Case C-​184/​99 Grzelczyk (n 86) para 31. Case C-​34/​09 Gerardo Ruiz Zambrano v Office national de l’emploi (onem) [2011] ecr I-​1177. 103 ibid para 42. 104 Case C-​85/​96 Sala (n 80). 105 C Tomuschat, ‘Annotation Martínez Sala’ (2000) 37 CMLRev 449, 449 and 457. 106 Or, perhaps, in their own Member State when they got back, as happened in the aforementioned Case C-​224/​98 D’Hoop (n 93).

222 Chapter 5 The suggestion from the previous paragraph seems to be proved right when one turns to the Court’s decisions in Dano,107 Alimanovic108 and García-​ Nieto,109 sequels of sorts to Vatsouras, which were handed down in 2014, 2015 and 2016 respectively. Ms Dano was a Romanian national who had gone to live with her sister in Germany; she also had a son, Florin. Ms Dano applied for the same social assistance benefit as Mr Vatsouras had done, although, unlike Mr Vatsouras, she had not worked or even looked for a job. When the award of this benefit was rejected, Ms Dano appealed, and a reference was made to the cjeu. The Court undertook a careful analysis of Ms Dano’s status under Directive 2004/​38, the composite Directive encompassing all of the rules on the free movement of workers, students, the economically inactive, and retirees, and repealing all of the separate instruments which had previously contained them.110 One of the rules from the old Students’ Directive, mentioned above in relation to Grzelczyk, and equally applicable to the other categories of non-​worker, namely, the economically inactive and retirees, was of course that that the individual concerned must have sufficient resources to avoid becoming a burden on the public finances of the host Member State, and they must be covered by sickness insurance. The Court held that Ms Dano, as someone who was economically inactive, was required to fulfil this rule, and, as she did not, she did not have any right of residence in Germany under Directive 2004/​38. Thus she could not make use of the principle of equal treatment contained within the Directive,111 and was not entitled to the benefits sought. The judgment is odd in that the Court went back to a traditional ‘free movement of workers’ analysis in examining Ms Dano’s situation, and did not 107 Judgment of 11 November 2014, Elisabeta Dano and Florin Dano v Jobcenter Leipzig, C-​333/​13, ECLI:EU:C:2014:2358. 108 Judgment of 15 September 2015, Jobcenter Berlin Neukölln v Nazifa Alimanovic and Others, C-​67/​14,ECLI:EU:C:2015:597. 109 Judgment of 25 February 2016, Vestische Arbeit Jobcenter Kreis Recklinghausen v Jovanna García-​Nieto and Others, C-​299/​14, ECLI:EU:C:2016:114. 110 Directive 2004/​38/​e c of the European Parliament and of the Council of 29 April 2004 on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States amending Regulation (eec) No 1612/​68 and repealing Directives 64/​221/​EEC, 68/​360/​EEC, 72/​194/​EEC, 73/​148/​EEC, 75/​34/​EEC, 75/​35/​EEC, 90/​364/​EEC, 90/​365/​EEC and 93/​96/​EEC [2004] oj L158/​77 (‘Directive 2004/​38’). 111 At ibid Article 24. Nor, presumably, could she make use of the principle of equal treatment available to job-​seekers under Article 45(2) tfeu, since it had been established that she was not looking for work. However, the Court does not say this expressly.

Nationality Discrimination

223

explore, certainly not in any depth, the possibilities which her eu citizenship brought with it (including equal treatment), possibilities which were not only open to workers like Mr Konstantinidis, but students like Mr Grzelczyk, and even accused persons like Mr Bickel and Mr Franz whose only activity in Italy, it seems, was to commit crime. Mr Vatsouras as an ex-​worker was allowed to retain his worker status for six months,112 and the Court even helpfully drew this to the national court’s attention should they wish to add this fact to their analysis,113 but, crucially, the national authorities including the national court had not appreciated that fact, had not dealt with Mr Vatsouras as anything other than a non-​worker, and had not (in the case of the court) asked questions of the cjeu on any other basis than that he was, in fact, a non-​worker. Thus most of the Court’s reply in Vatsouras concerns the situation of a non-​worker, and it is noteworthy that the Court did say that benefits previously off-​limits to job-​ seekers could be available to eu citizens if they satisfied the conditions set out in the citizenship case-​law, which included the establishment of a ‘real link’ with the host State as mentioned at the end of the previous section, but did not include the requirement of sufficient resources and health insurance.114 The Dano judgment therefore seems to show once again the Court of Justice’s inconsistency in discrimination cases, as the Luxembourg judges closed off an entire avenue of potential redress for Ms Dano which they had been prepared to open up for others who, like her, did not satisfy the residency rules of Directive 2004/​38. The result was that, with her citizenship status ignored, Ms Dano –​a non-​working eu migrant –​was not equal to a non-​working national. Similarly short shrift was given to Ms Alimanovic, a Swedish national with three children, also Swedish, resident in Germany, and claiming the same benefit as Mr Vatsouras and Ms Dano. The particular situation which Ms Alimanovic and her eldest daughter, Sonita, found themselves in was that of ex-​workers, and like Mr Vatsouras and his co-​claimant, their employment in Germany had been for less than one year, giving them the right to retain their worker status for six months after this employment had ended, although this fact was of little use to the two of them as this time period had elapsed by the time of the events in contention. The matter at issue was that, although both Directive 2004/​38 and the newly-​ modernised Social Security Regulation (Regulation 883/​ 2004)115 1 12 113 114 115

Pursuant to Directive 2004/​38 (n 110) Article 7(3)(c). Joined cases C-​22/​08 and C-​23/​08 Vatsouras (n 99) paras 23–​32. ibid paras 37–​40. Regulation (ec) No 883/​2004 of the European Parliament and of the Council of 29 April 2004 on the coordination of social security systems [2004] oj L 166/​1.

224 Chapter 5 contained rights to equal treatment with host State nationals, this did not extend, thanks to Article 24(2) of the Directive, to ‘entitlement to social assistance’ where those whose right of residence in the host State arose solely out of the search for employment (and their family members) were concerned. As the German benefits at stake had already been identified as being social assistance rather than social security, and as Ms Alimanovic and her daughter were now at best job-​seekers, they fell within the excluded category and could not claim the benefits at issue. But, unlike job-​seekers merely ‘trying their luck’ in another Member State, to whom Member States were particularly keen to deny social assistance, those who had already worked in the host state (long enough to deserve a retained worker-​status, but not long enough for that status not to be capped at six months) were surely entitled to an individual assessment to see whether the links which they had made during their longer stay in said State might have crystallized, via eu citizenship status, into entitlement to the help usually reserved for nationals, eu workers, eu self-​employed persons, eu ex-​workers who had retained their worker status indefinitely,116 and family members of same. Such was Advocate General Wathelet’s argument, anyway. He was prepared to discuss the new citizenship rights which Ms Alimonovic and Sonita (should have) enjoyed, opining that they were entitled to the same opportunities to demonstrate a ‘genuine link’ with Germany which Mr Bidar had had in relation to the uk.117 While the ‘social assistance’ at stake in Mr Bidar’s case, students’ maintenance costs, was admittedly different to that in Alimanovic, the right, as citizens of the eu, to claim equal treatment to nationals of the host State in the same situation was surely identical, and the only prerequisite which the Court had imposed on citizens wishing to invoke this right was the demonstration of a ‘genuine link.’ However, the Court did not see things the same way, and, having found Ms Alimanovic and her family’s right of residence (other than one based on job seeking) to have been lost, proceeded immediately to the conclusion that the conditions for the derogation from the principle of equal treatment found in Article 24(2) were met. Further, an individual assessment was not needed in their case, the Directive’s global assessment of ‘various factors characterising the individual situation of each applicant for social assistance’ apparently sufficing.118 As Hailbronner had put it in 2005,

1 16 Pursuant to Directive 2004/​38 (n 110) Article 7(3)(b). 117 Case C-​209/​03 Bidar (n 95). 118 Judgment in Case C-​67/​14 Alimanovic (n 108) para 60.

Nationality Discrimination

225

Article 24 unequivocally excludes job-​seekers from social assistance for the first three months of residence or where appropriate for a longer period of job-​seeking. No exception is made for a genuine link to the employment market.119 The trilogy was completed by García-​Nieto, which concerned Ms García-​Nieto, a Spanish national who had moved to Germany in April 2012 and found work, her partner, Mr Peña Cuevas (also a Spanish national), who joined her in Germany in June 2012, and their children. When Mr Peña Cuevas applied to the Employment Centre for subsistence benefits in July 2012, he was refused on account of the fact that, at the time of his application, he had resided in Germany for less than three months and did not have the status of a worker or a self-​employed person. The Court, as in Alimanovic, centred its judgment on an investigation into whether the claimant was lawfully resident as a preliminary to deciding whether or not they could plead the principle of equal treatment laid down in Article 24(1) of the Directive. First of all, then, the Court found that Mr Peña Cuevas could base a claim of lawful residence on Article 6(1). In many ways he was the very eu citizen for whom Article 6 –​ entitled Right of residence for up to three months –​was written. However, as the Court pointed out, this right of residence was conditional on the fact that the migrant in question did not become an unreasonable burden on the social assistance system of the host Member State (Article 14(1)), and, as already discussed above, the host Member State was not obliged to provide any social assistance to him or her at all (Article 24(2)). Mr Peña Cuevas, then, though lawfully resident, could not invoke the principle of equal treatment in Article 24(1), as he was one of those excluded from its ambit by Article 24(2). Again the Court declined to perform an individual assessment, relying on the same reasoning as it had used in Alimanovic. Mr Peña Cuevas, like Ms Alimanovic before him, both of them non-​working eu migrants, could not be equated with a non-​working national of the host State, Germany. The problem with all three of these cases is that, in closing the door to social assistance on job-​seekers, the Directive strangely chooses to define its excluded category by implication, using a sort of expressio rule,120 as follows: ‘persons 119 K Hailbronner, ‘Union citizenship and access to social benefits’ (2005) 42 CML Rev 1245, 1263 (emphasis added). This quotation was used by Wathelet AG in his Opinion of 4 June 2015, Vestische Arbeit Jobcenter Kreis Recklinghausen v Jovanna García-​Nieto and Others, C-​299/​14, ECLI:EU:C:2015:366 (footnote 33). 120 From the Latin tag Expressio unius est exclusio alterius (‘the express mention of one thing excludes all others’).

226 Chapter 5 other than workers, self-​employed persons, persons who retain such status and members of their families.’ Inevitably this shuts out individuals in a number of different situations, some of them perhaps intended to be subject to the ban,121 some perhaps not. Thus the Court has been faced with a sequence of cases in which these miscellaneous folk, denied the blanket protection of eu citizenship, have had to be inspected to see whether they are, in fact, in or out of the Directive. They have all been found to be out. It is perhaps the starkest example yet, played out over several judgments, of the manipulation inherent in the Aristotelian test, where the decision as to who is a ‘like’ and who is an ‘unlike’ is entirely dependent upon whether the judges consciously choose whether to take citizenship into account, or not. Had it remained true to the concept of eu citizenship, it might have been able to maintain the ‘Walzerian perspective’ detected at the end of the last section. By deploying it only arbitrarily, however, the Court has returned to the hit-​and-​miss myopia of simple equality. 5.14

A Walzerian Analysis

A pure Walzerian approach to the cases in the foregoing section would obviously turn on the shared meaning of social assistance/​subsistence benefits, rights of residence and work permits. However, because these things are distributed at national level, it would be the host Member States’ view of these goods, not the whole of the eu’s, which would count. And whether the nationality of the potential distribuend did or did not form part of that meaning would itself turn on one thing: solidarity. It is to this vital principle, therefore, that attention must turn. 5.15 Solidarity 5.15.1 What is Transnational Solidarity? Solidarity has been described as the ‘cohesion and commonality of purpose of a given group,’122 and solidarity rights have been defined as giving ‘legal

121 It is easy to think that the legislative draughtsperson had someone like Mr Peña Cuevas in mind when designing the ban on social assistance in the first three months of residence. 122 M Ferrera, ‘Towards an “Open” Social Citizenship? The New Boundaries of Welfare in the European Union’ in G de Búrca (ed), eu law and the welfare state: in search of solidarity (oup 2005) 19.

Nationality Discrimination

227

form to social relations of reliance and trust.’123 The bond of solidarity which a national of a given Member State feels towards his or her fellow nationals may be based on a variety of things, including shared culture, history, tradition, language, or even sporting prowess. As has been hinted earlier,124 solidarity would play a major role when a distributive community came to determine its shared understandings, if the cjeu were to adopt a Walzerian approach to equality. If it was proposed to distribute a certain benefit to a newly arrived migrant worker from another Member State, say, a lot would depend on whether the members of the distributive community considered that the solidarity which already existed between them extended to cover the newcomer. As things stand, the extension or otherwise of solidarity is not a matter for community-​members at all, but is decreed by the cjeu from Luxembourg in its ongoing attempt to integrate Europe by ‘redrawing [national] boundaries:’125 Especially in the field of social assistance (the sanctum sanctorum of national welfare […]), the orientation of the ecj has been very clear: its jurisprudence has tended to restrict the scope of discretionality of the Member States, contrasting their ‘closure’ tactics by appealing not only to the principle of non-​discrimination, but often by referring also to the need to promote transnational solidarity.126 In doing so, the Court is undoubtedly taking its cue from eu founding father Jean Monnet’s own integrationist ideal, described in one biography as simply ‘putting an end to foreignness.’127 Unfortunately, in pursuing its exploits in negative harmonisation, hurriedly establishing an Economic Union which has split European markets wide open, the Court has failed to wait for any legislatures, be they the European one or the individual national ones, to create what Tomuschat called a ‘Social Union,’128 whether as a simple auxiliary to the single market project (market-​ making), or whether as a separate, countervailing project in its own right

123

RM Unger, False Necessity:  Anti-​necessitarian Social Theory in the Service of Radical Democracy (cup 1987) 537, cited in Somek (n 40) 801. 124 Supra n 50 and n 52. 125 Ferrera (n 122) 24. 1 26 ibid 32. 1 27 F Duchêne,  Jean Monnet: The First Statesman of Interdependence (Norton 1994) 202. 128 Tomuschat (n 105) 454.

228 Chapter 5 (market-​correcting).129 Worse, this may not even be possible. As Sangiovanni has put it: Even if member states were to agree that competences over social policy should be transferred to the European level, fundamental conflicts of interest among member states—​produced by different modes of financing, running and organizing the welfare state—​make positive integration in employment, industrial relations and social policy all but impossible.130 The eu thus faces a solidarity crisis, or, as Maduro rightly predicted in 2000, even before the exacerbating interventions of terrorists and bankers, a crisis of social legitimacy.131 It is little wonder that empirical researchers looking at contemporary Europe are struggling to find any solidarity at all. In a fascinating article, Somek has investigated the natures of national solidarity and transnational solidarity, inter alia, to see if there is any connection between them.132 His initial finding is not very encouraging: ‘Far from being an extension of national solidarity and serving as a transmitter, transnational solidarity is clearly in opposition to it’.133 While the basis of national solidarity is fairly clear-​cut,134 Somek finds the root of transnational solidarity to be ‘mysterious.’135 Put another way, the link between the national and his nation (for example, Member State A) is reasonably obvious, but what is the link between the eu migrant and Member State A, be it the ‘real link’ referred to in D’Hoop136 and Collins,137 or the ‘genuine link’ referred to in Bidar?138 The Court is implying in these cases that there is a kind of continuum of linkage, with real or genuine links at one extreme and (presumably) false or bogus links at the other. 129 Opinions vary. For a helpful discussion on this, see J Shaw, J Hunt and C Wallace, Economic and social law of the European Union (Palgrave Macmillan 2007) 344–​5. 130 A Sangiovanni, ‘Solidarity in the European Union’ (2013) 33(2) Oxford Journal of Legal Studies 213, 224–​5. In a footnote to this passage, Sangiovanni cites Scharpf as principal proponent of this view: Fritz Scharpf, Governing in Europe: Effective and Democratic? (oup 1999). 131 See below. 132 Somek (n 40). 133 ibid 805. 134 See the first paragraph of this section. 135 Somek (n 40) 801. 136 Case C-​224/​98 D’Hoop (n 93). 137 Case C-​138/​02 Collins (n 94). 138 Case C-​209/​03 Bidar (n 95).

Nationality Discrimination

229

But what is the nature of the link? What is it based on? Somek rightly notes that the Court has been of very little help here, and that ultimately the link seems to be one of (what he calls) ‘being and time,’139 that is, that if a person is in a certain place for a certain length of time, they will develop a link with that place.140 He comments: This is not much. It is, in a sense, even amazingly meagre. One may wonder, indeed, whether being and time, taken by themselves, do suffice to establish a connection with any meaningful conception of solidarity.141 Sangiovanni, meanwhile, calls transnational solidarity ‘the fair return owed by eu citizens to one another,’142 thus introducing a transactional element to the discussion, which will be considered in the next section. 5.15.2 Consideration and Responsibility Somek has explained how at least one form of solidarity, national or transnational, ‘is mediated by the self-​interested motives of co-​operation for a mutual gain.’143 Where this form of solidarity is prevalent, life will be centred on the concept of transaction. When contemplating distributing a benefit to a migrant, members of a distributive community might employ a ‘consideration argument,’144 reasoning that only those who have paid for a certain benefit should be able to enjoy it.145 This approach to benefit-​distribution reduces all 1 39 Somek (n 40) 807. 140 Thus, the nexus between the migrant and the host Member State develops over time. The most well-​known instance of the Court’s saying that the link is formed by the passage of time is probably in Case C-​209/​03 Bidar (n 95), where it held that it was not illegitimate for a Member State to require of a student ‘a certain degree of integration’ before it would grant him or her financial assistance (ibid para 57), and that this ‘may be regarded as established by a finding that the student in question has resided in the host Member State for a certain length of time’ (ibid para 59). In the Bidar case itself, the Court was prepared to accept three years’ prior residence as a suitable prerequisite. Five years was also found to be acceptable in the later case of Förster: Case C-​158/​07 Jacqueline Förster v Hoofddirectie van de Informatie Beheer Groep [2008] ecr I-​8507. 141 Somek (n 40) 807. 142 Sangiovanni (n 130) 233. 143 Somek (n 40) 812. 144 Davies (n 14) 47. 145 In this section, the phrase ‘paid for’ covers not just direct contribution, but also indirect contribution via the payment of national tax. And see also rca White, ‘Citizenship of the Union, Governance, and Equality’ (2006) 29 Fordham International Law Journal

230 Chapter 5 social activity to an unbroken series of financial bargains –​an unceasing carousel which never lets new passengers on board. And those who do not (cannot) concede the quid may not receive the quo. Echoing Lord Atkin’s famous question ‘Who, then, in law is my neighbour?’146 citizens may decide that they need not love a neighbour who claimed rights while repudiating what they saw as the concomitant duties. As well as the fact that they have not ‘paid for’ the benefit in question, foreigners (especially indigent ones) may also face arguments based on the concept of responsibility. In deciding whether there are any ‘solidaristic’ bonds between them and an impecunious foreign migrant seeking a certain national benefit, the members of a distributive community may argue that this individual is the author of their own misfortune.147 The responsibility argument is thus a kind of worm in the bud of solidarity. Person A can distance him or herself from Person B by examining in ever greater detail the personal choices made by B, and reasoning that B’s plight is a direct result of those choices, for which B and B alone must be responsible; A, convinced that he or she would never make such choices, can no longer see the mutuality which formerly characterized his or her relationship with B, and so the bond of solidarity is ruptured. As Somek puts it, solidarity has ‘switch[ed] into a path-​dependent mode.’148 Path-​dependence is at least not nationality-​dependence; it is quite possible that a distributive community could take this approach towards a national whom it judged to be irresponsible. Similarly, a migrant who had made responsible choices may well find favour. But the fact remains that an applicant like Ms Sala,149 a Spaniard who, after a chequered employment history in Germany, had subsequently lapsed into (potentially) long-​term unemployment, may not endear themselves to a German distributive community looking to dispense child-​raising allowances.150

1 46 147 148

1 49 150

790 at 793, where he comments that, ‘[c]‌itizenship in its classical sense involves duties as well as rights’; he goes on to note that eu citizenship entails the second but not (at the moment) the first. In the Scottish case of Donoghue v Stevenson: [1932] ac 562, 580. ‘Solidaristic’ is Somek’s word: supra n 40, for example at 809. ibid 808. Equally destructive of solidarity is a dependence on what might be called future paths, in other words, risk. In this scenario, those who find themselves in a low-​risk category with regard to a certain matter may try to break with those whose equivalent risk is deemed to be higher. Referring again to Case C-​85/​96 Sala (n 80). A number of theorists, including Trappenberg, Elster and Gutmann, discuss the introduction of a responsibility element into the theory of complex equality: Margo Trappenburg, ‘In Defence of Pure Pluralism:  Two Readings of Walzer’s Spheres of Justice’ (2000) 8(3)

Nationality Discrimination

231

A community which takes an overly transactional view of benefits, and which is unduly concerned with the ‘link’ between benefit and recipient, is more likely to employ responsibility arguments, in parallel with consideration arguments, to query a given recipient’s entitlement. The thrust of the argument would be that the would-​be recipient has not come to the bargain with clean hands, that is, that they have vitiated the bargain via their behaviour. While a consideration argument questions whether there was ever any entitlement at all, then, a responsibility argument works on the basis that what entitlement there was has been lost.151 And just as the simple egalitarian need only treat another the same to the extent of their sameness, so the

Journal of Political Philosophy 343, 346 and 352; J Elster, ‘The Empirical Study of Justice’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) 98; A Gutmann, ‘Justice across the Spheres’ in ibid 99 and 112ff. As Gutmann points out, though, varying distributions in accordance with people’s voluntary behaviour itself leads to a certain degree of inequality/​discrimination: ibid 114. Walzer himself is not opposed to the taking into account of responsibility, although he says it should constitute part of the shared understanding of the good in question, rather than acting as a distributive criterion in its own right: M Walzer, ‘Response’ in ibid 94. This seems to be the correct approach. In the free movement of persons case of Tas-​Hagen, for example, it would be up to the distributive community to decide whether ‘not having chosen to live abroad’ formed part of its shared understanding of benefits for civilian war victims. Did Mr and Mrs Tas-​Hagen’s choice to move abroad indicate some kind of desertion of the national cause, or lapse of loyalty, such that, in making it, they had abandoned their entitlement to compensation for wartime (and immediate post-​war) hardships? Was the choice –​the conscious act of two responsible adults –​a severing of their connection to the Netherlands of such severity that it vitiated any earlier demonstrations of loyalty? The distributees’ responsibility (or lack thereof) for the move should not be the determinative criterion for the granting of the benefit (as it in fact had been), but should feature in the debate as to its meaning. In other words, was blamelessness for any move away from the Netherlands an essential part of how the community understood this particular benefit? See Case C-​192/​05 K Tas-​Hagen and RA Tas v Raadskamer wubo van de Pensioen-​en Uitkeringsraad [2006] ecr I-​10451. 151 Going back to Case C-​85/​96 Sala (n 80), Sangiovanni counterargues that the members of the German community would favour equal treatment for Ms Sala, because ‘long-​term lawful residents who are not economically active also contribute in a meaningful way to the reproduction and maintenance of the state (through taxation, compliance, and civic participation)’. However, it is hard to see how exactly the German state was raising tax revenue from Ms Sala, at least at the time that the case was brought. Thus, while Ms Sala might make an appeal to national solidarity on the strength of the taxation which she paid when she was employed, those members of the German distributive community adopting a responsibility argument might be more concerned with why the payments stopped.

232 Chapter 5 members of a community obsessed with transaction will feel a bond of solidarity only with those they regard as ‘the same;’ as soon as a difference can be identified, those manifesting the difference may be guiltlessly abandoned to their –​ different –​ fate. This last point raises the question as to whether sameness is the right basis for solidarity in the eu. Domurath has commented: Sameness is the most developed rationale of solidarity in the eu, although it is limited as a basis for transnational solidarity in a heterogeneous European community.152 She continues: Since solidarity is a sociological concept involving the feeling-​based ideas of sameness, community, and responsibility, we can question the suitability of a legislative or judicial approach for promoting solidarity with not same individuals, the environment or future generations. In order to advance a European solidarity that is based on plurality, a bottom-​ up approach deriving from acceptance by the European citizens could prove to be more suitable. Such bottom-​up approach can only be based on a societal understanding of a European community, which is, in fact, emerging.153 Ross has suggested that solidarity in the eu should in fact be based on the acceptance of differences,154 and Domurath herself has proposed a European solidarity based on education and shared experience which ‘could help to disconnect solidarity from the expectation of reciprocity and base it on the acceptance of a diverse, plural society.’155 This new conception of solidarity might at least ‘broaden the understanding of sameness,’156 even if it could not replace it altogether. Stjernø, meanwhile, has talked of a need for ‘enlightened self-​interest,’157 noting, in a manner reminiscent of Walzer, that 152 Irina Domurath, ‘The three dimensions of solidarity in the eu legal order: limits of the judicial and legal approach’ (2013) 35(4) Journal of European integration 459, 470. 153 ibid. 154 M Ross, ‘Solidarity  –​a new constitutional paradigm for the eu?’ in M Ross and Y Borgmann-​Prebil (eds), Promoting solidarity in the European Union (oup 2010) 23, 31. 155 Domurath (n 152) 471. 156 ibid. 157 S Stjernø, Solidarity in Europe –​the history of an idea (cup 2005) 201.

Nationality Discrimination

233

[s]‌olidarity has to be constructed socially and politically through the practice of individuals, groups, professional and political organisations, churches, and networks nationally and internationally.158 Davies accuses the Court of Justice of endorsing the consideration-​based approach in its equality case-​law and of ‘misunderstanding of the nature of community.’159 The Court’s ‘misguided philosophy’160 underestimates man’s capacity for benevolence towards his or her fellows, even where it is without reciprocation, or at least where the reciprocation is unseen (or hard to see). A rule (that old age pensioners may swim in the local pool at a reduced rate, for example) may form part of a complicated web of causes and effects, with one thing subsidizing another thing, so that the national may be at fourth or fifth remove from the benefit. But if the Court of Justice, seeing no link, forces the local authority to open the benefit out ‘to all Europe,’161 then the benefit will probably collapse under the weight of oversubscription, and be lost. In 2000, Maduro had pointed out the danger of the Union’s failure to engage in a proper debate on the role and means of distributive justice within the eu, warning that [w]‌ithout such a debate, there can be no true social contract capable of legitimizing the emerging European polity and the consequences would be either a return to a less advanced form of integration … or, if the current model continues to be stretched, a crisis of social legitimacy which may manifest itself in increased national challenges to European policies (whose redistributive effects are not understood and accepted).162 Arguably, such a crisis has now arrived, partly on the back of the financial crash of 2008, with its particularly devastating effects on some Member States of the eu, but also as part of a wider rise in intolerance and xenophobia throughout Europe, if not the world, prompted by global terror threats, wars and mass population displacements since 2001. Indeed the two fears –​for national security and financial stability -​have fed each other in a frightening feedback loop. The 158 ibid 325. 159 Davies (n 14) 48. 160 ibid. The particular case which Davies is criticizing is Case C-​388/​01 Commission v Italy (n 15). 161 ibid. 162 MP Maduro, ‘Europe’s Social Self: “The Sickness unto Death” ’ in Jo Shaw (ed), Social Law and Policy in an Evolving Eu (Hart 2000) 347.

234 Chapter 5 ‘increased national challenges to European policies’ have included Greece’s ‘no’ vote in its austerity referendum of 2015, the Netherlands’ ‘no’ vote in its plebiscite of 2016 on increased eu cooperation with Ukraine, and of course a protracted and noisy threat to quit the entire project  –​eventually carried out –​from the uk. None of these show eu citizens in any mood to love their neighbour, even within their own State borders,163 let alone beyond them. 5.15.3 Asymmetric Solidarity, Pan-​European Solidarity Publics and Walzer’s Position The point is that when the national pays their tax, they do not see clearly the return; they do not see clearly what, if anything, is ‘in it’ for them. And yet they pay anyway. Ferrera calls this ‘ “a-​symmetric” solidarity.’164 Is this a more optimistic vision of solidarity? It has been summed up by Davies as follows: We do not ask how much tax has been paid by their parents before allowing a child free education, or how much has been paid by a person needing healthcare or other services. Indeed, it is a premise of solidarity that there will be discrepancies: that many individuals will receive much from the authorities to which they have not financially contributed, and others will contribute more than they receive. This is justified by an underlying concept of membership, or perhaps of humanity, but it can also be imagined in terms of transactions that are more than material or financial; the primary payment of the community member is perhaps loyalty.165 It is suggested that distributive communities are just as likely to adopt this more positive interpretation of solidarity as they are the ‘meagre’ one proffered by Somek. The glass can also be viewed as half full.166 As Ferrera states, the entry of foreign workers into national sharing spaces will imply new contributions and not only new outlays.167 163 There are exceptions, as when Bulgarians of all ethnicities protested alongside Bulgaria’s Roma population against France’s expulsion of Romanies in 2010 under the banner, ‘We are all Roma’: -​-,​ “Home thoughts” The Economist (London, 25 September 2010) 50, 52. 164 Ferrera (n 122) 31. 165 Davies (n 14) 48 (emphasis added). 166 Banting and Kymlicka have shown empirically that multiculturalism does not threaten a country’s welfare state, and can even make it stronger:  K Banting and W Kymlicka, ‘Multiculturalism and Welfare’ (2003) 50 Dissent 59, throughout but see in particular 65. 167 Ferrera (n 122) 34.

Nationality Discrimination

235

What White calls ‘collective … solidarity,’168 referred to above as the Social Union, need not be a mere pipe-​dream. Already the necessary rights have been put in place, with an entire section dedicated to solidarity in the eu’s Charter of Fundamental Rights.169 The existence of ‘pan-​European solidarity publics’170 and a ‘pan-​European solidarity space’171 is starting to move from the realm of the impossible to the realm of the possible, thanks to many decades of ‘solidarity-​making’172 on the part of the eu and its Court, but also thanks, as Stjernø has pointed out, to the democratising effects of information technology.173 In addition, as Miller notes, ‘economic cooperation takes place increasingly in networks that run across state boundaries’.174 This has provided a further fillip to transnationalism, both hegemonic and counter-​hegemonic, the latter variety ‘shifting both upwards and downwards –​to transnational collectives and to cultural groups below the national level.’175 This double shift will bring with it ‘new kinds of distributive claim’ –​and one could add a new kind or new kinds of solidarity to underpin them (or not) –​thinking about which is still, according to Miller, ‘in its infancy.’176 What he does acknowledge is that particularists like him, and perhaps especially Walzer, have avoided some of the more difficult issues of ‘boundedness’ by relying on the easy in-​out delineation of the nation-​state.177 Others see in Walzer’s ‘drawing a boundary of boundaries around the state’ and in his emphasis on ‘the territorial state as the fundamental unit of political membership’178 a bid to ‘suppress (…) the value of transnational movements’ altogether.179 1 68 169 170 171 172 173 174 175 176 177 178 179

White (n 145) 806. Charter of fundamental rights of the European Union [2000] oj C364/​1, Chapter IV. Ferrera (n 122) 32. ibid. ibid. Stjernø (n 157) 325. D Miller,  Justice for Earthlings: Essays in Political Philosophy (cup 2012) 163. ibid. ibid, 164. ibid, 162. Both quotations from W Connolly, The Ethos of Pluralization (University of Minnesota Press 1995) xxviii and 148. AM Smith, Laclau and Mouffe: The radical democratic imaginary (Routledge 1998) 137. Sutch goes as far as to say that some critics view Walzer as a ‘statist with a “sovereignty fixation” ,’ although he defends him of the charge: P Sutch, ‘Reiterating rights: International society in transition’ in B Haddock and P Sutch (eds), Multiculturalism, identity and rights (Routledge 2003) 215.

236 Chapter 5 Transnational collectives, or movements, are increasingly common in today’s Europe, and indeed solidarity trans-​nationalization is an appealing prospect for many social groups who wish to strengthen cooperation in the field of social policy. Individuals seem prepared to travel large distances to protest alongside those whom they consider likeminded. These micro-​spheres or horizontal loyalties, stretching across national boundaries, are ellipses which straddle the spheres, taking in a few members from each. They are like the ‘point[s]‌of intersection’ identified by Smout in setting out his model of ‘concentric rings’ of identity: ‘shafts –​essentially non-​territorial –​that also define a person’s identity and may or may not reinforce one or more of the territorial identities.’180 However, the criticism of Walzer that he seeks to suppress these movements seems unfair as he has in fact championed transnational movements, which he calls ‘international civil society:’ ‘civic associations … operat[ing] across state borders and recruit[ing] activists and supporters without reference to nationality.’181 He gives Doctors without Borders as an example. He holds international civil society, in a version ‘stronger’ than today, as essential in creating an egalitarian and rights-​respecting global governance. In the same article, he also considers the ‘European Community’ model, but notes, correctly, that the ‘strength of the single cent[re] carries with it the threat of tyranny.’182 Interestingly, he suggests combining the pressurizing centre with ‘activism at … the grassroots:’ ‘Some combination of the two might work fairly well.’183 His most recent comment on the subject has been: The real work of democracy promotion should take place not in the society of states but in what is now called international civil society … I imagine a network of democratic political parties, labo[u]‌r unions, and non-​governmental organisations of many different sorts  –​defending human rights, religious freedom, gender equality  –​all working across borders, sharing resources, helping each other.184

180 TC Smout, ‘Perspectives on the Scottish identity’ [1994] 6 Scottish Affairs 101, 102ff (emphasis added). 181 M Walzer, ‘Governing the globe: what is the best we can do?’ (2000) 47(4) Dissent 44, 47–52. 182 ibid. 183 ibid. 184 M Walzer, ‘Response’ in Y Benbaji and N Sussmann (eds), Reading Walzer (Routledge 2014) 106–​7.

Nationality Discrimination

237

5.15.4 Solidarity and Complex Equality From the foregoing, it may be postulated that solidarity is itself a Walzerian concept. Davies has written that ‘membership is not monolithic, but in accordance with modern ideas about identity and belonging, a multi-​level, layered thing’.185 This is of course in keeping with a Walzerian world, where multiple understandings happily co-​exist. Dougan and Spaventa have declared that ‘the concept of social solidarity is not a constant or given, but dynamic and up for renegotiation’.186 Again this recalls Walzer’s world, where social meanings are continually shifting and where variations are ‘inevitable.’187 Different communities are like spheres, separate one from another, but with a high degree of connectedness as between the members of each. However, the separation is not set in stone and the boundaries are in constant flux. A fellow member of one’s sphere may have been a member of the neighbouring sphere yesterday, or he or she may even become a member of the neighbouring sphere tomorrow. Moreover, connectedness of migrant and State is not proven by a single ‘representative element,’ for example, the Member State in which the ‘migrant’ completed his or her full-​time secondary education. Such a stringent approach risks turning the element in question into a requisite entry-​ token which will in time become dominant, yielding access to a particular distribuend ‘to the exclusion of all other representative elements.’188 There may in fact be multiple ways in which such connectedness can be shown, as the Court seemed to accept, for example, in D’Hoop. It also implicitly accepted this fact in the case of Metock,189 where the Court was considering (not for the first time) the distribution of the right of residence to third country national spouses of eu citizens. It held that a condition of prior lawful residence in another Member State, imposed by Ireland on such spouses, was precluded by eu law. As Costello has nicely put it, the Court ‘support[ed] a vision of residence rights in which origins and belonging in the

1 85 Davies (n 14) 54. 186 M Dougan and E Spaventa, ‘ “ Wish You Weren’t Here …”. New Models of Social Solidarity in the European Union’ in E Spaventa and M Dougan (eds), Social Welfare and eu Law (Hart 2005) 216. And even Somek, following Unger, has commented that solidarity is ‘highly context-​sensitive:’ Somek (n 40) 801. 187 Walzer, Spheres of Justice (n 27) 226. 188 The quotation is from Case C-​224/​98 D’Hoop (n 93)  para 39. The word ‘migrant’ is in inverted commas because, in the D’Hoop case itself, the party claiming the Belgian benefit was in fact Belgian herself. 189 Case C-​127/​08 Blaise Baheten Metock and Others v Minister for Justice, Equality and Law Reform [2008] ecr I-​6241.

238 Chapter 5 eu are decoupled.’190 A distributee whose origin is a certain place can always assert, usually unchallenged, that they belong there, while a distributee whose origin is elsewhere has a much harder time making the same assertion. Origin, then, can be used tyrannically and, in situations where this is happening, the Court is right to begin to ‘decouple’ it from belonging, thus allowing claimants in the future a much broader range of means by which to demonstrate that a certain place is, or should be, their home. As Ferrera puts it, possession of ‘national roots’191 can be employed by potential distributees as a ‘marker of insiderhood,’192 and by the gatekeepers of the distributive sphere in question as a ‘filter for … rights;’193 those with the marker come in, while those without the marker stay out. But thanks to the Court, social rights (and the corresponding obligations) have been decoupled from national citizenship within the eu […]. [Consequently,] the underlying and ultimate filtering function performed by national citizenship qua overall and solid container of rights and basic instrument of closure is no longer there.194 In Walzerian terms, nationality is slowly being stripped of its (potential) dominance, making the distribution altogether fairer for distributees from other Member States. 5.16

Concluding Thoughts

The tentative conclusion must be that the cjeu’s applying Walzerian complex equality to its case-​law on nationality discrimination would be a positive development for eu law. In some cases it would produce fairer and more logical results  –​ Gilly, for example, or some of the cases arising from World War Two. However, even where the result would be the same as that arrived at under an Aristotelian regime, the reasoning is noticeably clearer; Walloon 190 C Costello, ‘Metock: Free Movement and “Normal Family Life” in the Union’ (2009) 46 CML Rev 587, 622. 191 Ferrera (n 122) 16. 192 ibid. 193 ibid. 194 ibid 22. The Court is not alone in pushing for a redefinition of the concept of nationality. See, for example, R Rubio-​Marín, Immigration as a Democratic Challenge (cup 2000) 245.

Nationality Discrimination

239

Waste and Angonese are two examples. This Walzerian reasoning avoids the tortuous intellectual gymnastics to which the Court must sometimes resort at present; such gymnastics lead to awkward, opaque judgments, unintelligible to the average eu citizen, who then feels further alienated from the Union and its law. In the later case-​law one can even detect a slight Walzerian dimension creeping into the Court’s thinking, which should be encouraged. On the other hand, the most recent developments in the citizenship case-​law have stretched the traditional equality principle to breaking point, and put notions of eu solidarity under colossal stress, prompting Domurath for one to comment, ‘it is questionable whether the ecj is the appropriate forum for advancing solidarity’.195 Clearly there is a need to considerably strengthen solidarity within the eu. Solidarity goes hand in hand with complex equality, and indeed the two concepts are related, as mentioned in the previous section. A complex egalitarian approach to free movement cases shines a light on local meanings, and the less solidaristic a community, the more likely it is that nationality, or some other place-​specific criterion, will still play a part in how it understands certain goods. Of course this still leaves the Court with the Override, and, exceptionally, the cjeu may be able to overrule the Member State if it is in the name of preserving co-​existential equilibrium, as described later in the book.196 However, it is submitted that the best way of ensuring that antiquated meanings do not poison distributive criteria is to work for change –​reinterpretation –​at the local level. Burying one’s head in the sand of simple equality, on the other hand, is not a viable option. 1 95 Domurath (n 152) 463. 196 See below, section 9.7. A  further option is use of the alternative theory  –​‘mediated ­complexity’ –​described in Chapter 8, which transplants the role of ascertaining meanings from the community to the Bench itself.

Chapter 6

Semi-​Suspect and Non-​Suspect Grounds 6.1

Introduction and Explanation of Terminology

The doctrine of suspect classifications is American in origin, although it now has a large following outside the us as well. If a law or executive action makes or relies on a classification which has been recognized as ‘suspect,’ then such law or action will be presumptively unconstitutional, will be subject to heightened or ‘strict’ scrutiny on the part of the courts, and is liable, unless a non-​ discriminatory objective can be shown, to be struck down. While it is difficult to pinpoint the exact moment of its birth, the 1944 case of Korematsu v us1 would seem to provide the earliest full statement of the doctrine: It should be noted, to begin with, that all legal restrictions which curtail the civil rights of a single racial group are immediately suspect. That is not to say that all such restrictions are unconstitutional. It is to say that courts must subject them to the most rigid scrutiny. Pressing public necessity may sometimes justify the existence of such restrictions; racial antagonism never can.2 Over the next twenty or so years, the doctrine grew in popularity until it became the Supreme Court’s principle method for dealing with discrimination cases, particularly in such decisions as McLaughlin v Florida3 and Loving v Virginia.4 The cjeu itself has never explicitly said that it is employing a theory of suspect classifications. However, a number of Advocates General

1 Korematsu v United States, 323 us 214 (1944) (reliance on racial classification acceptable in the context of war). 2 ibid 216. Although the concept of strict scrutiny (minus the concept of suspect classifications) can perhaps be traced back further. 3 McLaughlin v Florida, 379 us 184 (1964) (reliance on racial classification invidious). 4 Loving v Virginia, 388 us 1 (1967) (reliance on racial classification invidious). See JM Balkin, ‘Plessy, Brown, and Grutter: A play in three acts’ (2004–​2005) 26 Cardozo L Rev 1689 for a much fuller account of the development of the doctrine. Balkin moots that a foreshadowing of the doctrine can even be detected as far back as 1819: McCulloch v Maryland, 17 us 316. However, his basic thesis is that the system of suspect classifications arose out of the New Deal (1933–​36), which brought with it an increased involvement of the State in private citizens’ lives.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/​9 789004354265_​0 07

Semi-suspect and non-suspect grounds

241

have made mention of the doctrine (for example, Poiares Maduro AG in Coleman,5 and more recently in Arcelor,6 and Mazák AG in Age Concern).7 Furthermore, a number of commentators on eu law have referenced the doctrine. These include McCrudden, whose recent reanalysis of the principle of equality,8 at least as far as concerns the first three of his four categories or approaches, has, in his own words, ‘echoes in the United States Fourteenth Amendment context.’9 Martin believes that the Court is already using a type of strictness model, but calls for it to be tightened up, with both the suspect grounds and the levels of strictness properly classified.10 Koen Lenaerts, a cjeu judge but writing academically, has linked the cjeu’s indirect discrimination jurisprudence to the ‘suspect classification’ approach of the Supreme Court.11 Howard has also called for a version of the doctrine to be applied at the cjeu.12 It is true that the objective justification stage of the Aristotelian test used by the cjeu to some extent mirrors the ‘strict scrutiny’ technique pioneered 5 6

7

8

9 10 11 12

Case C-​303/​06 S Coleman v Attridge Law and Steve Law [2008] ecr I-​5603, Opinion of AG Poiares Maduro, para 7. Case C-​127/​07 Société Arcelor Atlantique et Lorraine and Others v Premier ministre, Ministre de l’Écologie et du Développement durable and Ministre de l’Économie, des Finances et de l’Industrie [2008] ecr I-​9895, Opinion of AG Poiares Maduro, para 31ff. Poiares Maduro avers here that the cjeu does use a suspect classification/​strict scrutiny approach, even if not in name. Case C-​388/​07 Incorporated Trustees of the National Council on Ageing (Age Concern England) v Secretary of State for Business, Enterprise and Regulatory Reform [2009] ecr I-​1569, Opinion of AG Mazák, para 71. To be found in C McCrudden, ‘Theorising European Equality Law’ in C Costello and E  Barry  (eds), Equality in Diversity (Irish Centre for European Law 2003)  and also in C McCrudden, ‘The new concept of equality’ (Paper prepared for the Academy of European Law conference, ‘Fight Against Discrimination:  The Race and Framework Employment Directives,’ 2–​3 June 2003)  accessed 20 May 2018. See also supra, Chapter 2, n 2. C McCrudden and H Kountouros, ‘Human rights and European equality law’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) 74. D Martin, Égalité et non-​discrimination dans la jurisprudence communautaire: étude critique à la lumière d’une approche comparatiste (Bruylant 2006) 587 ff. K Lenaerts, ‘L’égalité de traitement en droit communautaire:  un principe unique aux apparences multiples’ [1991] cde 3, 14. E Howard, ‘The case for a considered hierarchy of discrimination grounds in eu law’ (2006) 13(4) mj 445, 457-​8ff. It is interesting to note her view that, of the Article 19 grounds, only sex, race, religion and sexual orientation should be classified as suspect; disability and age would be non-​suspect.

