Justice and Egalitarian Relations 2020043862, 2020043863, 9780190084240, 9780190084264

Citation preview

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations

OX F O R D P O L I T IC A L P H I L O S O P H Y General Editor Samuel Freeman, University of Pennsylvania Oxford Political Philosophy publishes books on theoretical and applied political philosophy within the Anglo-​American tradition. Te series welcomes submissions on social, political, and global justice, individual rights, democracy, liberalism, socialism, and constitutionalism. N. Scott Arnold Imposing Values: An Essay on Liberalism and Regulation Peter de Marnefe Liberalism and Prostitution William J. Talbott Human Rights and Human Well-​being Iris Marion Young Responsibility for Justice Paul Weithman Why Political Liberalism? On John Rawls’s Political Turn Aaron James Fairness in Practice: A Social Contract for a Global Economy Margaret Moore A Political Teory of Territory

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Alan Tomas Republic of Equals: Predistribution and Property-​Owning Democracy Dan Moller Governing Least: A New England Libertarianism Christian Schemmel Justice and Egalitarian Relations

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations C H R I S T IA N S C H E M M E L

1

3 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2021 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Library of Congress Cataloging-in-Publication Data Names: Schemmel, Christian, 1977– author. Title: Justice and egalitarian relations / Christian Schemmel. Description: New York : Oxford University Press, 2021. | Series: Oxford political philosophy | Includes bibliographical references and index. Identifers: LCCN 2020043862 (print) | LCCN 2020043863 (ebook) | ISBN 9780190084240 (hardback) | ISBN 9780190084264 (epub) Subjects: LCSH: Equality. | Social status. | Social justice. | Discrimination—Moral and ethical aspects. | Liberalism. Classifcation: LCC HM821 .S24 2021 (print) | LCC HM821 (ebook) | DDC 303.3/72—dc23 LC record available at https://lccn.loc.gov/2020043862 LC ebook record available at https://lccn.loc.gov/2020043863 DOI: 10.1093/​oso/​9780190084240.001.0001 1 3 5 7 9 8 6 4 2 Printed by Integrated Books International, United States of America

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

For Miriam, Francesco, and Sara

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Contents Acknowledgements 

ix

PA RT I .   L I B E R A L R E L AT IO NA L E G A L I TA R IA N I SM 1. Justice and Egalitarian Relations 1.1. 1.2. 1.3. 1.4.

3 5 10 18

2. Distributive and Relational Equality

22

3. Liberal Non-​Domination

55

4. The Demands of Liberal Non-​Domination

94

2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 2.7.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Introduction Liberal Social Justice and Relational Egalitarianism: Te Project  Plan of the Argument  Relational Equality When and Where? 

3

3.1. 3.2. 3.3. 3.4. 4.1. 4.2. 4.3. 4.4. 4.5. 4.6.

Introduction Distributive Egalitarianism: Currency Teories of Equality Five Ways of Treating People A Preliminary Objection: Justice Is Not All Tat Matters Difering Institutional Causal Involvement Difering Institutional Attitudes Expressed in Treatment Conclusion Introduction Equality of Power and Social Cooperation Te Injustice of Domination Republican Conceptions of Justice as Non-​Domination Introduction Domination, Narrow and Wide Te Intensity of Non-​Domination Why Not Just Maximize (Equal) Non-​Domination? Realizing Liberal Non-​Domination Conclusion: Distributing Non-​Domination?

5. Relational Equality beyond Non-​Domination 5.1. 5.2. 5.3. 5.4. 5.5.

Introduction Te Insufciency of Non-​Domination Pluralist Social Egalitarianism  Relation-​Sensitive Metrics Conclusion

22 23 27 29 31 38 53 55 59 63 78

94 95 101 110 115 123

127 127 129 132 144 161

viii Contents

6. Social Status, Self-​Respect, and Opportunity 

6.1. 6.2. 6.3. 6.4. 6.5.

162

Introduction  162 Social Status Norms  164 Te Diverse Efects of Status Norms  173 Te Injustice of Status-​Induced Opportunity Loss  188 Conclusion: Treating Others as Inferior, With and Without Norms  193

PA RT I I .   I M P L IC AT IO N S 7. Political Equality 

199

8. Distributive Justice 

232

9. Health Inequality 

258

Conclusion 

288

Bibliography  Index 

301 315



Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.





7.1. 7.2. 7.3. 7.4. 7.5. 8.1. 8.2. 8.3. 8.4. 8.5. 8.6. 8.7. 9.1. 9.2. 9.3. 9.4.

Introduction  Two False Starts  Te Grounds and Shape of Political Equality  Restricting Political Equality  Conclusion: Dworkin and Democracy  Introduction  Relational Egalitarianism and Distributive Justice: Te Problem  Justifying Distributive Inequality  Distributive Inequality, Domination, and Political Inequality  Social Status and Distributive Inequality  Relational Equality and Equality of Opportunity  Conclusion  Introduction  Health and Social Cooperation  Unjust Health Inequalities  Objections: Undercoverage, Exclusion, and Discrimination? 

C.1. Challenging Liberal Relational Egalitarianism  C.2. Applying Relational Egalitarianism  C.3. International Justice and Beyond 

199 201 208 218 225 232 233 236 242 247 252 256 258 261 266 282

288 290 294

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Acknowledgements Tis book has changed shape and content so many times that it is a real challenge to encompass in my thanks all those who contributed to its development. Tanks for written comments, or extended personal conversations, on part of this manuscript, or on material that led up to it, go to Sara Amighetti, Peter Balint, Rainer Bauböck, Barbara Buckinx, Ian Carter, Emanuela Ceva, Francesco Chiesa, Richard Child, Stephanie Collins, Christine Chwaszcza, Julian Culp, Steve de Wijze, Holger Döring, Rainer Forst, Carina Fourie, Anca Gheaus, Sarah Gof, Stefan Gosepath, Robert Jubb, Andrew Lister, Daniel McDermott, Emily McTernan, David Miller, Darrel Moellendorf, Martin O’Neill, James Pattison, Tomas Pogge, Miriam Ronzoni, Andrea Sangiovanni, Rainer Schmalz-​ Bruns, Fabian Schuppert, Shlomi Segall, Liam Shields, Cynthia Stark, Hillel Steiner, Adam Swif, Patrick Tomlin, Laura Valentini, Bas van der Vossen, Kristin Voigt, Ivo Wallimann-​Helmer, Jo Wolf, two anonymous referees for Politics, Philosophy & Economics, two anonymous referees for Social Teory and Practice, and two anonymous referees for Oxford University Press, in particular for their excellent suggestions on how to restructure parts of the manuscript. Tanks also to my editor Peter Ohlin, and Sam Freeman, for their patience in waiting for it, and very helpful advice in the fnal stages. I have benefted from conversation with, and probing questions by, too many others to list here, including audiences in Amsterdam, Bad Homburg, Baltimore, Bergen, Bologna, Braga, Florence, Frankfurt, Göttingen, Jerusalem, London (UCL, KCL, and LSE), Manchester, Oxford, Pavia, Southampton, York, Warwick, and Zurich. Special thanks go to Miriam Ronzoni, for all her support, and to Francesco and Sara. While their arrival substantially delayed the birth of this book, they kept reminding me that there are much more important things in life. Parts of c­hapters  3, 4, and 6 were written during a Senior Research Fellowship at the Centre for Advanced Studies ‘Justitia Amplifcata—​ Rethinking Justice’ in Bad Homburg/​Frankfurt and Berlin. My thanks go to the Directors of the Centre, Stefan Gosepath and Rainer Forst, and to all of its staf.

x Acknowledgements

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Chapter  2 is an adapted and extended version of ‘Distributive and Relational Equality’, Politics, Philosophy & Economics 11 (2012), 123–​148. Chapter 8 is an adapted and revised version of ‘Why Relational Egalitarians Should Care About Distributions’, Social Teory and Practice 37 (2011), 365–​ 390; very small parts of it also appear in ­chapter 6. Chapter 6 (section 6.3.2) draws extensively on material frst published in ‘Real Self-​Respect and Its Social Bases’, Canadian Journal of Philosophy 49 (2019), 628–​651. Small parts of ‘Social Equality—​or Just Justice?’, in Social Equality—​On What It Means to Be Equals, eds. Carina Fourie, Fabian Schuppert, and Ivo Wallimann-​ Helmer (New York: Oxford University Press 2015), pp. 146–​166, appear in ­chapters 5 and 6.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

PART I

LIBE R A L R E L AT IONA L E GA L ITA R IA NI SM

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

1 Justice and Egalitarian Relations

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

1.1. Introduction Relations of unequal power and domination, and hierarchies of social status, characterize the societies we live in. Tey constitute the main obstacle to achieving a society of equals, and one that is diferent from economic inequality, even though the two ofen go hand in hand. Even if all had equal economic resources, this would not amount to a society of equals as long as some had arbitrary power over others and inegalitarian norms of social status deprived some of fair opportunities to acquire and enjoy social esteem for their traits, skills, pursuits, and projects. Tis book provides a theory of how concern for egalitarian relations of non-​domination and social status can be incorporated into a liberal conception of social justice, and lay a strong claim to constituting the most demandingly and stringently egalitarian components of such a conception. Its aim is to convince liberals that they should be relational egalitarians, and relational egalitarians that they should be liberals—​and to convince those who remain unpersuaded by well-​rehearsed philosophical arguments that justice requires people to get equal amounts of some good to be distributed, such as resources, or welfare, or equal opportunity to acquire these, that there is a better conception of egalitarian justice which they have strong reason to consider, and hopefully sign up to. Hardly anybody denies that people are fundamentally each other’s moral equals; and not many deny that this basic equality grounds a claim to stand, at the very least within the society of which one is a member, as a social and political equal. Taking these commitments seriously requires us to put stringent egalitarian constraints on the structure of important social and political relations of power and status, and thereby to make sure that the diferent social and economic outcomes these generate are signifcantly egalitarian, too—​or so the argument will seek to show. Te book argues that expressing respect for the freedom and equality of individuals in social cooperation requires stringent protections against domination. It develops a substantive, liberal conception of non-​domination Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0001