242 Chapter 6 in the us,13 allowing a discriminator (in theory, anyway) less and less scope to justify themselves, the more questionable the ground upon which they rely. This in turn should lead to a hierarchy of grounds (most strict to less strict, as it were), and some commentators believe they can see the formation of such a hierarchy in the cjeu’s case-​law: Martin and Warnier both insist, for example, that nationality discrimination is scrutinized in a stricter manner than gender discrimination (in other words, those discriminating on gender grounds will have their justifications more readily accepted by the Court than those discriminating on nationality grounds).14 Others regard the Court’s case-​law as too haphazard for a hierarchy of grounds to be properly discerned, certainly when compared with that of the us Supreme Court. Iliopoulou makes the interesting point that altering the degree of strictness also alters the very definition of equality itself.15 Nevertheless, the allusion to the doctrine of suspect classifications in the title of this chapter is not to be taken as a definitive assertion that the cjeu has been using this doctrine in its equality case-​law up till now, or as a call for it to do so in the future.16 Rather it is used simply as a convenient way to 13

14

15 16

Although, according to Costello and Davies, in the gender discrimination context, the cjeu’s scrutiny is strictest where employers’ practices are concerned, and is weaker in the case of domestic legislation, and weaker still in the case of Community legislation! C Costello and G Davies, ‘The case law of the Court of Justice in the field of sex equality since 2000’ (2006) 43 cml Rev 1567, 1588. That ‘employer made’ employment policies are scrutinized in a stricter manner than Member States’ employment policies is confirmed by Schiek: D Schiek, ‘A New Framework on Equal Treatment of Persons in Ec Law?’ (2002) 8(2) Eur LJ 290, 297. Meanwhile, Hilson detects differences in the severity of the Court’s review of laws and practices in the free movement field. He postulates, for example, that the cjeu employs stricter scrutiny in services cases than it does in establishment cases, and that the Court scrutinizes Member States’ laws on, say, insurance ‘more leniently’ than it scrutinizes laws on, say, tourism. See C Hilson, ‘Discrimination in Community free movement law’ (1999) 24(5) EL Rev 445, 461. Martin (n 10) 590; N Warnier, ‘Les discriminations directes and indirectes dans la domaine de l’égalité homme-​femme et de l’égalité nationaux-​non-​nationaux’ (2006) 84(2) Revue de droit international et de droit comparé 225, throughout, but see, for example, 285. A Iliopoulou, ‘Le principe d’égalité et de non-​discrimination’ in J-​B Auby and J Dutheil de La Rochère (eds), Droit administrative européen (Bruylant 2007) 448. Although it would perhaps be a step in the right direction. The doctrine of suspect classifications, via strict scrutiny, enables the Supreme Court to look behind the ostensible purpose of a law in order to tease out the real motive. In Korematsu, for example, the use of a racial criterion in the law in question triggered the need for strict scrutiny, which meant that the Court had to ascertain whether the law (which forbade us citizens of Japanese ancestry from being present in certain locations) was indeed aimed

Semi-suspect and non-suspect grounds

243

distinguish the grounds so far considered  –​gender, race, religion, disability, age, sexual orientation and, in the particular context of the eu, nationality –​ with others which the legislature has not seen fit to single out and which the Court has therefore not regarded as requiring particularly heightened review. While the elderly, or the gay, or nationals residing in another Member State may have been subjected to such a history of purposeful unequal treatment, or relegated to such a position of political powerlessness17 that the legislature and Court see the need to bestow on them ‘extraordinary protection,’18 persons and entities in ‘non-​suspect’ cases find themselves distinguished from other persons or entities on a ground which connotes neither historical subjection, nor political relegation. In fact, the cursory scrutiny given to such distinctions by the American courts is based on the assumption that the distinction in question will turn out to be merely the democratically fair outcome of struggles within the political process.19 As Poiares Maduro AG has put it, [A]‌ll legislative activity entails choices and involves the redistribution of interests:  in principle, although such choices and redistribution inevitably favour certain social and economic categories over others, they do not constitute discrimination and it is for the political process to discuss, define and determine the configuration of such redistribution.20

17 18 19 20

at meeting a ‘pressing public necessity,’ or whether it was merely an exercise in ‘racial antagonism’ –​see the quotation from the case set out in the text accompanying n 2 above. A  Walzerian approach too would involve looking behind the distribution of rights of entry to certain locations, in order to see whether race was or was not part of the shared social meanings of such rights. Both methods, then, entail consideration of a much wider context than is contemplated by the Aristotelian test, which requires only a surface review of the two comparators, without need of further contextualization at all. San Antonio School District v Rodríguez, 411 us 1 (1973) at page 28. ibid. Balkin (n 4) 1715. Case C-​127/​07 Arcelor (n 6) Opinion of AG Poiares Maduro, para 33.

244 Chapter 6 However, just because a distribution proceeds from a ground which is viewed as non-​suspect, does not mean that it is automatically legitimate, or, in Walzerian terms, just; it may still entail a boundary breach, and be in violation of complex equality. Nevertheless, this concept of non-​discriminatory redistributions is important from a Walzerian perspective. The waters are muddied slightly by the fact that the Treaty on the Functioning of the European Union does provide for a handful of other prohibitions on inequality, tailored to certain very specific situations or comparisons. The most important are Article 40(2) tfeu (ex Article 34(2) ec), prohibiting discrimination between producers and between consumers in the context of the Common Agricultural Policy, and Article 106 tfeu (ex Article 86(1) ec), requiring equal treatment as between public and private undertakings.21 Given, on the one hand, that the legislative draughtsperson saw fit to give these situations special mention, but, on the other hand, that the situations are very discreet and that the provisions concerned have to date generated relatively little case-​law, it is proposed to refer to the grounds of comparison outlawed by these two Articles as ‘semi-​suspect.’ They will be dealt with first.

21

Article 55 tfeu (ex Article 294 ec), which requires Member States to accord equal rights to all participants in the capital of companies or firms, is arguably just a rarefication of the prohibition on nationality discrimination in Article 18 tfeu (ex Article 12 ec), and will not be dealt with separately here. Other prohibitions on inequality, or requirements of equality, are to be found in secondary legislation, for example, the rule that all shareholders of a company ‘who are in the same position’ must be treated equally, which started life in Article 42 of Council Directive 77/​91:  Second Council Directive 77/​91/​e ec of 13 December 1976 on coordination of safeguards which, for the protection of the interests of members and others, are required by Member States of companies within the meaning of the second paragraph of Article 58 of the Treaty, in respect of the formation of public limited liability companies and the maintenance and alteration of their capital, with a view to making such safeguards equivalent [1977] oj L26/​1. This Directive has since been replaced twice. The first replacement was Directive 2012/​30/​E U of the European Parliament and of the Council of 25 October 2012 on coordination of safeguards which, for the protection of the interests of members and others, are required by Member States of companies within the meaning of the second paragraph of Article 54 of the Treaty on the Functioning of the European Union, in respect of the formation of public limited liability companies and the maintenance and alteration of their capital, with a view to making such safeguards equivalent [2012] oj L315/​74, with the equal treatment clause at Article 46. The second replacement was Directive (eu) 2017/​1132 of the European Parliament and of the Council of 14 June 2017 relating to certain aspects of company law [2017] oj L169/​46, with the equal treatment clause at Article 85.

Semi-suspect and non-suspect grounds

6.2

245

Semi-​Suspect Grounds

6.2.1 Article 40(2) tfeu: Producers and Consumers Within the context of the Common Agricultural Policy, perhaps the most contentious issue has been the quota system, an attempt to reduce the surplus of various agricultural products, itself brought about by the Community’s system of guaranteed prices at a level well above world market prices, which, depending on the product, is now either ended or coming to an end.22 This ‘capping’ took the form of an additional levy demanded, as a sanction, from all producers who had exceeded their quota (or ‘reference quantity’). Reference quantities for each Member State were fixed annually by the eu; it was then up to the individual Member State to divide the reference quantity between all of the national producers. The quota system has generated case-​law as long as it has existed. It is proposed to look at two cases –​an older one from 1988, and a more recent one from 2000  –​before briefly considering the two so-​called German Banana Cases.23 It is notable that the issues have hardly changed in twenty years. To bring things up to date, a case from the sugar sector, in the context of the eu’s dismantling of the quota system and restructuring of the markets, will be considered last of all. In Erpelding,24 a milk producer was granted an extra reference quantity by his Member State (Luxembourg), but brought an action because it had been calculated by reference to the period 1981–​1983, instead of the period 1975–​1979 as he had requested. He argued that his production had decreased between 1980 and 1983 due to an outbreak of mastitis, and again at the end of 1983 due to Canadian flu; it had since risen again. But a low reference quantity meant that he could produce less milk before incurring the additional levy, which meant lower profits. Was he being treated unfairly by comparison with a producer who had done well between 1981 and 1983? Advocate General Darmon did not think so, and the Court agreed. Darmon foresaw grave consequences if producers were accorded a wide choice of reference year: 22 23

24

Commission, ‘A short history of milk quotas’ (Europa, 22 February 2017)  accessed 20 May 2018. Case C-​280/​93 Federal Republic of Germany v Council of the European Union [1994] ecr I-​4973 –​‘German Banana Case (I)’; Case C-​122/​95 Federal Republic of Germany v Council of the European Union [1998] ecr I-​973 –​‘German Banana Case (II)’. Case 84/​87 Marcel Erpelding v Secrétaire d’État à l’Agriculture et à la Viticulture [1988] ecr 2647.

246 Chapter 6 [T]‌here is a risk that producers would choose a reference year which was not merely unrepresentative but which would ultimately reveal a truly remarkable yield. Given the number of Community producers such a consequence could not be regarded as negligible.25 The Court held that this was an instance of comparable situations being treated in a different manner, but that it was objectively justified by, firstly, the need for legal certainty, and secondly, the need to preserve the effectiveness of the system. Thus, there was no breach of Article 40(3) ec.26 A complex egalitarian, on the other hand, almost certainly would have found a breach. This case is in some ways a classic boundary breach: one producer having his success in, say, 1983 converted into success in the distribution of quotas, and one producer having his failure in 1983 converted into failure in the quota-​distribution. It was surely open to Mr Erpelding to argue that the fact of having done well (or badly) in 1983 was not germane to the sphere of quota-​distribution –​it was a sphere unto itself –​and those whose performance had been poor in 1983 should not have to wear this like a mill-​stone round their necks, the failure iterated and reiterated in all subsequent spheres from then onwards. On the other hand, perhaps success or failure in 1983 was the correct distributive principle, as agreed by all concerned parties. It would depend on what said parties understood to be the meaning of a quota. Was it a prize, to be awarded for good work in the past?27 Or was it part of an effort to safeguard the entire milk-​production industry by reducing surplus? It is logical that a producer’s (future) limit should be in some way related to his (past) capacity. But why should a producer be forced to have his limit calculated on the basis of a reference year which was not a true reflection, for him, and for

25 26 27

ibid, Opinion of AG Darmon, para 13. This was the old, pre-​Amsterdam numbering for Article 34(2) ec, which, since the coming into force of the Treaty of Lisbon, is now Article 40(2) tfeu. Contrast a case like Gascogne Limousin Viandes, where the issue was whether it was a breach of the principle of equal treatment for one Member State to receive an early marketing premium, while another did not. Here, it is quite obvious that the premium rewards effort (in reducing the slaughter-​weight of calves), so that it is just for a Member State which has put in the requisite effort to reap the requisite reward, and vice versa. The breeding of the calves and the distribution of the premium are still two separate spheres, but for the King of the one to rule in the other is not a boundary breach –​it is a legitimate crossing; the calves are part of the meaning of the premium. (The Court was also of the view that there was no discrimination.) See Case C-​56/​99 Gascogne Limousin viandes sa v Office national interprofessionnel des viandes de l’élevage et de l’aviculture (Ofival) [2000] ecr I-​3079.

Semi-suspect and non-suspect grounds

247

genuine reasons, of his past capacity? It is at least arguable that Luxembourg’s mechanical use of the period 1981–​1983 here28 was a boundary breach, and led to the artificial (and quite unnecessary) creation of a dominant. Besides, subject to the decision of the distributive community, the correct fundamentum distributionis would seem to be propensity for good (or bad) performance in the future; past performance is only one element which indicates that.29 In the 2000 case of Kjell Karlsson and others,30 a number of Swedish milk producers were also unhappy with their individual reference quantities. Towards the end of 1995, Sweden had decided to pass a law making a number of changes to the arrangements for the distribution of quotas obtaining up until then. For this purpose, it divided the distributees into four categories: milk producers whose production had not increased between 1991 and 1993, milk producers whose production had increased, so-​called ‘ecological’ milk producers (that is, those producing organic milk), and new producers. In relation to the members of the second category, that is, those who had increased their production, Sweden decided to increase their so-​called ‘own-​risk deduction.’ This meant, for example in the case of Mr Karlsson himself, that a reference quantity of roughly 48,500 kg was to be reduced by 55%, instead of the earlier 25%. A similar increase was imposed on members of the fourth category (new producers), whose quotas were to be reduced by 30% instead of 15%. Meanwhile, the new law made no change at all to the arrangements for producers whose production had stayed the same, and ecological producers even received benefits. Mr Karlsson and two others brought an action, claiming that several aspects of the new law breached the principle of equality in Article 34(2) ec (now Article 40(2) tfeu), in particular the differential treatment as between the second and fourth categories and the first and third categories, and the differential treatment as between the second category and the fourth category themselves. Advocate General Colomer found nothing wrong with Sweden’s law, and any inequalities that there were were objectively justified. The Court seemed to be of exactly the same opinion. 28

29 30

Although effectively dictated by the Community institutions. Note also that the governing Regulation did provide for some situations where alternative reference years could be used, but that none of these availed Mr Erpelding: Commission Regulation (eec) No 1371/​84 of 16 May 1984 laying down detailed rules for the application of the additional levy referred to in Article 5c of Regulation (eec) No 804/​68 [1984] oj L132/​11, Article 3. Erpelding also touches on the important topic of luck. This aspect of the case will be discussed below, at section 6.4. Case C-​292/​97 Kjell Karlsson and Others [2000] ecr I-​2737.

248 Chapter 6 Applying Walzer’s theory to the case, it would be entirely up to the distributive community whether the handling of each type of producer was or was not consistent with the meaning of the milk quota system. Was it just to grant new producers such small quotas, for example? Perhaps this served the purpose of trying to disincentivize newcomers, and perhaps, in an industry with chronic over-​production, that would be a distributive criterion which would be met with approval by the wider community. It must be remembered that, under Walzer’s theory of membership, members of a sphere have the exclusive power to decide who can or cannot join them.31 Meanwhile, was it just to reward the ecological producers? It seems fairly uncontroversial that those who make efforts towards the protection of the environment, for example, by cutting back on the use of pesticides, should be rewarded, but, again, it would be up to the distributive community. Neither of these categories, though, seems to have received treatment out of keeping with the industry under discussion, as understood by those who participate in it. Turning to the producers whose production had remained unchanged over the period 1991–​1993, this category raises the same question as that already discussed in Erpelding, namely whether past performance was or was not part of the meaning of a quota. Of course the final decision would be the community’s, but certainly there is some scope for the finding of a boundary breach here.32 Furthermore, turning from those whose performance was merely steady to those who actually managed to bring about an increase in production, it is striking to note that good performance in Karlsson is actually penalized, in contrast to Erpelding, where past success was the (potentially dominant) pass-​key to a higher quota going forward. As usual, it would be the wider community’s job to decide if it endorsed this distributive principle, which certainly seems unfair on the face of it. Perhaps it could be explained away as a kind of sanction for contributing, as the community saw it, to the glut in milk, which put everyone at risk. However, if Mr Karlsson had invested money in order to make improvements on his farm,33 it is particularly cruel to hamper his recouping 31

32

33

‘[W]‌e who are already members do the choosing, in accordance with our own understanding of what membership means in our community and of what sort of a community we want to have’: M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) 32. In theory, anyway. Perhaps in practice, though, the members of this category would not feel particularly hard-​done-​by, as their position vis-​à-​vis the distribution had not altered; they might not therefore feel a great need to complain of inequality in the first place. The Advocate General implies that he had bought five new cows: Case C-​292/​97 Karlsson (n 30) Opinion of AG Colomer, para 12.

Semi-suspect and non-suspect grounds

249

of it. Past success in this case, then, would appear to be acting as a negative dominant.34 Staying on the topic of quotas, it is instructive to look at the two so-​called German Banana Cases.35 The second of these provides a rare example of the situation where the Court saw a distinction which the complex egalitarian might not see, or at least might not regard as a breach. The first case is the less interesting of the two, from the Walzerian viewpoint. Here, tariff quotas were subdivided in favour of importers of Community or acp bananas.36 This meant that importers of this type received a larger quota, and could then import a larger quantity of bananas without incurring a penalty. Germany, which tended to import ‘third country’ bananas,37 and was then subject to a much harsher regime with much lower quotas, brought an action against the Council alleging a breach of the principle of equal treatment. The Court rejected the complaint, saying that the differential treatment was part and parcel of market-​integration (another example of inequality being used to justify itself). A complex egalitarian would of course invoke the Standard Contingent Reply; it is up to the members of the distributive community how they allot tariff quotas (and, indirectly, market share) amongst themselves. If Germany is unhappy, it must simply await the next meeting of the Forum to express its view and lobby for change. Prima facie, though, it does not seem to be contrary to the meaning of a tariff quota system to exclude or at least impede countries which are outside the protected bloc. In the second case, certain additional banana-​producing countries were accorded preferred-​nation privileges and duty-​free access to the Community,38 but only if the importing operator customarily marketed Community and/​or ‘traditional’ acp bananas. If the operator had previously marketed third country or ‘non-​traditional’ acp bananas, or if they had started marketing third 34

35 36 37 38

This case also touches on the issue of luck. That aspect of the case has not been mentioned here, but will be discussed at section 6.4 below. Analysis of another Article 40(2) case, Frico, will similarly be postponed until section 6.4: Joined cases 424/​85 and 425/​ 85 Coöperatieve Melkproducentenbedrijven Noord-​Nederland BA (‘Frico’) and others v Voedselvoorzienings In-​en Verkoopbureau [1987] ecr 2755. Case C-​280/​93 German Banana Case (I) and Case C-​122/​95 German Banana Case (II) (n 23). ‘ACP’ is the shorthand name for the African, Caribbean and Pacific countries with which the EU has a special agreement. I.e. bananas that were neither from other Community countries, nor from acp countries. These new countries were to be known as the non-​traditional acp countries, to contrast them with the traditional acp countries.

250 Chapter 6 country or ‘non-​traditional’ acp bananas since 1992, then they had to purchase costly export and import certificates, and pay duty on the incoming fruit. Again, Germany complained. In a judgment which was noteworthy for the way in which the Court was prepared to intervene in matters that were technically external to the Union, the cjeu this time upheld Germany’s complaint; there had been no need to penalize the latter two categories of operator and to do so was therefore discrimination. Utilizing a Walzerian form of reference, there would appear to be a connection between an operator’s choice of exporter on the one hand, and the distribution of (the obligation to buy) export certificates on the other. This would mean that there was no boundary breach here. The disparate treatment as between the first category of operator and the second two categories of operator was merely a ‘small inequality.’ In other words, a complex egalitarian would probably reach the same decision as was reached in the first case. If Germany was unhappy with the current arrangement, it could campaign for a new ‘shared meaning’ of an export certificate at the next Forum. As things stood, however, the distributive criterion being used did not appear to be out of keeping with the meaning of the distribuend at hand, and thus did not violate complex equality. Occasionally litigants will tag a complaint of unequal treatment onto the end of a case brought on other grounds, and they somewhat bring a negative outcome on themselves by insisting that unlikes have been treated like where the comparison on which they rely is with a party whose alleged difference (unalikeness) from them is arguably marginal and certainly irrelevant. This was the case for example in sfir v agea,39 where the complaint of unequal treatment, made in the context of a scheme for the restructuring of the sugar industry via dismantling of production facilities, turned on the fact that the same compensation was given to those sugar producers who had a silo on their property and those sugar producers who did not. It is hard to fault the Court in finding no breach of the principle of equal treatment here. In a Walzerian distribution too, whatever definition the relevant community arrived at for aid for the dismantling of sugar production facilities, it seems likely that possession of a silo would at most result in a deferred final payment to the owner of the silo, if the community was of the opinion (which the Court was not) that storing your own sugar in said silo, pending its eventual sale, was permissible and not an affront to the entire purpose of the scheme. However, it is hard to see how 39

Judgment of 14 November 2013, sfir –​Società Fondiaria Industriale Romagnola SpA, Italia Zuccheri SpA, Co.Pro.B.  –​Cooperativa Produttori Bieticoli Soc coop Agricola and Eridania Sadam SpA v agea –​Agenzia per le Erogazioni in Agricoltura, Ministero delle Politiche agricole, alimentari e forestali, C-​187/​12 –​C-​189/​12, ECLI:EU:C:2013:737.

Semi-suspect and non-suspect grounds

251

the meaning of the aid could be such that the owner of the silo would receive more, and they might even receive less if the community was of the view that this piece of quota-​hopping deserved a penalty. 6.2.2 Article 106(1) tfeu: Public and Private Undertakings Given that the entire goal of Article 106(1) tfeu (ex Article 86(1) ec) is to prevent Member States from distorting competition by using their greater resources to outdo private undertakings  –​what might be called legislative boundary defence –​the simple egalitarian will usually find himself on the same side as the complex egalitarian where this article is concerned. A straightforward example would be Asociación Profesional de Empresas de Reparto,40 where the Court held Spain to be in breach of Article 86(1) ec for having entrusted to the wholly State-​owned post office (Correos) not only permissible ‘reserved’ services, but also the non-​reserved services, and without having issued a call for tenders. A complex egalitarian too would find against Spain here; the distribution of the non-​reserved services was quite obviously flawed in favour of the public undertaking, with all of its State resources and backing, thus leaving any rival private undertakings with no hope of acquiring this work, and, it would seem, not even the chance to express their interest. However, more interesting cases, where a Walzerian analysis might produce a contrary result, may still be found. In Assurances de Crédit,41 two Belgian insurance companies found their return considerably reduced thanks to a new obligation, imposed by the Community via Council Directive, to set up an ‘equalization reserve’ (apparently to safeguard their solvency). In bringing an action, the applicants alleged a breach of the principle of equal treatment –​insurance undertakings in the public sector were excluded from the scope of the Directive. The Advocate General thought that this was a case of like (economic operators) being treated unlike, and, having rejected both of the proffered justifications, he found there to have been a breach of Article 86(1) ec (now Article 106(1) tfeu). The Court, so far as can be made out, took the situation to be one of unlikes being treated unlike; public undertakings were in an ‘objectively different situation,’42 and therefore there was no discrimination. 40 41

42

Case C-​220/​06 Asociación Profesional de Empresas de Reparto y Manipulado de Correspondencia v Administración General del Estado [2007] ecr I-​12175. Case C-​63/​89 Les Assurances du Crédit sa and Compagnie Belge d’Assurance Crédit sa v Council of the European Communities and Commission of the European Communities [1991] ecr I-​1799. ibid para 22.

252 Chapter 6 On the face of things, the private undertakings’ private status does seem to be something of a negative dominant here, a badge of dishonour causing them to be encumbered with burdens not imposed on holders of other badges. On the other hand, perhaps the obligation to set up the reserve has been justly distributed in this case. Whether a company is public or private might be the correct fundamentum distributionis; it could be argued that public companies do not need to put up the extra security because State ownership is already a sound enough guarantee.43 As usual, the Standard Contingent Reply would come into play, as the verdict of the distributive community itself is awaited. However, it is at least conceivable that whether or not a company is State-​ backed might be part of the meaning of an obligation to establish financial guarantees for the protection of the insured, and third parties. A second argument proffered by the defendant institutions, meanwhile, namely that they could not include public insurance companies in the Directive because they are too diverse, would almost certainly not stand up to Walzerian scrutiny. It is not at all clear how belonging to a category of companies wherein the members are homogeneous, or belonging to a category of companies wherein the members are heterogeneous, has anything to do with protecting the insured (and third parties) from loss of solvency on the part of their insurer. In Acoset,44 the Sicilian Regional Province of Ragusa, its Conference of Mayors, and twelve municipal councils of South-​East Sicily, established a body which was to be responsible for Regusa’s integrated water supply. This body then selected, as its form of management, a ‘semi-​private company with share capital which is predominantly publicly owned.’45 At this point, a competition was held to select the private participant for this company, which Acoset won fairly; it was clear from the contract notice that many of the works relating to the exclusive management of the water service would be the direct responsibility of the winning private participant, without need of further tendering procedures. However, a dispute then arose as to whether this process was compatible with ec law, and a reference was made. The Court and the Advocate General were unanimous that the process had indeed been correct. There was no need for two tendering procedures, one to select the private participant and another to award the service concession 43

44 45

Plus, as is pointed out a number of times in the Opinion and in the Judgment, the setting up, by the State, of a reserve for State-​owned companies might well violate the rules on State Aid. Case C-​196/​08 Acoset SpA v Conferenza Sindaci e Presidenza Prov. Reg. Ato Idrico Ragusa and Others [2009] ecr I-​9913. ibid para 17.

Semi-suspect and non-suspect grounds

253

itself. Such a ‘double competitive tendering procedure’46 would be a disincentive for private entities to come forward in the first place. Rather, the selection of the concessionaire can be regarded as an indirect result of the selection of [the] participant […], so that a second competitive tendering procedure for the selection of the concessionaire is unnecessary.47 Seen through complex egalitarian eyes, this case provides a neat example of the importance of defining the sphere or spheres at issue. Technically, there are two distributions at play here: the distribution of the chance to participate in the public-​private partnership, and the distribution of the management of the water service. The Court opted to kaleidoscope the two into one: the distribution of the chance to participate in a public-​private company which would then manage the water service. If this approach is followed, and all the inhabitants of the resultant single sphere acquiesce, then it would appear that Acoset’s fair winning of the ‘first’ (and now only) tender would be just. However, if the inhabitants insisted that the two distributions be kept apart, then two boundary breaches are immediately apparent. Firstly, the ‘winner’ in the first distributive sphere can smuggle this win into the second distributive sphere, in order to claim automatic victory there as well. Secondly, and perhaps more importantly, the public sector would be able to muscle in on the management of the water service, since, whichever private entity won in the first sphere, its (the public sector’s) own participation in the project was guaranteed. The public sector does not do everything better than the private sector, as a matter of course, and so it is not at all cut-​and-​dried that the new partnership would deliver a better water service than one of the other private entities, working alone. 6.3

Non-​Suspect Grounds

6.3.1 General Introduction: Omega The non-​suspect ground cases are the last stop in this case study. Certain sub-​ categories can be identified, and three of these, public procurement, human resources and competition law, will be considered below. However, aside from these, the list of non-​suspect grounds is non-​exhaustive, and the number of 46 47

ibid para 58. ibid para 60. Assuming of course that the first competitive tendering procedure had been conducted in accordance with the principles of ec law.

254 Chapter 6 cases is potentially infinite, encompassing every possible difference between all possible entities. That said, even the most minute and seemingly trivial difference may make a large difference to an applicant, even the difference between solvency and bankruptcy. As an illustration of this, it is proposed to start by looking at the Omega case where a company’s aeroplanes were discriminated against by the Council by reason of their low by-​pass ratios.48 In 1999, legislation was passed at Community level to address the problem of the noise pollution caused by commercial aeroplanes; aeroplanes the engines of which had a by-​pass ratio of lower than three were to be prohibited. The new law did not affect new planes, which were all built to the new specifications anyway. However, Omega was now prevented from using a number of older planes which it had refitted with newly manufactured engines, producing a by-​pass ratio of 1.74. The company brought an action against the authorities in its home Member State (the uk), in an attempt to have the legislation annulled. Before the European Court, on a reference from London, it insisted that the by-​pass ratio was not the only indicator of engine noise, and that its refitted planes were in fact slightly quieter than those with engines the by-​pass ratio of which was three or more. A first comparison, then, was between Omega’s re-​engined aeroplanes, and those complying letter-​for-​letter with the new law. If Omega objected to being treated differently from companies in compliance with the Regulation, it also objected to being treated the same as companies which had merely refitted their aeroplanes with so-​called ‘hush-​kits’ (a far inferior response to the problem, according to Omega); this, then, was the second comparison. However, neither the Advocate General nor the Court found there to have been a breach of the principle of non-​discrimination. Alber AG was of the opinion, in relation to the first comparison, that Omega and companies producing new, Regulation-​compliant planes could not be regarded as alike, especially considering that the latter had to be able to rely on their legitimate expectations that the new regime would be adhered to; in this way, unlikes were being treated unlike. The Court, meanwhile, preferred to see the situation as one of likes treated unlike, but with legitimate expectations acting as an objective justification for the unlike treatment. With regard to the second comparison, the Advocate General seemed to be saying that, despite the fact that they were different, the two comparators were the same –​likes had therefore been treated 48

Joined cases C-​27/​00 and C-​122/​00 The Queen v Secretary of State for the Environment, Transport and the Regions, ex parte Omega Air Ltd (C-​27/​00) and Omega Air Ltd, Aero Engines Ireland Ltd and Omega Aviation Services Ltd v Irish Aviation Authority (C-​122/​00) [2002] ecr I-​2569.

Semi-suspect and non-suspect grounds

255

alike. The Court reached the same conclusion. It can be seen, once again, how Aristotelian analysis tends to produce inconsistent or nonsensical results. A Walzerian analysis, on the other hand, would be relatively straight-​forward. What is being distributed is restrictions on the use of aeroplanes at European airports, ‘in the interests of protection against noise.’49 The companies producing all three types of plane –​planes reengined with engines having a low by-​ pass ratio, planes refitted with hush-​kits, and new planes the engines of which had a by-​pass ratio exceeding three –​are the distributees. On the face of it, the distribution would seem to be flawed in favour of companies producing aeroplanes of the first two types. Companies producing aeroplanes of the third type seem to be immune from restriction, as though their aeroplanes’ very newness was a dominant.50 However, it would as usual depend on the distributive community’s understanding of these anti-​noise restrictions. If newness formed part of the understanding then bringing newness into the distributive sphere would be a legitimate crossing, not a boundary breach. Yet there seems to be no reason why those living near airports should not be allowed into the Forum as recipients of, if not the main distribuend, then an important by-​product: peace and quiet. They might not care about the method by which a reduction in noise was achieved. Their distributive criterion might simply be:  those with the noisiest engines should receive the most restrictions. Did the Council draw up its thresholds in consultation with them? If they agreed, for example, that ‘on take-​off there would be no perceivable difference,’51 and judged that to be the decisive factor, then Omega’s planes should also have been exonerated and restrictions needed only be issued for planes refitted with hush-​kits. The danger is that the ‘by-​pass ratio of three’ is simply a line in the sand, mechanically applied by the Council and converted (also in the Walzerian sense) into a legislative death sentence for a company like Omega. One might hope that the cjeu judges, whose job it is to curb the excesses of the legislature, might take a fresh approach. But the yes-​no simplicity of the Aristotelian test enables them (or forces them, depending on their viewpoint) to merely rubber-​stamp the Council’s choice. This represents a jackpot for a company producing new planes, as the Council qua distributor bestows it with dominance, and the Court, unwilling or unable to make its own de novo assessment, is little more than a messenger-​boy. 49 50

51

ibid, Opinion of AG Alber, para 1. In this case, a negative dominant. If the distribution were viewed the other way around, for example, if the distribuend was read as being ‘permission to fly,’ then newness would be a positive dominant. Joined cases C-​27/​00 and C-​122/​00 Omega (n 48) Opinion of AG Alber, para 75.

256 Chapter 6 6.3.2 Public Procurement One context in which the principle of equal treatment is regularly invoked is that of public tendering procedures; an example has already been seen earlier with Acoset.52 In Acoset, however, one bidder was contesting the procedure itself, whereas more usually the issue is whether two bidders have had an equal chance. In Evropaiki Dynamiki,53 for example, the Commission launched a tendering procedure in early 2002 for the provision of support services in relation to cordis (an informatics tool for European research framework programmes); here services had been provided up until then by the existing contractor, Intrasoft International sa. Evropaiki Dynamiki (‘ed’) submitted a tender but ultimately failed, the contract instead going to a company called Trasys which had made clear from the outset its intention, if successful, to subcontract at least 35% of the work to the existing contractor. ed challenged this result on a number of grounds, of which the most important two were, firstly, that the running-​in phase was to be unpaid (which put a burden on everyone except the existing contractor), and secondly, that the Commission had initially failed to give out certain technical information (so that the only party who had this information from the beginning was the existing contractor). The Court of First Instance rejected the first of these pleas; the fact that an unpaid running-​in period disadvantaged all but the incumbent was not the Commission’s fault  –​it was merely, for the existing contractor, an ‘inherent de facto advantage.’54 Turning to the second plea, this was itself divided into two parts: the Commission’s alleged failure to inform the bidders that it had recently acquired the so-​called ‘Autonomy Software,’ and the Commission’s alleged failure to provide information regarding certain technical specifications and source codes for cordis. With regard to the first of these, the Court thought that there had been a procedural defect, but that it could not have made any difference to the tendering procedure; in other words, knowledge of the acquisition of the Autonomy Software was irrelevant anyway. With respect to the second, a similar result seemed to have been reached –​that there had been a defect but that ed failed on causation –​except for the argument that, had it had the pertinent source codes and so on, ed’s actual bid would have been lower. The Court accepted this argument; in its opinion, if it had had the source code information on time, ed would have tendered a lower price and

52 53 54

See supra, text accompanying n 44. Case T-​345/​03 Evropaïki Dynamiki –​Proigmena Systimata Tilepikoinonion Pliroforikis kai Tilematikis ae v Commission of the European Communities [2008] ecr II-​341. ibid para 70. Note how the cfi entertains here the Walzerian idea of the small inequality.

Semi-suspect and non-suspect grounds

257

thus would have fared much better in the competition. The Commission’s decision was therefore annulled. An illegitimate boundary crossing would most probably be the verdict of a complex egalitarian reading this case. In fact, it is a classic boundary breach for Intrasoft/​Trasys to bring ‘inside’ knowledge, gained in a different sphere, into the distributive sphere at issue (the tendering procedure) and thereby swing the distribution in their favour. Only if all candidates had been given the information55 would it lose its status as a monopolizable dominant and become simply part of the shared understanding of the procedure; in other words, only then would the crossing be legitimate. The outcome for the complex egalitarian would thus have been the same as that arrived at by the Court. However, on occasion, a different result may be reached. In Centro Studi,56 the Council issued a tendering procedure for the full management of a crèche. Centro Studi Antonio Manieri Srl (‘Centro Studi’), an Italian company, was accepted as a candidate for the procedure, and submitted its tender at the beginning of 2005. However, on 16 January 2006 the General Secretary notified Centro Studi by letter that he was abandoning the tendering procedure, and announced his decision to accept the proposal of the so-​called Office for Infrastructure and Logistics (oib), which formed part of the General Secretariat. Centro Studi brought an action consisting of several pleas, one of which was that the Council had breached the principle of equal treatment by evaluating the oib’s proposal independently of the tendering procedure. In its judgment, the Court of First Instance reiterated that tendering procedures were only obligatory where a contracting authority was intending to hire a supplier of assets or provider of services from outside of its own organization, not where the would-​be contractor was one of its own departments. In the cfi’s view, because the oib was a department of one of the Community institutions, it could not be compared to participants in a tendering procedure. Therefore, the Council’s separate evaluating of its proposal was not an infringement of the principle of equal treatment. The action was dismissed.57 55

56 57

Or, more properly, ‘had been given an equal chance to learn the information.’ Those who ignored it, or misunderstood it, or applied it wrongly, would not be treated unjustly if not selected. Case T-​125/​06 Centro Studi Antonio Manieri Srl v Council of the European Union [2009] ecr II-​69. Note that a plea made by Centro Studi on the basis of Article 106 tfeu (ex Article 86 ec) was also rejected because, while Centro Studi was a private undertaking, the oib was not a public undertaking.

258 Chapter 6 A supporter of complex equality would ask what the shared social meaning or understanding of the distribuend was. The work on offer, namely managing the Council’s crèche, was surely above all else about the looking after of children.58 From this meaning, the fundamentum distributionis would appear to be the ability to look after children, that is, whoever could best look after children should receive the contract. The automatic allocation of the work to the oib, then, whether or not a department of an institution, would now seem to be flawed. The oib had been allowed to use its connection to the Council as a ‘Master Good,’ which it could exploit to tyrannize neighbouring distributive spheres, in violation of complex equality. Thus a supporter of the theory would reach the opposite result to the one reached by the Court. Finally, in Højgaard, Bandanmark, the Danish rail infrastructure operator, was looking to build a new railway line between Copenhagen and Ringsted.59 In January 2013, it organised a tender for the work, with six places available for Round 1. In the event five candidates put themselves forward at the pre-​ selection stage, of which one was a group comprising MT Højgaard A/​S and Züblin A/​S (the two plaintiffs, ‘the Højgaard and Züblin group’) and one was a group comprising Per Aarsleff A/​S and E Pihl og Søn A/​S (‘the Aarsleff and Pihl group’). All five were therefore invited to submit a tender, although in June 2013 one of the three others withdrew. On 26 August 2013, E Pihl og Søn A/​S was declared insolvent by a Danish court, but nonetheless the next day the Aarsleff and Pihl group submitted a tender, signed by representatives of Per Aarsleff and Pihl, but not by Pihl’s liquidator. On 15 October 2013 Banedanmark informed all the tenderers of its decision to allow Per Aarsleff to continue to take part, alone, in the procedure, as the company satisfied the conditions required for participation in the procedure on its own. It is worth noting that Per Aarsleff had taken over the contracts of more than fifty salaried staff of E Pihl og Søn, including the individuals who were key to the implementation of the project concerned. Thus, Aarsleff submitted a tender in Round 2 in its own name. After the assessment of the second tenders received, Banedanmark decided, in accordance with the conditions of the contract notice, to choose three tenders with a view to determining which of them offered the best value for money and asked the three tenderers chosen to submit a third and final tender. The three tenderers included Per Aarsleff and the Højgaard and Züblin group. Those final tenders were lodged by the three tenderers on 12 December 2013. Finally, on 58 59

Taking the Standard Contingent Reply as read. Judgment of 24 May 2016, MT Højgaard A/​S and Züblin A/​S v Banedanmark, C-​396/​ 14,ECLI:EU:C:2016:347.

Semi-suspect and non-suspect grounds

259

20 December 2013 Banedanmark informed the three chosen tenderers that it had decided to award the contract, worth approximately eur 123 million, to Per Aarsleff. MT Højgaard and Züblin brought proceedings before the Public Procurement Complaints Board of Denmark, claiming inter alia that, by allowing Per Aarsleff to participate in the procedure in the place of the Aarsleff and Pihl group, although Per Aarsleff had not itself been pre-​selected, Banedanmark was in breach of the principle of equal treatment laid down in Article 10 of Directive 2004/​17.60 The Public Procurement Complaints Board, noting that Per Aarsleff could have been pre-​selected on its own without exceeding the set number of places available, made a reference to the Court of Justice, asking whether the Directive precluded the awarding of a contract to a tenderer who did not apply for pre-​selection, and who indeed was not pre-​selected. The Court’s judgment is less helpful than it might have been as it leaves much to be decided by the referring tribunal. Interestingly it comments that, while what it calls ‘strict application’ of the principle of equal treatment might have led to a finding of breach, the requirement of legal and substantive identity […] may be qualified in order to ensure, in a negotiated procedure, adequate competition.61 60

61

Directive 2004/​17/​EC of the European Parliament and of the Council of 31 March 2004 coordinating the procurement procedures of entities operating in the water, energy, transport and postal services sectors[2004] oj L134/​1. Article 20(2)(b) of Council Directive 90/​ 531/​E EC of 17 September 1990 on the procurement procedures of entities operating in the water, energy, transport and telecommunications sectors [1990] oj L297/​1 (‘Directive 90/​531’) had talked of tenderers being given ‘equal time to prepare and submit tenders’. This was succeeded, without change to the quoted text, by Article 26(2)(b) of Council Directive 93/​38/​E EC of 14 June 1993 coordinating the procurement procedures of entities operating in the water, energy, transport and telecommunications sectors [1993] oj L199/​84. Directive 98/​4/​EC of the European Parliament and of the Council of 16 February 1998 amending Directive 93/​38/​E EC coordinating the procurement procedures of entities operating in the water, energy, transport and telecommunications sectors [1998] oj L101/​1 amended Article 26, but not the phrase quoted. Article 5(1) of Directive 94/​22/​ EC of the European Parliament and of the Council of 30 May 1994 on the conditions for granting and using authorizations for the prospection, exploration and production of hydrocarbons [1994] oj L164/​3, also amended Directive 90/​531, including a provision that where two applications for authorisations were of equal merit having considered factors including capability, methods and price, then ‘other relevant objective and non-​ discriminatory criteria’ should be taken into account in order to make a final choice. Judgment in Case C-​396/​14 Højgaard (n 59) para 41. It is clear that the Court is comparing Per Aarsleff on its own with the dissolved partnership (i.e. the Aarsleff and Pihl

260 Chapter 6 The Court said that the principle as a whole was not infringed by Per Aarsleff continuing alone, despite its not having been pre-​selected, as long as firstly, Per Aarsleff on its own met the requirements of the tender, and secondly, its continuation did not put other tenderers at a competitive disadvantage. While happy to declare that the first of these conditions for equality had been met in Per Aarsleff’s case, the Court of Justice deferred to the referring tribunal on the second one, merely mentioning some of the matters which might sway the decision. These of course included the fact that the liquidator had not signed the original tender, and the fact Per Aarsleff had hired many of the liquidated company’s key workers. Advocate General Mengozzi had mentioned a third factor which might lead to the conclusion that the other tenderers had been put at a competitive disadvantage, and that is that Per Aarsleff on its own had new information which was not known to all of the candidates at the start of the procedure. For example, it knew that the dissolved partnership’s tender had been rated as the second best from Round 1. The Advocate General hints quite strongly that he considers this to be unjustified ‘different treatment’ as between Per Aarsleff and the other tenderers.62 Rather than the simple distribution of a contract to build a railway line, a Walzerian analyst would probably divide this into a number of sub-​ distributions corresponding to the different rounds: the distribution of places in Round 1, the distribution of places in Round 2, and so on. Taking the fundamentum distributionis of this latter distribution as success in Round 1, he or she might then wonder why Per Aarsleff on its own was granted a place in Round 2, when it had never submitted an independent bid to Round 1 and consequently could not possibly have succeeded in this round. That alone would seem to constitute a breach of complex equality sufficient to vitiate the distribution

62

group), presumably an alleged example of two unlikes being treated like, but Article 10 of Directive 2004/​17 (n 60) merely requires that ‘[c]‌ontracting entities shall treat economic operators equally and non-​discriminatorily’ and it is submitted that the correct comparison is as between Per Aarsleff and the Højgaard and Züblin group, the treatment in question being the awarding of the tender to the one rather than to the other, as opposed to the granting of permission to participate, at different stages of the competition, to both Per Aarsleff and to the dissolved partnership. Opinion of 25 November 2015, MT Højgaard A/​S and Züblin A/​S v Banedanmark, C-​396/​ 14, ECLI:EU:C:2015:774, para 82. See also his comment further up that the change from the dissolved partnership to Per Aarsleff on its own had been substantive, not purely formal: ibid para 77. It is obvious that, for the Advocate General, the correct comparison is as between the treatment in respect of Per Aarsleff and the ‘treatment in respect of the other tenderers:’ ibid para 82.

Semi-suspect and non-suspect grounds

261

in question, and thus it is impossible that this company could have gone on to win the tender overall. Even if somehow allowed to participate in Round 2, where the distribuend would be places in Round 3, one could just as easily find a breach in this distribution following Mengozzi AG’s logic. If Per Aarsleff had extra knowledge in preparing its bid for Round 2, and could tailor its bid accordingly (in respect of the composition of the project team, for example) in a way that the other candidates were no longer able to do, this extra, mid-​procedure knowledge, coupled with the facility to act on it, would appear to be a Master Good skewing the distribution in its favour; again, the result would be a disqualification of Per Aarsleff prior to Round 3. The emphasis by the Court on a kind of ‘Equality Lite’ in public procurement cases, making allowances (or rather overlooking discrepancies) for the assurance of adequate competition63 seems in and of itself an affront to complex equality  –​a taking into account of extraneous issues and imperatives –​not unlike Twenty first Century Europe’s strange insistence on subjugating equality of Member States64 under the yoke of economic advantage.65 While it is true that a disqualification of Per Aarsleff at the beginning of Round 3, leading to a two-​horse race between the Højgaard and Züblin group and the other remaining tenderer, would have led to a slight diminution of competition for the contract, it is submitted that that is a lesser affront to justice, in the Walzerian sense, than allowing Per Aarsleff to go forward with ‘insider’ knowledge and win. 6.3.3 Human Resources Another area in which it is commonplace to find allegations of unequal treatment is that of human resources, including the so-​called ‘staff cases.’ However, in addition to the individual staff-​members of the European institutions bringing actions against their employers, action is sometimes taken on an even larger scale, for example, when Italy sued the Commission for having decided to advertise vacancies in English, French and German only.66 The Court of 63

64 65 66

And even by the Advocate General, for example, ibid para 59. He cites Judgment of 10 October 2013, Ministeriet for Forskning, Innovation og Videregående Uddannelser v Manova A/​S, C-​336/​12, ECLI:EU:C:2013:647 as a possible source of this emphasis, but is clear that his view is that equality trumps adequate competition (for example, Opinion in Case C-​396/​14 Højgaard (n 62) para 67). Ultimately democracy itself. But crucially that of certain Member States, not of Europe as a whole. Case T-​185/​05 Italian Republic v Commission of the European Communities [2008] ecr II-​3207.

262 Chapter 6 First Instance held that, in a given competition, persons holding two of the excluded languages (knowledge of at least two languages being the minimum requirement for work at one of the European institutions), but who were nevertheless eligible for the post, would have no way to acquaint themselves with the vacancy notice, and would then be in a ‘less advantageous position’67 in relation to other candidates. This was discrimination on grounds of language. The Commission’s decision was annulled. A Walzerian student of this case would almost definitely agree with the Court, concluding that a candidate’s choice to have learnt English, French or German, or characteristic of knowing English, French or German,68 was exploitable here as a means of gaining privileged access to a separate sphere, where mastery of those three languages was not part of the meaning of (the possibility of) the post in question.69 Such spherical imperialism would represent a breach of complex equality. In many cases, officials of the European institutions complain of a perceived unfairness with regard to their pay, or other benefits. In Beltrante v Council,70 for example, the parents of various adult children, who for whatever reason were to be ‘treated as if they were dependent children,’ complained when travel expenses in respect of such children were scrapped, unless the child lived in the same city as the official. This created an inconsistency with the parents of dependent children stricto sensu, who could receive travel expenses for their children even where the child did not live in the same city as the official. In its judgment, the Court of First Instance demonstrated once again the illogical nature of the Aristotelian equality test. First, it seemed to hold that the two categories of parent were alike, but that their unlike treatment was objectively justified.71 However, in the next paragraph, it states clearly that the situation is one of unlikes treated unlike.72 In either case, the objective justification (version one) or differing circumstances (version two) come to the same thing –​ the Court is of the belief that dependent children stricto sensu are ‘part of the

67 68 69

70 71 72

ibid para 136. The analysis works in either event. See supra section 3.3. Taking the Standard Contingent Reply as read. Of course, the situation would be different where the post itself was reserved for those knowing two or all of the three languages mentioned. Then mastery of the languages would be part of the meaning of the distribuend, thus legitimizing the boundary crossing. Case T-​48/​89 Fernando Beltrante and others v Council of the European Communities [1990] ecr II-​493. ibid para 33. ibid para 34.

Semi-suspect and non-suspect grounds

263

family unit’73 and give rise to a presumption of cohabitation, while children ‘treated as if they were dependent children’ are ‘members of the family only in the broad sense.’74 This presumably means that a parent of the former type of child needs to be recompensed, if the child lives in a different place, for the rupture to the family circle caused by his or her work; but for the job, they would have enjoyed a greater amount of familial warmth because all members of the family would have been in the same place. The parent of the quasi-​dependent child, apparently, is not expected to miss their absent offspring. Again, the test is seen to be too black-​and-​white. It is true that in some circumstances the parents of children ‘treated as if they were dependent children’ –​adult children without gainful employment, for instance –​might not feel any especial need to have said children come and visit. But in other circumstances –​where the adult child is disabled, say –​the Court’s ruling seems unduly harsh and uncaring. A further criticism would be that, from the linguistic point-​of-​view, it seems particularly absurd to define a certain type of children, in the Staff Regulations themselves, as ‘treated as dependent children,’ and then, when asked, to declare that these same children, defined thus, are not to be treated as dependent children. A Walzerian dissection of the case avoids all of these pitfalls. There is a single, autonomous distributive sphere. Travel expenses are being distributed, by the Council, to various types of official. Prima facie, whether an official’s children are dependent stricto sensu or ‘treated as if dependent’ is not part of the meaning of travel expenses, and for the Council to use one or other of these attributes as a (positive or negative) dominant would appear to be a boundary breach. On the other hand, it would be open to the distributive community to legitimate such a border-​crossing by incorporating one or both of the attributes into its shared meaning of the benefit in question. Patterns of dominance across spheres are arguably most prevalent of all in the world of work, attainment in Sphere 1 often guaranteeing preference or even automatic selection in Sphere 2.75 On the other hand, to quote Walzer again: 73 74 75

ibid. ibid. A stark example is the tradition in the uk of automatically awarding barristers who are elected to parliament with the title of ‘Queen’s (or King’s) Counsel,’ an honour which ordinarily would only be acquired after many years of high-​level and successful practice at the Bar. Success in the political sphere can thus be leveraged unjustly to ‘buy’ success in the legal sphere, unless of course by coincidence the new mp was a barrister of repute worthy, due to relevant attributes, of the honour concerned.

264 Chapter 6 Failure pursues [excluded men and women] from sphere to sphere in the form of stereotyping, discrimination, and disregard, so that their condition is not in fact the product of a succession of autonomous decisions but of a single systemic decision or of an interconnected set.76 A further example could be the distinction often drawn at the eu institutions as between eu civil servants, that is, those who have passed the eu-​level competitive examination to become career bureaucrats (‘eu officials’) and seconded ‘national experts’ (those who have been recruited into the civil service of their home Member State, and whose presence at the eu institutions is limited to a maximum of ten years). In Dupré v eeas,77 Mr Dupré, a seconded national expert, had already completed a post at the European Commission and had then applied for the post of ‘Policy Officer –​Security’ at the European External Action Service (eeas). Put briefly, the vacancy notice promised that those eu officials who successfully applied to the job would retain their present grade, while seconded national experts would all be recruited to the apparently low grade of AD5. Mr Dupré felt that he had been discriminated against in relation to eu officials, a form of discrimination which was even expressly prohibited in the eeas’ foundational document, Decision 2010/​427, as follows: Officials of the Union and temporary agents coming from the diplomatic services of the Member States shall have the same rights and obligations and be treated equally, in particular as concerns their eligibility to assume all positions under equivalent conditions. No distinction shall be made between temporary agents coming from national diplomatic services and officials of the Union as regards the assignment of duties to perform in all areas of activities and policies implemented by the eeas.78 Mr Dupré therefore complained to the institution, and then challenged the resultant rejection in front of the European Union Civil Service Tribunal, which is part of the cjeu. This Tribunal threw out the challenge, having produced a bizarre sequence of arguments (not helped, for the English reader, by an obvious linguistic error), which is worth considering in full.

76 77 78

M Walzer, ‘Response’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) 291. Already quoted in Chapter 2 above. Judgment of 18 May 2015, Dupré v eeas, Case F-​11/​14, ECLI:EU:F:2015:47. Council Decision 2010/​427/​EU of 26 July 2010 establishing the organisation and functioning of the European External Action Service [2010] oj L201/​30, Article 6(7).

Semi-suspect and non-suspect grounds

265

First, of course, the Court stated the Aristotelian test for equality.79 It then went on to assert that there had been no unequal treatment ‘as concerns [officials’ and staff from national diplomatic services’] eligibility to assume all positions under equivalent conditions:’ It is clear from the wording of the vacancy notice that both the list of eligibility criteria it contained and the selection criteria were the same for eu officials and for staff from national diplomatic services of the Member States, the only difference being that the latter were required to meet nationality and education criteria which officials in function group ad were presumed to have met at the time of their recruitment. Accordingly, it is appropriate to conclude that the vacancy notice did not differentiate as regards access to the post according to the origin of the candidates.80 This seems like a very parsimonious reading of the word ‘assume’ from the Decision. One would have thought that assuming a position was a wider concept than just accessing it, taking in all manner of conditions of employment (the presence of the word ‘conditions’ in the Decision lending further support), including the grade at which the new employee is hired. Nevertheless, for the Court, this was a matter of likes treated like. Turning then specifically to the grading issue, the Court suddenly decides that eu officials and national experts are not like after all, certainly not, the reader is told at Paragraph 74, as regards the framework within which their career is supposed to progress or indeed the way in which their administration of origin assesses their seniority.81 The particulars of this alleged unalikeness are given in the next three paragraphs, by reference to the so-​called ‘ceos’ (Conditions of Employment of other Servants of the European Union).82 The first alleged difference, that relating to career progression, is described as follows: First, members of staff of the General Secretariat of the Council, the Commission or the eeas, who could apply for the post advertised in the 79 80 81 82

Judgment in Case F-​11/​14 Dupré (n 77) para 69. ibid para 71. ibid para 74. These derive from Council Regulation 723/​2004 of 22 March 2004 amending the Staff Regulations of officials of the European Communities and the Conditions of Employment

266 Chapter 6 vacancy notice if they met the grading criterion, were officials in function group ad, who, as such, are intended to follow a career path within the eu institutions, performing, inter alia, conceptual and analytical functions in grades ad 5 to ad 14. However, it is apparent from Article 50b of the ceos that staff from the national diplomatic services of the Member States, even if they are engaged as temporary staff under Article 2(e) of the ceos, are supposed to resume their career path subsequently in their administration of origin, since they may be engaged by the eeas only for a maximum period […] of ten years, with a guarantee of immediate reinstatement at the end of their period of service at the eeas.83 This difference as between permanent eu officials, who pursue their careers ‘within the eu institutions,’ and temporary seconded national experts, who ‘resume their career path … in their administration of origin,’ would seem to be so obvious as to render description redundant. A description of the second alleged difference, which relates to the assessment of seniority at the eeas vis-​ à-​vis that at the ‘administration[s]‌of origin,’ then begins, in almost unintelligible terms, as follows: Secondly, where they exercise their broad discretion in deciding on the correspondence between posts and grades, in the light of the importance of the duties assigned to the functions at issue and solely in the interest of the service, the eu institutions must take into account the fact that, under the Staff Regulations, identical or similar functions may be performed by persons in different grades, as is apparent from Annex I, Section A, to the Staff Regulations, which provides, in the case of most of the functions listed therein, that they may be performed by officials in different grades.84 This in fact tells the reader little about how seniority is assessed at the eeas, except perhaps to say that it does not really matter, because there is broadly no difference in function as between the grades. This may be an indirect way

83 84

of other servants of the European Communities [2004] oj L124/​1, as last amended by Council Regulation 1240/​2010 of 20 December 2010 adjusting, from 1 July 2010, the rate of contribution to the pension scheme of officials and other servants of the European Union [2010] oj L338/​7. ibid paras 75 and 76. ibid para 77.

Semi-suspect and non-suspect grounds

267

of knocking out Mr Dupré’s argument that the functions he was performing did not lend themselves to a grading of AD5. However, instead of going on to a description of how seniority is assessed at the national experts’ home institutions, so that a comparison might be drawn and a difference discerned, the Tribunal merely continues with a lengthy and repetitive list of how many eeas officials at each grade had been eligible for promotion to (presumably) the grade above, during the 2013 promotion round. The reader is then told that [these figures] provide adequate proof that, although the vacancy notice had confined the opportunity to apply for the vacant post solely to eeas officials in grade AD 5, nearly all eeas officials would have been excluded from the selection procedure, in view of the high grades occupied by all those officials.85 This finding of an apparently quite serious flaw in the promotion procedure, resulting in indirect discrimination to the detriment of the eu officials, would seem on closer inspection to be factually wrong; the vacancy notice had stated that ‘candidates must be officials of certain institutions of the European Union in one of the grades AD 5 to AD 14.’ Even if true, that would only help the reader to understand how Mr Dupré came to stand a good chance of getting the job, not why national experts would be compulsorily recruited at grade AD 5, while eu officials could retain their existing grade, the difference in treatment at issue. Rather, this would seem to be a translation error, with the word ‘although’ having been wrongly used in the place of ‘if.’86 The Tribunal’s analysis concludes as follows: In those circumstances, it was entirely legitimate for the vacancy notice to provide that the recruitment of a temporary agent would be at grade AD 5 whilst providing that an eu official in one of the grades AD 5 to AD 14 would be appointed in his existing grade, and the principle of equal treatment was not infringed so far as eligibility for eu officials and for staff from the national diplomatic services of the Member States to assume all positions under equivalent conditions and the grading of the successful candidate were concerned.87 85 86 87

ibid para 78. ‘[S]‌i l’avis de vacance avait limité …’ in the French version of the case; ‘si la convocatoria de vacante hubiera limitado …’ in the Spanish version. Judgment in Case F-​11/​14 Dupré (n 77) para 79.

268 Chapter 6 Ultimately then, despite an earlier attempt to brand them as alike,88 the Tribunal’s view was that eu officials and national experts were unlike, and thus merited unlike treatment. Even if one views the earlier finding of similarity as a means of holding there to be no direct discrimination against the national experts in this case, and the subsequent holding of difference to be part of an analysis into indirect discrimination, said analysis, resulting in a finding that no such discrimination had taken place, is problematic for two reasons. Firstly, the Tribunal’s assessment of the difference between eu officials and national experts reads more like a simple description of, rather than a genuine, material distinction between, the two categories, equivalent to saying:  ‘they are unlike because they are unlike’. At best, the Tribunal seemed to wish to call attention to the single-​track nature of the eu official’s career path, as distinct from the more circuitous nature of the national expert’s, but makes no attempt to explain how that distinction requires the national expert’s grade to be frozen at, or even reduced to, AD5 on his or her joining the eeas. The second problem concerns the difference in treatment. The Tribunal appears to wish to reframe the difference in treatment from the freezing or reduction of the national expert’s grade (as compared to the continuation quo ante of the eu official’s), to the expansion of the job opportunity to eu officials of all grades (as compared to the limitation of the opportunity, vis-​à-​vis the national experts, to those already at, or prepared to be demoted to, AD5). Such an expansion, according to the Tribunal, was justified by a scarcity of entry-​ level candidates amongst the eu officials. In other words, the unlike treatment is in fact blamed, not on the distinct career paths, but on the shortage of AD5 candidates amongst the ranks of the eu officials. The reader is left wondering: how are the differing career paths relevant to this particular grading decision? They might have been relevant, for example, to a decision to leave a newly recruited national expert’s grading up to their domestic administration. Similarly, they might have been relevant to a decision to grade an incoming national expert at a level as closely as possible commensurate with their past experience (twenty-​five years’ worth, in Mr Dupré’s case). However, the grading decision in fact taken by those drafting the vacancy notice, effectively to suspend any career progression at all in the case of seconded national experts, not only does not seem to flow in any way naturally 88

Possibly reprised in the second half of the paragraph quoted. After all, if ‘the principle of equal treatment was not infringed so far as eligibility for [the two categories] to assume all positions under equivalent conditions […] were concerned,’ in other words, if the principle of equal treatment was not infringed and an obligation of equal treatment was met, then it follows that the situation at hand must be one of likes having been treated like.

Semi-suspect and non-suspect grounds

269

from the distinct career structures of the two kinds of civil servant, but if anything emerges as a kind of perverse logic:  far from facilitating the entry of national bureaucrats into the eu civil service, by helping to smooth over such complications as were thrown up by the disjointedness of their career, the decision all but discouraged such entry in the first place, asserting that the two career paths were in reality too idiosyncratic to be reconciled at all. Since the difference in treatment does not fit the proposed ‘unalikeness’ of national experts vis-​à-​vis eu staff members, and having considered alternative differences in treatment which might have been a better fit, the only other option is to keep the difference in treatment, but to change the grounds upon which the two categories were held to be unequal. Had first the eeas, and later the Tribunal, declared the difference to have been that there were more (genuine) entry-​level candidates amongst the national experts then there were amongst the Union experts, then an expansion of the eligibility requirements to include Union experts from grades AD5 through to AD14 might have held water, as a means of equalizing the quality of the opportunity afforded by the advertisement to the two groups concerned. However, there was no such equalization of the quality of the opportunity where, in reality, great pains were taken by those drafting the advert to ensure that eu officials retained the professional advantage of their previous experience, while applicant national experts of any seniority (and it is hard to believe that a ‘sensitive’ post such as the one at issue would ever have been offered to a national expert with only two years of experience in their home civil service) were asked in effect to sacrifice the professional advantage of their previous experience as the price of their application. Relative size of recruitment pools, even if this had been used as the tertium comparationis, had nothing to do with it. Safeguarding of eu officials’ accrued rights to advancement seemed to have everything to do with it. While the exact motive for the disregard shown to national experts here is not clear, it cannot be justified by the differing make-​up of such experts’ career plans, or the bogus assertion that (unlike eu experts) they are all beginners with nothing to lose. 6.3.4 Competition Law Another field in which equality pleas are popular is that of competition law, where, after a number of undertakings have been found guilty of a competition offence, there are sometimes allegations amongst the culprits that the Commission has been unfair in the imposition of its sanctions, either treating different situations the same (for example, by ignoring differing levels of liability amongst the culprits) or treating comparable situations differently as in the example below.

270 Chapter 6 In the General Court case of Deltafina v Commission,89 the applicant, Deltafina SpA, was an Italian company whose main activities were the first processing of raw tobacco and the marketing of processed tobacco. In 2001, the Commission had begun to investigate possible horizontal collusion by tobacco processors in the eu, and this had resulted in the Commission’s adopting Decisions imposing fines on various processors in Italy (including Deltafina) and Spain.90 Deltafina challenged the Commission’s Decision, claiming unequal treatment as between it and the Spanish processors. This was because the Spanish processors had received a 40% reduction in their fine to reflect the (at that time) uncertain application of Article 81 ec to interprofessional agreements in the sector concerned, where the national law (according to Deltafina) ‘encouraged horizontal collusion between processors.’91 On the other hand, Deltafina had not received any reduction for this ‘attenuating circumstance,’ despite the similarity between the Spanish and Italian legislative frameworks. The General Court refused to uphold this argument. In its opinion, the situation of the Spanish processors was not comparable with the situation of the Italian processors. In the Spanish case, the public authorities had played a significant role in facilitating collusion between the processors, while no such role had been played by the public authorities in Italy. Since different situations had been treated differently, there was no breach of the Aristotelian principle of equal treatment. However, viewed through a Walzerian lens, the case in some ways resembles the second German banana case, discussed above.92 Both the Spanish and the Italian processors are legitimately within what might be called the ‘punishment sphere.’ They all bring their earlier behaviour or, more importantly, responsibility, with them, but that is not (on this occasion, anyway) a boundary breach; the two spheres are intimately related, the second designed to follow the first, with the result that any crossing from the one to the other is perfectly legitimate. It is part of the shared meaning of a fine93 that it is allotted –​ indeed, only allotted  –​to those who have, in the past, behaved in a certain way. But surely Deltafina is entitled to complain that, for the same offence, the distribution of punishments should also be the same? Indeed it is, but such 89 Judgment of 9 September 2011, Deltafina SpA v European Commission, T-​12/​06, ECLI:EU:T:2011:441. 90 In eu Competition Law, participation in a cartel is contrary to Article 101 tfeu, which at the time of the case was Article 81 ec. 91 Judgment in Case T-​12/​06 Deltafina (n 89) para 289. 92 See supra, text accompanying n 35. 93 Taking the Standard Contingent Reply as read.

Semi-suspect and non-suspect grounds

271

discussion is for the Forum; it is up to the members of the distributive community as a whole to decide, amongst themselves, what the shared meaning of this particular sanction is, and therefore how it should be distributed. The inequality being argued over is, at worst, a small inequality; no actual boundary has been breached.94 6.4

Luck: Boundary Breach or Small Inequality?

There is a certain type of equality case wherein it is difficult (for the complex egalitarian) to tell if a boundary has been crossed. Or put another way, there is a certain type of inequality which it is hard to categorize as either a small inequality or the result of a boundary breach. This is inequality caused by luck. Luck is not often seen in the suspect grounds, where, usually, a great deal of deliberation has gone into the production of the inequality. However, it appears quite regularly in the semi-​suspect and non-​suspect cases. Just for the sake of completeness, then, it is proposed to consider briefly the concept of luck as it applies or might apply to cases coming before the cjeu. In Frico,95 an agricultural case under Article 40(3) ec,96 butter-​ manufacturers from different Member States found themselves in different situations as a result of varying interest rates. Specifically, the national intervention agencies afforded their domestic butter-​ manufacturers financing to enable them to store surplus butter, or at least to recoup the costs of past storage. In the uk, payments were made with a 9.5% interest rate, while in Germany and the Netherlands, payments were made with a 7% interest rate. Frico, a Dutch butter-​manufacturer, was concerned that rivals in a high-​ interest Member State might obtain loans in a low-​interest Member State, so that when they were reimbursed they would actually make a profit; it therefore called for a uniform reimbursement rate of 7%. The Court was unsympathetic,

94

95 96

Space prohibits further examples, but other competition cases where equality was at issue include: Case T-​21/​99 Dansk Rørindustri A/​S v Commission of the European Communities [2002] ecr II-​1681; Case T-​354/​94 Stora Kopparbergs Bergslags ab v Commission of the European Communities [2002] ecr II-​843; Case C-​291/​98 P Sarrió sa v Commission of the European Communities [2000] ecr I-​9991; Case T-​308/​94 Cascades sa v Commission of the European Communities [2002] ecr II-​813; Case T-​86/​95 Compagnie générale maritime and Others v Commission of the European Communities [2002] ecr II-​1011. Joined cases 424/​85 and 425/​85 Frico (n 34). This was the old numbering for Article 34(2) ec, which, since the coming into force of the Treaty of Lisbon, is now Article 40(2) tfeu.

272 Chapter 6 holding that any such move by rivals was ‘offset by the risk, to which that practice exposes them, that the exchange rates of those currencies might change to their detriment,’97 and that the purpose of the scheme was to provide compensation for storage costs, not to improve competitiveness. From the point of view of complex equality, the arrangements with regard to the interest rate or rates used in the calculation of the repayments would certainly seem to be a matter for the distributive community. All of the national intervention agencies, the Commission (which issued the pertinent regulation), and the butter-​manufacturers, would have to decide between them how they wished the reimbursement payments to be worked out. However, assuming that something like the extant arrangement was arrived at, it is still a matter of luck for Frico if the interest rate in its (low-​interest) Member State falls to a point where it would become profitable (taking into account bank charges and so on) for a rival in a high-​interest Member State to cross the border to obtain its financing. The rate is set in a different sphere by the national bank of the Netherlands.98 But then the same could be said, mutatis mutandis, for the rival. A uk company, for example, would be dependent for its rate on the Bank of England. The ‘national rate’ cannot be a monopolizable dominant if everyone has one! The solution arrived at, for the purposes of this book at least, is to distinguish between the luck occurring at the moment of distribution (which will be referred to, simply for linguistic ease, as ‘today’s luck’), and luck occurring in the past (‘historical luck’); this seems the best resolution to the problem, and the most in keeping with Walzer’s theory in general. The coin-​toss at the beginning of the first football match of the season will determine which team plays in which direction, in the first and second halves, that day. Noone would seriously suggest that it was unjust. And so, technically, the same, perfectly fair coin-​toss could be used to determine the same matter for every game of the season; why bother with another at the beginning of the subsequent games? Fortuna has spoken, and that should be that. However, the more time passes, the more the team that first plays, say, left to right, can come to rely on this as an entrenched right. On a day when the sun would be shining particularly harshly on the team first playing right to left, the latter team may decide simply to default, without playing at all. The other team has now brought the result of the original coin-​toss with it into the changing rooms, and used it to claim .

97 98

Joined cases 424/ 85 and 425/ 85 Frico (n 34) para 14. Obviously the case, referred to the Court in 1985, predates the introduction of the euro, although if the rival is in the uk, inter-​currency speculation would still be possible today.

Semi-suspect and non-suspect grounds

273

instant victory. In other words, over time, it has become a dominant. Clearly a coin-​toss at the beginning of every game is fairer. It is still only luck which decides, but something about today’s luck is special. It is a new day and a new match; a new appeal to Fortuna is called for. In Walzerian terms, the new day is like a new sphere. The meaning of this for those attempting to use the theory in practice is reasonably straight-​forward. In no situation, and therefore in no distribution, can every eventuality be planned for and regulated. Something is always ‘left to chance.’ Everyone is equally affected by that ‘chance,’ whether it manifests itself on that particular day positively or negatively, or differently for different people. Everyone accepts it as an unavoidable risk inherent in human life.99 Therefore, for a given distribution, the ‘good’ or ‘bad’ luck100 (if it can even be characterized as such, a priori) brought into the distributive sphere by the various distributees, can in no way be regarded as a (positively or negatively) dominant good. Whatever result it produces, as in Frico, is at most a small inequality, impliedly part of the distributive principle, and intrinsic to that distribution, that day, in that sphere.101 However, as soon as it has manifested itself, it becomes vulnerable to being converted, or to converting itself, into a dominant, and over a longer period of time the specific piece of luck (being born to wealthy parents, for example) can most definitely be used tyrannically. Unless it is explicitly part of the shared meaning of a good, its presence in the sphere in which that good is being distributed would then represent a boundary breach.102

99

Although there are of course some philosophers, such as Dworkin, who do wish, if not to avoid luck altogether (of the type that he calls brute luck), then at least to mitigate its effects via compensation for the unlucky. Rakowski has developed an entire theory of equality of fortune: ‘[O]‌nly by eliminating unsolicited good and bad fortune from the forces shaping people’s holdings can true equality of station be attained, and the distribution of resources be made acceptable to a community of persons who regard one another as moral equals. […] Undeserved, unwagered, unchosen inequalities warrant redress’. Quotation from Eric Rakowski, Equal Justice (Clarendon Press 1991) 75–​76. 100 As in luckiness generally, not a specific piece of luck. 1 01 More recently, Walzer has confirmed that luck may act as a distributive criterion: M Walzer, ‘Feminism and me’ (2013) 60(1) Dissent 50, 51. 102 Already in this book it has been seen how a person’s attributes, which are another piece of historical luck, are to be regarded as having been separately distributed, in a separate sphere. The bringing of them into a second sphere represents a border-​crossing, legitimate or illegitimate, depending on the shared meaning of the good being distributed. See supra section 3.3.

274 Chapter 6 Returning to the case of Erpelding,103 the outbreaks of mastitis and Canadian flu which caused Mr Erpelding’s farm to do badly between 1981 and 1983, were matters of sheer luck; indeed, one of his arguments before the Court was force majeure. He has a good point here, even though the Advocate General accused him of trying to rely on ‘an “invented” provision.’104 Could he not say to the Luxembourg authorities that their perfunctory use of the period 1981–​ 1983, when distributing quotas, was unjust, as it forced him to bring into the distributive sphere a piece of historical bad luck, belonging to a separate time and a separate sphere? This piece of historical bad luck, he could continue, was unrelated to, and not part of the meaning of, a quota;105 its presence in the sphere governing the distribution of quotas was tyrannical. Alternatively, he could separate the historical bad luck from the history itself, and claim that, whilst farmers’ performances between 1981 and 1983 were part of the shared meaning, any luck elements were not. This would be the equivalent of having a defence of force majeure. The final outcome cannot be known until the distributive community has spoken, but it can at least be seen here how historical bad luck inhabits its own sphere, and that its importation into a second sphere is not automatically just, but is contingent on the say-​so of the community.106 6.5

Concluding Thoughts

During this final stop of the tour, the thesis that Michael Walzer’s theory of complex equality could be used by the Court of Justice when dealing with cases concerning equality, as a complement to the Aristotelian like-​for-​like formula, has been once again thoroughly tested via counterfactual analysis of the Court’s existing case-​law, firstly in respect of two types of discrimination explicitly outlawed by the Treaty which have been referred to as ‘semi-​suspect’ (discrimination between agricultural producers, and discrimination between public and private undertakings), and secondly deriving from three additional contexts where appeals to the general principle of equality are often made (public procurement, human resources and competition law). These three contexts, wherein discrimination is not explicitly banned by the Treaty but 103 104 105 106

Case 84/​87 Erpelding (n 24) and accompanying text. ibid, Opinion of AG Darmon, para 5. As explained above, that would be a matter for the entire community. The same analysis would apply to the milk producers whose production had not increased between 1991 and 1993 in Karlsson:  Case C-​292/​97 Karlsson (n 30)  and accompanying text.

Semi-suspect and non-suspect grounds

275

rather falls to be dealt with via the general principle, were referred to as ‘non-​ suspect.’ The section ended with a brief consideration of the role played by luck within Walzer’s theory. As with earlier chapters, the analysis showed that there were some cases where a Walzerian approach from the Court’s Bench would have produced a different result to that actually produced on the strength of Aristotelian reasoning alone. In the semi-​suspect section, for example, it was seen how Walzer’s nuanced approach would mitigate the harsher edges of some bad luck suffered on the part of Farmer Erpelding, a contextual nicety far too specific for Aristotle’s sledgehammer approach.107 Similarly, in the non-​suspect section, it was discovered that Walzer’s theory would help the parents of children ‘treated as if they were dependent children’ to enjoy the same rights to travel expenses for familial visits as the parents of children whose dependency was actual, rather than analogous.108 Again, the contextual finesse permitted by Walzer’s approach would allow the social understanding of such travel expenses to be placed front and centre, such that adult dependency brought about by disability would not need to be equated with adult dependency brought about by simple failure to secure employment, and the seemingly punitive rationale for shifting the burden of paying for the journey from the institution to the child themselves could be overcome in a situation where ‘blame’ for the continued dependency did not in fact lie with the child, or indeed with anyone. Finally, it seemed that Walzerian reasoning could help the unnecessarily demoted ‘national expert’ in Dupré to gain the same treatment as regular eu civil servants, in circumstances where the Aristotelian test was being used as little more than a judicial figleaf to legitimise an already questionable alterity between the two set out in the Staff Regulations, the judges corralled by the test’s harsh logic to endorse the inevitable disparate impact which such legislative prejudice produced.

1 07 Case 84/ 87 Erpelding (n 24). 108 Case T-​48/​89 Beltrante (n 70).

Chapter 7

Reflections 7.1 Introduction In order to reach a conclusion as to whether Michael Walzer’s theory of complex equality can be used by the Court of Justice of the European Union in its equality case-​law as a complement to the Aristotelian test, it is necessary to look first at three counter​arguments, which arise from the case-​study but which are not specific to any one type of ground. These are the need for judicial interventionism, the possible need for expert evidence, or even specialized courts, and the need to rewrite legislation. They will be dealt with in turn. 7.2

Three Counter​arguments

7.2.1 The Need for Judicial Interventionism The cjeu judge, at least when making a preliminary ruling pursuant to Article 267 tfeu (ex Article 234 ec, and before that Article 177 EEC), is conventionally required to pronounce upon the relevant law only and to distance himself or herself from the actual facts of the case. They are a matter for the referring judge, that is, the judge in the national court which made the reference. If the point comes up, the Court usually says something neutral, like: It is established case-​law that, in the procedure laid down by Article 177 of the Treaty providing for cooperation between national courts and the Court of Justice, it is for the latter to provide the referring court with an answer which will be of use to it and enable it to determine the case before it.1 However, sometimes it hints that the boundary between law (cjeu) and fact (national court) may not be so clear-​cut: It follows that in such a case it is for the national court, and if necessary, the Court of Justice to appraise the application of those provisions having 1 Case C-​334/​95 Krüger GmbH & Co. kg v Hauptzollamt Hamburg-​Jonas[1997] ecr I-​4517, para 22.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/​9 789004354265_​0 08

Reflections

277

regard to all the rules of Community law, including freedom of expression […] as a general principle of law the observance of which is ensured by the Court.2 For the Court, it is a ‘porous’ boundary, as Davies has noted.3 Meanwhile, the legislative draughtsperson is more decisive, for example: The appreciation of the facts from which it may be inferred that there has been direct or indirect discrimination is a matter for national judicial or other competent bodies, in accordance with rules of national law or practice.4 Under the Aristotelian regime, which, as has been seen many times, requires only a superficial analysis of the circumstances of the case, the allocation of competence as between the national and European courts has not usually been a problem, although the cjeu judge has from time to time strayed into the supposedly forbidden territory, often during the objective justification stage in a case of indirect discrimination. Such straying is a form of judicial activism –​a judge venturing into areas beyond his remit.5 Perhaps this particular form is better described as judicial ‘interventionism.’6 Under complex equality, meanwhile, the circumstances are all. What is being distributed, and by and to whom? What is the shared meaning of the distribuend to the distributive community? What distributive principle is being used and is it in keeping with the meaning ascertained in the previous question? All these matters need resolving, and they are unlikely to be resolved without 2 Case C-​260/​89 Elliniki Radiophonia Tiléorassi ae and Panellinia Omospondia Syllogon Prossopikou v Dimotiki Etairia Pliroforissis and Sotirios Kouvelas and Nicolaos Avdellas and others [1991] ecr I-​2925, para 44 (emphasis added). 3 G Davies, Nationality Discrimination in the European Internal Market (Kluwer 2003) 36. 4 Council Directive 2000/​78/​e c of 27 November 2000 establishing a general framework for equal treatment in employment and occupation [2000] oj L303/​16 (‘Directive 2000/​78’ or the ‘Framework Employment Directive’), recital 15. The question of which court is competent to assess the facts is further discussed in A Arnull, ‘Annotation Arsenal Football Club plc v Matthew Reed’ (2003) 40(3) cml Rev 753, and G Davies, ‘The Division of Powers between the European Court of Justice and the National Courts’ (Webpapers on Constitutionalism and Governance beyond the State No 3, conWEB 2004). Both of these contain further interesting references. 5 Although it is not the most notorious form, which of course is where the judge tries to usurp the role of the legislator. 6 Davies (n 4) 28.

278 Chapter 7 a thorough airing of the facts of the dispute. To take a concrete example, in the case of Gillard, which was briefly examined in Chapter 5,7 and where the good at issue was a special old age pension, Mayras AG put the matter very clearly: Naturally it is not the Court’s function to classify the French provision in question with regard to Community law or indeed to inquire whether the law in question is intended to confer benefits strictly in the nature of compensation or whether it also has the objective of compensating ex-​ servicemen and prisoners of war ‘for the hardships undergone by them and the services rendered by them to the country.’8 It was not for the cjeu, according to the Advocate General, to ask what the good was for, in other words, to enquire as to its meaning. But such an enquiry would be the first port of call for a Walzerian analysis. There is no doubt that a Walzerian disposal of this case would look quite different from an Aristotelian one. However, to take a more complex example, in Dirk Rüffert v Land Niedersachsen9 (a kind of sequel to the Viking and Laval cases),10 the ‘good’ was a German obligation to pay posted workers in Germany the same wages as local workers.11 Even for a Court utilizing an Aristotelian approach, the meaning of the good here was absolutely vital. Was it to protect the posted workers, to give them genuine additional help by ensuring that they received a wage that was significantly higher than their ‘home’ wage?12 Such beneficence on the part of the German authorities might well qualify as an objective justification (for breaching the incoming company’s free movement rights),13 allowing Land Niedersachsen to win the case. Or was it to protect German building undertakings from competition, by discouraging foreign undertakings from moving 7 8 9 10

11 12 13

Case 9/​78 Directeur régional de la Sécurité sociale de Nancy v Paulin Gillard and Caisse régionale d’assurance maladie du Nord-​Est, Nancy [1978] ecr 1661. See supra section 5.7. ibid, Opinion of AG Mayras, section II (Mayras’ emphasis). Case C-​346/​06 Dirk Rüffert v Land Niedersachsen [2008] ecr I-​1989. Case C-​438/​05 International Transport Workers’ Federation and Finnish Seamen’s Union v Viking Line abp and OÜ Viking Line Eesti [2007] ecr I-​10779; Case C-​341/​05 Laval un Partneri Ltd v Svenska Byggnadsarbetareförbundet, Svenska Byggnadsarbetareförbundets avdelning 1, Byggettan and Svenska Elektrikerförbundet [2007] ecr I-​11767. The distribuend could also be the equal wage itself. Once again, see how it is at the objective justification stage of a case that temptation to consider the facts is strongest. The Advocate General (Bot AG) thought it was.

Reflections

279

in the first place? Such a goal could never qualify as an objective justification (measures restricting free movement cannot be justified by economic aims), and for the Court to take this line would spell defeat for Land Niedersachsen. Thus the entire outcome of the case turned on the meaning of the good to the distributor and to the potential distributees –​a factual issue. In the course of his opinion, the Advocate General (Bot AG) often declared that this was a task for the domestic judge: It is for the national court, which enquires as to the true objective pursued by the legislature, to determine whether, viewed objectively, the rules in question ensure the protection of posted workers or, more generally, the prevention of social dumping.14 But he went on to offer his own painstaking analysis of the meaning, anyway. In its judgment, the cjeu did not even waste time with a bogus ascription of roles, but went through and pronounced upon all of the proposed justifications itself, ending with a proclamation (that the German measure was contrary to Community law) which would of course be binding on the referring court, leaving it with no say in the facts at all.15 The point is that judicial activism of this sort is already so prevalent,16 that it is sometimes hard to see how a switch to a Walzerian equality-​model would actually make anything worse. Certainly, use of Aristotelian reasoning did not prevent judicial interventionism in equality cases like Walloon Waste,17 or Lewark (where the Court indulged in a large amount of arguably inappropriate fact-​finding),18 or Bickel and Franz.19 However, it cannot be argued that, by using complex equality, the cjeu would

14 15

16 17 18 19

Case C-​346/​06 Rüffert (n 9) Opinion of AG Bot, para 115 –​footnote omitted. Bot AG even puts this allocation of duties into his final conclusion at para 136. The Court several times refers to the ‘case-​file submitted to the Court’ (for example, paras 40 and 42), as though to underline that it has looked at everything before drawing its conclusions. But were all of the factual data sent to Luxembourg in the first place? They were not in the Arsenal case, discussed by Arnull (n 4). Under a Walzerian regime they would be, as it would be clearly understood that they were required. For another example, see Arnull’s commentary on the Arsenal case: Arnull (n 4). Case C-​2/​90 Commission of the European Communities v Kingdom of Belgium [1992] ecr I-​4431. Case C-​457/​93 Kuratorium für Dialyse und Nierentransplantation eV v Johanna Lewark [1996] ecr I-​243. Case C-​274/​96 Criminal proceedings against Horst Otto Bickel and Ulrich Franz [1998] ecr I-​7637. See discussion of this case at supra, section 5.8.

280 Chapter 7 have to upset the jurisdictional balance as between it and the national court, any more than it is already upset. On the plus side, the issue is only relevant to the preliminary reference system. In challenges to Community legislation under Article 263 tfeu (ex Article 230 ec), for example, the cjeu itself would be the fact-​finding tribunal, and it could then investigate the facts of the case as thoroughly as it wanted without censure. Furthermore, Rasmussen has pointed out that forays by the cjeu into factual matters can have beneficial consequences: [Y]‌et another means of controlling the overall acceptability of judicial outcome might be the taking of steps, judicial or otherwise, aiming at enhancing the Court’s access to socio-​economic fact. […] [T]he absence of such fact[s] occasionally entails catastrophic consequences, and […] the usage of socio-​economic fact in recent case law ha[s] triggered important salutory effects. […] [O]nly by generalizing the availability of such fact might the Court, as a rule, make safe choices between various decisional alternatives because it will have been enabled to access fully the implications and ramifications of all of them.20 Thus, the Court’s using fact can have positive effects; in some cases, it could even be argued that, without access to fact, the Court’s contribution is deracinated and effectively worthless.21 But, if still thought necessary, solutions to the problem of judicial interventionism are available. A first possible solution can be extrapolated from Walzer’s general preference for particularism (not to mention the doctrine of subsidiarity). As it is the local community whose shared meaning is at issue (usually) in an Article 267 tfeu22 reference, it is the national court which is best placed to host the final adjudication, if such an adjudication is

20 21

22

H Rasmussen, On law and policy in the European Court of Justice: a comparative study in judicial policymaking (Kluwer 1986) 512. Certainly Hosking is of that view in relation to disability discrimination: ‘The justifications which will almost certainly be offered for discriminating against disabled people relate to safety, business necessity, the costs of accommodation and the rights of others. If the Court does not insist on a sufficiency of evidence to seriously look behind these justifications … the effectiveness of [Directive 2000/​78] will be considerably reduced even within the narrow scope of its intended reach’. See DL Hosking, ‘Great Expectations: Protection from discrimination because of disability in Community law’ (2006) 31(5) el Rev 667, 678. Ex Article 234 ec.

Reflections

281

needed.23 However, rather than shying away from the facts of the case and simply batting it back down to the referring court with a bland and possibly even useless ruling attached,24 it is submitted that the cjeu should have a sort of ‘first look’ at the facts and deliver an interim opinion or view (to put it no stronger) as to the nature of the shared meaning, the distributive principle arising therefrom, whether or not such distributive principle is in fact being used, and thus whether or not complex equality has been violated. While any legal rulings which it made would be supreme as usual (including with regard to the second of these four matters, where the issue of which distributive principle goes with which shared meaning might be regulated by law), the Court’s opinion or view on the first, third and fourth of these matters would be advisory only. The national court could use it as a guide, but would retain competence to make the final decision. The sort of mini-​trial which the cjeu would have to stage here is not uncommon in the law –​such hearings may be observed in many different courts, during the admissibility stage of an appeal, for example, or where a judge is called upon to rule on the need (or otherwise) for interim measures. In both these examples, courts have to steer a careful middle course between the necessity for a just decision based on all of the relevant issues, and the desire, usually on all sides, not to perform such an accurate dress-​rehearsal of the future (or possible future) trial that the need for such a trial is obviated altogether. There would seem to be no reason at all why the cjeu should not be able to steer such a course. A second possible solution is that only the national court should involve itself in the ascertaining of the shared meaning, with the cjeu perhaps given a new cassation-​style role (which would involve making it hierarchically superior to the national courts). It could then review the domestic court’s decision. If it felt that the national court had worked out the shared meaning or the distributive criterion wrongly, or that its conclusion as to whether or not there had been a breach of complex equality was flawed, it could send the matter back down to

23

24

For example, whether or not the German populace would be confused by a cosmetic product’s being called ‘Clinique’ is surely a question which only the German populace itself can definitively answer: Case C-​315/​92 Verband Sozialer Wettbewerb eV v Clinique Laboratoires snc et Estée Lauder Cosmetics GmbH [1994] ecr I-​317. But the reference in this case was still necessary to establish that the measure at issue was a product rule and not a selling arrangement. And of course the reverse is also true. If Community legislation or behaviour was at issue, it would be the cjeu judge who would be best placed to adjudicate. The cjeu would probably (although not necessarily) be the best choice in staff cases as well. Also a common occurrence under the Aristotelian test.