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4  Liberal Relational Egalitarianism tied to cooperation among free and equal persons, and argues that non-​ domination is a particularly urgent, but not the only, concern of social justice. Both its substantive nature and its account of the place of its demands within relational egalitarianism set it apart from, and render it superior to, neo-​republican accounts of non-​domination, which seek to reduce social justice to questions of non-​domination. It also develops an account of the wrong of norms of status inequality, showing how status-​induced foreclosure of important social opportunities is a social injustice in its own right, in addition to the role of status inequality in enabling domination, and the threats it poses to individuals’ self-​respect. It applies these core requirements of liberal relational egalitarianism to political, economic, and health justice, and works out its implications for these domains: they amount to demands for far-​reaching forms of equality in all three of them, which can rarely, if ever, be overridden by competing concerns. Tis systematic account and defence of liberal relational egalitarianism builds on an in-​depth engagement with several diferent literatures: literature on the kind of equality, if any, demanded by justice, which has mostly taken the form of a debate between relational and distributive egalitarians; literature about the nature of the value of social equality, which, in part, precedes the former, questions whether that value is best, or exclusively, accounted for by justice, and holds that reliance on diferent, fundamental values of equality and community is necessary; and neo-​republican literature about the nature and demands of non-​domination, and its connection to justice, which ofen proceeds by seeking to distance neo-​republicanism from liberal egalitarian theories of justice, while nevertheless claiming broadly liberal credentials. Tese literatures have mostly developed in isolation from each other. Recent scholarship has started to bring them into contact, but as of yet no systematic attempt to integrate them has been made. Te book takes full account of the signifcant advances recently made in each of them, brings them together, and thereby seeks to advance all of them. Tis completes the frst overview of the project. In order to give a fuller introduction to its key features, ambitions, and motivations, the remainder of this chapter proceeds as follows. Section 1.2 gives more depth to the theoretical challenges which justice-​based relational equality faces, by showing how it seeks to weave together two concerns—​social justice and social equality—​ which have, for the most part, been treated as separate matters in the history of modern political thought. Section 1.3 outlines how the book answers these challenges, by contrasting its ambitions to that of the three main families

Justice and Egalitarian Relations  5 of theoretical interlocutors just mentioned, and providing a more detailed outline of the arguments of each chapter and of how they contribute to the overall argument of the book. Section 1.4 concludes by delineating the scope of the theory to be developed, and the kind of social scenario it is intended to yield requirements of social justice for: in the frst instance, for a society characterized by a reasonably well-​functioning institutional structure capable of organizing social cooperation by enabling and sustaining a complex division of labour. Tis does not rule out that it can be extended and applied to matters of international and global justice, too, but that extension is a task for another day.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

1.2.  Liberal Social Justice and Relational Egalitarianism:  Te Project Political thought about justice, and later social and distributive justice, and political thought about relations of social equality have traditionally, and in some respects up to the present, been separate, and have travelled on different roads. Briefy taking stock of these diferences puts some of the main challenges, as well as prospects of relational equality as a central requirement of social justice, into sharper focus. If successful, such a view would be able to account properly for the objectionable nature of key inegalitarian social relations—​for their injustice—​and to explain, from this starting point also, when, and why, which kinds of unequal distributions of goods are unjust. However, unlike more traditional social egalitarian views, it would be able to do so while remaining a liberal theory which does not tell individuals how to lead their own lives, which does not put forward a vision of the good communal life directing and dictating the content of individual conceptions of the good, rather than merely constraining them in order to continuously safeguard everybody else’s freedom and equality. Te philosophical project of this book is thus that of showing that liberal social justice can be home to a demanding conception of egalitarian relations, and that such a conception can be arrived at from premises thinner than those traditionally endorsed by many social egalitarians—​whether or not such a conception also has good prospects of political success, in the sense of being capable of persuading as many people as possible, whatever their fundamental value outlooks, to sign up to it. A conception of egalitarian relations as a matter of social justice is not consonant with most of the history of

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6  Liberal Relational Egalitarianism political thought, in which the latter has generally been restricted to the protection, and later also distribution, of property, and the former connected to encompassing visions of the good society as one where positive relationships of equality prevail. Yet a connection of both within a specifcally liberal conception of social justice is an attractive and worthwhile prospect. It is a novel road in political thought, worth trying out. Te idea of social, or distributive, justice, according to which the primary task of justice for a state is to provide all individuals—​at least within a given society—​with both the liberty and the means needed to be a full participant in that society, is a relative latecomer in the history of Western political thought. Te demand that the state not only should ensure that people enjoy personal security and the protection of their property, but also should set up the societal order in such a way as to take appropriate care of everybody’s material situation frst rose to prominence in the socialist thinking of the nineteenth century which accompanied the acceleration of industrialization in key Western countries such as Britain, France, and, with some delay, Germany.1 Enlightenment thinking, the French Revolution of 1789, and the rapid displacement of traditional orders by industrialization made it possible to see the basic political and economic order of society as itself an appropriate object for deliberate political action and design, rather than simply having to be taken over by tradition. Tis is the birth of the contemporary concept of social justice, or at least of all its necessary presuppositions, as ranging comprehensively over the set-​up of a given society, being primarily entrusted to the state, and demanding not only that everybody’s personal freedom and property be preserved, but also that everybody be enabled, by that order, to command an appropriate share of resources (understood in the widest sense of the term) so as to be an active and free participant in it.2 Tought about social equality as a value governing social and political relations predates these developments considerably—​but without any clear

1 Te idea that the state may, and ought to, use coercive taxation to guarantee a basic minimum of material means for everybody appears somewhat earlier, in Kant’s work; Kant 1996 [1797], p. 101. However, Kant did not seem to regard this task as one of ‘distributive justice’; Kantian ‘distributive justice’ refers, somewhat idiosyncratically, to the determination and securing of rights (mainly to property) by enforceable and impartially administered public law, as opposed to a state of nature; ibid., pp. 86, 90. Most other thinkers understood distributive justice, in the Aristotelian tradition, to be about the distribution of ofces and privileges among those citizens of a polity, who, through personal excellence, merited them; Aristotle 2000, Book V. 2 See Fleischacker 2004, pp. 105, 125; Johnston 2011, pp. 3f and ch. 7; and Fleischacker 2004, chs. 1 and 2, on how, before, basic material provision was generally regarded as a matter for duties of charity, which did not translate into rights of recipients.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations  7 connection to justice. Renaissance writers such as Tomas More presented utopian societies characterized by simplicity of lifestyle and social equality as withering social criticism of what they saw as the decadence and corruption of the social and political life of their times.3 Tey did not argue for their ideals as conceptions of justice, as seeking to specify how social orders, or governments, ought to treat individuals by way of right. Later egalitarian movements, like the Levellers in seventeenth-​century England, fused Christian belief in equality before God and specifc egalitarian visions of political life. Te Diggers (or True Levellers), led by Gerrard Winstanley, especially stressed the connection between calls for greater political equality and greater socio-​economic equality as a necessary condition for it.4 Rousseau argued, on grounds of liberty, for a model of radical direct democracy, and for the claim that substantial social and material inequalities pervert societal relations5 and undermine the viability of democracy.6 In these movements in political thought, there was a development from utopian visions towards a conception of equality understood specifcally as a social and political value, tightly connected to republican and democratic forms of government, and viewed as both desirable, because liberty-​ preserving and -​enhancing, and, at least in principle, feasible. However, there was still little connection between such a value of equality and concern about socio-​economic inequality as a matter of justice. Rousseau’s view does make a case for an intrinsic connection between law-​making through egalitarian and direct democracy and the preservation of every citizen’s civil liberty, especially through the protection of property. But he regarded putting limits on socio-​economic inequality as important mainly to ensure proper democratic governance—​not because failing to do so also means negating individuals their fair share of the social product, in addition to political inequality.7 Later on, socialist thought transformed and widened such an ideal of equal political standing to a more encompassing and general ideal of living 3 More 2003 [1516]. Utopia also has some very inegalitarian features, such as a form of slavery for convicted criminals; ibid., p. 82. 4 Winstanley 1973; for discussion, see Fleischacker 2004, p. 43 n. 52. For recent discussions of Leveller thought about work relations and the proper place of markets, see Anderson 2017a, and the critical responses to her account included there. 5 Rousseau 1984 [1755]. 6 Rousseau 1968 [1762]; see especially p. 96: ‘[N]‌o citizen shall be rich enough to buy another, and none so poor as to be forced to sell himself.’ How much material inequality exactly was meant to be ruled out by this is an open question. 7 Rousseau nowhere discusses distributive justice in any detail, and understands it, along Aristotelian lines (see fn. 1) as requiring distributions according to personal merit (as measured, according to him, by services rendered to the state); 1984 [1755], p. 171.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

8  Liberal Relational Egalitarianism as social equals.8 It emphasized solidarity and communality, and refused to accept personal competition and confict as inevitable features of social life, against much of the liberal tradition, which saw the main task as that of managing, rather than eradicating, these features. Important parts of the socialist movement did regard the just distribution of the fruits of social labour as a primary socialist, or social democratic, concern. Marx subjected this focus on distribution to withering criticism for, in his view, failing to realize and stress how revolutionizing the modes of societal production would bring about a wholly diferent, much more desirable, set of social relations which would characterize the communist society to come.9 ‘Ofcial’ Marxist socialism eschewed framing this ideal in terms of an alternative conception of social justice, and generally privileged social scientifc analysis and theory of history over moral argument. However, ‘ethical’ (as opposed to ‘scientifc’) socialists did advocate a relationship ideal of social and political equality on moral grounds. Particularly notable is the work of the English socialist R. H. Tawney, who put forward an ideal of an egalitarian society, ultimately underwritten by Christian belief in the equal and infnite value of all individuals, as a society not characterized by hierarchical and divisive norms of social status, or dominatory power relations.10 Such a position is, in some respects, a forerunner of contemporary social egalitarian positions,11 and it could also be regarded as the basis for a distinct conception of social justice. However, that would be an ‘ethically socialist’ conception all the way down. It would rest on a perfectionist, comprehensive vision of solidarity among equals as the chief human good to be achieved in social life.12 Tat is not in line with a more liberal understanding of justice, which restricts itself to demanding that everybody’s rights be respected, and justifed claims honoured, because this is what individuals’ status as free and equal members of society demands, whatever more particular view of the human good, if any, we endorse. Much of contemporary liberal egalitarian theorizing has, in the wake of Rawls’s work, sought to justify egalitarian concern without any appeal to such perfectionist, communitarian visions of the individual and social good. Its distinctive reinterpretation and development of liberal thought stresses

8 For a reasonably representative early socialist view, see Proudhon 2011 [1846]. 9 Marx 2000 [1875]. See section 2.2.

10 Tawney 1931; for an overview of the main features of Tawney’s egalitarianism, see Wolf 2013. 11 See especially the ‘pluralist social egalitarian positions’ analysed in c ­ hapter 5. 12 See Wright 1987, p. 138.

Justice and Egalitarian Relations  9

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

the connection between an account of moral personhood which is intended to be more neutral, and broadly shareable, spelled out in terms of possession of the two fundamental moral powers of a capacity for a sense of justice, and the capacity to develop, pursue, and revise a conception of the good,13 and the task of social justice, which requires ensuring that each member of society is both free from obstruction by others and enjoys sufcient, and not too unequal, material means to live her life according to her own conception of the good. Tis is the bridge between liberal insistence on respect for personal autonomy and egalitarian conceptions of social justice in much of contemporary liberal egalitarianism: it is the former that ultimately underpins the latter, not any particular perfectionist vision of the individual and social good of living together as social equals, beyond continuously ensuring (roughly) equal substantive liberty to live one’s life as one sees ft.14 Initially, in the Teory of Justice, Rawls sought to spell out this link on the basis of a less well-​developed conception of the person, by drawing on a primarily distribution-​centred perspective on social justice, and emphasizing the question of which goods individuals ought to be entitled to get in a suitably equal manner. Indeed, the connections Rawls established between the requirement of respecting personal autonomy and distribution-​oriented welfare economics accounted for a signifcant part of the inaugural success of the theory.15 Other liberal egalitarians, most notably Dworkin in his 13 For a representative formulation of the second moral power and its implications, see Rawls 1996, p. 72: ‘Citizens think of themselves as free in three respects: frst, as having the moral power to form, to revise and rationally to pursue a conception of the good; second, as being self-​authenticating sources of valid claims; and third, as capable of taking responsibility for their ends.’ 14 A note on neutrality, non-​perfectionism, and the shareability of political ideals: a neutral, non-​ perfectionist view insists that the use of political power in pursuit of justice must not be justifed on the basis of its furthering any particular conception of what constitutes a good (individual or communal) life; a political, non-​comprehensive view further aims to deliver such requirements of justice on the basis of a set of moral ideas (about personhood, and society) which form, or are at least capable of forming, the basis of an overlapping consensus between diferent comprehensive philosophical doctrines about the good life, the nature of society, and the grounds of personhood; see Rawls 1996, and Quong 2010, pp. 15f, for a helpful taxonomy of diferent liberal views. Te view developed in this book certainly aims to be non-​perfectionist and compatible with a broad array of diferent, and potentially competing, conceptions of the good, and in this sense, to be broadly shareable. It also draws on ideas about personhood and the nature of society that have been developed, and refned, by Rawls in his search for a political conception of justice in the sense outlined earlier. However, it is not itself pitched, and developed, as a political ideal in this sense. It would certainly be nice if liberal relational egalitarianism could be the object of consensus among fundamentally diferent comprehensive philosophical doctrines, but no argument to this extent will be made, and its success does not crucially depend on it. 15 Rawls 1999a. In later works, Rawls’s focus shifs more clearly from the question of just distributions to the question of how free and equal members of a liberal democratic society should relate to each other; see Rawls 1996 (and 2001). It therefore ofers various points of departure for

10  Liberal Relational Egalitarianism contribution to the ‘equality of what’ debate, tied their ideal of justice even closer to the recipient-​oriented focus characterizing modern welfare economics.16 Much of contemporary liberal egalitarianism thus interpreted the political value of egalitarian justice in almost exclusively distributive terms, and therefore largely eclipsed the question of appropriately egalitarian social and political relations. It is this focus on distribution whose shortcomings will be analysed in the next chapter, and will be used as a starting point for the development of a liberal conception of egalitarian relations as matters of social justice. Yet, as seen, much of distributive egalitarianism is motivated by the aspiration to remain true to liberal conceptions of the person and of society, according to which it is up to individuals themselves to decide what is good for them, and up to them collectively to enable each other to do so (roughly) equally. Tis is an aspiration worth keeping.17

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

1.3.  Plan of the Argument Te aim of the book is constructive. It is to put forward a proposal for a liberal conception of social justice which accords centre stage to egalitarian relations, and to bring it into dialogue with rival theories of social justice and equality. Terefore, its aim is not to refute these theories, but to show what they are missing. As noted, three kinds of theories serve as primary interlocutors; it is worth introducing them in somewhat more detail to pin down the main contrasts with liberal relational egalitarianism, before outlining how the argument of the book will develop and defend this contrasting position. Te frst are views which regard the main task of theories of social justice as that of specifying a just distribution of goods. Te most important members of this family are luck egalitarian theories, whose main demand is to shield individuals from the unequal impact of all factors which are beyond relational egalitarian arguments. Some of these will be taken up, and discussed, in subsequent chapters. However, while the view to be developed in this book is certainly broadly Rawlsian (see fn. 14), Rawls scholarship is not its focus. 16 Dworkin 2000, chs. 1 and 2, which specify an auction of resources as the fairest initial situation for distributive justice, and draw on the works of Léon Walras. 17 Similar ambitions are shared by some contemporary neo-​republicans—​but not achieved, as ­chapters 3 and 4 will demonstrate.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations  11 their own control (‘brute luck’) on their lives, and to compensate them, up to equality, for the impact of those factors whose diferential impact cannot be fully neutralized.18 Luck egalitarianism is based on a specifc understanding of fairness: fairness requires equal distributive outcomes (unless individuals are themselves responsible for divergences from equality); luck egalitarians disagree among themselves about the proper currency for these outcomes (resources, opportunity for welfare, capabilities, or something else). In recent debates, the idea of relational egalitarianism mainly gained traction as a source of objections to luck egalitarianism.19 However, while there has been some positive development of diferent versions of the relational egalitarian ideal,20 there is as of yet no worked-​out proposal for a theory of social justice encompassing egalitarian relations. And there is no proposal at all as to how diferent kinds of them might ft within a specifcally liberal framework for social justice. For all the criticisms they received, distributive egalitarians, and luck egalitarians in particular, are right to want to know much more about that alternative; and especially about whether it really conficts and competes with their ideal as much as relational egalitarians tend to claim. Perhaps luck egalitarianism can itself sufciently account for relational egalitarian considerations.21 Piling up general objections to luck egalitarianism achieves little in answering these challenges; our focus will be on what exactly it is about relations of equality and the reasons for which they are demanded that distributive and luck egalitarian theories cannot capture. Te demands of liberal relational egalitarianism are themselves based on fairness, but on a diferent understanding of it. Ensuring fairness in the terms of social cooperation does not reduce to aiming at the right distributive outcomes, of whatever kind, but requires focusing on individuals’ position in relations of power and status, and putting stringent constraints on their structure, of the kind that will be outlined in a moment. Te second set of interlocutors are, as noted, neo-​republican theories of liberty, justice, and democracy, which have fourished over the last decades, 18 See Dworkin 2000, chs. 1 and 2; Cohen 1989; and Arneson 1989 for the probably best-​known developments of the luck egalitarian ideal. 19 See especially Anderson 1999 and 2010b; and Schefer 2003a, 2003b, and 2005. 20 See, for example, Fourie et al. 2015; and Mason 2012. 21 For a recent exploration of diferent versions of the relational egalitarian ideal which largely afrms its compatibility with luck egalitarianism, and argues that at least some such versions can be understood as variants of distributive egalitarianism, see Lippert-​Rasmussen 2018b (and further section 5.4).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

12  Liberal Relational Egalitarianism and ofen present themselves as alternatives to liberal egalitarianism—​in part because they draw on diferent sources of inspiration in the history of political thought. Te view developed in this book and neo-​republicanism have some signifcant overlaps.22 However, they construe the concept of domination diferently, and one of the main contentions of this book is that suitably wide-​ranging, equal, and deep protection against domination over central areas of individual lives can be better justifed on grounds of a liberal conception of society and the person; a conception which main proponents of neo-​republicanism, such as Pettit,23 explicitly seek to eschew. Because of its broader focus on the fair structuring of social cooperation, it is also able to account for other social egalitarian demands which neo-​republicans tend to ignore or neglect. Te third are theories of social equality which regard equality in social relations as a positive ideal of social life, as something that contributes value to individual lives, or instantiates a good society, or community24—​and not as requirements of social justice, or fairness. Tese views are the most clear-​cut contemporary successors of the historical ideals of social and political equality as a value apart from justice which were briefy surveyed in the preceding section. Here, we will see that the implications of a liberal, justice-​based, view of social equality focusing on the elimination of certain types of relational inequalities, rather than on celebrating the positive personal and impersonal value of social equality, are so far-​reaching that there is good reason to think that reliance on such a positive ideal is dispensable—​ especially as it incurs its own problems.25 However, just as with distributive and luck egalitarianism, the aim is not a comprehensive refutation of neo-​ republicanism and free-​standing social egalitarianism, but to use them as counterfoils for working out, and defending the credentials of, liberal relational egalitarianism. Since the book aims not only to develop a reasonably complete such liberal ideal, but also to show that it has far-​reaching and plausible implications for political and distributive justice, it is divided into two parts. Part I (­chapters 1–​ 6) works out core concepts and requirements:  the expressive concept of 22 Recent scholarship has begun to examine similarities as well as diferences between neo-​ republicanism and social egalitarianism; see especially Schuppert 2015; Laborde and Garrau 2015; Anderson 2017a; and Kolodny 2019. 23 Pettit 1999a, 2012, and 2014. 24 See, for example, Miller 1998; Mason 2012 and 2015; and Cohen 2009. 25 A note on terminology: throughout this book, ‘relational equality’ and ‘social equality’ are used interchangeably.