282 Chapter 7 the same court for a retrial. Minus the Walzerian element, this is Gareth Davies’ preferred remedy for the law-​fact problem.25 A third and final solution would be that only the cjeu should involve itself in the ascertaining of the shared meaning. This would only work if it was considered better for the adjudicator or referee himself to be from outside the distributive sphere concerned.26 As long as the Forum was faithfully recreated in the European Court, with all views heard (and examined) in full, then the foreign location should not alter anything. In fact, the ‘fairy-​tale Grand Duchy’27 might prove to be an ideal venue for a climactic showdown between two rival factions. Whichever solution is chosen, the new arrangement will have the added advantage of bringing to an end the current, intolerable situation. By spelling out once and for all the allocation of duties as between the cjeu and the national court, things might become clearer and more predictable. 7.2.2 The Possible Need for Expert Evidence, or Even Specialized Courts Following on from the previous argument, it is reasonably easy to imagine a judge at the cjeu deciding on the first question mentioned there, having heard counsel on both sides, or even by simply reading the pleadings. The second and third questions, though (the third by virtue of being contingent on the second), pose a problem. How is the shared meaning to be discovered? To be sure there is nothing to stop counsel reporting what they think it is, which would of course be the shared meaning most favourable to their client. But with both advocates and the judge or judges unlikely to be members of the relevant distributive community themselves, it would be a violation of complex equality to allow any of them to definitively pronounce upon the meaning:28 that is a job for the Forum and the Forum alone, assuming it can reach an agreement. However, if (or more likely when) there is a conflict, and the fact that an action

25 26 27

28

Davies (n 4) 26ff. The matter is debated at section 7.2.2, immediately below. GF Mancini, ‘The Making of a Constitution for Europe’ (1989) 26 cml Rev 595, 597, although Stein may have been the first to use such imagery, referring to the Court as ‘[t]‌ucked away in the fairyland Duchy of Luxembourg:’ E Stein, ‘Lawyers, judges and the making of a transnational constitution’ (1981) 75 American Journal of International Law 1, 1. It is true that the advocates’ contributions may simply be a repetition of the views of their clients, who almost certainly are members of the distributive community at issue. But they might just as easily be part of a legal strategy dreamt up by the advocates on their own.

Reflections

283

has been brought is almost definitely proof that there is indeed a conflict, it does not seem unreasonable for an arbiter (or ‘referee’) to be charged with resolving the argument one way or the other. Such an arbiter might come from inside the pertinent sphere, or he or she might come from outside it –​there are advantages and disadvantages to both situations. The internal arbiter would have local knowledge (he or she would be a kind of relative of Walzer’s ‘connected critic’), but both parties to the conflict might suspect him or her of bias. Meanwhile, the external arbiter is less knowledgeable, but more impartial. Nevertheless, if an external arbiter were to be used, he or she would still have to have the approval of the members of that sphere.29 Nor would any decision arrived at by the arbiter need to be the final word on the particular question at hand; debates on meanings (as has already been seen)30 are ongoing, so the arbiter’s ruling would really only be a way of settling, in the short term, the argument in front of him or her, and might well only hold until the next conflict arose.31 It follows that the cjeu judge or judges may legitimately hear submissions on the (disputed) meaning of the distribuend, prima facie without violating Walzer’s theory. However, it is also plain that these submissions should not just contain the advocates’ view on the meaning, or even their clients’ view on the meaning, but must consist of nothing less than a presentation, to the Court, of the distributive community’s view on the meaning (or at least said view as discerned by the party in question).32 This means that convincing evidence would have to be adduced, not just of the proposed meaning itself, but also of the number of supporters which that meaning had locally.33 Only after hearing 29 30 31 32

33

Quaere whether, in the case of the cjeu, this would lead to the politicization of judicial appointments. See supra section 2.4. But this is really not that dissimilar to the way in which the cjeu, and national courts within the eu, already operate today. In Grant, for example, the Court implied that there had been no advancement at Member State level of views on the equivalence of same-​sex partnerships and opposite-​sex partnerships: Case 249/​96 Lisa Jacqueline Grant v South-​West Trains Ltd [1998] ecr I-​621 –​ see in particular paras 32 and 35. As Mancini and O’Leary point out, ‘it will be regretted that further and more detailed evidence of this lack of evolution was not adduced’. See GF Mancini and S O’Leary, ‘The New Frontiers of Sex Equality Law in the European Union’ (1999) 24(4) el Rev 331, 351 (emphasis added). It is not just a numbers game, however. As mentioned at 4.2.2(a), advocates would also be able, if they wanted to, to use their written pleadings and oral submissions to request that the Bench apply the Override, possibly to the meaning which they themselves had adduced, but more likely to the meaning adduced by the other side, if and when the latter meaning was accepted. As the European Court of Human Rights itself has put it,

284 Chapter 7 both sides’ presentations, in the course of a fully adversarial proceeding, could the judge properly adjudicate as to which meaning was, for the time being anyway, the ‘true’ one.34 From all this, it is obvious that some expert evidence might need to be adduced. While representative members of the distributive community could be called to testify as to what Walzer might call the view from the cave,35 or indeed they might opt to appear of their own volition,36 more specialized witnesses

34

35 36

‘Although  individual interests must on occasion be subordinated to those of a group, democracy does not simply mean that the views of a majority must always prevail:  a balance must be achieved which ensures the fair and proper treatment of people from minorities and avoids any abuse of a dominant position’: Leyla Şahin v Turkey App no 44774/​98 (ECtHR, 10 November 2005) para 108. Common lawyers have for many years complained that the cjeu is not adversarial enough, and that it tends to adjudicate on points that were not argued before it: L Neville Brown and T Kennedy, The Court of Justice of the European Communities (Sweet & Maxwell 2000) 402. One of the early Advocates General (Verloren van Themaat AG) also called for an increased use of argument: Case 94/​84 Office national de l’emploi v Joszef Deak [1985] ecr 1873. In the Advocate General’s opinion, the Court should not take a position on whether the granting of a tideover allowance to Mr Deak (a Hungarian national who happened to also be the son of a Community migrant worker) would or would not break up the national system, without hearing argument first. The Court ignored the Advocate General, and clearly decided that it would not. Verloren van Themaat’s approach was arguably a proto-​Walzerian one, asking effectively for an investigation into the local community’s understanding of the tideover allowance (what and who was it for?), before deciding whether or not Mr Deak should receive it. He was correct that such an investigation involves the Court hearing arguments from all ‘interested parties’ (at section 2.3) and he included other Member States, whose nationals might of course be distributees, or, in Mr Deak’s case, parents of distributees. Such a debate is vital. The Court –​at least under orthodox complex egalitarian theory –​cannot simply invent an understanding off its own bat. See sections 9.4 and 9.5 below for more on the vexed question of truth. See the discussion at M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) xiv. Indeed, in the uk, such third party interventions are becoming quite common in judicial review proceedings:  Roger Smith, ‘Benign intervention’ ls Gaz (London 12 November 2009) 10. They are already common in the us. As Moore and Stambolis-​Ruhstorfer have written, ‘[t]‌he political and scientific are not unrelated’: MR Moore and M Stambolis-​ Ruhstorfer, ‘lgbt sexuality and families at the start of the twenty-​first century’ (2013) 39 Annu Rev Sociol 491, 502. Stambolis-​Ruhstorfer has spoken about the ‘scientificization’ of the rights debate in the us, and the huge impact which expert evidence has had and is continuing to have on the outcome of equality cases there:  M Stambolis-​ Ruhstorfer, ‘The scientific and legal recognition of gay and lesbian parenting in France and the us: How law and knowledge influence each other’ (Presentation given at the 22nd

Reflections

285

might be required, such as those whose job it is to take polls or surveys in the community in question, or, depending on the nature of the distribuend, historians, anthropologists, and so on. Some of the purported meanings themselves might be complex, for example, in the tax field, necessitating the appearance in court of economists, actuaries, and the like. Ultimately, it may even be thought better to have separate panels of specialist judges, each dealing with a different sort of distribution.37 The consequences of these developments, both in terms of expense and duration of cases, would be not insubstantial. It is thus an argument against the cjeu’s using complex equality that to do so would put financial pressure on both litigants and the eu taxpayer, and would exacerbate already lengthy backlogs. On the other hand, there are a number of arguments in favour of the admission of specialist evidence. Specialist witnesses can assist the judges on technical questions which may arise, thus leading to a better quality of assessment.

37

International Conference of Europeanists, Paris, 9 July 2015). A  major example is the case of Deboer v Snyder, in which a lesbian couple challenged Michigan’s ban on same-​ sex marriage, a challenge which ultimately brought about the legalisation of same-​sex marriage across the country. As perusal of the first instance judgment makes clear, a large number of experts testified, on both sides, at the trial, some more controversially than others: Deboer v Snyder, 973 F.Supp.2d 757 (United States District Court, ED Michigan, Southern Division, 21 March 2014). Such specialist judicial panels are already foreseen by Article 225a ec (introduced by the Nice Treaty), which, since the coming into force of the Treaty of Lisbon, is now Article 257 tfeu. The only such panel set up thus far is the European Union Civil Service Tribunal. For more on this, see N Lavranos, ‘The new specialised courts within the European judicial system’ (2005) 30(2) el Rev 261. The prospects for the establishment of further specialist courts seem bleak. The Court took the occasion of a mid-​2015 reform of the General Court to comment, in a press release, that ‘[t]‌he creation of specialized courts would not constitute a viable alternative [because they] are not flexible’: Council of the European Union, ‘Q&A on the reform of the General Court’ (Press release en 497/​15 of 23 June 2015) 1. By this, it seemed to mean the inflexibility of such courts in the face of a growing case-​load, not that specialist judges were peculiarly rigid in their development of the law. The press release went on to list three further disadvantages of the creation of specialized courts, namely, the attendant risk of inconsistency (as between courts, not as between judgments of the same court), as well as their complexity and their costliness: ibid. Nevertheless, some have described the Unified Patent Court, a specialist intellectual property court set up by international agreement in December 2016 and intended to have major divisions in Paris, Germany and the uk, as an ‘arm of the Court of Justice of the eu:’ M Walters, ‘ukip mp’s patent court warning fails to find support’ (LS Gaz, 24 February 2017)  accessed 21 May 2018.

286 Chapter 7 Furthermore, many other courts and tribunals around Europe already make use of experts; the wto Panel, for example, is in a regular dialogue with experts, and at the German Federal Patent Court in Munich, it is common to see chemists on the Bench. Even as things stand (that is, without the adoption of a Walzerian approach to equality), it is not certain how long the cjeu can maintain its current iura novit curia stance. The Court has already used its power to call for an expert opinion in staff and competition cases;38 why not cases concerning equality?39 As already mentioned, third parties from the distributive community who hear of the litigation might also apply to intervene in the case, to give the Court further guidance on the meaning of the distribuend.40 By widening the debate out, so that it (rightly) goes beyond the individual concerns of the two parties to the case, democracy too is arguably enhanced.

38

39

40

Case 12/​68 X v Audit Board of the European Communities [1969] ecr 109 (report ordered on official’s mental state); Case 48/​69 Imperial Chemical Industries Ltd v Commission of the European Communities [1972] ecr 619 (report ordered on the market in dyestuffs). See A Alemanno, ‘Science & eu Risk Regulation: The Role of Experts in Decision-​ Making and Judicial Review’ in E Vos (ed), European Risk Goveranance –​Its Science, Its Inclusiveness and Its Effectiveness (Connex Report Series No 6, 2008)  accessed 21 May 2018, 23 for further examples. Alemanno believes that, after many years of reticence, the Court is now showing ‘some readiness to become more involved in the examination of scientific evidence:’ ibid 20. Vos has written that the Court’s recent shift to a more interventionist review of the relevant science in a given case is to be welcomed, although, in her view, in the agriculture case of Pfizer, the Court ‘overstepp[ed] its mandate’ and became ‘immersed in the quagmire of scientific uncertainty and complexity surrounding [the] dispute,’ acting as a ‘kind of “super expert:” ’ E Vos, ‘The European Court of Justice in the face of scientific uncertainty and complexity’ in M Dawson, B de Witte and E Muir (eds), Judicial activism at the European Court of Justice (Elgar 2013)  142, 165, 145 and 144, referring to Case T-​13/​99 Pfizer Animal Health sa v Council [2002] ecr II-​3305. Rather than having the Court ‘plunge (…) into the science,’ she notes, it may be better to put up, from time to time, with ‘what is, from a scientific perspective, an incorrect result:’ ibid, 144 and 166. The us experience would suggest that the real challenge for the judge, in acting as scientific gatekeeper, comes in knowing when to allow an expert’s testimony and when to disallow it as potential ‘junk science.’ However, Breyer J’s concurring opinion in gec v Joiner, cited by Vos, concludes that, provided there is a ‘cooperative effort’ as between the scientific and legal communities, ‘[the] gatekeeping requirement will not prove inordinately difficult to implement’: gec v Joiner 118 S.Ct. 512 (Supreme Court of the United States, 15 December 1997); the concept of ‘junk science’ is discussed in Stevens J’s partial dissent. The cjeu already has power to allow third parties to intervene in cases before it. See, for example, Case T-​37/​04 Região autónoma dos Açores v Council [2008] ecr II-​103, where

Reflections

287

7.2.3 The Need to Rewrite Legislation Another consequence of the Court’s adopting a Walzerian approach to equality to sit alongside the current Aristotelian one would be that certain legislation might have to be rewritten, or even scrapped, especially where it sought to recreate or codify the Aristotelian test itself. Examples of word-​for-​word codifications are admittedly rare. One such was Article 60(1) of the now-​defunct Treaty establishing the European Coal and Steel Community: Pricing Practices contrary to Articles 2, 3 and 4 shall be prohibited, in particular:-​unfair competitive practices, especially temporary or purely local price reductions tending towards the acquisition of a monopoly position within the common market;-​discriminatory practices involving, within the common market, the application by a seller of dissimilar conditions to comparable transactions, especially on grounds of the nationality of the buyer.41 Another example is to be found in the tenth recital in the Preamble to the new Equal Treatment Directive:42 (10) […] According to the case-​law of the Court of Justice, discrimination involves the application of different rules to a comparable situation or the application of the same rule to different situations.43 However, other instruments, such as the Employment Framework Directive44 and the Race Directive,45 contain provisions which are Aristotelian in spirit

41 42

43 44

45

a number of local fishermen’s associations intervened in a challenge, by the Azores, of a Council Regulation restricting the islands’ exclusive fishing rights. And now see Article 101(1)(d) tfeu (formerly Article 81(1)(d) ec). Directive 2002/​73/​e c of the European Parliament and of the Council of 23 September 2002 amending Council Directive 76/​207/​e ec on the implementation of the principle of equal treatment for men and women as regards access to employment, vocational training and promotion, and working conditions [2002] oj L269/​ 15. Footnote omitted. Council Directive 2000/​78/​e c of 27 November 2000 establishing a general framework for equal treatment in employment and occupation [2000] oj L303/​16 (‘Directive 2000/​78’ or the ‘Framework Employment Directive’). Council Directive 2000/​43/​e c of 29 June 2000 implementing the principle of equal treatment between persons irrespective of racial or ethnic origin [2000] oj L 180/​22 (‘Directive 2000/​43’ or the ‘Race Equality Directive’).

288 Chapter 7 if not actually in letter, for example, the definition of direct discrimination at common Article 2(2)(a). That is not to say that such instruments do not serve a useful purpose, for example, by setting out the scope of Community competence, but simply that they should not seek to petrify evolving shared meanings, for example, the shared meaning of jobs, the shared meaning of training, and so on. In most cases, though, the provisions in question simply require rewording, leaving room for case-​by-​case decision-​making. At least some of the red lines should be converted into orange lines;46 legislation such as this should point, but not imprison.47 46

47

It is not the intention of this book to advocate the abolition of red lines, though. First of all, there should always be a minimum of red lines, which can then serve as the limit beyond which society will not allow people to fall (or be pushed). This is a boundary the need for which Walzer himself acknowledges, and which he labels the ‘safety net:’ M Walzer, ‘Exclusion, Injustice, and the Democratic State’ (1993) 40 Dissent 55, 57. However, towards the end of this important 1993 essay, he entertains the idea of even greater encroachment by the nation (or perhaps in this case the supra-​nation) into the spheres of justice, specifically when autonomy has failed: ‘[e]‌xclusion is a sign that the [citizens’] contests have gone badly –​an invitation, therefore, to the state to set them right’; ‘[t]he state –​or, at least, the modern democratic state –​must defend the values of complexity and equality on behalf of all its citizens’: ibid 63 and 64. So it is not anti-​ Walzerian for a state or quasi-​state to pass more thorough legislation (that is, to create a more extensive network of red lines) where it is intervening in a distributive community whose members are, for one reason or another, foundering. As Gardner has put it, ‘The law of discrimination sits comfortably in [a] non-​individualistic theory of autonomy, according to which the state has its own project of providing the conditions of valuable flourishing for its citizens. In this theory, prohibiting … discrimination is merely a central and straightforward example of the state’s legitimate role’: J Gardner, ‘Liberals and Unlawful Discrimination’ (1989) 9(1) ojls 1, 22. Even McCrudden, who is on the whole critical of communitarianism, has written that it is a mistake to ‘assume that a system of rights leaves no room for benevolence or communal sentiment … A system of legal rights is not necessarily inconsistent […] with a belief in the pre-​eminence of “the good,” including a belief in constitutive community’:  C McCrudden, ‘Community and Discrimination’ in J Eekelaar and J Bell (eds), Essays in Jurisprudence (Third Series) (Clarendon Press 1987) 227–​8. Harris makes a similar plea in the gender equality context: ‘Even a jurisprudence based on multiple consciousness must categorize; without categorization each individual is as isolated as [Borges’ character] Funes, and there can be no moral responsibility or social change. My suggestion is only that we make our categories explicitly tentative, relational, and unstable, and that to do so is all the more important in a discipline like law, where abstraction and “frozen” categories are the norm’. See Angela Harris, ‘Race and Essentialism in Feminist Legal Theory’ (1990) 42 Stanford Law Review 581, 586.

Chapter 8

Presenting a Theory of Mediated Complexity 8.1 Introduction The ascertainment of the shared meaning of a given distribuend, and in particular whose job this should be, remains the thorniest issue in the book thus far. There was some discussion in Chapter 2 of a ‘Forum’ consisting of all members of the relevant distributive community –​distributors and distributees alike –​ and these terms, ‘Forum’ and ‘distributive community,’ have been used in all subsequent argument. Up until now, no further attempt has been made to describe the precise modalities involved in the establishment, running or regulation of these fora, other than some basic observations, for example, that these would vary from distribuend to distribuend and from community to community. In this respect, it will be remembered, Walzer’s own lead is being followed, as he too preferred to leave his theory in the abstract, refusing to put any practical flesh onto the theoretical bones. Perhaps he worried that the flesh might crush the bones, and that complex equality, once removed from the rarefied and controlled atmosphere of the laboratory, would quickly disintegrate in the harsher conditions of the real world. Perhaps he simply wished to dodge the issue. Even Marx famously refused to set out in precise detail the exact logistics required for a real-​world implementation of his theory, dismissing this as: writing recipes … for the cook-​shops of the future.1 However, as this book set out to assist a genuine court with a genuine problem, it is appreciated that it will have reneged on its promise if it delivers only in word and not in deed. Some limited comments on the nature of the Forum will thus be made in the second section of this chapter: section 8.2. This will subdivide into subsections addressing the Forum’s possible design, and the Forum’s possible danger. Even then, there will be those who say that the theory remains too wooly to be of any practical use, at the cjeu or elsewhere. There will be still others who fear that complex equality would lead to a kind of Mob Rule, with a vociferous

1 K Marx, Capital: A critique of Political Economy Volume One (B Fowkes tr, Penguin 1976) 99 (this phrase appears in the Postface to the Second Edition, dated 24 January 1873).

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/​9 789004354265_​0 09

290 Chapter 8 majority bullying minorities into silence, or out of existence. What if the distributive community votes against benefits for migrant workers, picking at the stitches of the very Union itself? What if the distributive community votes against women on maternity leave receiving full pay? What if the Override is not strong enough to temper the wilder excesses of the demos? It is for these people that an alternative theory, that of mediated complexity, is proffered in the third section of this chapter, section 8.3. Since most versions of the two theories involve a degree of societal input, sometimes a significant degree, the important topic of the judge’s role vis-​à-​vis his or her society will be considered last of all (section 8.4). The reason why this issue comes particularly to the fore in this chapter is that, as will be seen, the alternative theory of mediated complexity brings with it a considerably enhanced role for the judge. 8.2

Further Consideration of the Forum, and Some Possible Problems

8.2.1 The Design of the Forum The first thing for a given distributive community to do, even if it is an entire Member State, is to design its Forum. For example, would all the citizens of the country gather together in one place, where the distribution under discussion was one which affected everyone, for example, in the case of a tax matter?2 Luckily, there a number of models already available, with several different names, including participatory democracy, deliberative democracy and informal politics, which will be considered below. But certainly the idea of a group, even a large group, of citizens coming together collectively to discuss and decide upon a given matter should not be dismissed out of hand. One needs only think of the ‘gacaca’ open-​air community trials in Rwanda,3 or President Obama’s Town-​Hall meetings in the us. In Australia there is a long tradition of public consultation on law reform, including public hearings.4 2 And what about all the citizens of twenty eight countries, if the distributive sphere were the whole of the eu? Nevertheless, in an interesting section of his book entitled, ‘Putting Europe in one room,’ Fishkin describes how a real-​life experiment was undertaken to try to do just that: JS Fishkin, When the People Speak: Deliberative Democracy and Public Consultation (oup 2009) 183 ff. 3 Although the system was revived specifically to deal with the issue of genocide, traditionally the Gacaca were village assemblies to settle all kinds of disputes. 4 M Zander, The Law-​Making Process (cup 2004)  507–​511. The hearings are informal, without rules of evidence or legal representation, and sometimes even take place in the evenings. Zander quotes Mr Justice Kirby as saying that ‘[t]‌he fears of irrelevant and long-​winded submissions or of hordes of unbalanced or nuisance witnesses has [sic] not been borne out’ (at 509).

Presenting a theory of mediated complexity

291

Meanwhile, Fishkin describes a number of other real-​life instances of what he calls ‘Deliberative Polling,’ including public consultations on wind power in Texas, on a budget crisis in Rome, on sewage treatment in China and on ethnic differences in Bulgaria.5 Of the many other books written on this topic,6 an important one is Graham Smith’s Democratic Innovations.7 Here, Smith conducts a comprehensive survey of a number of possible variations on participatory/​direct democracy, mapping each against six criteria which he refers to as the four Democratic Goods (inclusiveness, popular control, considered judgment and transparency), and the two Institutional Goods (efficiency and transferability); he makes clear that ‘the design of an innovation involves compromises between different goods.’8 At the risk of grossly oversimplifying Smith’s nuanced discussion, one learns for instance that mini-​publics, with assemblies by random selection, present pros and cons with regard to inclusiveness. While efforts can be made in their design to ‘ensur[e]‌the presence of a diversity of participants,’9 the fact remains that ‘[m]ost citizens do not feel able or willing to speak in front of large groups.’10 Mini-​publics also display potential weaknesses with regard to popular control, with some criticism that ‘the sponsoring body would be able to manipulate the process’,11 and considered judgment, with fear that questions were ‘too complex and abstract for participants.’12 From Smith’s research one could tentatively postulate that direct legislation (direct democracy through the ballot box) looks promising as a mechanism for the ascertaining of the shared meanings of a sizeable and diverse population –​ it is a mechanism commonly used for political decision-​making in California and Switzerland  –​though even here there are arguments on both sides. Direct legislation is inclusive in that it promises political equality, though this 5 6

7 8 9 10 11 12

Fishkin (n 2). ibid and the books which Fishkin refers to therein; A  Gutmann and D Thompson, Democracy and Disagreement (Harvard University Press 1996); HS Richardson, Democratic Autonomy: Public Reasoning About the Ends of Policy (oup 2002); JS Drysek, Deliberative Democracy and Beyond:  Liberals, Critics, Contestations (oup 2001); T Christiansen, Informal Governance in the European Union (Elgar 2004); S Besson and JL Martí (eds), Deliberative democracy and its discontents (Ashgate 2006). G Smith, Democratic Innovations (cup 2009). ibid 146. ibid 98. ibid 84. ibid 89. C Davies, M Wetherell and E Barnett, Citizens at the Centre: Deliberative participation in healthcare decisions (Polity Press 2006) 112, cited in ibid 97.

292 Chapter 8 inclusivity can be vitiated by electoral rules and boundary issues (the latter particularly relevant to Walzer).13 It is efficient in that the result comes rapidly (the decision reached in one day rather than after hundreds) but inefficient in terms of institutionalisation costs including the setting up of specialist agencies to run the votes and ensure fairness.14 It has particular weaknesses when it comes to considered judgment (for example, questions persist about the competence or even ignorance of citizens)15 and transparency (for example, deceptive literature and withholding of key information).16 It has particular strengths when it comes to popular control (fears of economic interest groups exerting, at least direct, influence unfounded)17 and transferability (it can operate across issues and across countries).18 8.2.2 The Danger of the Forum A first danger which the establishment of Walzerian Fora might provoke, and this is presuming that it is a design based on a physical debate which is settled upon, could be that gatherings of large numbers of people, all wishing 13 14 15

16 17 18

Smith (n 7) 113–​114. ibid 136–​138. ibid 125. If Tyler Burge is right that meaning is deferential, then, however competent he or she is, the ‘ordinary speaker’ at the Forum, when asked about the meaning of a word, may simply assign it the expert’s meaning automatically: D Marconi, ‘Reference and theories of meaning as use’ in A Bianchi (ed), On Reference (oup 2015) 40, 48, referring to T Burge, ‘Individualism and the Mental’ (1979) 4 Midwest Studies in Philosophy 73. According to Nordby’s account of Burge, it is particularly the ‘person with … partial understanding,’ a common attendee at the Forum one would imagine, who shows this kind of ‘deference-​ willingness,’ but he goes on to pose the interesting question as to whether this person, if confronted with a meaning closer to ‘his own,’ would choose that one instead: H Nordby, ‘Mental content externalism and social understanding’ (2012) 2(1) Open Journal of Philosophy 1, 2. If this is answered in the affirmative, then that would suggest that the willingness to defer is not so much automatic, as simply a product of the lack of an opportunity to choose one’s own (non-​expert) understanding. Post-​Marxists would take as their starting point that there is no correlation between language and the world at all: M Devenney, Ethics and Politics in Contemporary Theory: Between Critical Theory and Post-​Marxism (Routledge Innovations in Political Theory, 1st edn, Routledge 2014) 152. Thus the theory presented here may not be viewed as a post-​Marxist one. However, it may survive as a post-​structuralist one, bringing to the foreground ‘those structured totalities which are the unconscious backdrop … to our everyday life’ and engaging in a debate ‘immanent to these structures:’ ibid 153. Smith (n 7) 134. ibid 121. ibid 139.

Presenting a theory of mediated complexity

293

to make their divergent views known, can often descend into chaos and even violence. One could perhaps plead as examples the demonstrations held at the wto Summit in Seattle in 1999, the G8 Summit in Genoa in 2001, and the G20 Summit in London in 2009. Against this it could be counterargued that such ‘powder-​keg’ situations arise precisely because citizens are given too few opportunities to air their opinions, not because they are given too many. Once the novelty of permanent fora to determine shared meanings had worn off, it seems likely that things would settle down so that the only attendees would in fact be those with a genuine interest in discussing the matter at hand. As has sometimes worked in the sphere of criminal law, making something widely available leads to responsible usage tempered by self-​regulation.19 Following on from this last point, though, is a more urgent worry. If the only attendees at the fora are those ‘with a genuine interest’ in the topic at issue, might this not lead to a sort of tyranny-​of-​the-​cognoscenti in the future, with the majority foolishly entrusting serious decisions to a minority who in time will come to wield disproportionate power over them? One thinks here of the Soviet Union, or even Orwell’s Animal Farm, where ‘[t]‌he birds did not understand Snowball [the pig]’s long words, but they accepted his explanation’; by the end of the story, of course, the birds have become the pigs’ slaves.20 The job of representation should come at the price that those represented retain the right to vote their representative out of office. Unattended by such a democratic safeguard, representation becomes dangerous. If someone absents themselves from one of Walzer’s Fora (and it is unlikely that anyone would be able to attend them all), do they give away responsibility for the matter under discussion to others? And is that a gift which they will be able to take back later? Thus a major concern for this section is that the establishment of a Walzerian Forum might make a mockery of classic representative democracy. A counterargument, though, might be that representative democracy is outmoded in the age of the Internet, where many millions of citizens can vote simultaneously, on any matter and at any time, simply by pushing a button. In the time it takes the representative to travel to the capital city to attend at the parliament, 19

20

Walzer is also of the opinion that, while ‘impassioned, bloodthirsty mobs’ do exist, ‘organized parties and movements of many different sorts, good and bad, are far more common’: M Walzer, Politics and Passion: Toward a More Egalitarian Liberalism (Yale University Press 2004) 120. Happy, non-​violent mobilizations of citizens are possible in Europe; one thinks here of the rallies attending the Referendum in Ireland over gay marriage. G Orwell, Animal Farm (Penguin Student Editions, Penguin 1999) 21.

294 Chapter 8 those whom he or she represents could already have made their views known, in a way that is both more accurate and more democratic, on any number of issues.21 In a future world where such ‘online democracy’ was commonplace, the idea of multiple Fora debating and voting on shared meanings, day in and day out, would not seem strange at all.22 The underlying question here, though, is whether the Walzerian Forum is the same as (say) the Member State’s national parliament, or whether it is something else  –​a second entity. Given that the entire citizenry of the Member State is only one possible distributive grouping, the answer must surely be:  something else. Walzer himself seems to be implying that it is something else: Democracy puts a premium on speech, persuasion, rhetorical skill. Ideally, the citizen who makes the most persuasive argument … gets his way. But … he must talk about the issues at hand. And all the other citizens must talk, too, or at least have a chance to talk.23 This does not sound like a description of Parliamentary democracy in the classical sense. In his paper, ‘The long-​term perspective,’ presented two years after 21

22

23

A Hungarian party offering online democracy as a basic means of representation –​the ‘Party of Internet Democracy’ (ide) –​attempted to contest the 2009 European Parliament elections. Meanwhile, the Pirate Party of Sweden, also supporting direct democracy, won one seat in the European Parliament in 2009, later increased to two following the changes brought about by the Treaty of Lisbon. In the 2014 European Parliament elections, the Pirate Party of Sweden lost these two seats, but the Pirate Party of Germany won one. In a draft report of 2011, Andrew Duff MEP among other things proposed the election of an additional 25 MEPs in a Europe-​wide constituency: European Parliament, ‘Report on a proposal for a modification of the Act concerning the election of the Members of the European Parliament by direct universal suffrage of 20 September 1976’ (28 April 2011) A7-​0176/​2011 –​PE440.210v04-​00. However, although a similar report was eventually adopted by the Plenary in 2013, this proposal was dropped. Some have suggested that the Parliamentary seats vacated by the United Kingdom following her withdrawal from the Union might be put to a similar use e.g. R Kalcik and GB Wolf, ‘Is Brexit an opportunity to reform the European Parliament?’ [2017] 2 Bruegel Policy Contribution 1, 10. Even today, the millions of ‘blogs’ and other networking tools enable 24-​hour discussion on almost any topic imaginable. See, for a more thorough analysis, K McGrath and others, ‘Exploring the democratic potential of online social networking: The scope and limitations of e-​participation’ (2012) 30 Communications of the Association for Information Systems 239. M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) 304 (emphasis added).

Presenting a theory of mediated complexity

295

Spheres of Justice, he again gives the impression that, while society’s current method for working out shared meanings and patterns of distribution is via the ‘odd mix of individual choice, judicial ruling and the rough and tumble politics of budgetary decision-​making,’ the ‘more systematic effort’ which he proposes is something else entirely –​not ‘our government,’ then, but an as yet undefined second entity.24 Assuming that the Walzerian Forum is in fact a second entity, the question that then arises is whether this second entity would clash with the national parliament, and how any clashes that there were would be resolved. Naturally it is only once the model for the Forum has been settled upon that its relation to the local parliament can be worked out. This working out would certainly throw up some challenges, but it is submitted that these would not be insurmountable. One idea would be to give parliamentarians the ‘last word’ after the Forum (umpired if necessary by the judge) had delivered its view.25 Court cases could then act as a catalyst for legislative change, but then this is no different from the situation today, both at Member State level26 and at eu level.27 An important motivation for such an idea –​effectively a compromise between the Forum and Parliament –​would be that, if Parliament’s legislative authority is to be restricted at all, one should be slow to transfer it wholesale to

24 25

26

27

M Walzer, ‘The long-​term perspective’ (1986) 62(1) Bulletin of the New York Academy of Medicine 8, 13–​14. Excepting, perhaps, the timing of the courts’ interventions, this was to all intents and purposes the order of events following the uk’s referendum on membership of the European Union in June 2016. The result of the plebiscite of 23 June 2016 was to be given legal effect by an Act of Parliament to be called the European Union (Notification of Withdrawal) Act. In between these two events, the High Court and Supreme Court ruled that such an Act was necessary for the country to leave the Union in a constitutionally legal way: R (Miller) v Secretary of State for Exiting the European Union [2016] ewhc 2768 (Admin); R (on the application of Miller and another) v Secretary of State for Exiting the European Union [2017] uksc 5. In the uk, for example, where the Supreme Court can declare an Act of Parliament to be incompatible with the Human Rights Act 1998, effectively forcing Parliament to amend or even scrap the offending law. cjeu cases have often presaged changes to the Treaty, for example, Case 294/​83 Parti écologiste ‘Les Verts’ v European Parliament [1986] ecr 1339 and Case C-​70/​88 European Parliament v Council of the European Communities [1990] ecr I-​2041. O’Leary refers to the Court’s ‘policy-​making leadership’ and gives a large number of examples: S O’Leary, ‘The Free Movement of Persons and Services’ in P Craig and G de Búrca (eds), The Evolution of eu Law (oup 1999) 380.

296 Chapter 8 the populace –​a residue should be left with Parliament and perhaps some reserved for the courts. This might avoid what Letsas, following the uk’s ‘Brexit’ referendum of June 2016, called the ‘folly of majoritarianism.’28 He writes: MPs are representatives, not delegates of the people … Arguments about what ‘the people want’ are (…) politically suspect, because they bypass the need to provide reason-​based arguments of justice, as if the win of the majority who voted makes that need obsolete.29 It is to be expected that Walzer would regard the reason-​giving and the argument-​making as the proper job of the Forum.30 But, even if this process failed, the residue and/​or reservation referred to above could help to ensure that majoritarian decision-​making did not trump the Rule of Law. 8.2.3 Interim Conclusion It can be seen, then, that there are several concerns raised by the idea of the Forum, some against which counterarguments may be put forward, others not. Perhaps the greatest concern of all is that deliberative democracy itself is a form of regime, an institution with rules,31 and so might be as prone to hierarchisation as the very political system the hierarchies of which Walzer was trying to disrupt in the first place. The Community Councils championed by Hugo Chávez in Venezuela, for example, even if ‘horizontally structured,’ are still not entirely free of the corruption which plagued the workers’ cooperatives which they replaced.32 With the Forum being such a cause for worry, it seems sensible to at least countenance the idea of an alternative arrangement; one such arrangement is investigated in the next section.

28 G Letsas, ‘The constitution and the folly of majoritarianism’ (United Kingdom Constitutional Law Association Blog, 20 February 2017)  accessed 21 May 2018. 29 ibid. 30 Shklar notes his ‘relative silence about representative government:’ JN Shklar, Political thought and political thinkers (S Hoffmann ed, University of Chicago Press 1998) 384. 31 I am indebted to Clare Woodford for this point. 32 S Ellner, ‘A new model with rough edges: Venezuela’s Community Councils’ (North American Congress on Latin America, unknown date) accessed 21 May 2018.

Presenting a theory of mediated complexity

8.3

297

Mediated Complexity

If a continuum were drawn between the Court of Justice of the European Union as it operates today, and the same Court as it might operate under a Walzerian regime of complex equality, as described above, a stark difference would be noted between the two positions in respect of shared meanings. At the one extreme, the Court effectively ignores shared meanings altogether, judging the case without isolating the distribution at its heart, and thus not caring what the distribution is, what its shared meaning is, or whether the distributive criterion utilized is or is not in keeping with said meaning. At the other extreme, though, the dependence on authentic shared meanings is such that the Court is to all intents and purposes a slave to the distributive community,33 unable to judge at all until such community has communicated its shared meaning for the distribuend in question, or until this has been thrashed out in the course of a no doubt lengthy game of interpretative tennis with the judge as referee. Could there be a middle course? In what follows there is proposed just such a middle course –​a theory of mediated complexity. The central plank of the theory of mediated complexity would be that the judge would take charge of the ascertainment of shared meanings, not this time as a referee picking the side that made the best case, but literally discerning the meaning for him or her self. This of course goes against much of what Walzer stands for, and certainly violates the second tenet of the theory of complex equality as described in Chapter 2 above, so much so that the word ‘shared’ in ‘shared meanings’ now seems redundant. The alternative theory set out in this section is not, unlike the earlier sections, supposed to be a faithful translation of the theory of complex equality into the world of European law, though. Rather one could say that it borrows from, or pays homage to, Walzer’s theory. It would fall to the judge, then, to work out and declare what the distribuend at issue meant to the community in which it was distributed, or, if he or she thinks it more appropriate, to everybody. This last point is important. Under mediated complexity, it is open to the judge to dispense with Walzer’s ‘relativism’ altogether by deciding on a distribuend’s meaning erga omnes. The fear that strange local proclivities would produce abhorrent meanings which in turn would produce abhorrent distributive criteria is significantly quelled if the meaning is drawn from a much larger interpretative pool, as it were, to begin with. For example, in the field of gender discrimination, the fear that the members of a predominantly male selection committee at a bank in the City of London might include ‘ability

33

A matter discussed in fuller detail at the end of this chapter.

298 Chapter 8 to enjoy lap-​dancing outings with clients’ as part of the meaning of a job at such a bank (not forgetting that they would not be the only participants in the Forum at which the meaning was ascertained), would be allayed altogether if what mattered was not what a bank job meant to the members of that particular ‘culture,’ but what it meant to all the world.34 Complex e­ quality is thus distilled down to its bare essential: a theory of distributive justice, predicated on the interpretation of meanings. What something means (in the world generally) and how it is distributed (in the world generally) must match. Opening out the field of interpretation would also assuage fears of racist meanings being employed, as they might have been in apartheid South Africa. On this view of mediated complexity, the role of the Override is also reduced (almost) to nil, as applying a universal code to particularist meanings, and universalizing those meanings in the first place, amount to practically the same thing. Handing sole jurisdiction to decide on meanings to the cjeu judge need not be a recipe for arbitrariness and a return to (or continuation of) the unpredictability of the Aristotelian regime. It is submitted that there is also a substantive difference between a judge’s deciding on the ‘alikeness’ or ‘un-​alikeness’ of two entities, and a judge’s deciding on the meaning of a distributed good. The second task is more –​for want of a better word –​clear, while the first is obscure and intangible. And a task that is clearer to those doing justice is also clearer to those seeing that justice is done. The Walzerian test is more accessible to those observing the Court’s activities, its commentators and critics, not to mention the public, which in turn enables the judges to be put under more effective scrutiny. Furthermore, it is still possible to set rules for the judge to follow in carrying out his or her task. One model could be the rules of statutory interpretation in English law (and other systems based thereon), which are used –​in a loose and flexible way –​by English judges when construing an Act of Parliament or Statutory Instrument. Indeed, as a model, statutory interpretation provides a neat parallel for the ascertainment of ‘shared understandings’ in the Walzerian sense; both activities involve the teasing out of meanings, and both give rise to the fear, sometimes unfounded, of judicial caprice making outcomes unforeseeable and advice to litigants difficult. But just as the rules of statutory interpretation provide a check on judicial flights of fancy by limiting the judge to two or three easily-​anticipated options, so a set of rules for the ascertainment of meanings would act to curb the more outrageous indulgences of an overly creative Bench. 34

Likewise the fear that a distributive community might conclude that being male (or of a certain age) formed part of the common understanding of what it is to be the Chief Executive of a multinational company.

Presenting a theory of mediated complexity

299

The rules might be unwritten conventions, much like the English rules on statutory interpretation themselves. Alternatively, they might be codified in some way. Guidelines might be produced, along the lines of the Sentencing Guidelines in English law, or it is not uncommon for courts to produce Practice Directions to expand upon their basic rules of procedure. An important example is a Practice Direction at the International Criminal Tribunal for the former Yugoslavia, which elaborated a basic provision in the Rules of Procedure and Evidence to create a viable system for adducing witness statements where the witnesses themselves could not attend the hearing.35 Without such a system it is unlikely that any trials could have taken place. Other examples of Practice Directions include Practice Directions relating to direct actions and appeals (in eu law),36 and to the power of the House of Lords, as it then was, to depart from its own previous decisions (in English law).37 But what should the rules say? One possibility –​and what follows are possibilities only, not concrete proposals –​would be to ground the set of rules in the three tenets of the theory of complex equality as presented in Chapter 2. The judge would become, as it were, the guardian of the tenets, ensuring that the different spheres of justice are kept separate, that shared social meanings are respected, and that dominance is prohibited.38 His or her principle task in an equality case would be to detect any distribution flawed by reason of boundary breach. But of course to know whether an alien attribute was crossing a boundary legitimately or illegitimately, he or she would first have to work out the shared social meaning of the distribuend at hand. He or she could approach this enterprise in a number of ways, some of them already mentioned in Chapter 7, such as the calling of expert witnesses, or other new ones such as appointing an amicus curiae to advise on the particular point or even (assuming that argument was going to be heard in open court) empanelling a jury to 35 36 37 38

Practice Direction on Procedure for the Implementation of Rule 92 bis(b) of The Rules of Procedure and Evidence, it/​192, 20 July 2001. [2004] oj L361/​15. Note [1966] 3 All er 77. For another example from English law, see Practice Note [1980] 1 All er 555, on the possible sentencing penalties for unmeritorious criminal appeals. This is heresy from a pure Walzerian perspective. Walzer had stated that it was the State’s job to guard the boundaries of the spheres: Walzer, Spheres of Justice (n 23) 281. However, making the judge the guardian of the tenets would have another advantage, namely, that the judge would have to see to it not only that the second and third tenets were adhered to (distributive criteria respected/​dominance avoided), but also that the first tenet was adhered to (correct separation of spheres in the first place). Many if not most attempts to ‘cheat’ on Rules 2 and 3, by the original distributor or by the judge, would in fact be violations of Rule 1, and the judge (in quo or on appeal) could easily lay this bare.

300 Chapter 8 decide.39 In England anyway, it is the usual solution, at least in cases where a jury is already in place, to leave questions of fact, as opposed to questions of law, to a separate ‘tribunal of fact.’40 These ideas, though, take the matter further and further out of the judge’s own hands, and represent a kind of liberal extreme.41 At the other end of the spectrum, a judge could simply carry out his own research (just as English judges construing statutes are now allowed to consult Parliamentary records and so on). An international judge taking time to ascertain local views is not such an unusual concept; the European Court of Human Rights often undertakes such an exercise, and indeed has a Research Division for precisely that purpose.42 Alternatively, the judge could employ a ‘common sense approach;’ relying on abstract concepts like ‘natural law’ or some equivalent, the judge could fashion the meaning almost any way he or she wanted, but ensuring always that his or her view accorded with that which would be reached by one of English law’s favourite characters, the ‘reasonable man.’43 The judge might indeed refuse to hear evidence where a meaning ‘permit[ted] of ready 39

40

41

42

43

Hutchinson favours this: ‘By changing the civic members of the jury from passive observers to active participants, the legal process might begin to fulfill its democratic commitments. After all, law involves the lives of society’s citizens and is much too important to be left to lawyers’. See AC Hutchinson, ‘A Postmodern’s Hart: Taking Rules Sceptically’ (1995) 58 mlr 788, 817. Walzer’s main criticism of the courts (‘the least democratic rather than the most democratic branch of government’) is that judicial decisions refer ‘only to the entitlements of particular individuals.’ This leads to justice for them, but injustice ‘to all the rest of us.’ The judges do not ‘consult [the] people’ who, while not being parties, are nonetheless affected by the outcome. The empanelling of a jury, or one of the other similar expediencies described in this section, would bolster the democratic credentials of the courts in general, and the cjeu in particular. Thus, although what is being presented here is a departure from Walzerian orthodoxy, it does in some ways address Walzer’s principal objection to the juridification of allocative decision-​making:  M Walzer, ‘Justice Here and Now’ in FS Lucash (ed), Justice and Equality Here and Now (Cornell University Press 1986) 138–​9. Also at that extreme, one imagines, would be the ‘Lay-​Oriented Court’ envisaged by Nils Christie and which he intends mainly, though not exclusively, for criminal cases: N Christie, ‘Conflicts as property’ (1977) 17(1) British Journal of Criminology 1, 11ff. Discussed in K Dzehtsiarou, European Consensus and the Legitimacy of the European Court of Human Rights (cup 2016) 41, 72–​114 and in E Howard, ‘The case for a considered hierarchy of discrimination grounds in eu law’ (2006) 13(4) MJ 445, 458. See, for example, how Lord Morris in the case of Brutus v Cozens states that the magistrates who had earlier heard the case should have applied ‘rational judgment and common sense’ to reach a decision: Brutus v Cozens [1972] 2 All er 1297, 1300. The issue here was the meaning of the word ‘insulting’. The ‘common sense approach’ is viewed with suspicion by the critics of neoliberalism. This critique is considered at section 9.4 below.

Presenting a theory of mediated complexity

301

comprehension,’ ‘present[ed] no difficulty of application or understanding’ or was ‘ordinary [and] uncomplicated,’44 what Lord Diplock called ‘a matter of first impression.’45 In such a case, and following an equivalent of English law’s ‘literal rule’, the judge could simply give the distribuend its ‘ordinary natural meaning.’46 The judge could say, alternatively or in addition, that he or she was taking ‘judicial notice’ of the meaning. Then again, the judge could back this up with a presumption, or series of presumptions, as the European Court of Human Rights does, for example, in cases of torture or inhuman or degrading treatment.47 Thus, the ‘ordinary natural meaning’ would be presumed unless one or other party adduced convincing evidence of a special meaning. The onus would then be on counsel to bring to the Court’s attention, in particular via their pleadings, matters which they wanted it to consider, and which they knew it would not investigate proprio motu. It should not be forgotten, of course, that one possible meaning may already be on the table, namely the meaning utilized in the original, contested distribution. Even if this was not explicit it can usually be worked out, a contrario, from the fundamentum distributionis actually used. A  very cautious judge could perhaps limit themselves to upholding or striking down this original meaning, having set themselves (or had set for them) a threshold of acceptability, along the lines of the famous Wednesbury test –​in English law –​that the courts should only interfere when an authority has come to a conclusion ‘so unreasonable that no reasonable authority could ever have come to it.’48 In other words, the judge would only strike down an existing meaning where the distributive community has arrived at a meaning so unreasonable that no reasonable distributive community could ever have arrived at it.49 As an 44 45 46 47

48 49

All phrases from ibid 1300, 1301 and 1303. Garland v British Rail Engineering [1983] 2 ac 751, 771. Brutus v Cozens (n 43) 1302. Injury sustained while under State control means torture et cetera, unless an ‘alternative explanation … ha[s]‌been forwarded by the Government.’ This quotation from the case of Shishkovi v Bulgaria App no 17322/​04 (ECtHR, 25 March 2010), but the principle has been stated on many other occasions. Associated Provincial Picture Houses, Limited v Wednesbury Corporation [1948] 1 kb 223, 230. Under a strict regime of complex equality, if it transpired that the ‘shared’ meaning was not in fact that of the distributive community at all, for example, if it was the distributor’s own, then the meaning could already be invalidated, and the distribution could already be declared flawed, by dint of breach of the second tenet. However, in the alternative version of the theory being presented here, in which much less importance is being placed on the element of ‘sharing,’ perhaps the judge would allow any meaning to be

302 Chapter 8 alternative, the judge could interfere only when a meaning was perverse; this is a standard that has already been encountered in the book.50 As with the ascertainment of the meaning itself, the judge will have to find (or have found for them) a workable approach for identifying a perverse meaning, but it seems likely that, to slightly paraphrase Lord Reid in Brutus v Cozens, an ordinary sensible person knows a perverse meaning when he sees or hears it.51 The sexual orientation of the passenger’s partner, for example, has nothing to do with the meaning of a train ticket.52 Likewise, the nationality of the victim has nothing to do with the meaning of criminal compensation.53 In English judicial review cases, though, the Court does not try to stand in the place of the requisite authority, and make (or remake) the contested decision itself. Having struck a decision down, it sends the matter back to this authority so that it (the authority) can make a second attempt in the light of the Court’s ruling. Should the cjeu, hearing an equality case, and applying the theory of mediated complexity, follow suit, and, in the event that it invalidates an existing meaning, refuse to spell out the ‘correct’ meaning by itself? Again, if a strict Walzerian approach were being adopted, the answer to this question would probably be in the affirmative –​the matter should be sent back down to the distributive community for redetermination of the shared meaning –​ unless one of the models sketched in Chapter 7 has been accepted, allowing the Court (for example) to have a ‘first go’ at declaring the meaning, in a strictly non-​binding way. However, in the theory of mediated complexity, the answer must be in the negative; unlike a public authority, a distributive community is too ephemeral a concept to be communicated with, albeit implicitly, in this way. The judge, mediating the complexity on behalf of the community, must provide the recast meaning.54 Even then, though, the rejected meaning will

50 51 52 53 54

the meaning, as long as the reasonableness test (or equivalent) was satisfied. In such a case the test as stated in this sentence would have to be slightly rewritten, with a phrase like ‘decider as to meaning’ or ‘meaning-​interpreter’ replacing ‘distributive community.’ Allowing non-​shared meanings not just in Court but within the community as well represents, it is acknowledged, a further, massive break with orthodox Walzerian theory. An analogy could perhaps be drawn with the English ‘golden rule’ here, that interpretations which result in absurdities must be avoided. Brutus v Cozens (n 43) 1300. Case C-​249/​96 Lisa Jacqueline Grant v South-​West Trains Ltd [1998] ecr I-​621. Case 186/​87 Ian William Cowan v Trésor public [1989] ecr 195. Or the Court could plot a middle course, by sending the matter back on some occasions, and deciding itself on others. For example, it might choose to send back where the distribution had taken place in the public sphere, and decide itself where the distribution had taken place in the private sphere.

Presenting a theory of mediated complexity

303

no doubt provide some clues as to what the true meaning (for the moment, anyway) is or might be. It should be emphasized again that this section sets out a range of possibilities, not a concrete plan. Mediated complexity is the half-​way point on the continuum between the position today and the position as stated in the foregoing chapters, but even within mediated complexity itself there is a wide gamut of possible approaches for the cjeu to follow, depending which of the various options above it chooses. A number of different Courts of Justice might emerge. The Court might elect an accessible approach, throwing its doors open to a large number of helpers and advisers, or it might prefer to adopt a more remote position, standing aloof from the spheres themselves.55 Adopting a ‘common sense’ or equivalent approach, the Court might feel that in most cases it ‘knows best,’ and it might start to display the sort of paternalism of which the English Law Lords are sometimes accused. In contrast to such an imperious stance, the Court might borrow from the Continental traditions and choose to operate in a more inquisitorial manner. Like the English judges when they apply the mischief or purposive rules, or the more European teleological approach, the cjeu judge may decide to be more probing –​looking into methods, motives and causes, rather than just taking things at face value. And as the depth of the Court’s enquiry varies, so may the breadth. Perhaps the judge will wish to consider only the specific parties in front of him or her, producing judgments ‘confined to their particular facts,’ or perhaps they will opt to widen the focus, to produce judgments of more general application. A lot will depend on how much work the Court wants to take onto its own shoulders. At the moment, for example, the Court of Justice will not consider pleas other than those appearing in the parties’ previously-​lodged pleadings, while the European Court of Human Rights, for example, is much readier to pursue extraneous matters proprio motu. Put simply, adopting a regime of mediated complexity does not commit the cjeu to any one particular judicial style, but rather would offer the Court a 55

Another deviation from orthodox Walzerian theory. Such aloofness is a particular problem for Walzer, who disdains those who ‘fashion for [themselves] … an objective and universal standpoint’ (Walzer, Spheres of Justice (n 23) xiv). This has led him to a sustained critique of juridification, which perhaps reached its high-​water mark with his review of Ronald Dworkin’s Taking Rights Seriously (Duckworth 1977):  The New Republic (New  York, 25 June 1977) 28. More recently, though, he seems to accept that ‘[m]‌uch … political debate takes place in the courts’:  ‘not always the best place, in my view, but the place where the nuts and bolts of membership and equality are most often addressed.’ See Walzer, Politics and Passion (n 19) xiii-xiv, and below for more on Taking Rights Seriously.

304 Chapter 8 choice of different styles. When it comes to the trade-​off between interpretative complexity and judicial efficiency, it will be up to the Court to decide how much of the one it can cope with, without causing irreparable harm to the other. However, whichever possibility or style is elected, or approach is taken, the Walzerian judge practicing mediated complexity, as a fortiori the Walzerian judge practicing classic complex equality, will have to work out his or her relationship to society, and this means both society as represented in the national parliament, and society as represented or embodied in the Forum. The remainder of this chapter will therefore be spent considering the nature of this vital relationship. As has been seen, the former judge will have already chosen the level of societal involvement and, more importantly, influence which he or she will tolerate in his or her court, but, unless Walzerian theory is abandoned almost entirely, with the judge taking the erga omnes route and concentrating on universalised meanings only, this is likely to be somewhere between ‘some’ and ‘a lot.’ The latter judge, of course, is little more than society’s mouthpiece, as he or she merely arbitrates society’s interpretative disputes and pronounces the outcome. 8.4

The Judge vis-​à-​vis Society

One might do well to begin with the remark by Aharon Barak in his work, The Judge in a Democracy: The judge is a partner in creating [the] system of law. The extent of this partnership varies with the type of law being created. In creating common law, the judge is a senior partner. In creating enacted law, the judge is a junior partner. Nonetheless, he or she is a partner, and not merely an agent who carries out the orders of his or her principal.56 Throughout his book, Barak continues to assert the judge’s right to opine independently of the public’s ‘passing vogues,’ grounding his or her opinion instead on ‘principles and fundamentals,’57 ‘emphasising history and not hysteria.’58 The judge should certainly be a student of social consensus,59 but, at 56 57 58 59

A Barak, The judge in a democracy (Princeton University Press 2006) xviii; Judge Barak was president of the Supreme Court of Israel until his retirement in 2006. Both quotations from ibid. ibid 108. ibid.

Presenting a theory of mediated complexity

305

least according to Barak, not a slave to it. In this regard he opposes himself to the view of Antonin Scalia, who was at least prepared to entertain the idea of giving expression to modern developments, an evolutionary approach, as long as the consequences were understood, namely, the politicisation of the judiciary and the fact that ‘litigation becomes the prosecution of politics by other means’.60 However, he feels that, at least until recently, citizens have ignored these consequences and pretend instead that they have plac[ed] the evolution not in the hands of the majority but in the hands of the Supreme Court Justices (,)  entrust[ing] individual and minority rights to the care of a benign and sympathetic body insulated from majority control.61 How does this judge-​led evolution proceed? Scalia believes that historical judges have made ‘mistakes’ which need ‘updating,’ whether this is regarded as evolution or, as he prefers, a simple exercise of stare decisis (the doctrine of precedent). If a judge refuses to do the updating, though, he or she could be accused of ‘the worldwide promotion of honest, steadfast error.’62 However, he does sound a note of warning about this taking of an ‘evolutionary’ view of the law. If from generation to generation the law means simply ‘what it ought to mean’63 then: Justices are not so much lawyers as arbiters of the ‘ought.’64 Of course a judge who takes the evolutionary view also runs the risk of present-​day majoritarianism forcing him or her to collude in the harming of the minority, as warned by Letsas at the end of section 8.2.2 above. Scalia’s approach is more conservative here, opining that, were each evolutionary leap to be directed by the democratic majority, [t]‌hat would obviously frustrate the entire purpose of constitutional individual-​right guarantees, which is to protect individuals against the majority.65 60 61 62 63 64 65

A Scalia, ‘Modernity and the constitution’ in E Smith (ed), Constitutional justice under old constitutions (Kluwer 1995) 315. ibid. ibid 318. ibid 317. ibid. ibid 315 (Scalia’s emphasis).

306 Chapter 8 He thinks any evolutionary advance should be strictly a matter for the Bench. Barak, also keeping the matter to the Bench, circumvents the risk with this enigmatic remark: the fundamental values of the present are not necessarily the values that today’s majority accepts.66 He concludes, in a Walzerian vein: They are the deeply held values of the society that have developed over time.67 It is clear that the more control the judge takes of the situation, the more he or she asserts him or herself, the less likely he or she is to become a slave to the social consensus. In Barak’s analogy, he or she appoints him or herself the principal, not the agent. What has been called above the ‘danger of the Forum’ is thus averted. But it is equally clear that the Walzerian judge would not have the luxury of such self-​assertion –​they would not be, and should not be, to use Scalia’s analogy, insulated. Thus the danger lingers, and any proposal that the judges at the cjeu should adopt a Walzerian approach when deciding equality cases, will need to be accompanied by a twin proposal for a more ambitious reading of legal evolution than merely a matter of stare decisis, to be left entirely in the hands of the judge, as today, given that this model depends for its survival on a high level of judicial control and a low level of societal input. Thus the question of the judge’s role must and will be returned to in the section ‘Postmodernism and precedent’ below. 8.5 Conclusion What is being offered in this book is a rule or set of rules to help the European Court to decide the equality cases before it, not to conjure up the result as if from Aladdin’s lamp. The theory described herein does not strive to tell judges what to decide, merely to give them stepping stones to reach their own decision. Complex equality, and mediated complexity, are not touchstones, or yardsticks, or ready reckoners. Rather, they are ‘a common language and a shared set of concepts for talking and thinking 66 67

Barak (n 56) 108. ibid.

Presenting a theory of mediated complexity

307

about questions of … equality,’68 in other words, ‘a framework for debate.’69 The debate itself does not end, then, but its quality improves. It can be seen how the judge’s reasoning is much more considered and much less rigid. It is context-​sensitive; it sticks close to the subject-​matter of the case, placing it at the very heart of the debate. It resists deviations into less pertinent considerations and avoids reductionist, black-​or-​white rhetoric. Of course, policy questions remain difficult for a judge to call –​or indeed he or she may feel that they cannot call them at all and must leave them for the legislature. This is a problem which judges perennially face, and no amount of theories will solve it. A complex system of presumptions might bring some relief, but would inevitably just transfer the policy dilemma in question from the judge to whoever’s job it is to write the presumptions.70 Thus, 68

69 70

JM Balkin, ‘Plessy, Brown, and Grutter: A play in three acts’ (2004–​2005) 26 Cardozo L Rev 1689, 1699, referring to what he calls the ‘tripartite theory’ in the us, which divided the rights of citizens into civil, political and social rights, but offered equality in respect of the first category only. However, its replacement, the more familiar model of strict scrutiny and suspect classifications, is also described by Balkin as ‘a language for talking and thinking’ (at 1690). RB Siegel, ‘Why Equal Protection No Longer Protects: The Evolving Forms of Status-​Enforcing State Action’ (1997) 49 Stan L Rev 1111, 1120, also referring to the ‘tripartite theory.’ A concrete system of presumptions is not favoured as it would cost the theories their fluidity, which is one of their best qualities. However, for the sake of completeness, the matter is briefly considered here. As a preliminary point, the system would have to exclude all distributions of criminal sentences and other State-​endorsed sanctions. This should not prove too much of a problem at the cjeu, which does not hand down criminal penalties, although it does hand down some civil ones. The easiest presumptions would be along the lines of ‘in dubito pro distributor’ or ‘in dubito pro distributee,’ but these would most likely be condemned as too simplistic, given the wide variety of distributions at issue. A more useful set of presumptions would be: 1. The Reasonable Distributor distributes according to a meaning which combines the interests of the Distributor and the interests of the Distributee in a ratio of fifty to fifty. 2. Assuming that a distribution in line with Presumption 1 is impossible, the Reasonable Distributor gives additional weight (that is, from 51% of the total weight to 100% of the total weight) to the meaning of the party (‘the first party’) whose predicted position, following a distribution carried out according to the meaning of the other party (‘the second party’), would be worse than that of the second party following a distribution carried out according to the meaning of the first party. Elsewhere in this book, Presumption 2 has been referred to as a kind of ‘hardship rule.’ However, this would not settle everything and further presumptions would be needed to decide between, say, damage to life and limb and damage to material assets, in varying degrees and timeframes.

308 Chapter 8 some lines in the sand may remain, but there is a new concern for what Homi Bhabha has called the ‘border lives,’71 those caught in the ‘in-​between moment,’72 frozen as they attempt to cross from one sphere to the next. Focusing on this ‘intervening space’73 means more than just focusing on exceptional cases –​cases that are exceptional to some rule. It involves challenging the very rule itself: The hither and thither of the stairwell, the temporal movement and passage that it allows, prevents identities at either end of it from settling into primordial polarities. This interstitial passage between fixed identifications opens up the possibility of a cultural hybridity that entertains difference without an assumed or imposed hierarchy …74 Complex equality both ‘displays and displaces the binary logic through which identities of difference are often constructed.’75

71 72 73 74 75

HK Bhabha, The Location of Culture (Routledge 1993) 1. ibid 5. ibid 10. ibid 5. ibid (italics added).

Chapter 9

Evaluation 9.1

The Two Methods Compared: Erpelding

If one takes a case like Erpelding, step by step, one can see clearly where the Aristotelian test goes wrong. The first conclusion which the Aristotelian analysis produces in this case is that milk producers who did well between 1981 and 1983 and milk producers who did badly between 1981 and 1983 are ‘like.’ This seems uncontroversial, but already the Aristotelian analyst is guilty of superficiality. The Walzerian analyst, meanwhile, would have spotted that they are not alike at all –​one comes to the quota-​distribution empty-​handed, but the other, to make an analogy with a casino, brings with them a chip which they have already won at another table. The second Aristotelian conclusion here is that the two categories of milk producer, whilst like, were treated ‘unlike;’ the producer who did well between 1981 and 1983 received an additional individual milk reference quantity. Again, this analysis does not go far enough. The reason for this disparate treatment (having done well during the period 1981–​1983, or having done badly during the same period) seems completely inappropriate. In Walzerian terms, the reason for the distribution (or the distributive criterion) is ‘wrong.’1 But the Aristotelian analyst is powerless to do anything here. Restricted to a ‘yes or no answer,’ and faced, on the one hand, with the granting of a quota and, on the other hand, with a refusal to grant a quota, he or she can only state the obvious: that the two ‘treatments’ are dissimilar. He or she does not at any time, however, consider the reason for this, or answer the simple question: why? It is worth remembering Lord Walker’s comment in the uk case of Carson:2 One of the most powerful criticisms of a rigid, step by step approach based on comparators is, if I may respectfully say so, in the speech of my noble and learned friend Lord Nicholls of Birkenhead in Shamoon v Chief Constable of the Royal Ulster Constabulary […]. That was a case under the Sex Discrimination (Northern Ireland) Ord 1976 (si 1976/​1042) and this House had to grapple with the statutory definition of discrimination. 1 Keeping in mind the Standard Contingent Reply. 2 R (on the application of Carson) v Secretary of State for Work and Pensions and Conjoined cases [2005] ukhl 37.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/​9 789004354265_010

310 Chapter 9 Lord Nicholls demonstrated that a step by step approach was liable to obscure the real issue in the case, which was why the complainant had been treated as she had been treated. Until that question was answered, it was impossible to focus properly on the question of comparators. Lord Nicholls […] observed […] that: employment tribunals may sometimes be able to avoid arid and confusing disputes about the identification of the appropriate comparator by concentrating primarily on why the Claimant was treated as she was …3 Walzer, on the other hand, brings the reason or reasons into the middle of the debate, rather than relegating it or them to an afterthought: The phrase ‘for whatever reasons’ conceals a problem […], which philosophers who are quick with hypothetical examples are prone to ignore. […] Social meanings are constructed, accepted, and revised for reasons, and we have to engage those reasons.4 The third and final Aristotelian conclusion in Erpelding is that the unlike treatment of the two categories of milk producer is objectively justified on the grounds of certainty and effectiveness. The Court thus rubs salt in Erpelding’s wounds by announcing that the need to have only one rule justifies the rule. The fact that it is or might be the wrong rule is not mentioned, even here, at the one point in the test where it might have been. The wrongness of the rule is only revealed via Walzer’s root-​and-​branch scrutiny; the Aristotelian analyst satisfies him or herself with an inspection of the branch alone. In other words, applying Walzer’s doctrine to a case at the cjeu involving the principle of equality would facilitate a much more profound analysis of the original ‘distribution’ giving rise to the case. If a distribution is substantially flawed, more often than not an Aristotelian assessment will simply not reveal this (as in Erpelding). Another significant defect in the ‘symmetric’ approach to equality is its reliance on so-​called objective justification as a means of rectifying any inequalities discovered. As the consideration given to the two comparators is so superficial, the objective justification acts as little more than a ‘rubber stamp’ or ‘white-​washing exercise,’ often condoning, and consolidating, the flaw in the original distribution. 3 ibid [63] (Lord Walker’s emphasis). The reference for the Shamoon case is [2003] ukhl 11, [2003] icr 337. Lord Nicholls’ observation is at 342, [11]. 4 M Walzer, ‘Objectivity and Social Meaning’ in M Nussbaum and A Sen (eds), The Quality of Life (oup 1993) 173–​176.

Evaluation

311

The problem is that the Aristotelian test is confined in the present. Complex equality allows for a historical enquiry. To use a simplistic analogy, it is like replacing a photograph with a piece of film; Aristotle provides the judge with a highly decontextualized ‘snapshot’ of who he or she has in front of him, and what has (recently) happened to them. Walzer encourages the judge to ‘look behind’ this, revealing both the true identity of the comparators, and any secret advantages which they may enjoy, or disadvantages from which they may suffer, and the true nature of the grounds upon which the distribution was made. 9.2

Formulating the Argument

In the previous section, several metaphors illustrating the differences between the Aristotelian approach to equality and the Walzerian one were encountered (snapshot and film, branch and root, and so on). Of course, these metaphors cannot go far enough by themselves to convince a reader that Walzer’s theory of complex equality should be used by the Court of Justice as a complement to the Aristotelian test, or, if so, why. It is hoped in this section to delve a little deeper into the issues in order to formulate a more persuasive argument. A lot of the metaphors imply an absence of subtlety on the part of the Aristotelian test, and a concomitant presence of subtlety in Walzer’s technique. It is true that, where Aristotelian eyes can usually only distinguish the two poles, or extremes, of a given situation, the more focused eyes of the complex egalitarian can make out the intermediate degrees. The problem of intermediaries has been seen a few times in this book, for example, in the case of Beltrante,5 dealt with in the chapter on non-​suspect grounds. In this case, there are two ‘extremes’ (parents of dependent children and parents of quasi-​dependent children, that is, those who are above the age of majority but unemployed), but an applicant like Mr Beltrante can find himself caught between them, for example, as the parent of a quasi-​dependent child whose dependency  –​in adulthood  –​stems from disability.6 While there may be superficial similarities between him and the other members of the second category, the increased anxiety which he might feel for his child, if the child lived in another country, and his desire for filial visits whenever possible, would make him more akin to a parent in the first category. However, unable to deal 5 Case T-​48/​89 Fernando Beltrante and others v Council of the European Communities [1990] ecr II-​493. 6 This is theoretical only; it is not known if this was the actual situation that Mr Beltrante was in.

312 Chapter 9 with specificities or to take decisions on a case-​by-​case basis, an Aristotelian judge erects –​almost arbitrarily –​a barrier between the supposedly deserving and the supposedly undeserving. A Walzerian judge, on the other hand, judges who is deserving by reference to the meaning of the thing deserved, and accepts that there is no hard-​and-​fast answer to this because the meaning is always changing.7 This method is less elegant,8 but arguably much more just.9 9.3

In Search of Flexibility

The lurching between like and unlike which is symptomatic of the Aristotelian test is often caused by a sudden and dramatic shift in perspective. Many of those writing about equality these days are seeking a ‘happy medium’ between, to use cinematic terminology, the ‘wide shot’ which shows all people (or all objects) as the same, and the ‘close-​up’ which lays bare their individual differences. In his famous essay about the politics of difference, Charles Taylor declares that ‘[t]‌here must be something midway’ between the two approaches.10 However, what is sought may not be a single point at all, but rather acknowledgement of the continuum which exists between the two types of ‘shot;’ the lens in a camera or telescope would in fact be of little use if it could only discern those things right in front of it, or those things many miles away. The call, then, is for variability, for flexibility. Fittingly, what lies at the end of the rainbow is the ability to make out all the colours. 7

8

9 10

While a decision that two things are like –​or unlike –​may be much harder to reverse. A  rare exception is Maruko on the subject of same-​sex partnerships vis-​à-​vis (heterosexual) marriage:  Case C-​267/​06 Maruko [2008] ecr I-​1757. On the changeability of meanings, note the more progressive approach of US Supreme Court judge Sandra Day O’Connor who, having to define the word ‘accident,’ warned that any definition given had to be flexible: Air France v Saks, 470 US 392 (1985), 405. Later faced with the same task, the judges of the UK House of Lords (as it was) were more cautious: In re Deep Vein Thrombosis and Air Travel Group Litigation [2005] UKHL 72, [2006] 1 AC 495. Walzer would be the first to admit that complex equality is not an elegant theory. See M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) 21: ‘The theory that results is unlikely to be elegant’. Another example of an ‘intermediate’ case is Omega, also dealt with in the chapter on non-​suspect grounds: Joined cases C-​27/​00 and C-​122/​00 Omega [2002] ecr I-​2569. ‘There must be something midway between the inauthentic and homogenizing demand for recognition of equal worth, on the one hand, and the self-​immurement within ethnocentric standards, on the other’: C Taylor, ‘The Politics of Recognition’ in C Taylor, Multiculturalism: Examining the politics of recognition (Princeton University Press 1994) 72.

Evaluation

313

Among those hunting for flexibility, from amongst those writers already encountered in this book, are McCrudden, who supports a changeable approach to European anti-​discrimination law: What is necessary, then, is a recognition that different equalities are in play in different situations.11 O’Leary too, lambasting the Court’s ‘one solution fits all approach,’12 calls for variability: [N]‌ational authorities and, ultimately, the Court of Justice, may have to engage in a case-​by-​case assessment of whether denial of a benefit is reasonable given the individual circumstances of the claimant and the characteristics of the benefit.13 And, writing extrajudicially on the subject of (general) principles such as effectiveness and legal certainty, cjeu judge Sacha Prechal has commented: The concrete application of these principles may differ from case to case. Indeed, the approach of the Court is casuistic and cannot be otherwise.14 9.4

But Does Flexibility Lead to Uncertainty?

On the other hand, if meanings are always changing, could it be argued that a Walzerian approach might lead to uncertainty?15 As has already been seen countless times in the main case-​study chapters, any final conclusion as to 11

12

13

14 15

C McCrudden, ‘The new concept of equality’ (Paper prepared for the Academy of European Law conference, ‘Fight Against Discrimination:  The Race and Framework Employment Directives,’ 2–​3 June 2003)  accessed 20 May 2018, 24. S O’Leary, ‘Developing an ever closer union between the peoples of Europe? A reappraisal of the case law of the Court of Justice on the free movement of persons and eu citizenship’ (2008) 27 yb of Eur L 167, 193. ibid 192 (emphasis added). This approach is highly Walzerian, putting as it does the identity and situation of the distributee, and the nature of the distribuend, at the centre of the enquiry. S Prechal, Directives in ec Law (2nd edn, oup 2006) 75. Walzer himself has commented, ‘To my mind, certainty is always a fantasy’. See M Walzer, ‘Governing the globe: what is the best we can do?’ (2000) 47(4) Dissent 44, 52.

314 Chapter 9 whether a given distribuend was distributed correctly or not is contingent upon the discovery of its shared meaning, which in turn is contingent upon the say-​so of the distributive community itself. Even if it is accepted that the judge is the final arbiter and therefore that his or her conclusion is not contingent, the many others involved in the dispute, particularly the client being advised by his or her lawyer, must make do with at best a provisional, qualified response to the question. Such uncertainty could be disadvantageous to a client whose decision whether or not to proceed with their case requires a robust prediction of the outcome, not an interim hypothesis. However, under Walzer, the prognosis made in the lawyer’s office should not differ too much from the final decision reached in Luxembourg, since both lawyer and judge are basing their conclusion on the same thing, in contrast to the situation under Aristotle, where the result of the like-​for-​like test, like the famous ‘Chancellor’s foot’ in English law,16 varies from one beholder to the next. Such ‘precedents’ as there are under Aristotle are bogus; what was like today may be unlike tomorrow. Walzer at least acknowledges the ongoing nature of the debate and tries to incorporate it, but this in no way necessitates uncertainty. Although it may fluctuate, a doctor will still be able to take the temperature of his or her patient at any given moment and respond in the manner appropriate to that reading, and for that person, with that condition. The ‘manner appropriate’ is known due to years of experience (both his or hers, and others’), and the building up of genuine precedents. Thus, the doctor can usually predict the sequence of events with a great deal of certainty. If changeability were a bar to prediction, doctors could effectively be abolished. Similarly, in contemporary physics, irregularity is no longer viewed as totally separate from regularity, but as generated by the same kind of mathematics. Indeed, the idea of treating perfection and imperfection –​ chaos and pattern –​as two separate systems, as opposed to two parts of the same system, is distinctly old-​fashioned. The physicists’ breakthrough here, that is, to stop viewing the two phenomena as two distinct poles and to view them instead as two ends of the same spectrum, should set a precedent for lawyers grappling with sameness and difference. Indeed, the image of the continuum, or rainbow, has already been encountered above. The results of Walzer’s test may be variable, but the underlying system remains constant, and identifiable.

16

For a nice account of this story, see SH Bailey et al, Smith, Bailey and Gunn on the modern English legal system (Sweet & Maxwell 2002) 5, and there is an entire chapter on it in R E Megarry, Miscellany-​at-​Law (Stevens 1955) 139 ff.

Evaluation

315

The view that the law’s obsession with uniformity is outmoded has been expressed more and more frequently, and with greater and greater eloquence, in recent years. In the specific context of eu Law, Weatherill has written: Uniformity is today less apt … and … it is vulnerable to criticism as an inappropriate normative aspiration for the modern eu.17 Wielsch, although principally focusing on contract law, offers a fascinating vision of law embracing poly-​contextuality, with each contract its own social system; such a vision perhaps unsurprisingly is supportive, albeit conditionally, of Walzerian equality.18 However, perhaps in this context one should not ignore an equally prevalent viewpoint:  the critique of neoliberalism.19 This critique sees ‘flexibility’ as one of the hegemonic neoliberal state’s key weapons in its bid to reshape human beings and to create, and police, subjects. According to the critique, neoliberals contradict themselves and treat logical incoherence as a virtue. Mirowski cites a ‘confess[ion]’ of this made by Paul Klugman: ‘Inconsistency in the pursuit of useful guidance is no vice’.20 There would seem to be some worth to the claim that, in the hands of politicians, flexibility is a ‘foisting tool;’21 the neoliberal notion of transferable skills, for example, a mere ploy more easily to move workers from one money-​making context to the next. It brings with it dispensability,22 allowing, of two aspects, the one more favoured by the politicians to be selected and the one championed by the heterodox to be dispensed 17 18

19 20 21 22

S Weatherill, Law and values in the European Union (oup 2016) 258. D Wielsch, ‘Relational Justice’ (2013) 76 Law and Contemporary Problems 191; see especially footnote 50 at 203 (the term poly-​contextuality is attributed to Gotthard Günther at 199). However, the implication in the text that constitutional courts operating the like-​ for-​like test, and in particular the justificatory tool which goes with it, are already embracing Walzerian sphere-​plurality is respectfully disagreed with. The argument that x and y are treated like in one case, but unlike in another case, owing to an appreciation by the court of the differential contexts runs aground, in reference to the cjeu anyway, where the context in both cases is one and the same. For one example, W Brown, Undoing the Demos: Neoliberalism’s Stealth Revolution (Zone Books 2015). P Mirowski, Never Let a Serious Crisis Go to Waste: How Neoliberalism Survived the Financial Meltdown (Verso 2014) 21. B Brecher, ‘Universities and the Neoliberal Agenda’ (Lecture given at the Old Courtroom Lecture Theatre, Brighton, 10 November 2015). For example, Mirsowski (n 20) 110.

316 Chapter 9 with, with the thing most flexible of all being the truth. That would naturally include the true sayings –​or verdicts –​of judges.23 However, it is crucial to underline that what is being proposed here is not flexibility of verdicts, but flexibility of meanings of goods. As the meaning shifts, the set of potential distributees expands or contracts. Shifts in the meaning are brought about through a meaning-​ascertaining process, and (under pure Walzerian theory) the meanings are ascertained within the set, not without. If the meaning-​ascertaining process is being abused to redraw boundaries to exclude the unwanted, rather than the undeserving (although perhaps labeling them as such for the sake of appearances), then it may be transferred to an external ‘critic’ in the Walzerian sense. At that point it transforms into more of a meaning-​discerning process, but one carried out through examination of objective evidence, not recourse to subjective bias via arbitrary whim. Again, the shifts in meaning are brought about through the meaning-​discerning process, and so the meanings are flexible only in the sense that that process is capable of producing, indeed logically must produce, varying results.24 A table is a table in a kitchen and an altar in a church because of the way people who go to church understand it in its clerical, as opposed to its culinary, context. Those who do not go to church, but who are charged with discerning the meaning of the table having heard the evidence of those who do, must necessarily come to the same conclusion. A table is not a table in a kitchen and an altar in a church because those who go to church, or the non-​church goers charged with the discernment, want rid of its culinary meaning.25 On the contrary, flexibility of verdicts is the stock-​in-​trade of the Aristotelian (or purely Aristotelian) judge.26 And it is the Aristotelian judge, who insists on seeing only two poles, who can afford, under fake duress, to choose one and discard the other. Easy and guilt-​free dispensability of unpalatable options 23

24 25 26

Kaupa gives as one of the possible locations for the cause of the Court’s assumed bias ‘the personal beliefs of the judges:’ C Kaupa, ‘Maybe not activist enough? On the Court’s alleged neoliberal bias in its recent labor cases’ in M Dawson, B de Witte and E Muir (eds), Judicial activism at the European Court of Justice (Elgar 2013) 57. Note Hart’s comment that ‘a margin of uncertainty should be tolerated and indeed welcomed’: HLA Hart, The concept of law (3rd edn, oup 2012) 251–​2. This example taken from M Walzer, ‘Objectivity and Social Meaning’ in M Nussbaum and A Sen (eds), The Quality of Life (oup 1993). For Walzer, too, it is society’s current allocative/​distributive arrangements, including the courts, which give rise to ‘radical uncertainties,’ and his own alternatives which are the ‘more systematic:’ M Walzer, ‘The long-​term perspective’ (1986) 62(1) Bulletin of the New York Academy of Medicine 8, 14.

Evaluation

317

is the Court’s very problem, and any certainty which the current system has merely reflects consistency of judicial taste (and distaste).27 9.5

Judge Right Now

As seen at the end of the last section, in the heads-​and-​tails world of Aristotle, the judge’s role seems to be one of ‘calling’ the coin toss. He or she takes a view as to the likeness or unalikeness of two things or people, A and B: a ‘likeness call.’ However, he or she sometimes changes his or her view in a later case, as was seen in Maruko28 when the Court of Justice ruled that life partnerships were ‘like’ marriages between opposite sex partners, contrary to what had been decided in, for example, D, where the two types of union had been ‘unlike.’29 Where this happens, however happy the outcome, there is an affront to logic,

27

28 29

In a masterful essay, Somek has set out a theory of the Court’s current approach to equality and non-​discrimination cases, and that theory includes more than a hint of an allegation of neoliberal bias on the Court’s part. According to Somek, at the risk of oversimplifying his thesis and impoverishing the richness of its construction, ‘anti-​ discrimination law is sufficiently flexible to have, where necessary, protection from discrimination yield to imperatives of economic rationality’:  A Somek, Engineering equality:  An essay on European anti-​discrimination law (oup 2011)  158. Business-​ friendly defences to claims of gender and age discrimination are examples, and the cjeu’s judges can exclude or permit such defences at their discretion simply by use of the like-​for-​like test to either ‘create […] comparability’ or ‘invok[e]‌[…] difference,’ which in turn determines the case’s classification as either one of direct discrimination (defences excluded) or indirect discrimination (defences permitted): ibid 124. Of course, if it should turn out that the Court’s alleged bias is present in its judgments independently of the like-​for-​like test, as Somek suggests for example in relation to Case C-​303/​06 S Coleman v Attridge Law and Steve Law [2008] ecr I-​5603 (withholding from Ms Coleman of benefit of accommodation rules ring-fenced for the disabled), then changing or tempering the test will make no difference: ibid 182. This case is dealt with in Chapter 4 above, text relating to n 280; the paragraph of the judgment which Somek refers to is para 42. The point is respectfully disagreed with, however, because it is clear from the description of the facts at para 26 that Ms Coleman was not seeking (associative) accommodation, that is, receipt of treatment over and above what employees with non-​disabled children received, but merely to be brought up to the same level of treatment as them. Case C-​267/​06 Maruko (n 7), discussed above, section 4.5.1. Joined cases C-​122/​99 P and C-​125/​99 P D and Kingdom of Sweden v Council of the European Union [2001] ecr I-​4319. Discussed above in the same section.

318 Chapter 9 unless, of course, A and B have actually changed.30 But in this case it is hard to see how marriage certificates, or certificates attesting to life partnerships, or the relationships they stand for, had changed one jot. Thus it seems that it might at least be more intellectually honest for the Court of Justice to eschew a method of adjudicating claims of unequal treatment as between A and B based on questionable exclamations of purported likeness (or unalikeness) between the two, and instead to embrace a method based on periodic appeals to society with a view to ascertaining whether its view of the meaning of a certain good or goods (the certificates, say) has or has not changed since the last such ascertainment. Even if not getting rid of the exclamations themselves, such a method would at least give them a stronger theoretical, or indeed empirical, underpinning. In philosophy, foundationalist thinkers like to believe in something which Sellars called ‘the given,’31 or a ‘non-​inferential knowing […].’32 Rorty called it an ‘a priori constraint.’33 However, other philosophers struggle with, or even despair of, such a thing. First to do this, perhaps, was Hume, and then later Sellars himself, who declared the Given to be a ‘Myth.’34 Unfortunately what seems to happen in the Law is that precedents are afforded the same level of reverence as the Given, and become an a priori constraint on the judges of subsequent cases. It would be instructive, perhaps, for lawyers to remember Rorty’s famous quip: [Truth is] what your contemporaries let you get away with.35 If he is right, then it would follow –​in micro –​that when a judge gives his or her verdict, his or her ‘true speech,’ it has no special claim to rightness, legitimacy or believability; it could be argued that it is just (to use Rorty’s provocative 30

31 32 33 34 35

Oddly, the gap between the two viewings does have philosophical significance. Certain logicians, going back as far as Hume, believed that the ‘vulgar[‘s]’ insistence that A continued as A, and that B continued as B, during the gap, or ‘interruption,’ in observation was a ‘gross illusion.’ This belief lies at the heart of the empiricism for which he is famous. It calls, of course, for fresh observation from moment to moment: D Hume, A Treatise of Human Nature (DF Norton and MJ Norton eds, oup 2007) Book I, Part IV, sec 2 (Of skepticism with regard to the senses) 125, 128, 139, 144. W Sellars, Empiricism and the Philosophy of Mind (Harvard University Press 1997) 13. ibid 22. R Rorty, Philosophy and the Mirror of Nature:  Thirtieth-​Anniversary Edition (Princeton University Press 2009) 9. Sellars (n 31) 33 and throughout. Rorty (33) 176.

Evaluation

319

phrase) whatever he or she could get away with. In the uk, Baroness Hale at least seemed to acknowledge the rightness/​legitimacy point, when she said, We are not the final court of appeal because we are always right. We are right because we are final.36 She is correct; the buck does have to stop somewhere. And it is the rightness/​ legitimacy point which lies at the heart of the anxiety which some may feel about the thesis contained within this book. This anxiety proceeds along the following lines: Who will decide the meaning? The Forum? But they may be an angry bigoted mob, brainwashed with majoritarian illiberal thoughts, hostile to justice or change! The judge? But they may base their ‘meaning’ on what the matter means to them –​ personally –​and this personal meaning would in turn be the product of their environment, upbringing and preexisting belief system. Bias and a failure to appreciate the Other’s standpoint are inevitable. So how is this theory better than, or even as good as, what we had before? It is better for two reasons. Firstly, because the judge is asking himself or herself a different question: he or she is pronouncing on meaning, not equivalence. (And there is no ‘nature’ meaning, only a ‘nurture’ one.) And, unlike before, where meaning ascertainment came up randomly in the course of proceedings and was met with ‘common sense’ responses, here meaning would have been the point of proceedings and evidence from all ‘interpreters’ would have been heard. Secondly, there is the difference in the number of interpretations or ascertainments. As Walzer put it, [T]‌he arguing and the fighting have no visible end.37 It is intellectually unnerving when someone changes their mind about an equivalence, like announcing that one and one no longer equal two. ‘Changing one’s mind’ about a meaning is innate, second nature. Speakers use words with dual meanings all the time  –​there is no anxiety here because, almost from the cradle, they appreciate the need for (or resign themselves to) the duality and proceed to use deduction from context to discern the other speaker’s intended meaning. Meanings are discerned, rather than chosen, and so there is 36

37

Baroness Hale of Richmond, ‘Welcome to the uk Supreme Court?’ (2008 Denning Lecture) accessed 22 May 2018, 22. M Walzer, ‘Liberalism and the Art of Separation’ (1984) 12(3) Political Theory 315, 328.

320 Chapter 9 no ‘changing of mind’ the next time, just a rediscernment. Discern, and discern again. Under a more postmodern system, such as the one set out in this book, the doctrine of precedent would not set the meaning in concrete. And neither is the new concrete meaning hung around the necks of the losing ‘interpreters’ forever more. The conversation about meanings –​the interpretive effort –​goes on, case to case, context to context.38 And Walzer says that this effort ought to be understood ‘less by analogy with what philosophers do’ and more by analogy with, say, architects: ‘No architect … aims to design the last building’.39 Postmodernism and precedent will be investigated further below. But in the Aristotelian world, treating one distributee differently from another similarly-​situated distributee is automatically exceptional. The big guns have to be rolled out to make the situation good: the doctrine of proportionality, for example (‘it is unequal but only a little’), or higher order norms such as fundamental rights (‘it is unequal but the rights of the one for whom the deal turned out well simply trumped the rights of the one for whom it turned out badly’). The first of these approaches  –​applying the doctrine of proportionality –​makes little sense as a salve for inequality when one considers that, from the Walzerian point of view, a purported ‘small inequality’ may be tiny, yet still fail as a small inequality by dint of a boundary breach. Likewise, a purported ‘small inequality’ may be significant in size yet still be perfectly apt within its/​the relevant sphere. The bigness or smallness is not the right test, the aptness is. The second –​checking for a breach of fundamental rights –​is better. Avoiding a breach of rights is a good reason for acting unequally, but that is because such an approach acknowledges that there are multiple interpretations of the act in question, and then favours (contextually, not automatically) one of the interpretations over the other: a Walzerian approach. Walzer has the advantage, however, by running this comparison-​ of-​ interpretations as the main event, not as the hurried encore to a doomed comparison-​of-​actors. Put another way, Aristotle can only ever see prelitigation deeds as misdeeds (to act unequally is always a breach of the Aristotelian equality principle). Litigation is about the perpetrators of misdeeds explaining themselves. Walzer is different –​prelitigation deeds are just contested distributions, of themselves neither good nor bad (to act unequally does not necessarily breach the Walzerian equality principle); litigation is simply the contest 38

39

According to Bauman, those living through ‘liquid modernity’ must in any case get used to this kind of ‘travel[ling] without an idea of destination:’ Z Bauman, Liquid Modernity (Polity 2000) 134. M Walzer, Thick and Thin:  Moral Argument at Home and Abroad (University of Notre Dame Press 1994) 52.