Justice and Egalitarian Relations  13 justice that the ideal relies upon; its interplay with liberal conceptions of society and the person; the concept of domination and its rightful place within liberal social justice; the resulting array of requirements to equal protection against domination; and, fnally, the role of egalitarian norms of social status in warding of domination, securing self-​respect, and keeping important social opportunities open to all. Part II (­chapters 7–​9) develops the implications of these relational egalitarian requirements for three domains of social justice, broadly conceived. Te frst two are rights to participate as an equal in institutions of political decision-​making, and requirements of distributive justice governing the distribution of income, wealth, and opportunities to attain generally favourable social positions in society. Tese domains merit a central place: any conception of justice that is to count as a serious contender has to have determinate and plausible implications for at least these two areas, so it has to be shown that this is the case for liberal relational egalitarianism. To these, the second part adds an investigation of its implications for health and healthcare. Tis is a particularly important topic, as one might worry that an ideal of relational equality is too narrowly concerned with the quality of social and political relations to be able to satisfactorily capture our concern with these goods, and with health inequalities, in particular. If it can, however, it has a strong claim to being able to serve as a reasonably complete conception of the dimensions in which social justice should be stringently and demandingly egalitarian.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Part I:  Liberal Relational Egalitarianism Chapter  2 develops the expressive perspective. It shows that the way social and political institutions treat individuals and groups is of irreducible importance to justice, and that this consideration cannot be satisfactorily accounted for by more traditional distributive theories of egalitarian justice, which focus on according individuals equal shares of justice-​relevant goods; paradigmatically goods such as resources, opportunity for welfare, or basic capabilities. It makes a case for the special relevance for justice of the attitudes expressed by institutions in the treatment of those subject to their power, as that expression constitutes its meaning. Tat meaning is particularly salient where the treatment gives rise to, or shores up, power and status hierarchies between diferent individuals and groups.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

14  Liberal Relational Egalitarianism While ­chapter 2 makes clear that power relations between individuals, and institutions and individuals, are a central subject of social justice, the expressive perspective does not, by itself, yield determinate answers to the question of which kind of power relations exactly are required by justice, and which are unjust. Chapters 3 and 4 tackle this question. Chapter 3 connects the expressive perspective to a liberal framework for social justice aiming at fair cooperation between individuals as free and equal, and derives a resulting liberal conception of non-​domination from it, as demanded by respect for such freedom and equality. Domination consists of the asymmetrical capacity of one agent to arbitrarily interfere in the choices of another; interference is arbitrary when it is not forced to respect others’ prima facie relevant claims arising out of cooperation. Tis conception of domination is indebted to recent neo-​republican scholarship, but it improves neo-​republican frameworks for theorizing non-​domination as a matter of social justice, in two important respects. First, it shows that non-​domination is not the only concern of relevance to social justice, but only one among others—​albeit one with a justifed claim to priority. Second, it yields a substantive conception of non-​domination which gives a more determinate and plausible account of which choices and interests protection against domination has to range over, and shows how the liberal core ideas of personhood and society are essential for ensuring requirements of far-​reaching, and intensive, protection. Chapter  4 develops these requirements of liberal non-​domination. It shows how exactly they extend to protection against dominatory groups as well as against power relations which are not mediated by any kind of authority. It then demonstrates how diferent choices and interests, such as those falling under the basic liberties, connected to intimate personal relationships, or at stake in resource-​intensive programs and policies to enhance life options, call for diferent kinds and thresholds of protection, and how the liberal conception is better placed to account for this than republican conceptions demanding the maximization of non-​domination (independently of their disagreements about the construal of the concept of domination and its scope, tackled in the preceding chapter). It also requires that all protection be itself appropriately respectful of people’s moral agency. Te chapter concludes by demonstrating how these liberal requirements of non-​domination give a wide policy mandate to combat domination not only by setting up the right kinds of formal institutions, but also by fostering a societal ethos, and social norms, of non-​domination, and argues that liberals have no good reason to be worried by such a wide mandate.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations  15 Chapters 5 and 6 then take on the extension of liberal relational egalitarianism beyond non-​domination. Chapter 5 compares the liberal approach to two important and powerful rival views which seek to argue, and account for, social and political relations of equality in a diferent way—​and could, if successful, cover a wider array of such relations. Te frst is the pluralist, free-​standing social egalitarian approach, which regards the value of social equality as going beyond, and sometimes perhaps even conficting with, social justice. Te second are relation-​sensitive distributive egalitarian views, which enlarge the metrics of distributive justice beyond the traditional conceptions criticized in ­chapter 2, so as to incorporate fair shares of the various goods at stake in egalitarian relations, because of the distinctive contribution to the individual good, or to opportunities for it, that these are supposed to make. It shows that the best versions of pluralist social egalitarianism fail to give liberals a mandate to seek to shape society according to its demands, while relation-​sensitive distributive conceptions either fail to capture what is distinctive about relational goods, or fail to yield demands which are recognizably egalitarian. Both results corroborate the case for the liberal approach to relational equality based on respect for freedom and equality in cooperation. However, the analysis also confrms that this conception must demand more than non-​domination, and, in particular, be able to account for the wrong of inegalitarian norms of social status, as well as the connection between (in)egalitarian relations and the crucial psychological good of self-​respect. Chapter 6 undertakes the required extension. It develops an account of esteem-​based norms of social status, and analyses the kinds of injustices that inegalitarian such norms may engender, or constitute. Tere are three of them. First, norms of social status can enable, or aggravate, domination, because there is ofen a tight link between perceived authority and social status. Second, they can harm self-​respect. Where they do, this is a particularly stringent reason to combat them. However, closer analysis of self-​ respect and its crucial role in underpinning individual autonomy reveals that not all inegalitarian norms of social status can be classifed as threats to self-​respect without threatening precisely that role: self-​respecting individuals are capable of resisting at least some threats to their self-​worth, so requirements to protect self-​respect have to aim at shoring up this capacity, not at shielding individuals from all possible threats. Self-​respect thus yields particularly stringent requirements, but not all inegalitarian norms of social status violate them. Tird, such norms can be unjust, even when not

16  Liberal Relational Egalitarianism engendering domination, or harming self-​respect, where they deprive individuals of important social opportunities. Losing such opportunities due to norm-​coordinated, self-​sustaining disesteem by others is a threat to one’s equal standing in social cooperation which is not present when these opportunities go missing due to other causes. Tis is thus a genuinely relational egalitarian rationale for combating these norms which is both accessible to liberals and does not encounter any of the problems besetting the rival views analysed in the preceding chapter.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Part II:   Implications Chapter 7 investigates which requirements of political equality liberal relational egalitarianism gives rise to, and which status these should have vis à vis other requirements of social justice. It argues that it requires, for reasons of both non-​domination and equal respect for persons’ frst moral power of a sense of justice, a demandingly egalitarian conception of rights to political participation, along the lines of the Rawlsian requirement of the ‘fair value of all of the equal political liberties’,26 and that this requirement enjoys a qualifed priority over other requirements of justice. However, as liberal relational egalitarianism is, as demonstrated in Part I, a substantive conception of social justice, it must, in principle, be open to some restriction, should this be strictly necessary to fulfl such other requirements. Te chapter goes on to identify the conditions governing the justifcation of such restrictions. It argues that, while some special, circumscribed institutional practices such as constitutional review stand a good chance of fulflling them, they very likely rule out any ground-​level political inequality between citizens in salient real-​world circumstances. Tis is because other, egalitarian strategies for improving the quality of democratic input by citizens have to be tried frst, and such strategies will almost always be available. Te chapter concludes with a critical analysis of Ronald Dworkin’s account of political equality, which is the most worked-​out such account to be found among primarily distribution-​centred theories of justice. It shows that, despite claims to the contrary, it cannot do justice to the expressive value of democratic rights, and thus fatally underplays the importance of political



26 Rawls 1999a, p. 198.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations  17 equality. Tis strengthens the case against such distribution-​centred theories made in Part I. Chapter 8 develops the implications of liberal relational egalitarianism for the distribution of goods produced by social cooperation. It shows that there are not only strong instrumental reasons to set stringent limits to inequality of income, wealth, and opportunity, on grounds of both non-​ domination and social status, but also, contrary to what both many critics and proponents of relational equality argue, strong non-​ instrumental, expressive reasons to do so:  since participants in social cooperation are equals, all inequalities in social goods need to be justifed by justice-​relevant reasons even where they do not lead to domination between individuals, or the emergence of inegalitarian status norms. As the book delivers an account of the place of requirements of relational equality within liberal social justice, and not a complete account of the latter, it is consistent with some disagreement about other components of justice, and thus also with some disagreement about which reasons to permit distributive inequality are good enough where domination and status inequality are not at stake. However, this does not mean that the requirement is weak, but only that it awaits further confrontation with such candidate reasons for inequality. So, unlike many distributive views, such as luck egalitarianism, liberal relational egalitarianism does not object directly to any kind of material inequality between people, whatever its cause (apart from responsible individual choices). But it does undertake a concentric attack on material inequality in society as well as on its sources in power inequality, through a plurality of rationales. Chapter 9 takes on the special goods of health and healthcare. As noted, this is a particularly important area because it might be thought that relational egalitarianism sufers from a kind of ‘social relation fetishism’, and has little to say about this good, and of the good of healthcare, whose importance for individual lives is evident. Te chapter argues that it has distinctive, and plausible, implications for these goods, too. First, it yields a clear ordering of the injustice of diferent kinds of health inequalities: inequalities caused by inegalitarian relations which are independently unjust are more unjust than those caused by other social processes, which in turn are more unjust (unless justifed according to the model developed in the preceding chapter) than those not so caused. Te resulting requirements ft well with important strands in public health research on the social determinants of health. Second, the cooperation-​based framework underlying the conception also

18  Liberal Relational Egalitarianism requires universal provision of healthcare, independently of the causes of individual health defcits, and relational egalitarians need not, and should not, be committed to prioritizing patients with socially caused health defciencies at the point of delivery of treatment. Prioritizing the fght against socially caused health inequalities, in terms of public health spending, research, and overall coverage of the healthcare system is, against that, not only permitted, but required. If the arguments of Parts II are sound, they signifcantly strengthen the argument for a liberal justice-​based conception of egalitarian relations made in Part I. In political theory and philosophy, detailed analysis of what the adoption of a conception of social justice and equality would amount to, overall and in real world societies, drawing also on research in the social sciences, is an essential stage in its justifcation.27 Working out the resulting demands for concrete institutional and policy domains, about which many of us have strong political convictions, thus makes, if things go well, for a much stronger case for the conception. If things do not go well, it establishes at least that it is worth engaging and disagreeing with, because it is relevant.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

1.4.  Relational Equality When and Where? To conclude this introductory chapter, we need to delineate with some more precision the target of the relational egalitarian requirements of liberal social justice to be developed—​that is, we need to specify the context for which these requirements are meant to deliver direct and stringent guidance for social and political action. Tis clarifes, inter alia, where we should locate the theory on a spectrum of ‘ideal’ versus less ideal, or non-​ideal theories.28 As we will see, subsequent chapters develop these requirements by seeking to answer the questions of which role these should play for fairly structuring social cooperation among individuals conceived of as free and equal, and which form they have to take in order to play it. Tese kinds of questions 27 Tis kind of justifcatory methodology, according to which the justifcation of principles of political morality draws not only on moral judgements and intuitions, but also on social theories, is that of ‘wide refective equilibrium’; see Daniels 1996, chs. 1–​3. 28 Te debate about the distinction between ideal and non-​ideal theory has reached dizzying heights of conceptual sophistication, and these remarks are not in any way intended as an original contribution to it. Tey only serve to specify the scope and primary target of liberal relational egalitarianism. For two overviews, see Stemplowska and Swif 2012, and Valentini 2012; for recent criticism of the distinction, see Levy 2016; and for a prominent criticism of ideal theory as liable to serve the group interests of the privileged, see Mills 2017, ch. 5.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Justice and Egalitarian Relations  19 are natural to ask for domestic societies as we know them, characterized by at least basically functioning state institutions; that is, by an institutional order capable of coordinating diferent agents and institutions to implement a fairly comprehensive set of diferent policies governing a complex social division of labour—​and capable of reforming such societies towards more justice. But this is not just, as a matter of fact, the scenario that many readers will fnd themselves in. Such an institutional order, and division of labour, are essential for liberal social justice, for the general requirement of enabling all to fully participate in society to take the liberal form of fairly enabling all to develop, pursue, and revise their own conception of the good, against the background of a society ofering a very wide and diverse range of opportunities for doing so (­chapter 3, section 3.2). Terefore, it will be taken as the paradigm scenario in subsequent chapters, to which the resulting requirements are most directly applicable, and where they are stringent. Tis is, of course, in at least one sense, quite an ideal scenario. It may not be fully ideal because, for the presence of such an order, it is not required that all individuals within such societies comply with its demands.29 Functioning institutional orders are robust enough to deal with some, non-​trivial, amount of individual non-​compliance. Still, such conditions are by far not fulflled in all societies around the globe; nor is a fully functional international order governing the relations between diferent societies present. Developing relational egalitarian requirements for global, transnational, international (and intergenerational) justice is a task for another day. Te shape of that task as it presents itself to liberal relational egalitarians will be discussed only briefy in the Conclusion. By then, all substantive relational egalitarian requirements for the original target scenario will have been put on the table, so it is easier to devise hypotheses about how they may apply to, or have to be modifed for, these domains. Furthermore, even when we restrict our view to single societies, there are other, signifcantly less ideal scenarios for theorizing justice that we can think of. What counts as a more or less urgent matter of justice will change with these. If functioning institutions are missing, then, arguably, the frst demand of justice is to put these in place, and undertaking this efort could require, at least for a time, signifcant violations of the relational egalitarian requirements this book puts forward. Similarly, it will argue throughout that relational egalitarian requirements have a strong claim to be the most

29 ‘Full compliance’ is an important feature of Rawls’s defnition of ideal theory; 1999a, p. 6.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

20  Liberal Relational Egalitarianism stringent requirements of equality (see, for example, sections 4.3.3, 7.3, 8.4–​8.6, 9.3), compared to other egalitarian demands, such as distributive demands of no consequence for the structure of social and political relations (insofar as there is any case for these at all). But this does not mean that they cannot be trumped by some more urgent non-​egalitarian requirements, such as requirements of basic sufciency.30 However, if such reasonably favourable conditions are in place, then the contention of this book is that relational egalitarian requirements are indeed urgent requirements:  that they are action-​guiding, and identify pressing kinds of injustice in need of social and political counteraction, right here and now. Tis is illustrated also by the fact that they are, mostly, negative, reactive requirements: they condemn domination and inequality of social status, and defne a situation of their absence as just. Tey focus on a set of social and political evils with which we are familiar in contemporary societies, and call for reactions to these. Tis is another possible sense in which we can call the theory ‘non-​ideal’, and one of the reasons why it is consonant with the demands of many real-​world egalitarian movements, such as movements for gender, sexual, racial, and ethnic equality, and for the abolition of hierarchies of social class. Tese movements take the evil of unequal social and political standing as their starting point. Tey tend to focus less on describing the desirable nature of positive relations of equality. Of course, one particular substantive reason for the negative formulation of many of the demands of the liberal conception is precisely its liberal nature: once a set of demanding constraints ruling out structural inequality is in place, it is up to individuals themselves to decide how they want to conduct their relations within them (see especially section 4.3 in ­chapter 4).31 Because of this structural, and largely negative, focus, the theory also has signifcant diagnostic value:32 it specifes not only which requirements a society has to fulfl in order to instantiate relational equality, as a matter of social justice. Tese requirements can also serve, in many cases, to understand 30 For example, if some urgent needs are unfulflled, such as basic nutritional needs, and we face a choice of either fulflling these or combating some domination, then we should arguably, in many cases, choose the former even if the domination in question is more than trivial (basic needs shortfalls could, of course, be due to domination, and regularly cause further domination). 31 Some argue that it is generally easier to identify injustice than justice (see Sen 2009), and therefore to identify just social equality as absence of certain social inequalities; Wolf 2015, pp. 213f. Identifying severe injustice certainly seems easier than identifying precise justice, at least most of the time. I remain agnostic about whether there is anything in a negative focus as such that makes such identifcation easier. 32 For an account of diagnostic non-​ideal theory, see Anderson 2010a, pp. 3f.

Justice and Egalitarian Relations  21

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

the specifc nature, gravity, and difering dimensions of injustice, and more so than the requirements yielded by ideal theories of distributive justice specifying perfectly just distributions (see especially ­chapter 2, and section 4.6 in ­chapter  4).33 Unequal subjection to the arbitrary power of others, and to status norms decreeing one’s social inferiority, have, in their injurious meaning, a special dimension of injustice: they constitute expressive wrongs,34 and their gravity does not only depend on the material importance of the goods, or opportunities, which individuals might thereby be denied. To an account of this expressive perspective we now turn.

33 See Bidadanure and Axelsen 2019, pp. 343f. on this contrast and the diagnostic focus of important strands of social egalitarianism. 34 For the distinction between expressive wrongs and (mere) expressive harms, see Voigt 2018, p. 446.

2 Distributive and Relational Equality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.1.  Introduction Many of the liberal egalitarian theories of justice that have been dominant in political philosophy since the publication of Rawls’s A Teory of Justice conceive of equality as a distributive ideal. In particular, luck egalitarianism—​ the view according to which the role of brute luck in people’s life ought to be equalized as a matter of justice (see section 1.3 in c­ hapter 1)—​has been prominent in discussions of justice and equality over the last three decades or so. Tis book makes a case for relational equality instead; for a conception of egalitarian social and political relations as central concerns of social justice.1 Tis chapter motivates the search for such a conception: it argues that distributive views of equality cannot account for the specifc importance to justice of the way that social institutions create or maintain inequalities between individuals in society—​how institutions treat individuals, as opposed to which patterns of distribution they bring about. Tat treatment sets people up as social and political equals, when things go well, and slots them into hierarchies, when they do not. In developing this objection, it makes clear what is distinctive about a relational outlook on equality, and how it is diferent—​already at the level of the concept of justice employed, not only at the level of difering conceptions of justice—​from distributive views. It thus prepares the ground for the development of a justice-​based conception of relational equality in the subsequent chapters. Te argument draws on, and further develops, Tomas Pogge’s critique of ‘purely recipient-​ oriented’ conceptions of justice.2 Such conceptions view social justice as a matter of what individuals are entitled to get, rather than of how social and political institutions are to treat those to whom they apply. Section 2.2 argues that distributive conceptions of equality are purely 1 For main contributions to the debate between relational and distributive egalitarians, see Anderson 1999 and 2010a; Schefer 2003a, 2003b, and 2005; and responses by Arneson 2000; Dworkin 2002, especially pp. 113–​118, and 2003. 2 Pogge 1995 and 2003.

Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0002

Distributive and Relational Equality  23

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

recipient-​oriented theories. Section 2.3 presents a counterexample to purely recipient-​oriented theories devised by Pogge. Tis example constitutes the basis for the following discussion. Section 2.4 raises, and wards of, a peremptory objection that concedes that the ways institutions treat people have intrinsic moral signifcance, but disputes that what is at stake in such cases is justice. Section 2.5 explores a frst possible interpretation of the critique of distributive theories qua purely recipient-​oriented theories: that they neglect the intrinsic importance to justice of the way institutions cause advantages and disadvantages. It argues that this causal interpretation, while not wholly unsuccessful, has only limited reach. Section 2.6 argues that a diferent interpretation of the critique is more successful: according to it, what is primarily justice-​relevant about the way institutions treat people is the attitude towards individuals and groups, and their standing towards each other, that is expressed in institutional action. Te section brings together and links recent research on distributive justice, on collective responsibility, and on expressive theories of law and state action. It discusses what it means for an institution to express attitudes that are relevant to justice and why distributive views cannot account for such attitudes. If the argument of the chapter is successful, it shows that the weaknesses of distributive egalitarianism make the development of such a justice-​based conception of relational equality a worthwhile enterprise, and clarifes some of the conditions that such a conception has to fulfl in order to be successful. Te subsequent chapters then develop this conception.

2.2.  Distributive Egalitarianism: Currency Teories of Equality In a general form, the objection to distributive egalitarianism that this chapter discusses is that it cannot account properly for the relational dimension of justice: it fails to object to unequal relations as a potential problem of justice in its own right, that is, to power and status diferences independently of their distributive consequences. Variants of this complaint have been brought forward as a criticism of liberal distributive egalitarianism by many writers, recently especially by feminists. Perhaps the most famous of these recent criticisms was formulated by Iris Marion Young, who argues that the liberal ‘distributive paradigm’ does not properly recognize and tackle structural injustice as exemplifed in the ‘fve faces of oppression’: exploitation,