Evaluation

321

which ensues (an ‘appeal’ for a disgruntled distributee, as it were). The Court of Justice with its current approach shoots first and asks questions later. Walzer favours conversation over condemnation. To ask questions first and (if necessary) shoot later not only seems ‘fairer,’ it also pays a more honest tribute to the principle of equality itself. To breach this principle is no longer just to draw a distinction, but to flaw a distribution. And while there are only bad discriminations (albeit subdivided into the justified and the unjustified), there are no bad distributions, only bad distributive criteria. It is not hard for a court to investigate whether distributive criteria are bad or not. The Walzerian judge holds a trial and then convicts or acquits. But all discriminators are convicted on the simple strength of having discriminated, leaving the Aristotelian judge nothing to do except to hear endless pleas in mitigation. As stated at the beginning of this section, lawyers struggle with a kind of Myth of the Adjudged, such that a judicial decision stands singular, victorious over all other decisions which could have been made, either the truth, or something very close to it. And like the truth, only one decision, or foundation, or postulate, call it what one will, may obtain at any one time: The rest is falsehood. In the Law, it is submitted, the time is right to drop the flawed and refutable idea of Decisions as Truths requiring Belief. It is from a bygone era, and Aristotelian likeness and unalikeness,40 similarly belongs in that era. Walzerian meaning –​ascertainment or discernment, requiring a richer empirical enquiry, need not be held to be true, only empirically adequate. Aristotle’s ‘either/​or’ decisions smack of absolutism and the Myth of the Given. They demand expressio unius est exclusio alterius belief.41 Empirically-​rich Walzerian decisions would leave logical wiggle-​room for the later arrival of new and different experiments, or reruns of old ones, but with different results. And while those who object to an Aristotelian decision need to allege falsehood and attempt to rewrite a myth, those who object to a Walzerian discernment need only allege an inaccuracy, and then seek new discernment armed with new, better and/​or more persuasive data. They say that if one does not like the answer, one should change the question. The judge at the Court of Justice, asking a more appropriate question and taking a new view with regard to the permanence of the answer, can act as a filter42 or funnel43 of the relevant distributive community’s meanings, and then act as a bulwark against bigotry or brainwashing. 40 41 42 43

Lending itself to belief as –​like belief/​a thing believed –​it is given its identity by what it excludes. Quine’s law of excluded middle. Pure Walzer, plus Override. Mediated complexity.

322 Chapter 9 9.6

Postmodernism and Precedent

It sometimes feels like law was the only field which did not have a ‘postmodern turn.’44 Postmodernism, defined by Lyotard as ‘incredulity toward metanarratives,’45 has been described as ‘the Swiss Army Knife of critical concepts.’46 With its erosion of disciplinary lines and its ‘relaxation of categories,’ it defied and perhaps continues to defy the ‘orderly impulse’ of the lawyer.47 The matter has been investigated at length elsewhere.48 However, as this book proposes the replacement, or at least the complementing, of Aristotle’s ‘either/​ or’49 with Walzer’s ‘both/​and,’50 the discreet question of ‘fit’ might be briefly addressed. 9.6.1 Dworkin on Precedent, and A Critique Certainly from the late 1960s, Ronald Dworkin was the preeminent keeper of law’s ‘grand narrative.’ In the ‘Hard cases’ chapter of his Taking rights seriously, he patiently explains how his chosen judge, Hercules, must first interpret statutes and then, in an even more difficult labour, interpret his and other courts’ own previous decisions in execution of the doctrine of precedent. He is at pains to point out that, even if such a decision may be said to have ‘enactment force,’51 it is not a statute emanating via Parliament from the will

44 45 46 47 48

49 50

51

G Adamson and J Pavitt, ‘Postmodernism: Style and subversion’ in G Adamson and J Pavitt (eds), Postmodernism: Style and subversion, 1970–​1990 (V&A Publishing 2011) 13. J-​F Lyotard, The postmodern condition:  A report on knowledge (G Bennington and B Massumi trs, Manchester University Press 1984) xxiv. Louis Menand, ‘Saved from drowning: Barthelme reconsidered’ The New Yorker (New York, 23 February 2009) 68. -​-​, ‘Curators’ Foreword’ in Adamson and Pavitt (n 44) 9–10. AC Hutchinson, ‘A Postmodern’s Hart: Taking Rules Sceptically’ (1995) 58 mlr 788; HM Stacy, Postmodernism and law: Jurisprudence in a fragmenting world (Ashgate 2001). While Ladeur’s proposals (an adaptation to the logic of plural legal regimes) are very different from the proposals in this book, his postmodern understanding of law and descriptions of the Court of Justice’s singular failure to appreciate this are also very instructive: K-​H Ladeur, ‘European Law as Transnational Law –​Europe has to be conceived as an heterarchical network and not as a superstate!’ (2009) 10 German Law Journal 1358. -​-​, ‘Curators’ Foreword’ (n 47) 10. ibid. In a section in his book On Toleration entitled ‘Postmodernity?’ - in which he contemplates what he calls the ‘postmodern project’ - Walzer comments: ‘We still know ourselves to be this or that, but the knowledge is uncertain, for we are also this and that’: M Walzer, On Toleration (Yale University Press 1997) 88, 90. R Dworkin, Taking Rights Seriously (Duckworth 1977) 111.

Evaluation

323

of the democratic majority  –​a vocalisation not just of that majority’s principle, but also of its policy –​like the statute the interpretation of which was at stake in the case being decided. Rather, the decision is a creature of principle alone.52 If a highly simplified example may be forgiven, suppose that one had a Government of one stripe which pursued policy X, then one of the opposite stripe which pursued policy Y, and finally one of the first stripe again pursuing, once again, policy X.  The progress of policy over those fifteen or so years would be XYX. Dworkin’s point is that, over the same time period, a judge acting according to the doctrine of precedent, and facing cases in the same or analogous fields, may not take cavalier policy decisions in the same way as the various Governments,53 and may look only to the principle arising from the precedent under consideration. If the first judgment of a three-​ judgment sequence thus gave rise to principle X and was given in the context of policy X, then the judge must follow principle X, even when the context shifts to policy Y. All other things being equal, the progress of principle over that time period would thus be XXX. Dworkin, drawing on the celebrated legal positivist hla Hart, calls this a ‘doctrine of fairness:’54 The practices of precedent do not suppose that the rationales that recommend judicial decisions can be served piecemeal …55 However, if Scalia is right that the evolutionary approach is in fact merely adherence to the doctrine of precedent,56 then the focus must shift to those occasions where the judges do ‘recant (…)’57 the earlier decision and move the law along: when X becomes Y. For him or her to depart from a previous judgment, according to Dworkin, Hercules must construct a whole ‘theory of mistakes.’58 Part of this process will involve Hercules assessing what Dworkin calls the ‘gravitational force’59 of the judgment the overruling of which is under consideration. Was this judgment ‘limited to its enactment force?’60 If it was,

52 53 54 55 56 57 58 59 60

ibid. He may not be a knight-​errant, in Cardozo’s memorable phrase: BN Cardozo, The nature of the judicial process (Dover 2005) 137. Dworkin, Taking rights seriously (n 51) 113. ibid 115. Above, section 8.4. Dworkin, Taking rights seriously (n 51) 115. ibid 121. ibid. ibid 113.

324 Chapter 9 then the overruling, in a later case occurring in a different context, was justified. But there was a second possibility. As he had noted earlier: Judges and lawyers do not think that the force of precedents is exhausted, as a statute would be, by the linguistic limits of some particular phrase.61 So was the precedent under consideration perhaps a precedent of that sort, ‘exert[ing] a gravitational force on later decisions even when these later decisions [lay] outside its particular orbit?’62 If it was, in other words, if its underlying principle could be transplanted to the new context, then Hercules was obliged to carry out this transplant. As mentioned above, judges dealt in principle only; it was legislators who dealt in policy and principle both. Dworkin’s concept of the ‘gravitational force’ of judgments merits a closer investigation, though. In physics, the gravitational force exerted by one planet over another would depend on the size of the planet  –​it must be assumed that all judgments are the same ‘size’ –​and the distance between the two planets concerned. However, Hercules seems to be gauging this distance, and the resultant gravitational pull of the earlier decision, him or herself. The fact that the judge can declare the distance between the planets him or herself makes a big difference to the analogy and robs it of some of its power; if the force of gravity depended on distance-​perception then people could commit suicide by jumping out of a ground floor window. So what Hercules is really doing is making a decision as to whether the two contexts, the present one and that of the earlier case, are like or unlike, in order to decide if the two cases should be decided in like or unlike fashion. That the Aristotelian test for equality is entrenched in Dworkin’s theory of Hard Cases is made explicit: The gravitational force of a precedent may be explained by appeal, not just to the wisdom of enforcing enactments, but to the fairness of treating like cases alike.63 A second part of the theory of mistakes concerns the role or rather non-​role of Hercules’ community. This links back to the discussion in section 8.4 above about the way in which a judge relates to his or her society. However, according 61 62 63

ibid 111. ibid. ibid 113.

Evaluation

325

to Dworkin, if the earlier decision with which Hercules is faced had in the meantime become unpopular, as a matter of popular morality, or community morality, this need not sway him or her. Rather, he or she should concern him or herself only with constitutional/​institutional, or political, morality, which apparently trumps the other kinds in the event of inconsistency. According to Dworkin, the ordinary man’s political preferences have not been subjected to the appropriate ‘dialectical[ly] skill[ful]’ examination.64 Thus Hercules (as opposed to Herbert, his or her less clever rival) would not consider him or herself ‘the agent of the majority’65 in the same way that Judge Barak did not, in the discussion above.66 If the evolutionary approach is circumscribed in this way, so that the stasis is only punctuated where constitutional, not merely popular, morality calls for such punctuation, such calling determined by the judge alone using their dialectic skill, then Scalia can be agreed with that the evolutionary approach is no more than the normal execution of the doctrine of precedent. Evolution in this book is defined more ambitiously, and may require more frequent breaches in the doctrine of precedent, perhaps even an abandonment of that doctrine itself, as citizens change their mind about meanings.67 Such changes of mind may have come about because an alternative meaning to the one currently prevalent has been more persuasively argued in the Forum, or its popularity more persuasively demonstrated in front of the Court. Thus, Dworkin’s characterisation of the doctrine of precedent as a ‘chain novel’ is rejected here in favour of that of the ‘series of independent short stories,’68 or even better, something like Eve Kosofsky Sedgwick’s ‘varied, contingent, recalcitrant but re-​forming seriality.’69 A  Walzerian judge would have no qualms bringing policy into his or her courtroom  –​arguably the current, hidden

64 65 66 67

68 69

ibid 129. ibid. Above, section 8.4. Cardozo, while stopping short of calling for the abandonment of the doctrine, nevertheless takes a progressive view of precedent and is not afraid to take the social aspect of a case into account: ‘when the social needs demand one settlement rather than another, there are times when we must bend symmetry, ignore history and sacrifice custom in the pursuit of other and larger ends’. See BN Cardozo, ‘Lecture II: The methods of history, tradition and sociology’ in C Morris (ed), The great legal philosophers: Selected readings in jurisprudence (12th edn, University of Pennsylvania Press 1997) 519. R Dworkin, A Matter of Principle (Clarendon 1986) 159, 161. Eve Kosofsky Sedgwick, ‘Tales of the Avunculate:  Queer Tutelage in The Importance of Being Earnest’ in EK Sedgwick, Tendencies (Routledge 1998) 63.

326 Chapter 9 practice anyway –​and his or her decisions might look decidedly piecemeal, postmodernism of course recognising flexibility as strength.70 9.6.2 Postmodernism Challenges Neutrality and Positivism Besides flexibility, itself threatening, from a Dworkinian perspective, the achievement of fair outcomes for litigants, what else does postmodernism have to offer law that might recommend the introduction, into the law as currently practiced, of Walzer’s ‘postmodern project?’ ‘Post’ implies replacement. There was, and is, certainly a lot in the law that one might wish to replace, including firstly its insistence on Enlightenment objectivity, complete with neutral vantage point, and secondly its overreliance on positivism as its most favoured interpretive method. As Stacy puts it, the latter method provides ‘a comforting appeal to linearity and the scientific method:’71 its self-​reiteration provides the very basis for its legitimacy in the eyes of the general population: Despite evidence that the social world rarely acts in linear ways, positivism is the legal default when human conflict becomes messy, chaotic and unpredictable.72 Critical theorists without doubt did a lot to address the first of these problems, which Stacy calls ‘law’s foundational claim to disinterested objectivity.’73 The cold rationality of judges was exposed as a means of concretising patterns of dominance in society, their supposedly neutral rules as biased benchmarks forcing those who deviated from the norm into line. However, the critical theorists still utilised a neutral vantage point and so ultimately were simply espousing ‘novel applications’ of the ‘good-​old fashioned lawyer-​ academics,’74 that is, a belief in ‘monolithic process’ leading to single ‘right answers.’75 Postmodernism adopts the subject-​centred reasoning that these theorists fought hard for, but attempts to take it further still. The litigants’ 70

71 72 73 74 75

It was clear then that the two theorists would, and did, argue about what was or was not in fact fair, and what justice/​fairness in fact required. Dworkin pointedly gave one description of Walzer’s work the title, ‘What Justice Isn’t:’ Dworkin, Principle (n 68) 214. Stacy (n 48) 47. ibid. ibid 44. ibid 46. G Minda, Postmodern legal movements: Law and jurisprudence at century’s end (nyu Press 1995) 58–​9.

Evaluation

327

and other interested parties’ multiple vantage points are the only ones that matter, and, as will be seen in section 9.7 below, it is they who must view themselves. 9.6.3 Saying Yes to ‘We’ Having declared, then, that ‘[t]‌he grand narrative has lost its credibility,’76 Lyotard went on to warn that in its absence the postmodern world would have to learn to live with incomplete information, tolerating the incommensurable and the undecidables.77 Tests such as Aristotle’s for equality, involving ‘oppositional thinking’78 and ‘input/​output … system[s],’79 come in for harsh criticism in his book, The Postmodern Condition. But apart perhaps from a rallying call to ‘wage a war on totality,’80 he does not especially help his reader with a way forwards. It is this apparent ambivalence as to the direction which society now takes  –​any as good as any other  –​which causes Habermas to accuse Lyotard and others of neoconservatism.81 According to Rorty, in critiquing modern science’s metanarratives, effectively calling for ‘permanent revolution,’82 Lyotard disconnects philosophy from society, and cuts himself off from his community: It is as if thinkers like … Lyotard were so afraid of being caught up in one more metanarrative about the fortunes of ‘the subject’ that they cannot bring themselves to say ‘we’ long enough to identify with the culture of the generation to which they belong.83 Rorty himself, citing Dewey, calls for an engagement with, even an embrace of, ‘the meaning of the daily detail.’84 76 77 78 79 80 81 82 83 84

Lyotard (n 45) 37. ibid xxv (inter alia). ibid 14. ibid 54 ibid 82. J Habermas, ‘The entwinement of myth and enlightenment:  re-​reading Dialectic of Enlightenment’ (1982) 26 New German Critique 13, 28–​9. R Rorty, ‘Habermas and Lyotard on Postmodernity’ in RJ Bernstein (ed), Habermas and Modernity (Polity 1985) 163. ibid, 172. ibid 174–​5. The phrase is from J Dewey, Reconstruction in Philosophy (Boston 1957) 164. It was also from Dewey that Rorty drew one of his most famous mantras, that language was for ‘coping rather than copying:’ R Rorty, ‘Dewey’s metaphysics’ in SM Cahn (ed), New Studies in the Philosophy of John Dewey (University Press of New England 1977) 70.

328 Chapter 9 The best one can do, perhaps, is to take Lyotard’s idea of ‘local determinism,’85 very much elaborated upon by Walzer, as a starting point for how a Walzerian equality case might proceed. To redeploy Dworkin’s own analogy, but reiterating it rather than extrapolating from it, each judge is a sports referee in a particular tournament, determining the outcome of a dispute by reference to the local context and local rules: chess one day, certainly, but draughts the next, and then dominos, and then bridge, and so on.86 Of course the dispute is not between two people playing a sport. It is between two people arguing about a meaning, playing what Lyotard would call a ‘language game.’ He continues: A recognition of the heteromorphous nature of language games is a first step [towards justice] … The second step is the principle that any consensus on the rules defining a game and the ‘moves’ playable within it must be local, in other words, agreed on by its present players and subject to eventual cancellation. The orientation then favo[u]‌rs a multiplicity of finite meta-​arguments …87 This is very close to Walzer’s concept of the continual debate, which was first encountered in Chapter  2 and is central to the construction of a system of spherical justice, even if that must be a series of temporary constructions and reconstructions in a courtroom. However, one must be clear about two things. First, it is not being argued that each context has a meta-​law –​the law can and must remain universal –​merely that each context has a meta-​ethics,88 which informs distributive decisions within that context. Second, one must be clear that the conduct of the court case itself is not a game, and nor are its highly necessary procedural rules subject to cancellation. The language game at issue, and the one requiring local input in its playing and ‘winning,’89 is the working 85 86 87

88

89

Lyotard (n 45) xxiv. Dworkin, Taking rights seriously (n 51) 101ff. Lyotard (n 45) 66 (Lyotard’s emphasis). Note Hutchinson also regards lawyers as being ‘engaged in a language game of rhetorical politics,’ necessitating ‘a healthy sceptical attitude’ from judges:  Hutchinson (n 48)  818. However, he is quick to point out that ‘the sceptical claim is not that “anything goes,” but that “anything might go” ’: ibid 798. This phrase coined by Sussmann:  N Sussmann, ‘Introduction:  The substantive unity of Michael Walzer’s pluralism’ in Y Benbaji and N Sussmann (eds), Reading Walzer (Routledge 2014) 3. Ethics is used here in a general sense to mean ‘rules,’ in the same way that Dworkin talks about the rules of chess, for example: Dworkin, Taking rights seriously (n 51) 101. ‘Winning’ is in inverted commas because those whose favoured interpretation of a given good or utterance is rejected do not, and should not, consider the matter closed or the

Evaluation

329

out of the shifts in social meanings of the good or goods whose distribution is being contested. 9.7

Macro and Micro: The Equilibrium of Co-​Existence

The main source of problems, from the eu legal point of view, is the European Union’s split-​level structure, mentioned above:90 the Member States on the one hand, and the Union itself on the other.91 This can mean that there are two co-​existent distributive communities, at least in areas where the eu does not yet have total competence, but even in some areas where it does –​a micro sphere within a macro sphere.92 And it is not just different goods which may be distributed at the two levels, but also different aspects of the same good. To understand this, it is helpful to recall a slogan that was popular in the arena of ec monetary policy in the Seventies: the Snake in the Tunnel. This refers to the experiment whereby the Member States attempted to prevent wild fluctuations between the European currencies by setting them a maximum of 1.125 percent above the dollar and a minimum of 1.125 percent below the dollar (the ‘Tunnel’), leaving individual currencies free to move about within these, but not beyond them (the ‘Snake’). This imagery could be adapted to represent the situation when, for example, the Union institutions have decided to legislate on a certain matter, say, by means of a Directive. The macro sphere (the Union itself, made up of all of the Member States) is responsible for drafting the Directive, that is, for setting out the basic coordinates of the new law, what it is hoped will be achieved, and, most importantly, the parameters within which the Member States will be expected to act. This is the Tunnel.

90 91 92

chosen interpretation the true one and theirs the false. Durant and Izarra present a nice real-​world example of this, although ultimately the case did not go to court, in respect of the Rushdie controversy in the uk: A Durant and L Izarra, ‘Reading mixed reception: The case of The Satanic Verses’ (2001) 24 Revista de filología y su didáctica 653. See supra section 5.1, last paragraph. Although of course in many cjeu equality cases, only the micro level is implicated. It is mainly in the field of nationality discrimination that the macro level comes into play. The same phenomenon, though from a Scots point of view, is described in TC Smout, ‘Perspectives on the Scottish identity’ [1994] 6 Scottish Affairs 101. However, Smout goes further, holding that there are no fewer than seven ‘concentric rings of territorial identity’ circumscribing every inhabitant of Scotland: ibid 102ff. Nussbaum traces the idea right back to the Stoic philosophers: M Nussbaum, ‘Patriotism and Cosmopolitanism’ in M Nussbaum and J Cohen (eds), For Love of Country, Debating the Limits of Patriotism (Beacon Press 1996) 9.

330 Chapter 9 The distributive community for the Tunnel is the whole of the eu: all citizens of all Member States have the right to participate in the debate, either via their mep’s in the Parliament and the Ministers of their Government (elected by them) in the Council, as at present, or else via some new means, yet to be invented. Thus, the Tunnel will be distributed, to the Member States, in accordance with what it means, collectively, to all of them. The micro level (the individual Member State) is responsible for implementing the Directive within the national borders, in other words, deciding on the best ways to achieve the goals set out in the Directive, and then putting those decisions into practice. This is the Snake: the Member State must work within the parameters set by the Union but must not go beyond them. The distributive community for the Snake is the whole of the Member State concerned. The point is nicely illustrated by the fedesa case.93 Here, the Council approved a Directive prohibiting the use in livestock farming of certain substances having a hormonal action. Fedesa complained that the Directive had had unequal consequences. The new obligations to be fulfilled were the same for all Member States. That meant that the differentials between levels of protection94 in the different Member States remained identical, even after implementation of the Directive: Member States with a high level of existing protection still had a high level of protection, and Member States with a low level of existing protection still had a low level of protection. However, the Court had no sympathy for fedesa’s argument. It was true that all Member States had to fulfill the same new obligations (Tunnel), but, as long as that was done, continuing divergences between them did not breach the principle of equal treatment (Snake). In other words, the new obligations were a product of the macro distributive community –​a single sphere with no internal boundaries. But the Court was happy to re-​erect the boundaries as between the Member States when it came to the Member States’ overall policies on this issue; whether they wanted a high or low level of overall protection was a decision for each of them to make separately. Litigants will sometimes champion the micro sphere, and other times the macro sphere, depending on their interest. Fedesa was fighting for a macro sphere on that occasion; many product-​manufacturers looking to conquer the whole of the eu market with a single design will (usually rightly) do the same. In the competition case of Distillers,95 meanwhile, the litigant bemoaned the 93 94 95

Case C-​331/​88 The Queen v Minister of Agriculture, Fisheries and Food and Secretary of State for Health, ex parte: Fedesa and others [1990] ecr I-​4023. Against hormones finding their way into foodstuffs. The Distillers Company Limited [1978] oj L50/​16, [1978] 1 cmlr 400; on appeal Case 30/​ 78 Distillers Company v Commission [1980] ecr 2229, [1980] 3 cmlr 121.

Evaluation

331

new eec obligation to charge the same price for whisky in every Member State, because, according to it, the customs and traditions pertaining to the consumption of whisky were different from one State to the next. It was arguing, unsuccessfully as it turned out, for the Community to be viewed as nine96 micro spheres, each distributing whisky after its own fashion. What is sometimes hard for litigants (and others) to grasp is that the two models –​one large sphere, many small spheres –​co-​exist. Which one is to the fore depends on the circumstances, hence the absolute necessity for the Court to adopt a variable approach.97 Walzer expresses this well, using the metaphor of the mirror which Hamlet holds up to his mother.98 The viewer looks in the mirror and sees what is wrong, where the definition of ‘wrongness’ is the viewer’s own. Each viewer or group of viewers has their or its own mirror. But sometimes the definitions will coincide. As he puts it, Nor is the meaning of critical exposure always a local and particular meaning. If what is rotten in the state of Denmark includes murder and betrayal, then all of us can recognize the rottenness. For some purposes, we all stand in front of the same mirror. But only for some purposes …99 For some purposes, then, the citizens of the European Union ‘stand in front of the same mirror.’ For others, ‘local and particular’ meanings return, and the one mirror becomes twenty eight separate ones. Difficulties arise, though, where the ‘rottenness,’ the legal mischief requiring resolution, is not visible in the smaller mirrors, but only in the bigger one, from which Member States then deliberately avert their gaze. Diversity issues inevitably give rise to this type of situation. As Kjaer and Adamo have written: Most eu Member States seem to endorse the view that diversity is valuable only if they are in charge of that diversity, defining its meaning and limits.100

96 97 98 99

The case was first brought in 1978. See supra. W Shakespeare, Hamlet, III.iv. M Walzer, The Company of Critics:  Social Criticism and Political Commitment in the Twentieth Century (2nd edn, Basic Books 2002) 233. 100 AL Kjaer and S Adamo, ‘Linguistic diversity and European democracy: Introduction and overview’ in AL Kjaer and S Adamo (eds), Linguistic diversity and European democracy (Ashgate 2011) 10.

332 Chapter 9 This in turn gives rise to a clash of the models. An oft-​quoted example is the Danish Bottles case.101 Here, a local law establishing a complex ‘return’ system for bottles, the aim of which was the reduction of waste and the protection of the environment, was alleged to be a hindrance to the free movement of goods. Although it accepted that protection of the environment was, in principle, a justification for the action (a so-​ called ‘mandatory requirement’),102 the Court went on to hold that Denmark’s behaviour was –​at least partially –​disproportionate to this aim. While some have praised the decision, given that a system such as the Danish one was not in theory declared to be inimical to the free movement of goods,103 Denmark’s defeat in practice here spells a victory for the macro sphere over the micro one. Was permission to trade in bottles to be distributed at micro level, where, in this case, environmental concerns were the dominant consideration? Or was it to be distributed at macro level, where economics took greater priority? And if the latter, is the Court itself guilty of a boundary breach here –​dethroning the local Distributor and rewriting the distributive criteria after its own fashion, ignoring those arrived at by the populace? Although a cursory study of Walzer’s work might suggest that micro should always triumph over macro, particular over universal, this is not in fact strictly the case. As he made clear in a famous essay of 1990, and reiterated in a book of 2004, the outer sphere must sometimes take precedence. Writing in the context of the State’s relation to its citizens, he observes as many others have104 that classic liberalism produces a society of bloodless, abstract individuals, each free to choose their own conception of ‘the good,’ but, in relation to one another, isolated, divided, fragmented, and undetermined:

101

Case 302/​86 Commission of the European Communities v Kingdom of Denmark [1988] ecr 4607. 102 ibid paras 8 and 9. 1 03 Sexton goes as far as to call Commission v Denmark ‘a landmark decision for environmentally minded States:’ ‘This decision represents the first time the Court has allowed Member States to enact environmental protection measures contrary to economic integration’:  TRF Sexton, ‘Enacting national environmental laws more stringent than other States’ laws in the European Community: Re Disposable Beer Cans: Commission v Denmark’ (1991) 24(3) Cornell Intl LJ 563, 593 and 564. 104 Charles Taylor’s essay ‘Atomism’ is one of the most well-​known versions of this observation:  C Taylor, ‘Atomism’ in C Taylor, Philosophical Papers Vol 2:  Philosophy and the Human Sciences (cup 1985). But see also J Waldron, ‘The Cosmopolitan Alternative’ in W Kymlicka (ed) The Rights of Minority Cultures (oup 1987) for a celebrated rebuttal of Taylor and others.

Evaluation

333

The members of liberal society share no political or religious traditions; they can tell only one story about themselves and that is the story of ex nihilo creation, which begins in the state of nature or the original position. Each individual imagines himself absolutely free, unencumbered, and on his own.105 Walzer’s ‘communitarian critique,’ however, requires States to remedy this by ‘sponsor[ing] certain sorts of communal identification’106 and ‘foster[ing] associative activities:’107 [T]‌he communitarian correction does require … a state that is, at least over some part of the terrain of sovereignty, deliberately nonneutral.108 The State, then, must sometimes step in to protect those groups within the community which are ‘at risk.’109 It follows that this nonneutral State must from time to time lend its support to a certain identity, in the process withdrawing its support from another: it must ‘discriminate among … assemblies,’110 the macro sphere –​on that particular occasion –​overruling the micro one. So although by nature hostile to what might be called ‘outside intervention,’ outsiders deciding something which it is the insiders’ right to decide, Walzer does endorse it when it is needed precisely to uphold the system which bestows this right to begin with: being cruel to be kind.111 In Politics and Passion, Walzer again examines the situation of a smaller group within a larger one, and reaches similar conclusions. Where the larger one is the State, its need occasionally to overrule the smaller one, in the name 105 M Walzer, ‘The Communitarian Critique of Liberalism’ (1990) 18(1) Political Theory 6, 7–​8 (Walzer’s emphasis). 106 ibid 7. 107 ibid 16. 108 ibid. 109 ibid 17. 110 ibid 19. 111 As he puts it, ‘the only theory that is necessary to the communitarian critique of liberalism is liberalism itself’: ibid 8. State intervention, although not unfettered state intervention, was also discussed by Walzer in a 1993 essay: M Walzer, ‘Exclusion, Injustice, and the Democratic State’ (1993) 40 Dissent 55. This is considered in more detail supra at section 7.2.3. An example of a sphere-​within-​a-​sphere utilized by Walzer in Spheres of Justice is that of a church within a State (n 8, at 35–​42, for example). Such a church could enact its own rules on membership, content of services, upkeep of buildings, and so on, but it would still be subject to the laws of the land vis-​à-​vis crime, tax, et cetera.

334 Chapter 9 of ‘assist[ing]’ other smaller groups which find themselves under threat, is seen again.112 The same sense of having to be cruel to be kind is encountered: the cultural community at issue must engage with the larger entity, and vice versa, precisely to ensure the relevant culture’s perpetuation: So the internal hierarchies of group life are tolerated by this version of egalitarianism but also simultaneously subverted.113 And in the event of a clash, at least where the cultural community poses a threat to the larger entity, he is clear that the balance should be tilted against the community.114 However, he warns that this is advisory only: This liberal tilt is simply a guideline for decisionmaking in a political crisis. It doesn’t solve the problem of day-​to-​day coexistence. For that there is no theoretical solution, no deduction from a set of principles, only a long and unstable series of compromises … universal happiness is not a plausible political project.115 This caveat should be taken as applying to this book as well. The Walzerian model –​with the general rule of micro trumping macro, and the exceptional ‘rule’ of macro trumping micro –​should be seen as nothing more than a ‘guideline for decisionmaking.’116 112

M Walzer, Politics and Passion: Toward a More Egalitarian Liberalism (Yale University Press 2004) 40: ‘A form of state provision designed to assist the weaker groups will be a necessary –​ and permanent –​feature of any egalitarian multiculturalism’ (Walzer’s emphasis). 113 ibid 57. 114 ibid 64. This does seem to be a softening of his line in Spheres of Justice. 1 15 ibid 65. 116 It is not the intention of the foregoing to promote the eu as a whole to the role of ‘State,’ and to relegate its current Member States to that of ‘secondary associations,’ or to insinuate that the latter pose ‘threats’ to the former, and so on. These two pieces of Walzer’s are considered only for the insight they can give into the troublesome question –​not dealt with in Spheres of Justice –​of how an alleged boundary breach is to be dealt with where the spheres are not adjacent but concentric. Are the borders of an inner sphere impenetrable, or porous? As usual, it will depend on whether the business of the outer sphere (for want of a better term) is part of the shared meaning of the distribuend at issue in the inner sphere –​shared by the members of the inner sphere, that is. But these pieces rightly draw attention to the atypical nature of the situation. The same rules as would apply were the two spheres side by side cannot apply here; new rights and responsibilities flow from the fact of the one being inside the other. Returning to the question of State versus individual(s), Walzer has in fact written of his ‘dilemma’ in trying to find a midway point

Evaluation

335

This model can be extrapolated outwards –​mutatis mutandis –​to create, in Walzer’s own phrase, a ‘republic of republics.’117 Whether this could ever work at the global level is questionable,118 but the European Union is certainly very close to a working model of the republic of republics, which, as Walzer mentions in a footnote, goes hand in hand with complex equality.119 So while Danish shared meanings should be respected ordinarily,120 a case can be made for the macro sphere –​the eu as a whole –​intervening via its Court to check Denmark’s environmentalism where the ‘bigger picture’ may be at stake.121 9.8

Complex Equality and European Union

The situation seen at the end of the previous section is, it is stressed, the exception. Under normal circumstances, the theory of complex equality favours local norms and local meanings, and to apply a local meaning outside its sphere may constitute a boundary breach. Each sphere has its own meanings and between the porous inner border and the impenetrable one  –​full intervention in the inner sphere and none: Politics and Passion (n 112) 171 (footnote 8 to c­ hapter 3). If he were prepared to move one level of analysis up, he might find the eu to provide just such a creature, at least in embryonic form. 117 Walzer, Communitarian Critique (n 105) 20. 118 See the section entitled ‘Conclusion: Global Equality’ in Politics and Passion (n 112) 131ff, where Walzer looks at some of the difficulties which the global version might encounter. 1 19 Walzer, Communitarian Critique (n 105) 23 (footnote 21). 120 This is also in keeping with the principle of subsidiarity, where action at the higher level should only be contemplated if there is no solution to be found at the lower level. See below for further discussion of this point. 121 In this case –​perhaps –​the eu’s need for a functioning system of trade, and efficient collection of vat monies which then feed Union projects, including environmental protection. Again there is a sense here of the macro sphere’s being ‘cruel to be kind,’ denying pluralism to sustain it, suppressing environmentalism to promote it, but the environmentalism promoted is likely to offer greater protection than the environmentalism suppressed. Of course being part of the eu requires Member States to accept that there is a ‘picture’ bigger than their own, and if the acceptance comes hard this is almost certainly an occurrence or reoccurrence of the demos problem so masterfully identified by Weiler in his seminal paper: JHH Weiler, ‘Does Europe Need a Constitution? Demos, Telos and the German Maastricht Decision’ (1995) 1(3) European Law Journal 219. As he puts it, ‘majority rule is only legitimate within a demos … a parliament without a demos is conceptually impossible, practically despotic’ (at 228–​231). Such questions, and their elusive answers, are regrettably beyond the scope of this book.

336 Chapter 9 those visiting a sphere from outside should ordinarily respect these: ‘When in Rome, do as the Romans do’. The question is, does such an approach suit the European Union and its law, and more particularly its Court and its Court’s approach to equality? Of course, one of the chief missions of the eu is to prevent European markets from being ring-​fenced along national borders; this inevitably means that a fence erected by a Member State –​usually in the form of a national law –​will be struck down as a hindrance to free movement. In this specific sense, it could be said that local meanings are being routinely jettisoned in favour of non-​local meanings in which the Member State concerned had no say whatsoever, contrary to complex equality.122 The Belgian community may have felt that it was having Spanish rules on surnames foisted upon it in García Avello, for example, and that its view on the meaning of a surname was being ignored.123 More recently, though, the Court has displayed a new-​found ‘respect for Member State constitutional identity’124 in a case like Sayn-​Wittgenstein, where it was the migrant who had to conform to the host Member State’s surname rules.125 However, many aspects of European Union do follow the ‘When in Rome’ precept, making complex equality a good match for the eu project 122 Although this is to ignore the second, ‘macro’ sphere which encompasses all Europeans (see supra section 5.1). This distributive community, having agreed, via election of representatives and sometimes even via referendum, to create/​join the eu, has also agreed to the principle of non-​discrimination on the grounds of nationality, at least as a grundnorm for its regional legislation. So the members of the micro distributive community (qua members of the macro distributive community) have had a bigger ‘say’ in things than they may think. Of course, that still does not mean that they had a say in their neighbour’s law. But, by agreeing to equality on grounds of nationality, they accepted a ‘tunnel’ (see the discussion at supra section 9.7.), which foreclosed certain options to them in the drafting of their own law. The point is that this foreclosure was something which they distributed themselves, to themselves, in accordance with a meaning which they shared, even if they later chose to forget it. Maduro AG attempts to remind them in the last sentence of Case C-524/06 Heinz Huber v Bundesrepublik Deutschland [2008] ECR I-9705, Opinion of AG Maduro, para 18. 123 Case C-​148/​02 Carlos Garcia Avello v Belgian State [2003] ecr I-​11613. Similarly, in the later case of Grunkin, the German community may have felt that it was having Danish rules on surnames foisted upon it, in disregard of its view on the meaning of a surname: Case C-​353/​06 Stefan Grunkin and Dorothee Regina Paul [2008] ecr I-​7639. 124 LFM Besselink, ‘National and constitutional identity before and after Lisbon’ (2010) Utrecht Law Review 6(3) 36, 36. 125 Case C-​208/​09 Ilonka Sayn-​Wittgenstein v Landeshauptmann von Wien [2010] ecr I-​13693. See, also by Besselink, LFM Besselink, ‘Annotation Sayn-​Wittgenstein’ (2012) 49(2) CML Rev 671, where he talks about there being ‘a new equilibrium between the constitutional orders of the eu and the Member States:’ ibid 688.

Evaluation

337

generally.126 In cases like Baumbast127 and Collins,128 for example, the migrants actually sought the host State’s benefits, and, certainly in Baumbast, the Court took a very Walzerian line by insisting that the uk accede to their request; Mr Baumbast was in Rome, and was entitled to be treated as other Romans. In cases like Konstantinidis129 and Laval,130 the basic rule seemed to be that the migrant worker or service provider should take the host State rules as he or she found them, unless they were overridden by Community primary or secondary legislation.131 Similarly, in Vlassopolou,132 the host State’s rules on professional qualifications took precedence, just as long as it did not ignore whatever skills or knowledge the migrant worker had brought with him or her, and, if they were deemed insufficient, was prepared to offer him or her an opportunity to prove that he or she in fact possessed the knowledge and qualifications which were lacking (as opposed to making him or her start from scratch). It is probably in the sphere of the free movement of goods and services that eu law could be said to be at its most anti-​Walzerian, with the doctrine of ‘mutual recognition’133 effectively forcing a host Member State to accept the decisions of an alien distributive community as valid for its own. A  foreign 126 Tsakatika, although undertaking a very different project to the present one, has also found communitarianism useful in analysing the eu (communitarianism as a whole, that is, not any one particular theory): M Tsakatika, Political responsibility and the European Union (Manchester University Press 2008). 127 Case C-​413/​99 Baumbast and R v Secretary of State for the Home Department [2002] ecr I-​7091. 128 Case C-​138/​02 Brian Francis Collins v Secretary of State for Work and Pensions [2004] ecr I-​2703. 129 Case C-​168/​91 Christos Konstantinidis v Stadt Altensteig [1993] ecr I-​1191. 130 Case C-​341/​05 Laval un Partneri Ltd v Svenska Byggnadsarbetareförbundet, Svenska Byggnadsarbetareförbundets avdelning 1, Byggettan and Svenska Elektrikerförbundet [2007] ecr I-​11767. 131 As Advocate General Bot said in the Dirk Rüffert case (a follow-​up to Laval), ‘as a general rule, Community law does not preclude Member States from applying their legislation… to any person who is employed, even temporarily, within their territory, no matter in which country the employer is established’: Case C-​346/​06 Dirk Rüffert v Land Niedersachsen [2008] ecr I-​ 1989, Opinion of AG Bot, para 69. As authority for this proposition he uses Cases 62/​81, 63/​ 81 Société anonyme de droit français Seco and Société anonyme de droit français Desquenne & Giral v Etablissement d’assurance contre la vieillesse et l’invalidité [1982] ecr 223. 132 Case C-​340/​89 Irène Vlassopoulou v Ministerium für Justiz, Bundes-​und Europaangelegenheiten Baden-​Württemberg [1991] ecr I-​2357. 133 Deriving from the case of ‘Cassis de Dijon:’ Case 120/​78 Rewe-​Zentral AG v Bundesmono­ polverwaltung für Branntwein [1979] ecr 649.

338 Chapter 9 meaning is thus imposed on a Member State by one of its neighbours  –​‘If it’s good enough for the home Member State, it’s good enough for the host Member State’. 134 However, even in this area the tide is beginning to turn, with the Keck case135 signaling a partial return to local regulatory autonomy (or the so-​called ‘country of destination principle’). As Weiler has commented, the situation now is that a foreign seller, in return for not being excluded from the host State’s market, must respect the host State’s selling arrangements, in exactly the same way as a foreign company would have to respect the host State’s tax arrangements, or a foreign visitor would have to respect the host State’s criminal law.136 It is true that some commentators announced the obsolescence of Keck with the arrival at the cjeu of a series of ‘restrictions on use’ cases, including Commission v Italy (Trailers)137 and Mickelsson and Roos,138 134 This result can also be brought about legislatively, via the so-​called ‘country of origin principle.’ One of the principle’s first usages was in fact in the field of the free movement of services, with Directive 89/​552/​e ec, the old ‘Television without frontiers’ directive: Council Directive 89/552/EEC of 3 October 1989 on the coordination of certain provisions laid down by Law, Regulation or Administrative Action in Member States concerning the pursuit of television broadcasting activities [1989] OJ L298/23. Under this directive, the Member State into which a broadcast was being transmitted could not impose its (stricter) laws on a broadcaster located in a different Member State. However, the directive turned out to have many limitations, especially once the Court had decided to take a ‘differentiated approach’ and to give the directive a ‘flexible reading,’ and particularly vis-​à-​vis its scope rationae materiae: N Reich and others, Understanding eu internal market law (Intersentia 2015)  205–​7. Thus in Case C-​429/​02 Bacardi France [2004] ecr I-​6617, France was allowed to impose her indirect advertising law on incoming sports transmissions, thereby hampering Bacardi’s attempt to use trackside billboards in a motor race broadcast from Belgium. See Reich, Understanding eu internal market law, 207–​213 for examples of subsequent usages of the principle, most attenuated in one way or another. 135 Joined cases C-​267/​91 and C-​268/​91 Keck and Mithouard [1993] ecr I-​6097. 136 Joseph Weiler, ‘Towards a Principle of Economic Comity’ (Lecture given at Bentham House, ucl, 23 May 2001). A selling arrangement may ‘reflect deeply-​held policy choices about the character of national life:’ Daniel Wilsher, ‘Does Keck discrimination make any sense? An assessment of the non-​discrimination principle within the European single market’ (2008) 33(1) el Rev 3, 21. Indeed, the Court has acknowledged this, describing the consumption of alcoholic beverages in the Gourmet Foods case as being ‘linked to traditional social practices and to local habits and customs.’ See Case C-​405/​98 Konsumentombudsmannen (ko) v Gourmet International Products ab (gip) [2001] ecr I-​1795, para 21. 137 Case C-​110/​05 Commission v Italian Republic [2009] ecr I-​519 (Trailers). 138 Case C-​142/​05 Åklagaren v Percy Mickelsson and Joakim Roos [2009] ecr I-​4273. Spaventa, for example, referred to the case as having ‘passed its sell-​by date’ and of being ‘no longer relevant:’ E Spaventa, ‘Leaving Keck behind? The free movement of goods after the rulings in Commission v Italy and Mickelsson and Roos’ (2009) 24(6) EL Rev 914, 923 and 929.