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

24  Liberal Relational Egalitarianism marginalization, powerlessness, cultural imperialism, and subjection to systemic violence.3 Elizabeth Anderson has taken up this point in her criticism of luck egalitarianism.4 Te concern is not new. Indeed, the frst to bring it forward was Marx, who argued that the Gotha Programme of the German Social Democratic Party from 1875 wrongly concentrated on the ‘fair distribution’ of material benefts in society, rather than on the qualitative transformation of social relations at which communism aims.5 But it is a concern that has many facets, and it is not clear what precisely its core is. For example, some criticisms are concerned with the supposed commitment of distributive egalitarianism to tackle injustices solely by means of cash transfers, and a consequent failure to take into account other remedies, such as changes in the social environment of disadvantaged individuals. In this form, it is an objection to an exclusive focus on redistribution. For example, some of Anderson’s objections to luck egalitarianism depend on taking luck egalitarians to be committed to cash transfers as the only remedy to injustice, and to be unable to recommend, for example, measures that would enhance the social inclusion of the disabled.6 Tis is not, however, a promising version of the relational objection. Distributive egalitarians, such as Ronald Dworkin, G. A. Cohen, or Richard Arneson, are not committed to redistribution as the only remedy to injustice. Setting up a specifc ideal of distributive equality as a target a society should aim at implies no such commitment; it only implies a commitment to such measures as will best bring about the preferred form of equality. Tat might, depending on the case, well be a change in social circumstances rather than individualized cash transfers. Concern about redistribution as the only remedy to injustice thus does not go to the heart of the matter. Te core of the relational objection to distributivism is not that the latter ofers the wrong remedies to injustice, but, at a prior stage, that it is not able to identify properly the intrinsic moral importance of the way social and political institutions act. Tat is to say that how institutions treat people has relevance to social justice that is independent of, or at least not reducible to, the distributive efects of such treatment. Tis is still a broad way of putting the objection, and I will further narrow it down as the argument proceeds. 3 Young 1990, pp. 48–​63. 4 Anderson 1999, p. 312. She replaces ‘powerlessness’ and ‘systemic violence’ with ‘domination’ and ‘status hierarchy’. 5 Marx 2000 [1875]. See 1.2. 6 See Anderson 1999, pp. 305f., and p. 333.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  25 To do that, it frst needs to be specifed in a more precise manner what I mean by a distributive egalitarian theory, and which theories fall in this category. Distributive egalitarian theories share a commitment to three principles of equality. Te frst principle of equality is the abstract principle of the equal moral worth of persons:  Persons, qua being persons, belong to the same moral category, so that diferences in their entitlements of justice cannot be justifed by arguing that they have intrinsically diferent moral status, for example, because some are born aristocrats, or into purportedly higher or lower castes. Te second principle is more concrete, and mandates, in Dworkin’s famous formulation, that persons are entitled to equal concern by social and political institutions in the assignment of benefts and burdens. Finally, the third principle of equality spells out the distributive requirement according to which people are entitled to some form of equality in the distribution of a certain currency, such as resources, or opportunity for welfare. Te frst two egalitarian principles are ofen said to form the ‘egalitarian plateau’ on which debates about justice in contemporary moral and political philosophy take place.7 Te interesting step for the present argument follows now: distributive egalitarian theories, such as Dworkin’s ‘equality of resources’ and (the earlier) Arneson’s ‘equal opportunity for welfare’ go on to argue that the best interpretation of the ideal of equal concern is to regard people as entitled to equal shares of a distribuendum. What it is that is to be distributed equally—​which metric or currency to adopt—​is the subject of the ‘equality of what’ debate that was sparked by Sen’s lecture of the same name.8 Luck egalitarian theories are motivated by the principle that bad brute luck inequalities—​that is, inequalities in life factors that are not due to circumstances that people can reasonably be held responsible for—​are unjust. Hence, as a matter of justice, brute luck ought to be equalized, or at least signifcantly mitigated. What counts as a relevant brute luck disadvantage is determined by reference to the respective currency of equality. Diferent currencies, or metrics, have been proposed by diferent members of the distributive egalitarian camp, such as resources,9 opportunity for welfare,10 or a broader notion of access to advantage, a mixture of welfare and resources.11 7 Kymlicka 2002, p. 5. 8 Sen 1979. 9 Dworkin 2000. 10 Arneson 1989 and 1990. 11 Cohen 1989 and 2004; see also Cohen 1993, p. 25, for a related notion of ‘midfare’. Te possibility of distributive theories incorporating concern for equal social relations alongside more traditional individualist currencies, such as resources, or opportunity for welfare will be discussed in 2.6.2, and important recent attempts to do so analysed in 5.4 (­chapter 5).

26  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Diferent luck egalitarians arrive at diferent accounts of what individuals are distributively responsible for, and therefore at diferently demanding egalitarian theories. But, insofar as these diferences can be traced back to different proposed currencies of equality, they show an underlying agreement that what is required is equality in some currency. Furthermore, distributive egalitarian theories in the sense used here do not need to demand strict distributive equality, or perfect equality of opportunity of some form.12 What is needed is merely that a third-​stage principle of distributive equality is part of its foundations, so that distributive equality in some currency is taken as the baseline, departures from which can be justifed either because they are due to individual choices, or for certain other qualifed reasons, for example, because they are overall better for everybody.13 Te smooth transition from the second to the third principle of equality makes clear that participants in this debate share a consensus about the underlying concept of justice:  they agree that fnding out what social justice requires is a matter of fnding out what people are entitled to get. Pogge calls this underlying concept of justice ‘purely recipient-​oriented’;14 I adopt this term. Tis, rather than a focus on redistribution, is the core feature of distributive theories. Such theories could also be called allocative, currency, or metric theories; I keep the term ‘distributive’ because this is the feature of such views that its critics have the most fundamental objection to. It is essential to such theories that they have to measure individual advantage or disadvantage according to the respective metric in order to ascertain whether unjust inequality exists; individual shares are of ultimate concern under such theories.

12 In his later works, Dworkin claims that a principle of full equalization of brute luck does not form part of his theory; it merely requires that ‘people be made equal, so far as this is possible, in their opportunity to insure or provide against bad luck’; Dworkin 2003, p. 191. Tis does not make the preceding characterization of distributive egalitarianism inapplicable to his theory, because the deviation from strict brute luck equality may be due to the excessive costs of trying to bring it about, which is itself a distributive consideration; see Dworkin 2002 (and see 2.5). However, Dworkin also argues that the consolidated version of his theory regards distribution merely as one dimension of equality among others of independent importance, political equality in particular, which might confict with the distributive dimension; Dworkin 2000, ch. 4, and 2003, pp. 195f. For discussion of this point, see fn. 69 in this chapter. 13 Te distributive viewpoint is, of course, shared by a large variety of non-​egalitarian theories as well, such as prioritarianism; Parft 1991 (to a variant of which Arneson has converted afer his initial focus on equal opportunity for welfare; see Arneson 2000 for his endorsement of ‘responsibility-​ catering prioritarianism’), or utilitarianism. I focus on distributive egalitarian views because holding a commitment to some kind of equality constant makes it easier to show that the relational objection is indeed an objection to distributivism as such. 14 Pogge 2003, p. 143.

Distributive and Relational Equality  27 Terefore, when it comes to the question of how to achieve equality, distributive egalitarians might well be concerned about transforming social structures, institutions, and relations. But their reason for this is based on the distributive state of afairs of justice-​relevant goods that such transformations would generate. Other justice or injustice judgements are derivative from this; actions and agents are just or unjust insofar as they promote just or unjust distributions.15 It is because of this feature that they cannot capture what is distinctively unjust about relations of inequality, and will therefore also ofen recommend the wrong remedies, and overlook the right ones—​as we will see later in this chapter, and in subsequent chapters focusing on these remedies. However, initially, the case for such distributive accounts seems very strong. Tey seem to have the attractive feature of ofering a simple and rational method for assessing the justice of social and political institutions:16 it is rational to look at how their performance afects individuals’ distributive shares, and to regard institutional design as a matter of achieving the best possible distribution, according to the respective theory. Afer all, it seems plausible that, from a justice perspective, social and political institutions are merely instrumental to the goal of achieving social justice.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.3.  Five Ways of Treating People Drawing on the account of distributive theories as purely recipient-​oriented theories, the remainder of this chapter explores the variant of the relational critique according to which such theories problematically ignore the way that institutions treat individuals. Consider the following example, devised by Pogge. Assume, plausibly, that health is a basic good that a distributive theory regards individuals as entitled to. Now [distinguish fve] diferent scenarios in which, owing to the arrangement of social institutions, a certain group of innocent persons is avoidably deprived of some vital nutrient V—​the vitamins contained in fresh fruit, say, which are essential to good health. Te [fve] scenarios are arranged in order of their injustice, according to my preliminary intuitive judgment.

15 See ibid., p. 147.

16 Pogge 1995, pp. 242–​246.

28  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

In scenario 1, the shortfall is ofcially mandated, paradigmatically by the law: legal restrictions bar certain persons from buying foodstufs containing V. In scenario 2, the shortfall results from legally authorized conduct of private subjects:  sellers of foodstufs containing V lawfully refuse to sell to certain persons. In scenario 3, social institutions foreseeably and avoidably engender (but do not specifcally require or authorize) the shortfall through the conduct they stimulate: certain persons, sufering severe poverty within an ill-​conceived economic order, cannot aford to buy foodstufs containing V. In scenario 4, the shortfall arises from private conduct that is legally prohibited but barely deterred: sellers of foodstufs containing V illegally refuse to sell to certain persons, but enforcement is lax and penalties are mild. In scenario 5, the shortfall arises from social institutions avoidably leaving unmitigated the efects of a natural defect: certain persons are unable to metabolize V owing to a treatable genetic defect, but they avoidably lack access to the treatment that would correct their handicap.17

Call this the V example. Tis example is supposed to appeal to our intuitions about the justice-​relevance of the way institutions treat people. Imagine also that the extent of the health shortfall and the number of deprived people are exactly the same in all fve scenarios; this is essential to the claim that the quality of institutional treatment is of moral importance in itself, independently of its efects. I fnd myself in rough intuitive agreement with Pogge. In any case, what is needed for the argument of this chapter is not that the proposed preliminary ordering is exactly right, but that some such ordering is right—​for example, that scenario 1 is indeed much more unjust than scenarios 4 and 5—​and that distributive theories cannot account for this.18 If the V example has signifcance, one ought to be able to explain what it is exactly in the institutional treatment of individuals that has such justice-​ relevance independent of its distributive efects. Pogge does not analyse the dimensions of the diferent scenarios that account for the intuitive diference between them in any detail. He merely suggests that the intuition that the 17 Pogge 2008, pp. 47f., his emphases (also, with slight verbal diferences, in Pogge 2003, p. 156). Te original quotation also lists a sixth scenario, in which certain people cause their V deprivation themselves, and institutions fail to react to this. I leave it aside, because this chapter focuses on the specifc shortcomings of luck egalitarian theories qua distributive theories, and does not discuss the relationship between responsibility for personal choices and distributive justice; for tensions between such responsibility for choice and maintaining relational equality, and possible ways to resolve them, see 5.4.2 in ­chapter 5. 18 For example, we will see, in c­ hapters 3 and 4 on non-​domination, that, depending on just how defcient protection against discrimination is in scenario 4, it may be more unjust than scenario 3.