Evaluation

339

and it is correct that these cases did put in issue again exactly how much regulatory autonomy was now able to be exercised at Member State level: slightly less than previously thought, it turned out. However, these cases did not overrule Keck.139 Indeed, the judgments appear to endorse the Court’s position in Keck.140 This taking back of sovereignty by Member States is also in keeping with the doctrine of subsidiarity, by which the Union should take action ‘only if and insofar as the objectives of the proposed action cannot be sufficiently achieved by the Member States.’141 And since the cjeu already has to operate a kind of ‘jurisdictional subsidiarity,’142 complex equality would not represent a particularly big leap forward. Even theorists have spotted the link between complex equality and subsidiarity; Den Hartogh refers to it as ‘a principle to which Walzer often appeals:’143

139 S Weatherill, ‘The road to ruin: ‘restrictions on use’ and the circular lifecycle of Article 34 tfeu’ [2012] 2 European Journal of Consumer Law 359, 368. 140 Oddly, cases like Case C-​110/​05 Trailers (n 137), where the Court goes ‘beyond discrimination,’ are not really about non-​discrimination on grounds of nationality, or equality generally, but more about, as Spaventa puts it, ‘freedom to trade:’ Spaventa (n 138) 928. Viewed at the moment that the Commission brought its action, it is hard to see how the putative Italian seller of motorcycle trailers, and the hypothetical seller from another Member State, are unequal; they would seem to be equally without a market, anyway. The Commission pursues such actions, it would seem, in defence of potential future market entrants, and has been doing so for years, as witness the occasion when it opened up the fois gras industry to all-​comers even though there was little desire to make this product anywhere outside France: Case C-​184/​96 Commission v France [1998] ecr I-​6197. However, at least in the latter case there was a threat of discrimination. In equality terms, Trailers feels more like, as Davies has pointed out in relation to another ‘restriction on use’ case, an attempt at ‘pure affirmative action:’ G Davies, ‘The Court’s jurisprudence on free movement of goods: pragmatic presumptions, not philosophical principles’ [2012] 2 European Journal of Consumer Law 217, 229. 141 Art 5 ec (now Art 5(3) teu). McCrudden has commented that, as long as there is a continued dialogue between the Member States and the Commission, ‘the strategy of devolving responsibility back to the Member States … has much to commend it’. See C McCrudden, ‘The new concept of equality’ (Paper prepared for the Academy of European Law conference, ‘Fight Against Discrimination:  The Race and Framework Employment Directives’, 2–​3 June 2003)  accessed 20 May 2018, 23. 142 A Iliopoulou, ‘Le principe d’égalité et de non-​discrimination’ in J-​B Auby and J Dutheil de La Rochère (eds), Droit administrative européen (Bruylant 2007) 449. 143 G den Hartogh, ‘The Architectonic of Michael Walzer’s Theory of Justice’ (1999) 27(4) Political Theory 491, 508.

340 Chapter 9 Principles of justice should be applied by collective decisions of the smallest circle of people who can apply them efficiently.144 The Keck judgment and the doctrine of subsidiarity may have what Bernard has called ‘a decentralising effect,’145 but as he rightly goes on to stress, centralization is not a sine qua non for integration, and ‘decentralisation is not synonymous with fragmentation’.146 That the Court and other Union actors have come to understand this is indeed ‘a sign of maturity.’147 It is submitted that the Court’s adoption of a complex egalitarian theory of equality would fit perfectly with this new, mature approach to eu law.148 Gareth Davies has called for a more sophisticated, less formalistic and above all adaptable approach to equality. Again, Walzer’s theory would seem to fit the bill. In particular, the way that the theory of complex equality permits ‘small inequalities’ within a given sphere –​perhaps Walzer’s most innovative proposal and the facet of the theory which marks its point of departure from simple equality149 –​would seem to meet Davies’ requirement that the notion of equality be separated from that of uniformity (‘the principle of one rule for all’):150 ‘The tendency to treat [equality] as synonymous with uniformity is logically false, and can often be a route to intolerance’.151 144 145 146 147 148

ibid. Nicolas Bernard, ‘Discrimination and Free Movement in ec Law’ (1996) 45 iclq 82, 108. ibid. ibid. In a recent book, though, Muniz-​Fraticelli notes that the principle of subsidiarity is not neutral on the issue of whether the group or the state has the ultimate claim to authority, but it ‘assumes a unity of purpose’ thus becoming in effect a ‘tool for co-​opting diverse authorities to a single purpose:’ VM Muniz-​Fraticelli, The structure of pluralism (oup 2014)  79, 80. Ultimately it ‘conscript[s]‌associations and render[s] them mere organs of the state:’ ibid 56. In reaching this conclusion he draws on support from Davies: G Davies, ‘Subsidiarity: The wrong idea, in the wrong place, at the wrong time’ (2006) 43(1) cml Rev 63. 149 Also the facet of the theory which has most confused Walzer’s critics, for example, Richard Arneson, who lambasts complex equality as ‘a very weak brew, in which any element of anything that could plausibly be identified with egalitarianism is so diluted as to be virtually undetectable:’ RJ Arneson, ‘Against “Complex” Equality’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) 226. 1 50 G Davies, Nationality Discrimination in the European Internal Market (European Monographs, Kluwer 2003) 201. 151 G Davies, ‘Higher education, equal access, and residence conditions:  Does eu law allow Member States to charge higher fees to students not previously resident?’ 12 mj 3 (2005) 240.

Evaluation

341

The distribution of student maintenance grants, for example, takes place within its own, autonomous sphere and in accordance with the relevant distributive community’s shared understanding of such grants. But that does not mean that every single student qualifies, at exactly the same moment, for exactly the same grant (as Davies puts it, ‘[t]‌he prohibition on nationality discrimination does not demand equal fees’).152 To go back to the education example in Chapter 2,153 that would be the equivalent of every student in the class receiving the same grade. Under a Walzerian regime, variety is allowed; in Davies’ terms, Walzer can ‘cope with difference.’154 While Aristotle presents the Court with a black-​and-​white choice –​insiders versus outsiders, for e­ xample –​ Walzer’s more detailed analysis enables the Bench to place the choice in its wider context and to ‘recogni[ze] … what has occurred outside the jurisdiction.’155 This allows for a more nuanced decision –​a dimmer switch rather than one that can only go on or off. 9.9

Complex Equality for European Courts?

Before proceeding to a final verification or falsification of this book’s thesis, that Michael Walzer’s theory of complex equality could be used by the Court of Justice when dealing with cases concerning equality, as a complement to the Aristotelian like-​for-​like formula, it would be worth considering whether any of the national courts in eu Member States, which after all are the cjeu’s mainstay of case-​law, have themselves ventured outside the confines of Aristotle’s formula when dealing with equality?156 Looking around the eu, for the most part, one searches in vain for domestic courts which have adopted any approach to equality cases which could be characterised as outside the traditional ‘Aristotelian’ approach. One reason for this is that the domestic law which these courts are interpreting is itself based on eu Law, which as has been seen earlier concretises the Aristotelian test in a number of places.157 Thus the national judge must run a comparison in certain circumstances 152 153 154 155 156

ibid. See supra, section 2.2. Davies (n 150) 202. ibid 201. Not including well-​established exceptions to/​departures from this approach to equality, or instances of substantive equality, such as non-​comparative equality in cases of pregnancy, positive action in the gender context or accommodation in the disability context. 157 For example, in Article 2(2) of Directive 2000/​78.

342 Chapter 9 simply to comply with his or her own country’s legal system. However, this would not stop a national court operating an alternative system alongside the Aristotelian one. Coming close to this, it seems, is Denmark. Particularly instructive are two cases from the most recent country report produced by the European Commission’s European Network of Legal Experts in the Non-​Discrimination Field: A Western High Court judgment dealt with a young Muslim woman who had been studying for nutrition assistant at a vocational school. The case was an appeal of a city court ruling from 2013. The Muslim woman had to quit her education because of the fact that the school would not exempt her from the requirement to taste pork. In her case, the court concluded that it would be against her religion to taste dishes of pork. According to the court the vocational school could not document that it was necessary for the complainant to taste pork for her to complete the education as nutrition assistant. The judgment does not illustrate a duty to provide reasonable accommodation in a traditional sense. However, the reasoning of the court in obligating the school to make exceptions for a student like this Muslim woman is similar to the general argumentation of reasonable accommodation. Another decision from 2015 by the Board of Equal Treatment illustrates a similar arguing [sic]. In this decision the Board concluded that it was a violation of the law to require that a job applicant shake hands. In the case the employer had argued that one of the reasons for the complainant not to get the job was the fact that he did not want to shake hands with female customers. The Board held that the employer had not established a legitimate purpose of the requirement to shake hands. Thus the Board concluded that the complainant had been indirectly discriminated against based on his religion.158 In analyzing these two decisions, it is helpful to start, as in fact the report does in the paragraph immediately preceding the two reproduced, with the rule, 158

Pia Justesen, Country report Non-​discrimination Denmark (Publications Office of the European Union 2016) 51. The three footnotes from the original text give the two case references for the first case described (first instance and appeal), and the case reference for the second case described, and are reproduced here: Holstebro city court judgment in case No. BS 7–​189/​2012 of 23 April 2013; Western High Court judgment in case No. B-​1213-​13 of 5 May 2015. Printed in U2015.2984V; Board of Equal Treatment, Decision No. 149/​2015 of 23 September 2015.

Evaluation

343

which exists in many countries, that male Sikhs are exempted from wearing a crash helmet when riding a motorbike since they are obliged to wear the turban outside at all times. Were this rule not to exist, and looking at the situation as one of indirect discrimination rather than reasonable accommodation as the report seems to, one would simply note that the State, by making the wearing of crash helmets mandatory, was imposing a uniform requirement on two different types of motorcyclist, but that one type had significantly more difficulty fulfilling it than the other. The focus of the inquiry is, rightly, on the plaintiff and their hypothetical or indeed real comparator. Noone questions the requirement. That is how the Aristotelian inquiry always proceeds, and so for example in the Prais case which was examined in Chapter 4,159 neither the Court nor the Advocate General questioned the need for the Council to hold the concours on the particular day which coincided with Pentecost. However, in the Western High Court judgment and the Board of Equal Treatment judgment described above, a subtle shifting of focus is taking place. With regard to the Western High Court judgment, the rule that students must taste pork in order to qualify as nutrition assistants is not only questioned, but the employer is asked to produce supporting evidence. Thus, although ultimately it is used alongside the traditional Aristotelian approach of comparing the Muslim student with the non-​Muslim ones, a Walzerian approach is visible here too, the Danish court clearly wishing to establish if the rule about tasting pork was needed at all, for anyone: What was the meaning of a qualification as a nutrition assistant in Denmark? Was the experience of having tasted pork part of it, or not? The Board of Equal Treatment judgment goes further still. Here, the court sees its only task as establishing the need for the employer’s rule in this case, namely, that employees must shake hands with customers. The complainant’s religion is barely mentioned, which means that they are never really compared with other employees, similarly situated but for their religion, at all. Rather the court contents itself with the question: What is the meaning of this particular job in Denmark? Is shaking hands with customers part of it, or not? The negative answer to the second question, in both judgments, means that the uniform requirement imposed on students or applicants is in violation of complex equality, rendering those able to fulfill it dominant over those not. This has a knock-​on effect for the outcome, which shifts from being the making of an exception (and in Prais not even that), to the removal of the unnecessary requirement. This is not only a benefit for everyone, but avoids

159

Case 130/​75 Vivien Prais v Council of the European Communities [1976] ecr 1589.

344 Chapter 9 the subtle perpetuation of, if not the discrimination, then at least the distinction upon which it is based, or, more precisely, the attention paid to it. It is no wonder many do not complain of discriminatory treatment at all, but see it merely as a short term hurdle to ‘get over,’160 where to complain, even successfully, would isolate them still further from the group of which they simply wish to be part. On the other hand, making the employer/​distributor remove the requirement washes away this discriminatory aftertaste, and, what is more, forces them in future to think about the ramifications of their rules for those –​all of those –​to be affected by them. Is this the judiciary taking agency away from employers and distributors? It could be argued that yes, it is, but if it creates a deterrent effect, prompting the latter to think ex ante about the implications of requirements, and saving those in a minority from having to ‘transact’ their difference in a way they would rather not, then that must be seen as a good thing. 9.10

Parting Words: To Thine Own Self Be True

In Chapter 2, it was discussed how human rights law, whether in the guise of a ‘Minimal Morality’ or an Override, or some other guise, had the capacity to fix distributive meanings. As this book draws to a close it should not be thought that the message has been that this, in and of itself, is a bad thing. It is just that too much fixing, of the sort lawyers and litigants are faced with now, stultification of meanings, is bad for justice as it involves judges first creating, and then perpetuating, patterns of dominance: Law-​as-​Hegemon.161 Although not a radical counter hegemonic project, Walzer’s project skewers the patterns of stultification which do so much to reify, and reiterate, injustice and inequality.162 His Forum, if actualized correctly, enables citizens 160 See the extremely revealing research presented in a special issue of Transfer, entitled ‘Challenging discrimination at work against ethnic minority and migrant workers,’ referred to supra in Chapter 4. 161 This phrase coined by Professor Mann of Harvard Law School: BH Mann, ‘Afterword’ in CL Tomlins and BH Mann (eds), The Many Legalities of Early America (University of North Carolina Press 2001) 447. 162 He has of course been accused of being part of the problem, not part of the solution, but, in defending himself, he asserts accurately that the ‘ “hegemonic” cultural and political discourse’ is his starting point, not his endpoint: ‘my purpose in everything I have written over many years is to challenge [historical dominance], and every other (…) form of dominance’. See M Walzer, ‘Shared meanings in a poly-​ethnic democratic setting: A response’ (1994) 22(2) Journal of Religious Ethics 401, 402.

Evaluation

345

to equalize themselves, even if the gift of choosing the meaning of spherical membership needs to be accompanied by an enforced ‘Returns Policy,’ the judge retaining a reservoir of competence from case to case,163 or in every case,164 to rule given meanings as ‘beyond the pale.’165 However, this ‘last resort’ role of the judge remains problematic in what ought to be, as Sellars puts it, a ‘self-​correcting exercise.’166 If there is no ‘true speech,’ no ‘verdict,’ how can one meaning be judged beyond the pale and another not? To be sure, a purely Rortian approach would do away with the pale altogether and the tolerance of one’s fellows would indeed be the only limit to what might conceivably be true. But it is submitted that society should retain a pale, placing it somewhere, even somewhere far away, because tolerance can be enforced through fear. At the end of Stoppard’s play Every Good Boy Deserves Favour, a dying Soviet dissident, ‘admitted’ to a mental hospital for failing to speak the ‘truth’ about the Soviet Union, tells his son, as he departs from the hospital for the last time, To thine own self be true, one and one is always two.167 It would be worth society picking a few Master Truths to be going on with, for coping, but regrettably also for copying –​stents, albeit artificial, to tie the ever-​shifting, multiperspectival morass of modern life to the hidden subvenient core of truth.168 Such Master Truths would in effect be Fixed Meanings, or parts of meanings  –​‘pre-​interpret[ations],’ Walzer might call them169  –​ that part of the meaning of human life is its inviolability, for example, or that part of the meaning of torture is its impermissibility. Permanently, and fairly, contested truths –​of the sort envisaged by Walzer and endorsed by this book –​may be the dream. But meanings dictated by those with the most guns, with those disagreeing locked up or worse, is –​unquestionably –​a nightmare. 163 The Override. 164 Mediated complexity. 165 The ‘pale’ in old English is a pointed piece of wood used in fencing –​it survives in the verb ‘to impale’ –​and in time came to mean the safe area enclosed by the fence made of pales. 166 Sellars (31) 78–​9. 167 T Stoppard, Every Good Boy Deserves Favour and Professional Foul (Faber & Faber 1978) 36. 168 For the origins of Rorty’s coping/copying distinction, see n 84 above. 169 M Walzer, ‘Feminism and me’ (2013) 60(1) Dissent 50, 52.

346 Chapter 9 It is submitted that a slight loss of complex egalitarian purity would be a small price to pay to avoid it. If it is accepted, with Deleuze and Guattari, that hapless man ‘no longer believe[s]‌(…) in a final totality that awaits [him or her],’170 and this book has certainly played its small part in debunking that final totality, it does not mean that he or she has stopped believing in anything. 170 G Deleuze and F Guattari, Oedipus complex: capitalism and schizophrenia (tr R Hurley, Viking 1977) 42.

chapter 10

Conclusion The Walzerian approach to equality exposes the Aristotelian test for all of its limitedness and shortsightedness, and admits a world of possibilities beyond the playing-​board. Likes within the same sphere may be treated unlike unjustly (boundary breach) or justly (‘small inequality’). Unlikes within the same sphere may be treated unlike justly (‘small inequality’) or unjustly (boundary breach).1 Likes within the same sphere may be treated like justly (correct placement of boundaries, correct distributive criterion) or unjustly (boundary breach, or wrongful placement of boundaries in the first place). Unlikes within the same sphere may be treated like unjustly (boundary breach) or justly (‘small inequality’). A similar list could be drawn up for likes or unlikes who or which find themselves in separate spheres,2 a boundary breach or wrongful (initial) placement of boundaries usually being at the root of any and all injustice. Like treatment could now be merely a coincidence, or it could perhaps indicate a boundary breach by the distributor. Even within this paragraph, it has been difficult to ‘shoe-​horn’ the myriad possibilities into the strictures, physical, temporal and perspectival, of the terms ‘like’ and ‘unlike,’ but it is only intended to provide a rough indication of the poverty of the Aristotelian test when faced with the richness of human experience. It is submitted, then, that the thesis is verified and that Michael Walzer’s theory of complex equality may be used by the Court of Justice of the European Union when dealing with cases concerning equality, as a complement (at the very least) to the Aristotelian ‘test’ that likes should be treated in like fashion and unlikes in unlike fashion. Like a cataracts operation for the cjeu judge,

1 A serf and a King, for example, receive unequal treatment justly under normal circumstances, but unjustly if the King has no business wearing his crown in that sphere. Even then, that does not mean that he is to be stripped of his crown permanently, just that he must only wear it in the sphere in which he was coronated. 2 While the latter –​unlikes in separate spheres –​might seem to be the more obvious scenario, an example of the former –​likes in separate spheres –​would be where two farmers were receiving quotas for different crops, or where two manufacturers were receiving finance for different products. This situation came up, in one of the pleas, in Joined cases 424/​85 and 425/​85 Coöperatieve Melkproducentenbedrijven Noord-​Nederland Ba (‘Frico’) and others v Voedselvoorzienings In-​en Verkoopbureau [1987] ecr 2755, discussed at supra section 6.4.

© Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/​9 789004354265_011

348 Chapter 10 Walzer’s ‘radically particularist’3 theory finally restores his or her full sight and enables them for the first time to see the special cases, the exceptions, the substance of the dispute;4 the ‘dilemma of difference,’ referred to in the Introduction,5 is resolved.6 Attempting to apply overly reductive and unitary rules and tests, from the top of the mountain (to use one of Walzer’s favourite images), to involved and complex situations on the ground (or ‘in the cave’), produced at best hit-​and-​ miss results –​the misses often being catastrophically bad, and the hits more the result of good luck than design. It was like trying to wind an intricate watch while wearing an oven glove. Particularism allows the judge to contextualize each comparison. With its emphasis on proceedings in the Forum, complex equality effectively allows citizens to equalize themselves.7 Critics of the theory, though, see danger lurking in the Forum. Some have tried to fix the problem by combining complex equality and its context-​ dependency with something else, some universalist element that will (in their view) provide the missing safeguard; in this regard, Habermas’ ideal speech 3 M Walzer, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) xiv. 4 For a real-​life Walzerian court case, albeit one abandoned at the last minute by the pharmaceutical companies, see the account of the former president of Doctors Without Borders, James Orbinski, of his attempts to change the distributive criterion of aids drugs in South Africa from wealth to need:  J Orbinski, ‘Keynote address:  Justice and global health’ in M Freeman, S Hawkes and B Bennett (eds), Law and Global Health (Current legal issues volume 16, oup 2014). Interestingly, he describes the case as a ‘communicative act that challenged the … law as it was being interpreted’ and states that law is at its ‘most powerful when it is a function of a communicative process over time’ and ‘an ongoing redefinition or reshaping of the community’s constitutive principles:’ ibid 11, 15–​20. 5 See supra section 1.2. 6 As Armstrong puts it, Walzer offers ‘a highly positive intervention into the apparent impasse between equality and difference’ and indeed a ‘synthesis’ of the two ideas: C Armstrong, ‘Complex equality:  Beyond equality and difference’ (2002) 3(1) Feminist Theory 67, 68 and 80. 7 Jürgen Habermas was perhaps thinking along the same lines when he wrote, ‘For in the final analysis, private legal persons cannot even attain the enjoyment of equal individual liberties unless they themselves, by jointly exercising their autonomy as citizens, arrive at a clear understanding about what interests and criteria are justified and in what respects equal things will be treated equally and unequal things unequally in any particular case’. See J Habermas, ‘Struggles for Recognition in the Democratic Constitutional State’ in C Taylor, Multiculturalism: Examining the politics of recognition (Princeton up 1994) 113. In a similar vein, Hannah Arendt wrote, ‘We are not born equal; we become equal as members of a group on the strength of our decision to guarantee ourselves mutually equal rights’:  H Arendt, Imperialism: Part Two of The Origins of Totalitarianism (Harcourt 1968) 181.

Conclusion

349

theory is especially popular.8 O’Neill describes this hybridity particularly well, referring to his own approach as ‘contextual impartialism.’9 In this book, this kind of synthesizing of complex equality with distinct theories has not been thought necessary. Walzer’s partiality –​should it be considered dangerous –​is sufficiently offset, it is submitted, firstly, by the Override, and secondly, by the doctrine of co-​existential equilibrium, both of these traceable to or rooted in his own writing. It is important, though, that complex equality is not made to stand or fall on the question of who should ascertain the meanings of distributed goods. The doctrine still has much to offer whoever does the ascertaining. If use of the Walzerian Forum is thought to pose a threat, then this aspect of the theory may be shed while still preserving the kernel of complex equality  –​analyzing the allegedly discriminatory act, not in terms of comparisons of the actors involved, but rather in terms of distributions and meanings of goods. This is the true heart of the theory –​its ability to separate the people from the problem –​ and it is not lost if the task of ascertaining the meanings is entrusted to the Bench. The alternative theory of mediated complexity –​admittedly representing a more significant departure from orthodox Walzerian doctrine –​would be another way to rein in complex equality, if such reining in were needed. Aristotelian ‘like for like’ equality constrains those who apply it, and ultimately those in relation to whom it is applied. On the other hand, as Walzer has commented, ‘complexity is free:’ ‘the more complex the [social] construction [of goods] the more room there is for cultural difference’.10 As the European Union grows wider, as the activities of its citizens –​both individual and collective –​grow more complicated and their relations one with another more sophisticated, this is a freedom which they already deserve, and which they will increasingly demand. The Court of Justice should be ready. 8

9 10

S O’Neill, Impartiality in Context:  Grounding Justice in a Pluralist World (suny Press 1997); L Atkinson, ‘Coming to Terms with Procedure: The Potential of the “Ideal Speech Situation” for Michael Walzer’s Communitarian Justice’ (1998) 56 U Toronto Fac L Rev 223; ST Johansson, ‘Towards spherical justice: a critical theoretical defence of the idea of complex equality’ (PhD thesis, University of Southampton 2003, unpublished). O’Neill (n 8) 201. M Walzer, ‘Objectivity and Social Meaning’ in M Nussbaum and A Sen (eds) The Quality of Life (oup 1993) 171.

Appendix I Comparator 1 and Comparator 2

Treated

Outcome

Like

Like

Not discrimination

Like

Unlike

Discrimination

Unlike

Like

Discrimination

Unlike

Unlike

Not discrimination

©  Koninklijke Brill NV, Leiden, 2018 | DOI 10.1163/9789004354265_​0 12

Bibliography —​—​, ‘Case Comment. Nepal: Supreme Court –​equality rights of ‘third sex’ (lesbian, gay, bisexual, transsexual and inter-​sex)’ [2008] pl 606 —​—​, ‘Curators’ Foreword’ in G Adamson and J Pavitt (eds), Postmodernism: Style and subversion, 1970–​1990 (V&A Publishing 2011) 9 —​—​, ‘Home thoughts’ The Economist (London, 25 September 2010) 50 —​—,​ ‘Relocation:  Commission refers the Czech Republic, Hungary and Poland to the Court of Justice’ (European Commission Press Release, 7 December 2017) accessed 7 May 2018 —​—​, ‘Summary of Workshops: Themes and Issues Raised’ in J Cormack (ed), Proving discrimination –​the dynamic implementation of eu anti-​discrimination law: the role of specialised bodies. Report of the first experts’ meeting, 14–​15 January 2003 hosted by the Belgian Centre for Equal Opportunities and Opposition to Racism (European Commission 2003) Adams KD, ‘Can Promise Enforcement Save Affordable Housing in the United States?’ (2004) 41 San Diego L Rev 643 Adamson G and Pavitt J, ‘Postmodernism:  Style and subversion’ in G Adamson and J Pavitt (eds), Postmodernism:  Style and subversion, 1970–​ 1990 (V&A Publishing 2011) Alemanno A, ‘Science & eu Risk Regulation: The Role of Experts in Decision-​Making and Judicial Review’ in E Vos (ed) European Risk Goveranance  –​Its Science, Its Inclusiveness and Its Effectiveness (Connex Report Series No 6, 2008) accessed 21 May 2018 Allen D, ‘Equality and American democracy: why politics trumps economics’ (2016) 95(1) Foreign affairs 23 Allen R, ‘Article 13 ec, evolution and current contexts’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Anderson JA, and Weijer C, ‘The Research Subject as Wage Earner’ (2002) 23 Theoretical Medicine 359 Andre J, ‘Blocked Exchanges: A Taxonomy’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Andrews J, ‘National and International Sources of Women’s Right to Equal Employment Opportunities:  Equality in Law Versus Equality in Fact’ (1994) 14 Northwestern Journal of International Law and Business 413 Arendt H, Imperialism: Part Two of The Origins of Totalitarianism (Harcourt 1968) Aristotle, The Nicomachean Ethics (T Irwin tr, 2nd edn, Hackett 1999) Aristotle, The Nicomachean Ethics (WD Ross tr, oup 1980)

Bibliography

353

Armstrong C, ‘Philosophical Interpretation in the Work of Michael Walzer’ (2000) 20(2) Politics 87 Armstrong C, ‘Complex equality: Beyond equality and difference’ (2002) 3(1) Feminist Theory 67 Arneson RJ, ‘Equality and Equal Opportunity For Welfare’ (1989) 56 Philosophical Studies 77 Arneson RJ, ‘Against “Complex” Equality’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Arnull A, ‘The European Court and judicial objectivity: a reply to Professor Hartley’ (1996) 112 lqr 411 Arnull A, ‘Annotation Arsenal Football Club plc v Matthew Reed’ (2003) 40 CML Rev 753 Arnull A, ‘Editorial: Out with the old ...’ (2006) 31 EL Rev 1 Atkinson L, ‘Coming to Terms with Procedure:  The Potential of the “Ideal Speech Situation” for Michael Walzer’s Communitarian Justice’ (1998) 56 U Toronto Fac L Rev 223 Atrey S, ‘Redefining frontiers of eu discrimination law’ [2017] pl 185 Bailey SH and others, Smith, Bailey and Gunn on the modern English legal system (Sweet & Maxwell 2002) Balkin JM, ‘Plessy, Brown, and Grutter: A play in three acts’ (2004–​2005) 26 Cardozo L Rev 1689 Ball CA, ‘Communitarianism and Gay Rights’ (1999–​2000) 85 Cornell L Rev 443 Bammens N, The principle of non-​discrimination in international and European tax law (ibfd 2012) Banting K and Kymlicka W, ‘Multiculturalism and Welfare’ (2003) 50 Dissent 59 Barak A, The judge in a democracy (Princeton University Press 2006) Barbera M, ‘Not the same? The Judicial Role in the New Community Anti-​Discrimination Law Context’ (2002) 31 Indus LJ 82 Barnard C, ‘The Economic Objectives of Article 119’ in TK Hervey and D O’Keeffe (eds), Sex equality law in the European Union (Wiley 1996) Barnard C, ‘The Principle of Equality in the Community Context: P, Grant, Kalanke and Marschall: Four Uneasy Bedfellows?’ (1998) 57 clj 352 Barnard C, ‘Some are more equal than others: the decision of the Court of Justice in Grant v South-​West Trains’ (1999) 1 cyels 147 Barron A, ‘Foucault and Law’ in J Penner, D Schiff and R Nobles (eds), Introduction to Legal Theory and Jurisprudence (Butterworths 2002) Barry B, ‘Intimations of Justice’ (1984) 84 Columbia L Rev 806 Barry B, Liberty and Justice –​Essays in Political Theory 2 (Clarendon 1991) Barry B, ‘Spherical Justice and Global Injustice’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995)

354 Bibliography Barry E, ‘Different Hierarchies  –​Enforcing Equality Law’ in C Costello and E Barry (eds), Equality in Diversity –​The New Equality Directives (Ashfield 2003) Bauman Z, Liquid Modernity (Polity 2000) Bell M, ‘Shifting Conceptions of Sexual Discrimination at the Court of Justice: from P v S to Grant v swt’ (1999) 5(1) Eur LJ 63 Bell M, Anti-​Discrimination Law and the European Union (Oxford Studies in European Law, oup 2002) Bell M, ‘eu anti-​racism policy: the leader of the pack?’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Bell M, ‘Publication review  –​Sexual Orientation Discrimination in the European Union: National Laws and the Employment Equality Directive’ (2007) 4 ehrlr 481 Bell M and ILGA-​Europe, Equality for lesbians and gay men: a relevant issue in the civil and social dialogue (ILGA-​Europe, Brussels 1998) Bell M and Waddington L, ‘Reflecting on inequalities in European equality law’ (2003) 28 EL Rev 349 Bellamy R, ‘Justice in the community: Walzer on pluralism, equality and democracy’ in D Boucher and PJ Kelly (eds), Social Justice: From Hume to Walzer (Routledge 1998) Benhabib S, The rights of others (cup 2004) Berlin I, ‘Equality as an Ideal’ in FA Olafson, Justice and Social Policy: A collection of essays (Prentice-​Hall 1961) Bernard N, ‘Discrimination and Free Movement in ec Law’ [1996] 45 iclq 82 Bernard N, ‘What are the purposes of ec discrimination law?’ in J Dine and B Watt (eds), Discrimination Law:  Concepts, Limitations and Justifications (Longman 1996) Besselink LFM, ‘National and constitutional identity before and after Lisbon’ (2010) 6(3) Utrecht L Rev 36 Besselink LFM, ‘Annotation Sayn-​Wittgenstein’(2012) 49 CML Rev 671 Besson S and Martí JL (eds), Deliberative democracy and its discontents (Ashgate 2006) Bhabha HK, The Location of Culture (Routledge 1993) Bickenbach JE, ‘Disability and Equality’ (2003) 2 JL & Equal 7 Blakely R, ‘Race activist continues to insist white is black’ The Times (London, 17 June 2015) Boele-​ Woelki K, ‘The Legal Recognition of Same-​ Sex Relationships within the European Union’ (2007–​8) 82 Tul L Rev 1949 Boltanski L and Browne C, ‘ “ Whatever works:”  Political philosophy and sociology  –​ Luc Boltanski in conversation with Craig Browne’ in S Susen and BS Turner (eds), The spirit of Luc Boltanski: Essays on the ‘Pragmatic sociology of critique’ (Anthem Press 2014) Bosniak L, The citizen and the alien: dilemmas of contemporary membership (Princeton up 2006)

Bibliography

355

Boujeka A, ‘La gestation pour autrui et le handicap confrontés au principe de non-​ discrimination en droit de l’Union’ (2014) 31(7612) Le Dalloz 1811 Brecher B, ‘Universities and the Neoliberal Agenda’ (Lecture given at the Old Courtroom Lecture Theatre, Brighton, 10 November 2015) Bribosia E and Bombois T, ‘Interdiction de la discrimination en raison de l’âge: du principe, de ses exceptions at de quelque hésitations –​Réflexions sur des arrêts Wolf, Petersen et Kücükdeveci de la Cour de justice de l’Union européenne’ (2011) 47(1) Revue trimestrielle du droit européen 41 Brillat M, Le principe de non-​discrimination à l’épreuve des rapports entre les droits européens (Fondation Varenne 2015) Brown C, ‘The race directive: towards equality for “all” the peoples of Europe?’ (2001–​ 2002) 21 YB of Eur L 195 Brown W, Undoingthe Demos: Neoliberalism’s Stealth Revolution (Zone Books 2015) Burge T, ‘Individualism and the Mental’ (1979) 4 Midwest Studies in Philosophy 73 Cain PA, ‘Feminism and the limits of equality’ (1989–​1990) 24 Ga L Rev 803 Calderón J and Baez A, ‘The Columbus Container Services ecj case and its consequences: a lost opportunity to shed light on the scope of the non-​discrimination principle’ (2009) 37(4) Intertax 212 Callinicos AT, Equality (Polity Press 2000) Canetti E, Crowds and Power (C Stewart tr, Victor Gollancz 1962) Canor I, ‘Equality for Lesbians and Gay Men in the European Community Legal Order –​ “they shall be male and female”?’ (2000) 7 mj 273 Caracciolo di Torella E, ‘Gender equality after Test Achats’ (2012) 13 ERA Forum 59 Caracciolo di Torella E, ‘No sex please, we’re insurers’ (2013) 38 EL Rev 638 Caracciolo di Torella E and Masselot A, ‘Pregnancy, maternity and the organisation of family life: an attempt to classify the case law of the Court of Justice’ (2001) 26(3) EL Rev 239 Cardozo BN, ‘Lecture II: The methods of history, tradition and sociology’ in C Morris (ed), The great legal philosophers:  Selected readings in jurisprudence (12th edn, University of Pennsylvania Press 1997) Cardozo BN, The nature of the judicial process (Dover 2005) Carens J, ‘Complex Justice, Cultural Difference, and Political Community’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Carens J, ‘Citizenship and Civil Society: What Rights for Residents?’ in R Hansen and P Weil (eds), Dual Nationality, Social Rights and Federal Citizenship in the us and Europe: The Reinvention of Citizenship (Berghahn Books 2002) Carey N, ‘From obloquy to equality: in the shadow of abnormal situations’ (2001) 20 YB of Eur L 79 Chettiparamb A, ‘Articulating “public interest” through complexity theory’ (2016) 34(7) Environment and Planning C: Government and Policy 1284

356 Bibliography Chong Ho Shon P, ‘ “Now You Got a Dead Baby on Your Hands”: Discursive Tyranny in “Cop Talk”’ (1998) XI(33) International Journal for the Semiotics of Law 275 Christiansen T, Informal Governance in the European Union (Elgar 2004) Christie N, ‘Conflicts as property’ (1977) 17(1) British Journal of Criminology 1 Cillo R and Perocco F, ‘Outsourced racism in Italy: discrimination at work and trade union responses in three sectors’ (2015) 21(1) Transfer: European Review of Labour and Research 51 Clark S, ‘Society against Societies: The possibility of transcultural criticism’ (2007) 13 Res Publica 107 Clayton J, ‘Running without legs? That’s not fair on all the other athletes’ The Times (London, 15 January 2008) 30 Cohen J, Review of Spheres of Justice (1986) 83 The Journal of Philosophy 457 Commission, ‘Proposal for a Council Directive on implementing the principle of equal treatment between persons irrespective of religion or belief, disability, age or sexual orientation’ com (2008) 426 final Commission, ‘Proposal for a Directive of the European Parliament and of the Council amending Council Directive 92/​85/​eec on the introduction of measures to encourage improvements in the safety and health at work of pregnant workers  and workers who have recently given birth or are breastfeeding’ com (2008)637 Commission, ‘Proposal for a Directive of the European Parliament and of the Council on work-​life balance for parents and carers and repealing Council Directive 2010/​ 18/​e u’ com (2017)253 final Commission, ‘A short history of milk quotas’ (Europa, 22 February 2017) accessed 20 May 2018 Connolly M, ‘The coalition government and age discrimination’ [2012] 2 Journal of Business Law 144 Connolly M, ‘The ecj signals a light touch towards age discrimination and compulsory retirement’ (2007) 81 Emp LB 1 Connolly M, ‘The Heyday case: much ado about very little’ (2009) 90 Emp LB 2 Connolly W, The Ethos of Pluralization (University of Minnesota Press 1995) Cordewener A, Kofler G and van Thiel S, ‘The clash between European freedoms and national direct tax law:  public interest defences available to the Member States’ (2009) 46 CML Rev 1951 Costello C, ‘Metock: Free Movement and “Normal Family Life” in the Union’ (2009) 46 CML Rev 587 Costello C and Davies G, ‘The case law of the Court of Justice in the field of sex equality since 2000’ (2006) 43 CML Rev 1567 Council of the European Union, ‘Q&A on the reform of the General Court’ (Press release en 497/​15, 23 June 2015) 1

Bibliography

357

Croon J, ‘Comparative institutional analysis, the European Court of Justice and the general principle of non-​discrimination –​or –​alternative tales on equality reasoning’ (2013) 19(2) European Law Journal 153 Danchin PG, ‘Suspect Symbols: Value Pluralism as a Theory of Religious Freedom in International Law’ (2008) 33(1) Yale J Intl L 1 Davies C, Wetherell M and Barnett E, Citizens at the Centre: Deliberative participation in healthcare decisions (Polity Press 2006) Davies G, Nationality Discrimination in the European Internal Market (Kluwer 2003) Davies G, ‘The Division of Powers between the European Court of Justice and the National Courts’ (Webpapers on Constitutionalism and Governance beyond the State No 3, conWEB 2004) Davies G, ‘ “Any Place I Hang My Hat?” or: Residence is the New Nationality’ (2005) 11(1) Eur LJ 43 Davies G, ‘Higher education, equal access, and residence conditions:  Does eu law allow Member States to charge higher fees to students not previously resident?’ (2005) 12(3) mj 227 Davies G, ‘Subsidiarity: The wrong idea, in the wrong place, at the wrong time’ (2006) 43 CML Rev 63 Davies G, ‘The Court’s jurisprudence on free movement of goods:  pragmatic presumptions, not philosophical principles’ [2012] 2 European Journal of Consumer Law 217 Davies G, ‘Achbita v G4S: Religious Equality Squeezed between Profit and Prejudice’ (European Law Blog, 6 April 2017) accessed 14 May 2018 de Búrca G, ‘The Role of Equality in European Community Law’ in A Dashwood and S O’Leary (eds), The Principle of Equal Treatment in ec Law (Sweet & Maxwell 1997) de Dominicis N, ‘La tension entre l’égalité de traitement et le droit à la vie privée: la protection des données à caractère personnel et l’emploi des personnes souffrant d’un handicap’ [2014] 2 Revue du droit de l’Union européenne 299 de Schutter O, ‘Methods of proof in the context of combating discrimination’ in J  Cormack (ed), Proving discrimination –​the dynamic implementation of eu anti-​ discrimination law:  the role of specialised bodies. Report of the first experts’ meeting, 14–​15 January 2003 hosted by the Belgian Centre for Equal Opportunities and Opposition to Racism (European Commission 2003) de Tocqueville A, Democracy in America (GE Bevan tr, Penguin Classics 2003) Dean J, Crowds and Party (Verso 2016) Deleuze G and Guattari F, Oedipus complex: capitalism and schizophrenia (tr R Hurley, Viking 1977)

358 Bibliography den Hartogh G, ‘The Architectonic of Michael Walzer’s Theory of Justice’ (1999) 27(4) Political Theory 491 Derrida J, ‘Force of Law: the “Mystical Foundations of Authority” ’ (1990) 11 Cardozo Law School 919 Derrida J, The Politics of Friendship (G Collins tr, Verso 1997) Devenney M, Ethics and Politics in Contemporary Theory:  Between Critical Theory and Post-​Marxism (Routledge Innovations in Political Theory, 1st edn, Routledge 2014) Dewey J, Reconstruction in Philosophy (Boston 1957)  Dine J and Watt B, ‘Introduction’ in Janet Dine and Bob Watt (eds), Discrimination Law: Concepts, Limitations and Justifications (Longman 1996) Docksey C, ‘The European Community and the promotion of equality’ in C  McCrudden (ed), Women, Employment and European Equality Law (Eclipse 1987) Domurath I, ‘The three dimensions of solidarity in the eu legal order:  limits of the judicial and legal approach’ (2013) 35(4) Journal of European integration 459 Dougan M and Spaventa E, ‘ “ Wish You Weren’t Here …”. New Models of Social Solidarity in the European Union’ in E Spaventa and M Dougan (eds), Social Welfare and Eu Law (Hart 2005) Drysek JS, Deliberative Democracy and Beyond: Liberals, Critics, Contestations (oup 2001) Duchêne F, JeanMonnet: The First Statesman of Interdependence (Norton 1994) Durant A and Izarra L, ‘Reading mixed reception: The case of The Satanic Verses’ (2001) 24 Revista de filología y su didáctica 653 Dworkin R, Taking Rights Seriously (Duckworth 1977) Dworkin R, A Matter of Principle (Clarendon 1986) Dworkin R, Sovereign Virtue: The Theory and Practice of Equality (Harvard up 2001) Dyevre A, ‘The constitutionalisation of the European Union: discourse, present, future and facts’ (2005) 30(2) EL Rev 165 Dzehtsiarou K, European Consensus and the Legitimacy of the European Court of Human Rights (Cambridge University Press 2015) Edward D and Lane R, European Union Law (Elgar 2013) Eliot TS, Murder in the Cathedral (Faber and Faber 1935) Ellis E, ‘The definition of discrimination in European Community sex equality law’ (1994) 19(6) EL Rev 563 Ellis E, ‘Recent developments in European Community sex equality law’ (1998) 35 CML Rev 379 Ellis E, ‘The recent jurisprudence of the Court of Justice in the field of sex equality’ (2000) 37 CML Rev 1403 Ellis E, eu Anti-​Discrimination Law (Oxford ec Law Library, oup 2005)

Bibliography

359

Ellner S, ‘A new model with rough edges:  Venezuela’s Community Councils’ (North American Congress on Latin America, unknown date) accessed 21 May 2018 Elster J, Local Justice:  How Institutions Allocate Scarce Goods and Necessary Burdens (Russell Sage Foundation 1992) Elster J, ‘The Empirical Study of Justice’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) European Parliament, ‘European Parliament resolution of 27 September 2007 on the application of Council Directive 2000/​43/​e c of 29 June 2000 implementing the principle of equal treatment between persons irrespective of racial or ethnic origin (2007/​2094(ini))’ (P6_​TA(2007)0422, 2007) European Parliament, ‘Report on a proposal for a modification of the Act concerning the election of the Members of the European Parliament by direct universal suffrage of 20 September 1976’ (28 April 2011) A7-​0176/​2011 –​PE440.210v04-​00 Evans A, ‘Union Citizenship and the Constitutionalization of Equality in eu Law’ in M La Torre (ed), European Citizenship: An Institutional Challenge (Kluwer 1998) Everson M, ‘Women and Citizenship of the European Union’ in TK Hervey and D O’Keeffe (eds), Sex equality law in the European Union (Wiley 1996) Ewald F, ‘Justice, Equality, Judgement:  On “Social Justice” ’ in G Teubner (ed), Juridification of Social Spheres (de Gruyter 1987) Fallon M and Martin D, ‘Dessine-​moi une discrimination’ (2010) 18 Journal de droit européen 165 Fenwick H and Hervey T, ‘Sex equality in the single market:  New directions for the European Court of Justice’ (1995) 32 CML Rev 443 Ferrera M, ‘Towards an “Open” Social Citizenship? The New Boundaries of Welfare in the European Union’ in G de Búrca (ed), eu law and the welfare state: in search of solidarity (oup 2005) Fishkin JS, When the People Speak:  Deliberative Democracy and Public Consultation (oup 2009) Fitzpatrick B, ‘The “mainstreaming” of sexual orientation into European equality law’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Flindall SA, ‘In the spirit of equality: breastfeeding leave for men’ [2010] 11 European Law Reporter 348 Flynn L, ‘Annotation P v S and Cornwall County Council’ (1997) 34 CML Rev 367 Fox JW, Jr, ‘Relational Contract Theory and Democratic Citizenship’ (2003–​2004) 54 Case W Res L Rev 1 Frankfurt H, ‘Equality as a Moral Ideal’ (1987) 98 Ethics 21 Franzese D, ‘The Gender Curve:  An Analysis of Colleges’ Use of Affirmative Action Policies to Benefit Male Applicants’ (2007) 56 Am UL Rev 719

360 Bibliography Fraser N, ‘Rethinking the Public Sphere:  A Contribution to the Critique of Actually Existing Democracy’ (1990) 25/​26 Social Text 56 Fredman S, ‘The Age of Equality’ in S Fredman and S Spencer (eds), Age as an equality issue: legal and policy perspectives (Hart 2003) Fredman S, ‘Changing the Norm:  Positive Duties In Equal Treatment Legislation’ (2005) 12 mj 369 Fredman S, ‘Reversing roles: Bringing men into the frame’ (2014) 10(4) International Journal of Law in Context 442 Gardner J, ‘Liberals and Unlawful Discrimination’ (1989) 9(1) ojls 1 Gilden A, ‘Toward a more transformative approach: The limits of transgender formal equality’ (2008) 23 Berkeley J Gender L & Just 83 Gutmann A, ‘Justice across the Spheres’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Gutmann A and Thompson D, Democracy and Disagreement (Harvard up 1996) Gyulavári T, ‘Age discrimination:  recent case law of the European Court of Justice’ (2013) 14 ERA Forum 377 Habermas J, ‘The entwinement of myth and enlightenment:  re-​reading Dialectic of Enlightenment’ (1982) 26 New German Critique 13 Habermas J, ‘Struggles for Recognition in the Democratic Constitutional State’ in C Taylor, Multiculturalism: Examining the politics of recognition (Princeton up 1994) Hagen B, ‘Schumacker Comparability –​the Conditions of the Schumacker Doctrine’ in K Dziurdz and C Marchgraber (eds), Non-​Discrimination in European and Tax Treaty Law: Open Issues and Recent Challenges (Linde Verlag 2015) Hailbronner K, ‘Union citizenship and access to social benefits’ (2005) 42 CML Rev 1263 Hale BM, Baroness Hale of Richmond, ‘Welcome to the uk Supreme Court?’ (2008 Denning Lecture) accessed 22 May 2018 Harris A, ‘Race and Essentialism in Feminist Legal Theory’ (1990) 42 Stanford L Rev 581 Hart HLA, The concept of law (3rd edn, oup 2012) Hartley TC, ‘The European Court, judicial objectivity and the constitution of the European Union’ (1996) 112 lqr 95 Heide I, ‘Sex equality and social security: selected rulings of the European Court of Justice’ (2004) 143(4) Intl Labour Rev 299 Hellman D, When is discrimination wrong? (Harvard University Press 2011) Hervey T, ‘Putting Europe’s House in Order:  Racism, Race Discrimination and Xenophobia after the Treaty of Amsterdam’ in D O’Keeffe and P Twomey (eds), Legal issues of the Amsterdam Treaty (Hart 1999) Hervey T, ‘Thirty Years of eu Sex Equality Law: Looking Backwards, Looking Forwards’ (2005) 12(4) mj 307

Bibliography

361

Hervey T and Shaw J, ‘Women, work and care: women’s dual role and double burden in ec sex equality law’ (1998) 8(1) Journal of European Social Policy 43 Hilson C, ‘Discrimination in Community free movement law’ (1999) 24(5) EL Rev 445 Holmes E, ‘Anti-​Discrimination Rights Without Equality’ (2005) 68(2) mlr 175 Holtmaat R and Tobler C, ‘cedaw and the eu’s Policy in the Field of Combating Gender Discrimination’ (2005) 12(4) mj 399 Honneth A, ‘Critical Theory’ in D Moran (ed), The Routledge companion to twentieth century philosophy (Routledge 2008) hooks b, ‘Dig Deep:  Beyond Lean In’ (The Feminist Wire, 28 October 2013) accessed 22 May 2018 Hosking DL, ‘Great Expectations: Protection from discrimination because of disability in Community law’ (2006) 31(5) EL Rev 667 Howard E, ‘The case for a considered hierarchy of discrimination grounds in eu law’ (2006) 13(4) mj 445 Hume D, A Treatise of Human Nature (DF Norton and MJ Norton eds, oup 2007) Huxley A, Brave New World (Vintage 2004) Iliopoulou A, ‘Le principe d’égalité et de non-​discrimination’ in J-​B Auby and J Dutheil de La Rochère (eds), Droit administrative européen (Bruylant 2007) Ingram J, Radical Cosmopolitics:  The Ethics and Politics of Democratic Universalism (Columbia up 2013) Johansson ST, ‘Towards spherical justice: a critical theoretical defence of the idea of complex equality’ (PhD thesis, University of Southampton 2003, unpublished) Joly L, L’emploi des personnes handicapées entre discrimination et égalité (Dalloz 2013) Justesen P, Country report Non-​discrimination Denmark (Publications Office of the European Union 2016) Kalcik R and Wolf GB, ‘Is Brexit an opportunity to reform the European Parliament?’ [2017] 2 Bruegel Policy Contribution 1  Kaupa C, ‘Maybe not activist enough? On the Court’s alleged neoliberal bias in its recent labor cases’ in M Dawson, B de Witte and E Muir (eds), Judicial activism at the European Court of Justice (Elgar 2013) Kautz S, Liberalism and Community (Cornell up 1995) Keat R, ‘Colonisation by the Market:  Walzer on Recognition’ (1997) 5(1) Journal of Political Philosophy 93 Kjaer AL and Adamo S, ‘Linguistic diversity and European democracy: Introduction and overview’ in AL Kjaer and S Adamo (eds), Linguistic diversity and European democracy (Ashgate 2011) Koldinská K, ‘Case law of the European Court of Justice on sex discrimination 2006–​ 2011’ (2011) 48 CML Rev 1599 Kommers DP, The Constitutional Jurisprudence of the Federal Republic of Germany (2nd edn, Duke up 1997)

362 Bibliography Koppelman A, Antidiscrimination Law and Social Equality (Yale up 1996) Koppelman A, ‘The Miscegenation Analogy in Europe, or, Lisa Grant meets Adolf Hitler’ in R Wintemute and M Andenæs (eds), Legal recognition of same-​sex partnerships: A study of national, European and international law (Hart 2001) Lacey N, ‘From Individual to Group?’ in B Hepple and EM Szyszcak (eds), Discrimination: The Limits of Law (Mansell 1992) Ladeur K-​H, ‘European Law as Transnational Law –​Europe has to be conceived as an heterarchical network and not as a superstate!’ (2009) 10 German Law Journal 1358 Laski HJ, A Grammar of Politics (George Allen & Unwin 1925) Lavranos N, ‘The new specialised courts within the European judicial system’ (2005) 30(2) EL Rev 261 Lawson D, ‘From Pentecost island to modern Britain, the futility of trying to measure happiness’ The Independent (London 6 July 2007) 33 Lembke U, ‘Tackling sex discrimination to achieve gender equality? Conceptions of sex and gender in eu non-​discrimination law and policies’ [2016] 2 European Equality Law Review 46 Lenaerts K, ‘L’égalité de traitement en droit communautaire: un principe unique aux apparences multiples’ [1991] cde 3 Lenin VI, The deception of the people by the slogans of equality and freedom (Lawrence and Wishart 1940) Lester A, ‘The Uncertain Trumpet –​References to the Court of Justice from the United Kingdom:  Equal Pay and Equal Treatment without Sex Discrimination’ in HG Schemers (ed), Article 177 eec: Experiences and Problems (North Holland 1987) Lester A and Uccellari P, ‘Extending the equality duty to religion, conscience and belief: proceed with caution’ (2008) 5 ehrlr 567 Letsas G, ‘The constitution and the folly of majoritarianism’ (United Kingdom Constitutional Law Association Blog, 20 February 2017) accessed 21 May 2018 Lieber DL, Etz Hayim:  A Torah Commentary (Jewish Publication Society of America 2001) Locke J, The Second Treatise of Civil Government (Basil Blackwell 1946) Lucas JR, ‘Against Equality’ (1965) XL Philosophy 296 Lyotard J-​F, The postmodern condition:  A report on knowledge (G Bennington and B Massumi trs, Manchester University Press 1984) Macdonald RA, ‘Access to Justice and Law Reform’ (1990) 10 Windsor YB Access Just 287 MacKinnon C, ‘Feminism, Marxism, Method, and the State:  Toward Feminist Jurisprudence’ (1983) 8 Signs 635 MacKinnon C, Feminism Unmodified: Discourses on Life and Law (Harvard up 1987)

Bibliography

363

MacKinnon C, ‘Reflections on Sex Equality Under the Law’ (1990–​91) 100 Yale LJ 1281 Macklem T, Beyond Comparison: Sex and Discrimination (cup 2003) Maduro MP, ‘Europe’s Social Self: “The Sickness unto Death” ’ in Jo Shaw (ed), Social Law and Policy in an Evolving eu (Hart 2000) Mancini GF, ‘The Making of a Constitution for Europe’ (1989) 26 CML Rev 595 Mancini GF and O’Leary S, ‘The New Frontiers of Sex Equality Law in the European Union’ (1999) 24(4) EL Rev 331 Mann BH, ‘Afterword’ in CL Tomlins and BH Mann (eds), The Many Legalities of Early America (University of North Carolina Press 2001) Marconi D, ‘Reference and theories of meaning as use’ in A Bianchi (ed), On Reference (oup 2015) Marriott L, ‘Justice and the Justice System: A Comparison of Tax Evasion and Welfare Fraud in Australia and New Zealand’ (2013) 22(2) Griffith Law Review 403 Martens K, ‘Justice in transport as justice in accessibility: Applying Walzer’s “Spheres of Justice” to the transport sector’ (2012) 39(6) Transportation 1035 Martin D, Égalité et non-​discrimination dans la jurisprudence communautaire: étude critique à la lumière d’une approche comparatiste (Bruylant 2006) Martin D, ‘L’interdiction de discriminer en raison de la nationalité:  quelques arrêts troublants’ [2009–​2010] 2 rae 237 Martin D, ‘La libre circulation des personnes: au-​delà de l’évolution et des révolutions, la perpétuelle quête de sens’ [2012] 1 rae 85 Martin LH, Gutman H and Hutton PH (eds), Technologies of the self: a seminar with Michel Foucault (University of Massachusetts Press 1988) Marx K, Capital: A critique of Political Economy Volume One (B Fowkes tr, Penguin 1976) Marx K, Marx: Later political writings (T Carver ed, cup 1996) Masselot A, ‘The State of Gender Equality Law in the European Union’ (2007) 13(2) Eur LJ 152 Matson W, ‘Justice: A Funeral Oration’ (1983) I Social Philosophy and Policy 94 May Theresa, ‘Theresa May’s Brexit speech in full’ (17 January 2017) accessed 7 May 2018 May Todd, Nonviolent resistance: a philosophical introduction (Polity 2015) Mayer R, ‘Michael Walzer, Industrial Democracy, and Complex Equality’ (2001) 29(2) Political Theory 237 McColgan A, Discrimination Law: Text, Cases and Materials (2nd edn, Hart 2005) McColgan A, ‘Cracking the comparator problem: Discrimination, “Equal” Treatment and the Role of Comparisons’ (2006) 6 ehrlr 650 McCrudden C, ‘Community and Discrimination’ in J Eekelaar and J Bell (eds), Essays in Jurisprudence (Third Series) (Clarendon 1987)

364 Bibliography McCrudden C, ‘The new concept of equality’ (Paper prepared for the Academy of European Law conference, ‘Fight Against Discrimination:  The Race and Framework Employment Directives,’ 2–​3 June 2003) accessed 20 May 2018 McCrudden C, ‘The new architecture of eu Equality Law after chez: Did the Court of Justice reconceptualise direct and indirect discrimination?’ [2016] 1 European Equality Law Review 1 McCrudden C and Kountouros H, ‘Human rights and European equality law’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) McGlynn C, ‘Social Policy: Equality, Maternity and Questions of Pay’ (1996) 21(4) EL Rev 327 McGlynn C, ‘ec legislation prohibiting age discrimination: “Towards a Europe for All Ages”?’ (2000) 3 cyels 279 McGrath K and others, ‘Exploring the democratic potential of online social networking: The scope and limitations of e-​participation’ (2012) 30 Communications of the Association for Information Systems 239 Meenan H, ‘Age equality after the employment directive’ (2003) 10(1) mj 9 Meenan H, ‘Age discrimination –​Of Cinderella and The Golden Bough’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Meenan H, ‘Conclusion’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Meenan H (ed), Equality Law in an Enlarged European Union (cup 2007) Meenan H, ‘Introduction’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Megarry RE, Miscellany-​at-​Law (Stevens 1955) Menand L, ‘Saved from drowning: Barthelme reconsidered’ The New Yorker (New York, 23 February 2009) 68 Merkel W, ‘Social justice and the three worlds of welfare capitalism’ (2002) 43 European Journal of Sociology 59 Mešić N and Woolfson C, ‘Roma berry pickers in Sweden: economic crises and new contingents of the austeriat’ (2015) 21(1) Transfer: European Review of Labour and Research 37 Miller D, ‘Complex Equality’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Miller D, ‘Introduction’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Miller D, Justice for Earthlings: Essays in Political Philosophy (cup 2012) Millett T, ‘Sex equality: the influence of Community law in Great Britain’ (1986) 6 YB of Eur L 219

Bibliography

365

Millett T, ‘European Community law: sex equality and retirement age’ (1987) 36(3) iclq 616 Minda G, Postmodern legal movements: Law and jurisprudence at century’s end (nyu Press 1995) Mirowski P, The effortless economy of science? (Duke University Press 2004) Mirowski P, Never Let a Serious Crisis Go to Waste:  How Neoliberalism Survived the Financial Meltdown (Verso 2014) Moon G and Allen R, ‘Dignity discourse in discrimination law: a better route to equality?’ (2006) 6 ehrlr 610 Moore MR and Stambolis-​Ruhstorfer M, ‘lgbt sexuality and families at the start of the twenty-​first century’ (2013) 39 Annu Rev Sociol 491 More G, ‘ “Equal Treatment” of the Sexes in European Community Law:  What Does “Equal” Mean?’ (1993) 1 Feminist LS 45 More G, ‘The Principle of Equal Treatment:  From Market Unifier to Fundamental Right?’ in P Craig and G de Búrca (eds), The Evolution of Eu Law (oup 1999) Moreau S, ‘What is discrimination?’ (2010) 38(2) Philosophy & Public Affairs 143 Moreau S, ‘In defense of a liberty-​based account of discrimination’ in D Hellman and S Moreau (eds), Philosophical foundations of discrimination law (oup 2013) Muir E, ‘Enhancing the effects of Community law on national employment policies: the Mangold case’ (2006) 31(6) EL Rev 879 Mulder J, ‘Some more equal than others? Matrimonial benefits and the cjeu’s case law on discrimination on the grounds of sexual orientation’ (2012) 19 mj 505 Mullenix LS, ‘The limits of “complex equality” ’ (1983) 97 Harv L Rev 1801 Muniz-​Fraticelli VM, The structure of pluralism (oup 2014) Narveson J, ‘Libertarianism v Marxism:  Reflections on GA Cohen’s Self-​Ownership, Freedom and Equality’ (1998) 2(1) Journal of Ethics 1 Neruda P, ‘Oda a la crítica’ Neville Brown L and Kennedy T, The Court of Justice of the European Communities (Sweet & Maxwell 2000) Nordby H, ‘Mental content externalism and social understanding’ (2012) 2(1) Open Journal of Philosophy 1 Nozick R, Anarchy, State, and Utopia (Basic Books 1974) Numhauser-​Henning A, ‘eu sex equality law post-​Amsterdam’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Nussbaum M, ‘Patriotism and Cosmopolitanism’ in M Nussbaum and J Cohen (eds), For Love of Country, Debating the Limits of Patriotism (Beacon Press 1996) O’Cinneide C, ‘Age Discrimination and Mandatory Retirement’ (2008) 6/​7 Eur Anti-​ discrimination L Rev 13 O’Cinneide C, ‘The Right to Equality: A Substantive Legal Norm or Vacuous Rhetoric?’ (2008) 1 UCL Human Rights Rev 80

366 Bibliography O’Leary S, ‘The Free Movement of Persons and Services’ in P Craig and G de Búrca (eds), The Evolution of Eu Law (oup 1999) O’Leary S, ‘Developing an ever closer union between the peoples of Europe? A reappraisal of the case law of the Court of Justice on the free movement of persons and eu citizenship’ (2008) 27 YB of Eur L 167 O’Neill S, Impartiality in Context:  Grounding Justice in a Pluralist World (suny Press 1997) Okin SM, Justice, gender, and the family (Basic Books 1989) Onions CT (ed), The Shorter Oxford English Dictionary (3rd rev edn, Clarendon 1973) Opsahl T, ‘Equality and non-​discrimination’ in T Opsahl, Law and equality:  Selected articles on human rights (Ad Notam Gyldendal 1996) Orbinski J, ‘Keynote address:  Justice and global health’ in M Freeman, S Hawkes and B Bennett (eds), Law and Global Health (Current legal issues volume 16, oup 2014) Orwell G, Animal Farm (Penguin Student Editions, Penguin 1999) Osbourne T, ‘Will the European Union Directive on equal treatment fulfil its purpose of combating age discrimination in employment?’ (2004) 38(3) The International Lawyer 867 Pigott C, ‘ecj passes the buck’ (2009) 159(7361) nlj 407 Pilgerstorfer M and Forshaw S, ‘Transferred discrimination in European law’ (2008) 37(4) Indus LJ 384 Pitt G, ‘Religion or belief: aiming at the right target?’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Posner RA and Rosenfield AM, ‘Impossibility and Related Doctrines in Contract Law: An Economic Analysis’ (1977) 6 jls 83 Prechal S, ‘Equal Treatment, Non-​Discrimination and Social Policy:  Achievement in Three Themes’ (2004) 41 CML Rev 533 Quinn G, ‘Disability discrimination law in the European Union’ in H Meenan (ed), Equality Law in an Enlarged European Union (cup 2007) Rakowski E, Equal Justice (Clarendon 1991) Rancière J, The ignorant schoolmaster: five lessons in intellectual emancipation (K Ross tr, Stanford University Press 1991) Rancière J, On the Shores of Politics (L Heron tr, Verso 1995) Rasmussen H, On law and policy in the European Court of Justice: a comparative study in judicial policymaking (Kluwer 1986) Read K, ‘When Is a Kid a Kid? Negotiating Children’s Rights in El Salvador’s Civil War’ (2002) 41(4) History of Religions 391 Richardson HS, Democratic Autonomy:  Public Reasoning About the Ends of Policy (oup 2002) Ricoeur P, The Just (D Pellauer tr, University of Chicago Press 2003)

Bibliography

367

Ritchie J and others, Qualitative Research Practice (Sage 2003) Robertson AH and Merrills JG, Human Rights in the World: An introduction to the study of the international protection of human rights (4th edn, mup 1996) Robson R, ‘Third parties and the third sex:  Child custody and lesbian legal theory’ (1994) 26 Conn L Rev 1377 Rochel J, ‘Welcome to the eu: Immigration and Non-​discrimination’ in S Besson and AR Ziegler, Egalité et non-​discrimination en droit international et européen: Equality and non-​discrimination in international and European law (Schulthess 2014) Rodgers DT, ‘Contesting equality’ (2014) 33(4) Raritan 19 Rodgers L, ‘Labour law and the “public interest”: discrimination and beyond’ (2011) 2(4) European Labour Law Journal 302 Rorty R, ‘Dewey’s metaphysics’ in SM Cahn (ed), New Studies in the Philosophy of John Dewey (University Press of New England 1977) Rorty R, ‘Habermas and Lyotard on Postmodernity’ in RJ Bernstein (ed), Habermas and Modernity (Polity 1985) Rorty R, Philosophy and the Mirror of Nature: Thirtieth-​Anniversary Edition (Princeton up 2009) Rosenblum NL, ‘Moral Membership in a Postliberal State’ (1984) 36(4) World Politics 581 Ross M, ‘Solidarity  –​a new constitutional paradigm for the eu?’ in M Ross and Y Borgmann-​Prebil (eds), Promoting solidarity in the European Union (Oxford University Press 2010) Rubio-​Marín R, Immigration as a Democratic Challenge (cup 2000) Ruddick G, ‘Tesco proposes deal with dairy farmers after milk price protests’ (Guardian, 4 November 2015) accessed 6 May 2018 Russell B, History of Western Philosophy (rev edn, Routledge 2004) Rustin M, ‘Equality in Post-​Modern Times’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Sabbagh C and Resh N, ‘Unfolding Justice Research in the Realm of Education’ (2016) 29(1) Social Justice Research 1 Sangiovanni A, ‘Solidarity in the European Union’ (2013) 33(2) ojls 213 Scalia A, ‘Modernity and the constitution’ in E Smith (ed), Constitutional justice under old constitutions (Kluwer 1995) Scanlon TM, What we owe to each other (Bellknap Press of Harvard up 1998) Scanlon TM, The difficulty of tolerance: Essays in political philosophy (cup 2003) Scharpf F, Governing in Europe: Effective and Democratic? (oup 1999) Schiek D, ‘A New Framework on Equal Treatment of Persons in ec Law?’ (2002) 8(2) Eur LJ 290 Schiek D, ‘Age discrimination before the ecj: Conceptual and theoretical issues’ (2011) 48 CML Rev 777

368 Bibliography Schiek D, Waddington L and Bell M (eds), Cases, Materials and Text on National, Supranational and International Non-​Discrimination Law (Ius Commune Casebooks for the Common Law of Europe, Hart 2007) Schmitt C, Politische Theologie:  Vier Kapitel zur Lehre von der Souveränität [Political Theology:  Four Chapters towards a Theory of Sovereignty] (Duncker & Humblot 1990) Schwarze J, European Administrative Law (Rev 1st edn, Sweet & Maxwell and Office for Official Publications of the European Communities 2006) Sebok AJ, ‘Should the law preserve party control? Litigation investment, insurance law, and double standards’ (2015) 56(3) William and Mary law review 833 Sedgwick EK, ‘Tales of the Avunculate:  Queer Tutelage in The Importance of Being Earnest’ in EK Sedgwick, Tendencies (Routledge 1998) Sellars W, Empiricism and the Philosophy of Mind (Harvard up 1997) Sen P and Al-​Khalili J, Atom (Television series, bbc 2007) Sevón L, ‘General Principles of Community Law –​Concluding Remarks’ in U Bernitz and J Nergelius (eds), General Principles of European Community Law (Kluwer Law International 2000) Sexton TRF, ‘Enacting national environmental laws more stringent than other States’ laws in the European Community: Re Disposable Beer Cans: Commission v Denmark’ (1991) 24(3) Cornell Intl LJ 563 Shakespeare W, Hamlet Shakespeare W, Twelfth Night (The New Cambridge Shakespeare, cup 1985) Shklar JN, Political thought and political thinkers (S Hoffmann ed, University of Chicago Press 1998) Siegel RB, ‘Why Equal Protection No Longer Protects: The Evolving Forms of Status-​ Enforcing State Action’ (1997) 49 Stan L Rev 1111 Skidmore P, ‘Sex, Gender and Comparators in Employment Discrimination’ (1997) 26 Indus LJ 51 Skidmore P, ‘ec Framework Directive on Equal Treatment in Employment: Towards a Comprehensive Community Anti-​Discrimination Policy?’ (2001) 30 Indus LJ 126 Smith AM, Laclau and Mouffe: The radical democratic imaginary (Routledge 1998) Smith G, Democratic Innovations (cup 2009) Smith R, ‘Benign intervention’ LS Gaz (London, 12 November 2009) 10 Smout TC, ‘Perspectives on the Scottish identity’ [1994] 6 Scottish Affairs 101 Somek A, ‘Solidarity Decomposed: Being and Time in European Citizenship’ (2007) 32(6) EL Rev 787 Somek A, Engineering equality:  An essay on European anti-​ discrimination law (oup 2011) Spaventa E, ‘Leaving Keck behind? The free movement of goods after the rulings in Commission v Italy and Mickelsson and Roos’ (2009) 24 EL Rev 914

Bibliography

369

Stambolis-​Ruhstorfer M, ‘The scientific and legal recognition of gay and lesbian parenting in France and the us: How law and knowledge influence each other’ (Presentation given at the 22nd International Conference of Europeanists, Paris, 9 July 2015) Stein E, ‘Lawyers, judges and the making of a transnational constitution’ (1981) 75 American Journal of International Law 1 Stjernø S, Solidarity in Europe –​the history of an idea (cup 2005) Stokke H, ‘Reasonable Discrimination? Affirming Access to Higher Education in Malaysia’ [1999–​2000] Hum Rts Dev YB 189 Stoppard T, Every Good Boy Deserves Favour and Professional Foul (Faber & Faber 1978) Stryker S, ‘My words to Victor Frankenstein above the village of Chamounix: Performing transgender rage’ [1994] 1 GLQ: A Journal of Lesbian and Gay Studies 237 Sussmann N, ‘Introduction:  The substantive unity of Michael Walzer’s pluralism’ in Y Benbaji and N Sussmann (eds), Reading Walzer (Routledge 2014) Sutch P, ‘Reiterating rights:  International society in transition’ in B Haddock and P Sutch (eds), Multiculturalism, identity and rights (Routledge 2003) Taylor C, ‘Atomism’ in C Taylor, Philosophical Papers Vol 2: Philosophy and the Human Sciences (cup 1985) Taylor C, ‘The Politics of Recognition’ in C Taylor, Multiculturalism:  Examining the ­politics of recognition (Princeton up 1994) Thigpen RB and Downing LA, ‘Liberal and Communitarian Approaches to Justification’ (1989) 51(4) Review of Politics 533 Thómqvist C and Bernhardsson S, ‘Their own stories –​how Polish construction w ­ orkers posted to Sweden experience their job situation, or resistance versus life projects’ (2015) 21(1) Transfer: European Review of Labour and Research 23 Tickner JA, Gendering world politics:  issues and approaches in the post-​Cold War era (Columbia University Press 2001) Tobler C, Indirect discrimination: a case study into the development of the legal concept of indirect discrimination under ec law (Intersentia 2005) Tobler C and Waaldijk K, ‘Annotation Maruko’ (2009) 46 CML Rev 723 Tomei M, ‘Discrimination and equality at work:  A review of the concepts’ (2003) 142(4) Intl Labour Rev 401 Tomuschat C, ‘Annotation Martínez Sala’ (2000) 37 CML Rev 449 Trappenburg M, ‘In Defence of Pure Pluralism: Two Readings of Walzer’s Spheres of Justice’ (2000) 8(3) Journal of Political Philosophy 343 Tridimas T, ‘The Court of Justice and judicial activism’ (1996) 21(3) EL Rev 199 Tridimas T, ‘The Application of the Principle of Equality to Community Measures’ in A Dashwood and S O’Leary (eds), The Principle of Equal Treatment in ec Law (Sweet & Maxwell 1997) Tsakatika M, Political responsibility and the European Union (mup 2008)

370 Bibliography Tye L, The Father of Spin: Edward L Bernays and The Birth of Public Relations (Crown Publications 1998) Unger RM, False Necessity:  Anti-​necessitarian Social Theory in the Service of Radical Democracy (cup 1987) van den Hoven J and Vermaas PE, ‘Nano-​Technology and Privacy:  On Continuous Surveillance Outside the Panopticon’ (2007) 32 Journal of Medicine and Philosophy 283 van der Mei AP, Free Movement of Persons within the European Community (Hart 2003) van der Veen RJ, ‘The Adjudicating Citizen: On Equal Membership in Walzer’s Theory of Justice’ (1999) 29(2) British Journal of Political Science 225 Vanistendael F, ‘Annotation Gilly’ (2000) 37 CML Rev 167 Villiers C, ‘Employees as creditors:  a challenge for justice in insolvency law’ (1999) 20(7) Company Lawyer 222 Vlastos G, ‘Justice and Equality’ in R Brandt (ed), Social Justice (Prentice Hall 1962) Vos E, ‘The European Court of Justice in the face of scientific uncertainty and complexity’ in M Dawson, B de Witte and E Muir (eds), Judicial activism at the European Court of Justice (Elgar 2013) Waaldijk K and Bonini-​Baraldi M, Sexual orientation discrimination in the European Union: national laws and the Employment Equality Directive (tmc Asser Press 2006) Waddington L, ‘Article 13 ec:  Setting Priorities in the Proposal for a Horizontal Employment Directive’ (2000) 29(2) ilj 176 Waddington L, ‘The development of a new generation of sex equality directives’ (2004) 11(1) mj 3 Waddington L, ‘Annotation Coleman’ (2009) 46 CML Rev 665 Waddington L, ‘Future prospects for eu equality law:  lessons to be learnt from the proposed Equal Treatment Directive’ (2011) 36 EL Rev 163 Waddington L and Bell M, ‘More equal than others: distinguishing European Union equality directives’ (2001) 38 CML Rev 587 Waldron J, ‘The Cosmopolitan Alternative’ in W Kymlicka (ed), The Rights of Minority Cultures (oup 1987) Walters M, ‘ukip mp’s patent court warning fails to find support’ LS Gaz (London, 24 February 2017) accessed 21 May 2018 Walzer M, ‘Review of Taking Rights Seriously by Ronald Dworkin’ The New Republic (New York, 25 June 1977) 28 Walzer M, ‘Philosophy and Democracy’ (1981) 9(3) Political Theory 379 Walzer M, Spheres of Justice –​A Defense of Pluralism and Equality (Basic Books 1983) Walzer M, ‘Liberalism and the Art of Separation’ (1984) 12(3) Political Theory 315 Walzer M, ‘Justice Here and Now’ in FS Lucash (ed), Justice and Equality Here and Now (Cornell University Press 1986)

Bibliography

371

Walzer M, ‘The long-​term perspective’ (1986) 62(1) Bulletin of the New York Academy of Medicine 8 Walzer M, ‘Interpretation and Social Criticism’ in SM McMurrin (ed), The Tanner Lectures on Human Values, viii (cup 1988) Walzer M, ‘The Communitarian Critique of Liberalism’ (1990) 18(1) Political Theory 6 Walzer M, ‘Exclusion, Injustice, and the Democratic State’ (1993) 40 Dissent 55 Walzer M, ‘Objectivity and Social Meaning’ in M Nussbaum and A Sen (eds), The Quality of Life (oup 1993) Walzer M, ‘Shared meanings in a poly-​ethnic democratic setting: A response’ (1994) 22(2) Journal of Religious Ethics 401 Walzer M, Thick and Thin: Moral Argument at Home and Abroad (University of Notre Dame Press 1994) Walzer M, ‘Response’ in D Miller and M Walzer (eds), Pluralism, Justice, and Equality (oup 1995) Walzer M, ‘Spheres of Affection’ in M Nussbaum and J Cohen (ed), For Love of Country, Debating the Limits of Patriotism (Beacon Press 1996) Walzer M, On Toleration (Yale up 1997) Walzer M, ‘Governing the globe: what is the best we can do?’ (2000) 47(4) Dissent 44 Walzer M, The Company of Critics:  Social Criticism and Political Commitment in the Twentieth Century (2nd edn, Basic Books 2002) Walzer M, Politics and Passion: Toward a More Egalitarian Liberalism (Yale up 2004) Walzer M, ‘Feminism and me’ (2013) 60(1) Dissent 50 Walzer M, ‘Response’ in Y Benbaji and N Sussmann (eds), Reading Walzer (Routledge 2014) Walzer M and Dworkin R, ‘Spheres of Justice: An Exchange’ New York Review of Books (New York, July 1983) 43 Warnier N, ‘Les discriminations directes and indirectes dans la domaine de l’égalité homme-​femme et de l’égalité nationaux-​non-​nationaux’ (2006) 84(2) Revue de droit international et de droit comparé 225 Warren ME, ‘What Can Democratic Participation Mean Today?’ (2002) 30(5) Political Theory 677 Watson P, ‘Equality, fundamental rights and the limits of legislative discretion’ (2011) 36 EL Rev 896 Weatherill S, ‘The road to ruin: “restrictions on use” and the circular lifecycle of Article 34 tfeu’ [2012] 2 European Journal of Consumer Law 359 Weatherill S, Law and values in the European Union (oup 2016) Weiler JHH, ‘Does Europe Need a Constitution? Demos, Telos and the German Maastricht Decision’ (1995) 1(3) Eur LJ 219 Weiler JHH, ‘Towards a Principle of Economic Comity’ (Lecture given at Bentham House, ucl, 23 May 2001)

372 Bibliography Wentholt K, ‘Formal and Substantive Equal Treatment:  the Limitations and the Potential of the Legal Concept of Equality’ in T Loenen and PR Rodrigues (eds), Non-​discrimination law: comparative perspectives (Kluwer Law International 1999) Westen P, ‘The Empty Idea of Equality’ (1982) 95 Harv L Rev 537 White RCA, ‘Citizenship of the Union, Governance, and Equality’ (2006) 29 Fordham International Law Journal 790 White SK, Political Theory and Postmodernism (cup 1991) Wielsch D, ‘Relational Justice’ (2013) 76 Law and Contemporary Problems 191 Williams BAO, ‘The Idea of Equality’ in P Laslett and WG Runciman (eds), Philosophy, Politics and Society, Series II (Basil Blackwell 1972) Wilsher D, ‘Does Keck discrimination make any sense? An assessment of the non-​ discrimination principle within the European single market’ (2008) 33(1) EL Rev 3 Wintemute R, ‘Recognising New Kinds of Direct Sex Discrimination: Transsexualism, Sexual Orientation and Dress Codes’ (1997) 60 mlr 334 Wintemute R, ‘When is Pregnancy Discrimination Indirect Sex Discrimination?’ (1998) 27(1) Indus LJ 23 Wirszubski C, Libertas as a Political Idea at Rome During the Late Republic and Early Principate (cup 1968) Zander M, The Law-​Making Process (cup 2004)

Index Age discrimination See Discrimination Anti-​discrimination law See Discrimination Aristotelian test for equality 1, 178, 189 n301, 203, 315 n18, 317 n27 Absurdity/​illogicality of 194, 255, 263 Circularity of 158 Difference-​blindness of 6 Inconsistency of 7, 72, 83, 223 Malleability of 5, 8, 46, 62 Reinforcement of inequality by 158, 202, 205 Yes-​no simplicity of 1, 8, 226 Aristotle 1 n2, 1–​3, 6, 8–​12, 74, 81, 111, 138, 160, 179 n266, 180 n266, 195, 202, 206, 275, 311, 314, 317, 320, 341 See Aristotelian test for equality ‘Article 19’ grounds, the 9 n34, 118, 144–​5, 145 n126, 165, 179, 182, 196, 241 n12 Autonomy of spheres  Maintenance of See Spheres Boundary breach See Complex equality Brexit 50 n18, 294 n21, 295 n25, 294–​6 cjeu  And like-​for-​like equality 2, 3, 3 n8, 61, 75, 84, 111, 163–​4, 172, 185, 200–​ 3, 201 n19, 203 n25, 206, 212, 246, 251, 254, 265, 317 n27 And objective justifiation 4–​5, 5 n15, 102, 125, 133–​4, 144–​8, 151, 153, 153 n160, 157–​9, 185 n289, 189 n301, 241, 246, 279 And proportionality 147, 148 n141, 149, 152, 125, 185 n289 And scientific evidence 286, 286 n38, 286 n39 And solidarity 227 And substantive/​noncomparative equality 54, 59, 61, 64, 70, 70 n93, 91, 104

And suspect classifications 241–​242, 242 n13 And the echr 43–​4, 128 And third party intervention 286 n40 As potential ascertainer of shared meanings 297, 302 As potential reviewer of shared meanings 176, 181, 190, 281–​3, 281 n23, 285 n29 Handling/​mishandling of equality 2 Judicial activism/​interventionism 276–​7, 279–​80, 282 Jurisdictional subsidiarity 339 Maintaining diversity/​union balance 44 Market-​driven approach 19 n76, 93, 93 n193 Perpetuation of stereotype/​unfairness by 157, 159, 180 n267, 202 Political/​policy-​making role 166, 206, 295 n27 Possible equality strategy for 12, 23 Understanding of disability by 183, 187, 189, 195 Co-​existential equilibrium, doctrine of 329, 349 Competition law 10, 33, 253, 269, 274 See Non-​suspect grounds Complex equality 1, 9, 22–​8, 28 n28, 30 n37, 34 n55, 34–​8, 44, 47 n9, 47–​9, 63, 64 n78, 67, 82–​5, 96 n204, 94–​8, 106, 106 n246, 108, 116, 136, 174, 189, 199 n7, 203, 208, 213, 230 n150, 237–​9, 244, 250, 258, 262, 272, 276–​7, 279, 281–​2, 285, 289, 297, 299, 301 n49, 311, 312 n8, 335–​6, 339–​40, 340 n149, 343, 348–​9, 349 n8 Blocked exchanges  Definition 24 Boundary breach 27, 42 n118, 48–​9, 49 n15, 51, 52 n30, 52–​4, 66–​7, 82, 94, 96–​7, 106, 131, 156, 174–​5, 181, 208, 217, 219, 244, 246 n27, 246–​8, 250, 255, 257, 263, 270–​1, 273, 299, 320, 332, 334 n116, 334–​5, 347

374 Index Complex equality (cont.) Definition 24 Luck as 271 Shared understandings See Shared understandings Small inequality See Small inequality Three tenets of 26 Consumers, discrimination between See Semi-​suspect grounds Court of Justice of the European Union See CJEU Dilemma of difference, the 6, 348 Disability, discrimination on grounds of See Discrimination Discrimination  Age, based on 140 Disability, based on 182 Gender, based on See Gender discrimination law Indirect 6, 11, 76 n112, 80 n134, 85, 86 n153, 87 n165, 89, 93, 102, 124–​5, 129–​31, 133–​4, 136–​8, 171–​3, 173 n244, 174 n246, 179 n266, 187, 241, 267–​8, 277, 317 n27, 343 Intersectional 140 n94 Non-​suspect grounds, based on See Non-​suspect grounds Racial or ethnic origin, based on 119 Religion or belief, based on 131 Semi-​suspect grounds, based on See Semi-​suspect grounds Between consumers vis-à-vis the CAP See Semi-​suspect grounds Between producers vis-à-vis the CAP See Semi-​suspect grounds Between public and private undertakings See Semi-​suspect grounds Sexual Orientation, based on 160 Distribution  Distributive communities 10, 199, 234, 329 Distributive principle 28, 32–​3, 64, 66, 94, 98, 219, 246–​8, 252, 258, 260, 273, 277, 281, 301

Flawed distributions 24, 49, 94, 129, 136 Of biological characteristics 49, 53 Dominance  Negative dominance 27, 27 n26, 64, 96, 130–​1, 136 n79, 176, 190, 194, 207, 249, 252, 255 n50 Prohibition on  As third tenet of complex equality 24 Dominant good 24–​7, 35, 52 n24, 136 n79, 156, 175 n252, 175–​6, 178, 190 n304, 273 See Dominance Equality  Complex equality See Complex equality General principle of 2, 118, 121 n12, 143–​4, 180, 181 n268, 189 n302, 212, 274 Rough equality 25 Simple equality See Aristotelian test for equality Substantive 6, 6 n19, 6 n21, 55, 59, 70, 78, 80, 93, 99, 103–​4, 341 n156 Vs freedom 20 Ethnic origin, discrimination on grounds of See Discrimination European Community Treaty 2, 102 n228, 118, 122 European Convention on Human Rights 42 n117, 42 n118, 43 n120, 43 n120, 43 n121, 42–​4, 128, 128 n48, 132, 144, 168, 213, 215, 220 Expert evidence  Need for 282 Flexibility 1, 98 n208, 146, 152, 159, 183, 188–​90, 190 n305, 312–​13, 315–​16 Forum, the 8 n32, 8–​9, 29, 96 n204, 114, 120 n11, 190 n305, 192 n307, 190–​4, 289–​90, 294–​6, 298, 344, 348–​9 Design of 290 Vs national parliament 294–296 Freedom  Vs equality 20

375

Index Fundamentum distributionis See Distribution Gender discrimination law 3, 12, 46, 51 n22, 109, 117, 120, 161, 163, 188, 242, 242 n13, 297 Insurance premiums and benefits 108 Paternity leave 68 Pensions 74 Positive action 99 Pregnancy 54 The question of part-​time and full-​time work 85 Human resources See Non-​suspect grounds Judaism  Pentacost 132, 343 Sabbatical 13 Judicial interventionism  Need for 276 Laïcité 133 Legislation  Need to rewrite 287 Libertarianism  Vs Marxism 12 Like-​for-​like test, the See Aristotelian test for equality, See CJEU Luck  Boundary breach vs small inequality 271 Marx, Karl 12–​14, 289, 289 n1 Mediated complexity, theory of 9, 11 n37, 11–​12, 190 n305, 239 n196, 297, 321 n43, 345 n164 Minimal morality, the 41, 42 n118, 45, 128, 213, 215 Neoconservatism 327 Neoliberalism 12 n42, 315, 317 n27 Non-​suspect grounds 253 Override, the 36, 38, 42 n118, 42–​4, 101, 128, 129 n49, 182 n270, 220 Explained 42

Post-​Marxism 292 n15 Postmodernism 51–​2, 320, 322, 322 n44, 322 n45, 322 n47, 322 n48, 326–​7 Post-​structuralism 292 n15 See Postmodernism Principles  Equality See Equality Producers, discrimination between See Semi-​suspect grounds Public procurement See Non-​suspect grounds Race, discrimination on grounds of See Discrimination Rancière, Jacques 17, 17 n68, 139 n93 Regulatory autonomy 338–​9 Relativism 36–​7, 37 n72, 297 Religion, discrimination on grounds of See Discrimination Semi-​suspect grounds  Producers and consumers 245 Public and private undertakings 251 Separation, art of 20, 20 n79, 35 n63, 139–​40, 319 n37 See Spheres Sexual orientation, discrimination on grounds of See Discrimination Shared meanings See Shared understandings Shared understandings  Concerns about ascertainment 30, 32, 36, 63, 67–​8, 137, 194, 197, 277, 281–​2, 289, 291, 297–​8, 302, 316, 318–​19, 321, 349 Definition 24 Explained 28 Respecting of  As second tenet of complex equality 24 Risk of cruelty/​barbarity 37, 297 Role in complex equality of 24 Role of Forum See Forum, the Small inequality 25, 48–​9, 66, 82, 250, 256 n54, 271, 273, 320, 340, 347 Definition 25 Luck as 271

376 Index Specialised courts  Need for 282 Spheres  Separation of  As first tenet of complex equality 23 Taxation  Discriminatory 10 Double 202 TFEU 9 n34, 9–​10, 12, 31, 57 n52, 57 n56, 57–​8, 69, 78 n120, 78 n123, 78–​9, 79 n130, 80 n137, 86 n154, 86–​7, 88 n168, 89 n172, 92–​3, 102 n228, 118, 145 n126, 162 n196, 168 n228, 199, 199 n8, 199 n9, 199 n10, 212, 215, 222 n111, 244 n21, 244–​5, 246 n26, 246–​7, 251, 257 n57, 270 n90, 271 n96, 276, 280, 285 n37, 287 n41, 339 n139 Article 19 of. See ‘Article 19’ grounds, the  Provisions giving rise to semi-​suspect grounds  Consumers and producers, equality of guaranteed by Article 40(2) See Semi-​suspect grounds Public and private undertakings, equality of guaranteed by Article 106 See Semi-​suspect grounds Transsexualism 53, 161, 167–​9, 176–​7 Treaty of Amsterdam 27 n27, 93 n196, 102 n228, 118, 144, 246 n26 Insertion into EC Treaty of Article 13 by 118 Treaty on the Functioning of the European Union 102 n228, 244 See TFEU Tyranny See Dominance Prohibition of  As third tenet of complex equality 24

Uncertainty 8 n28, 189 n301, 270, 313–​14 Undertakings  Private See Semi-​suspect grounds Public See Semi-​suspect grounds United States, the 22, 48, 106, 204, 219, 240, 242, 290 Voluntary choices, pre-​distributive making of 49 Walzer, Michael 1, 14 And complex equality 22–​45 And industrial democracy 50 n18 And liberalism 17 n68, 332–​4 And multicultural empires 199 n7 And particularism 19, 41, 128, 213, 235, 280, 348 And the ‘communitarian critique’ 333 And the ‘connected critic’ 283, 316 And the art of separation 19–​20, 35 n63, 139–​40 As applied by Bosniak 204 As applied by Okin 31–​34, 107–​108 Critique of courts 300 n40 Disagreement with Dworkin 25 n19, 30 n37, 37 n72, 303 n55, 326 n70 Reasons for choosing 18–​21 Walzer’s relativism 36–​41, 297 Walzerian society 10 Weakness on positive action 115–​117 Walzerian test for equality  Counter​arguments to See Expert evidence, See Judicial Interventionism, See Legislation, See Specialised courts Zuccotti Park/​Occupy Wall Street 17 n64