Distributive and Relational Equality  29 V example seeks to engage has two dimensions: the way institutions cause goods shortfalls, and the attitudes that are implicit in institutional action.19 Sections 2.5 and 2.6 will discuss these in turn. But before that, a possible rejoinder on the part of distributive egalitarianism has to be addressed.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.4.  A Preliminary Objection: Justice Is Not All Tat Matters Tis rejoinder is that examples such as the V scenarios indeed succeed in pointing out that diferent modes of institutional treatment have intrinsic moral signifcance; but that this signifcance is not a matter of justice, but of diferent moral considerations. Tus, they do not present an objection to distributive theories, because such theories are only intended to account for the requirements of justice on institutions, not for all moral considerations that may apply to them.20 Two points can be made in response to this objection. Te frst insists that what is at stake in treatment cases such as the one introduced earlier is intuitively well described as a justice concern. Te second argues that an approach that integrates the treatment dimension into an account of the requirements of social justice on institutions has important theoretical advantages over an approach that regards the treatment and distributive dimensions as fundamentally distinct. To the frst point, then: Few people would want to deny that it is important that one not only receives one’s fair share of goods, but also that one is treated justly by others. For example, if you and I are to divide a bundle of resources among ourselves, and I do not consult you as to how you think they should be distributed, but just go ahead distributing it the way I think is fair, it seems plausible to hold that I am treating you unjustly, because you have a claim to be heard on issues that matter to you: I am violating a procedural right of yours. Similarly, the V scenarios may be described as involving violations of rights—​for instance, in scenario 1, a right not to be discriminated against. At the very least, it seems natural to describe institutional action in these scenarios as wronging the members of the concerned group. Institutions are

19 Pogge 2008, p. 48. 20 For an argument against regarding justice as the only value applying to institutions, see Goodin 2007.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

30  Liberal Relational Egalitarianism violating what looks like legitimate claims the group members have. To be sure, these are only preliminary refections about some intuitively plausible ways of relating justice, wronging, and treatment. But they sufce to shif the burden of proof back to the objector who seeks to argue that what is at stake here cannot, in principle, be about justice, since it is not about distributions.21 Second, a distributive egalitarian might react to this counter-​reply by proposing distribution and treatment as diferent spheres within justice. Along these lines, G. A. Cohen, in his fnal works, mentions that there are matters of justice that are nevertheless ‘outside distributive justice, such as the just and unjust treatment of individuals with respect to their liberty and privacy’.22 Tis response has the advantage that it accounts for the intuition that treatment has a justice dimension. However, it has the disadvantage that it is not clear, on this view, how the treatment and distributive dimensions of justice are to be related. Note how having to balance the distributive outcome dimension with the treatment dimension of social justice would do away with one of the advantages of purely recipient-​oriented theories discussed previously: the rationality of focusing on the outcomes produced by institutions when assessing their justice performance. Partisans of the distributive model may retort that this rationality is not a reasonable expectation on their views, and that the relational objection takes them to argue for a more ambitious position than they actually do. However, the likely upshot of a view that regards distributive and non-​distributive justice as two separate subjects of investigation is that the two dimensions have to be balanced case by case, without any further general theoretical guidelines as to how to undertake this balancing.23 An integrated relational view that managed to unite these two dimensions under an overarching framework for social justice would be a theory with greater explanatory power. It would have the theoretical advantage of accounting for more of our convictions about justice, by making its outcome-​and treatment-​oriented dimensions 21 See Bidadanure and Axelsen 2019, p. 337: a further argument showing ‘why one should prefer a narrow conception of egalitarian justice to an inclusive and ramifed one’ is needed. 22 Cohen 2008, p. 6. Cohen 2011 further suggests that there may be unjust distributions against which, however, nobody has a ‘legitimate complaint’, p. 129. 23 See Meijer and Vandamme 2019, p. 321, for resulting practical disadvantages in terms of loss of action-​guiding value. However, the relevant disadvantage is not merely practical, as isolating diferent domains also means forgoing opportunities to examine their connections, which could strengthen justifcation; see what follows in the main text. Cohen does not discuss the question whether, and how, on his account, distributive and non-​distributive justice are related in any detail, but his commitment to ‘radical pluralism’ speaks for his endorsing case by case balancing; 2008, pp. 4f. Similarly, Cohen 2011 contains no attempt to spell out how distributive justice and ‘legitimacy’ in distribution (see fn. 22) are related.

Distributive and Relational Equality  31 shed light on each other. Furthermore, it would have the practical advantage, by accounting for these two dimensions of justice and their interplay in one unifed set of principles, of having a better claim to practical pre-​eminence, or priority, of these principles for assessing social and political institutions over other considerations—​and therefore of also ftting better with the widespread conviction that justice, as a political value, enjoys a special stringency not shared by other values.24 So there is a reason to look for such an integrated overarching conception. Of course, it cannot be guaranteed that this enterprise of integration will succeed; similarly, it cannot be guaranteed that any pre-​theoretical intuitions about the special stringency of justice do not simply have to be given up, in the end. Te issue will reappear, in diferent guises, in subsequent chapters (­chapter 3, 3.2, and ­chapter 5, 5.3); all that this section has established is that there is good reason to try, and hence to doubt the peremptory force of the rejoinder that the V example does not apply to distributive egalitarian theories.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.5.  Difering Institutional Causal Involvement So what is it about the institutional treatment of people that has independent justice-​relevance: the way they cause distributive outcomes, or the attitudes they express in their actions? Tis section discusses the causal interpretation of the relational claim: other things being equal, good shortfalls are more objectionable the more institutions are ‘materially involved’25 in causing them, and distributive egalitarianism cannot account for that. In the V example, this contrast is exemplifed by the low injustice rank of scenario 5, in which institutions merely fail to react to a genetically caused health inequality; from there, the degree of causal involvement of social institutions increases up to scenario 1, which exemplifes maximum such involvement:  actively depriving the individuals in question of V through an explicitly discriminatory law. Or so the claim has to go. Te remainder of this section will argue that this characterization of the relational dimension of justice is of too limited reach to do all, or even most, of the work required to sustain the relational objection. 24 See, for example, Nagel 1997, p. 303, and Arneson 2008a, p. 378: ‘In ordinary English usage, the term ‘justice’ tends to be applied to what the speaker regards as a paramount value and also an all-​ things-​considered value’. 25 Pogge 2003, p. 157.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

32  Liberal Relational Egalitarianism Te example appeals to a morally salient distinction akin to the familiar distinction between negative duties and positive duties. Tis distinction mandates that, others being equal, actively causing an objectionable outcome is worse than merely letting it happen. What is involved in the example must be congenial in spirit to that distinction, but more complex, since it invites us to rank the difering scenarios on a scale of assessment, on which particular cases can allegedly be ordered according to the degree, or quality, of causal involvement.26 Te distinction between negative and positive duties certainly has great intuitive weight in cases of individual conduct. We regularly object more strongly to a positive, intentional individual action that brings about a certain outcome than to mere inaction in the face of such an outcome—​to use a worn-​out example, throwing a child that cannot swim into a lake is more objectionable than not saving it when it merely fell into the lake, even though the outcome is the same in both cases: death. In the individual case, we regard individual omissions as equally problematic as individual action leading to the same outcome only if we think that the individual in question had a special responsibility to prevent that outcome; for instance, a parent who lets her small child starve to death in order to get rid of it is precisely as responsible for its death as if she had shot it. Nothing similar is true for state action. Fulflling the standards of justice is not a special responsibility a state has on top of its personal life; from the point of view of theories of justice, it is its very purpose of existence.27 Putting these two considerations together, it is plausible to suppose that inaction on the part of individuals is less problematic especially if there is a state that does in fact take care of maintaining standards of social justice. Tus, the moral signifcance of this distinction seems to be much less in cases of institutional action.28 From the perspective of social justice, social and political institutions exist primarily to implement standards of justice; so why should it matter in itself whether institutions actively bring about goods shortfalls, from the point of view of the recipients of these goods, or merely fail to react to such outcomes? As mentioned, Pogge concedes the plausibility of 26 Ibid. 27 See Nagel 1991, p. 100. 28 Hosein argues that the special responsibilities of states cannot override pre-​existing rights of individuals not to be harmed, so that the distinction has signifcance where these are at stake; 2014, p. 258. Tis seems correct, but we are here interested in the theoretically prior question of what the distinction itself contributes to answering the question of which rights individuals have against the state, and which are more important than others.

Distributive and Relational Equality  33 this perspective,29 but still thinks the causal interpretation strong enough to overcome it. Against that, I shall argue that the question of causation has weight only in the special case of sorting naturally generated from socially generated inequality—​scenario 5 versus scenarios 1–​4. It cannot account for the intuitive diferences between the exclusively social scenarios 1–​4.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.5.1.  Natural and Social Inequalities Te V example ranks scenario 5, the failure of social institutions to treat a genetic defect causing V defciency, as the least unjust, and that seems plausible. It also mentions the reason for that: the inequality in question is natural, and hence not, or at least less, within the domain of responsibility of social institutions. Te issue of natural versus social inequalities has received a fair bit of attention in recent literature about equality. For example, Elizabeth Anderson claims that luck egalitarians have missed the point of equality by focusing on compensation for natural inequalities, such as inequalities in talents or physical attractiveness, instead of objecting to social hierarchy, which is by defnition socially caused.30 It is also of importance in Rawls’s theory of social justice, even though his comments about the ‘arbitrary efects of the natural lottery’31 have inspired luck egalitarians: Rawls proposes a principle of fair equality of opportunity requiring that persons of equal natural talent, and of the same willingness to use them, have equal opportunity to attain desirable social positions, regardless of their class background. And this principle enjoys lexical priority over the diference principle, which takes care of the case of naturally unfortunate individuals, who lack talents.32 Tus, he regards inequalities due to social background as more unjust than social inequalities of equal extent that are due to underlying natural inequalities.33

29 Pogge 1995, pp. 241–​247. 30 See Anderson 1999, pp. 288, 309, 312f. 31 Rawls 1999a, p. 64. 32 Ibid., pp. 62–​65. 33 See Nagel 1997, p. 310. See also Pogge 1995, pp. 247–​250. Te tortuous formulation is necessary because Rawls argues that natural inequalities per se are not unjust at all: ‘the natural distribution is neither just nor unjust’; 1999a, p. 87. In later work, Rawls entertains the suggestion that the priority ordering between these two principles should perhaps be weaker than lexical; 2001, p. 165, fn. 44; but any form of priority sufces to illustrate the general point.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

34  Liberal Relational Egalitarianism As Nagel puts it, ‘every society is in the business of transcending the state of nature, but how far it is obliged to resist the diferential impact of fate and natural variety is a difcult question.’34 Te V example plausibly assumes that tackling natural health defciencies and vulnerabilities is one of the most prominent aims of society, because all humans are vulnerable to health risks to some extent. But there is a good reason for holding that justice should be less concerned with health inequalities that do not stem from social processes, as in scenario 5, than with those that do, as in scenarios 1–​4 (see c­ hapter 9, section 9.3). Tis reason is that there is an irremediable tension between the liberal idea that societal cooperation has to guarantee to all those who are under a duty to cooperate the social conditions, especially the liberty, necessary to choose a conception of the good from a suitably wide array of possibilities, and to pursue and modify it, and the idea that people ought to be compensated for disadvantages, whatever their cause (apart from their own choices), up to equality. Duties to such compensation may unduly restrict the array of possible conceptions of the good open to cooperators, and take on the character of a substantive goal for their lives, rather than embodying merely constraints on the pursuit of their own conception of the good. If natural inequality is both pervasive and difcult to remedy, such compensation might amount to a requirement of self-​sacrifce on the part of the initially advantaged. Terefore, in order to avoid this confict, a liberal conception of justice has to limit its concern with distributive inequalities in some principled manner. Limiting concern with natural inequality seems one good way of doing this, since it achieves the desired aim while at the same time fulflling the intuitive requirement that recourse to such a restriction is not available if a society has itself caused a morally relevant inequality: if I am causally responsible for your disadvantage, I cannot claim that compensating you would unduly restrict my opportunities. At this point, it is not necessary to validate one particular liberal theory relying on a version of the natural-​social distinction.35 It is enough to note that the general idea of such theories is sound. As seen, one theory that does limit the scope of objectionable inequality in such a way is Rawls’s, which limits egalitarian concern to socially produced goods.36 Distributive egalitarianism misses the importance of the cause of 34 Nagel 1997, p. 305. 35 See further ­chapter 9, and Aas and Wassermann 2016 for an instructive discussion building on, and improving, Nagel’s 1997 account. 36 Pogge calls Rawls’s view ‘semi-​consequentialist’, because it restricts the scope of distributivism in such a manner; 2003, pp. 154f. I am not convinced that Rawls’s theory in its fnal form lends itself to a

Distributive and Relational Equality  35

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

morally salient inequalities by focusing exclusively on how to bring about an optimal egalitarian distribution. Treating people justly cannot simply mean ensuring such a distribution in cases where inequality is naturally caused. Such a restriction on duties of social justice hence seems able to provide a coherent rationale for the intuitive salience of the natural-​social distinction. To be sure, this is only a preliminary result, and the issue of avoiding the confict mentioned—​in the sense of keeping justice and/​or equality from becoming mandatory aims for personal lives, as opposed to just constraints on them—​will continue to occupy us in subsequent chapters. Distributive egalitarians might seek to argue that there are other ways of safeguarding against this problem without drawing this particular distinction. One possibility could be the one mentioned in section 2.4: distributive justice might not be sensitive to it, but non-​distributive justice—​especially the treatment of people with regard to their liberty—​might set limits to the former. Another one is to incorporate an ‘agent-​centred prerogative’ into a distributive egalitarian framework, according to which people may give some degree of priority to certain personal projects over the duty to follow impersonal, egalitarian principles.37 However, at this point, it is more important to note that the argument so far, if it is sound, supports Pogge’s causal claim only to a very limited extent, anyway. It accounts only for the low injustice ranking of scenario 5, the case of failure to treat a genetic health defect causing V deprivation, vis à vis scenarios 1–​4. Tat is not enough to vindicate the relational claim that there is a scalar ranking of injustice tracking the way institutions treat individuals.

clear and unambiguous characterization as wholly distributive about social goods, but cannot pursue this disagreement here (but see ­chapter 1 fn. 15, and ­chapter 8, 8.3, on Rawls’s distinction between ‘allocative justice’, and distributive justice understood as ‘pure background procedural justice’. 37 For the concept of a personal prerogative, see Schefer 1994. Cohen mentions it as a restriction on his luck egalitarian view; Cohen 2008, pp. 10, 61. However, he suggests that ‘we cannot say where the limit of the prerogative lies’, so that ‘everyone must make her or his own principled decision’; ibid., p. 220; the task of striking the right balance between impersonal demands and self-​indulgence is thus lef to individual conscience. Striking this balance in a general fashion is, against that, a crucial part of what theorizing social justice is about, for liberals such as Rawls.

36  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.5.2.  Causal Diferences between Exclusively Social Scenarios To be more precise:  the problem is not that the natural-​social distinction is binary, because that does not mean that it could not justify scalar justice assessments. It could do so in cases where natural and social causes interact, and account for the outcome to diferent degrees. Social structures can have very diferent efects depending on people’s difering natural endowments, and the fact that endowment inequality may be natural does not by itself justify compounding it by social means. Furthermore, socially caused pollution of the environment causes genetic mutations,38 and class and caste structures infuence the mating patterns that create the genetic endowments of the next generation39—​to name just some examples where it is difcult to assess to which extent inequalities are really properly regarded as natural. But scenarios 1–​4 are all exclusively socially caused. Te causal scale that would allow us to rank scenarios 1–​4 according to their injustice would have to be of a diferent nature. Te claim about the intrinsic importance of causes must here rely on distinctions between the diferent intrinsic qualities of different ways of social causing, such as degrees of immediacy: Scenario 1, deprivation mandated by law, must somehow represent a more immediate causal chain than scenario 2—​merely authorizing private individuals to refuse the sale of V—​which in turn must be more immediate than scenario 3, where V-​ deprivation comes about as an efect of an ill-​conceived economic order, and so on. Or at least some such ordering according to the nature of the causal chain in question must be right. But here, distributivists can pass the buck right back and ask: But what is it about causal chains that makes them intrinsically relevant in this way, if we are dealing with exclusively social causation? A principled explanation akin to the one just given for the salience of the natural-​social distinction has to be provided. And they can bolster this challenge by pointing out two additional features of the V scenarios that may account for some of the intuitive diferences between them without thereby endangering the distributive point of view. First, the V scenarios could prompt one to make diferent assumptions regarding the knowledge of institutions about the goods shortfalls in

38 Pogge 2003, p. 155. 39 Rawls 1999a, p. 92.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  37 question: where an agent actively and intentionally brings about an outcome, as in scenario 1, no further investigation is needed to ascertain that this course of events was avoidable and foreseeable for the agent. Te less actively involved the agent in question is, the more difcult it becomes to ascertain whether this is the case (for example, in scenario 4).40 Second, information about the costs, including risks, of institutional remedial action in each scenario is missing. Tese would have to be equal, too, to make a case against distributivism, because there is no reason to suppose that they cannot, as a matter of principle, be incorporated into distributive metrics; this would seem to depend on the particular metric that is chosen. However, the diferent scenarios may lead us to think that diferent costs are involved. For example, it seems highly plausible that it would cost very little, if anything at all, to stop ofcially mandated discrimination against a group of people like in scenario 1, but considerably more to gear the economic order towards avoiding the poverty due to which the people in scenario 3 are deprived of V—​all kinds of things can go wrong in the attempt to reform an economic order. Similarly, where diferent agents are involved, like in scenarios 2 and 4, in which private individuals refuse to sell V, it is plausible to suppose that it would cost much more to enforce a prohibition of their behaviour than it would to simply repeal a law and call back the police, like in scenario 1. Tese rejoinders underline the strength of the ‘purely recipient-​oriented’ view: If A and B sufer a goods deprivation of precisely equal extent, as a result of two diferent institutional measures, and it was equally easy or diffcult for the institutions to foresee in both cases that the measures would have this efect, and if it had been precisely equally costly to avoid the disadvantage in both cases—​then, why should we think that one has been treated more unjustly than the other, simply because the causal chains in question happened to take diferent routes? As mentioned earlier, we might have good reasons to assess individual responsibility for particular outcomes according to the precise nature of the causal connection between an individual and that outcome—​for example, in matters of criminal justice. But it is not clear what such reasons could be in the case of institutions.41 40 One may point out that, strictly speaking, avoidability and foreseeability should not infuence the justice of a state of afairs, from a distributive point of view, but only institutional blameworthiness for that state of afairs. But even then, it is not self-​evident that institutional blameworthiness is not the right level to focus on in the explanation of some aspects of the V example. 41 Accordingly, there is reason to think that the force of the causal interpretation of the relational claim in scenarios of global (in)justice, on which most of Pogge’s work focuses, has to be due to its successfully disputing the belief that global inequality is relevantly akin to natural inequality (scenario 5). Citizens of afuent states cannot claim that global inequalities are of no concern to them if

38  Liberal Relational Egalitarianism But still, the V example has bite. Tere is a justice diference between a state legally authorizing private discrimination (scenario 2), and the state merely failing to enforce the prohibition of such behaviour (scenario 4), even if all the dimensions of assessment discussed so far—​extent of goods shortfall, number of victims, foreseeability and avoidability of outcome, and costs of remedy—​are exactly equal in both scenarios. Instead of focusing on supposed qualitative diferences between types of causal chains, I  shall argue that this can be better explained by focusing on the quality of the attitude that is expressed in such diferent modes of actions.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.6.  Difering Institutional Attitudes Expressed in Treatment Tis second interpretation of the relational claim goes as follows: Te attitudes of social and political institutions towards people expressed in the way such institutions treat them are relevant to justice, and they are not ultimately reducible to distributive considerations. According to this interpretation, the intuitively diferent degrees of injustice in the example are accounted for by the fact that, in each case, there is a diferent normative attitude implicit in the diferent ways social and political institutions treat people. Here are some suggestions as to how an assessment of the expressive dimension of institutional treatment in the cases in question could look like (to confrm them, more information than given in the example would be necessary, especially about the reasons institutions were acting on in each scenario): In scenario 1, in which the V deprivation of a certain group is mandated by the law, social and political institutions express outright hostility towards the group in question, because the state aims at bringing about the deprivation: it regards itself as required to prosecute and sanction members of that group who seek to procure themselves V against the prohibition. In scenario 2, in which sellers of foodstufs may lawfully refuse to sell food containing V to members of the group, what is expressed is plausibly described as institutional contempt: the state allows other individuals they are caused by institutions, or other social processes, that their states control; the degree of causal involvement in bringing about or maintaining global inequality is relevant to justice, just as the degree to which social measures exacerbate natural inequality is. Nothing similar is true where the inequality in question undisputedly lies within the scope of the institutions responsible for justice, as in scenarios  1–​4.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  39 to discriminate against members of the group; so it regards members of the group as not worthy of its assistance in such cases. Tat, however, does not happen in the more negligent and clandestine way of scenario 4, in which the state ofcially recognizes that such conduct by sellers of foodstufs is illegal, but then does not do enough against it, but in an open and public way: in scenario 2, it is explicitly declared permissible to deprive group members of V, so the state is conniving with the discriminators. Scenario 5, in which the state fails to ofer treatment of a genetic defect leading to V deprivation, plausibly expresses only neglect, in cases where the state, for example, avoidably fails to gather information about people sufering from the defect and is hence unaware of their existence. In all scenarios, insofar as they successfully single out an injustice, institutions demonstrate an attitude of disregard, or disrespect, towards the unjustly disadvantaged. What accounts for the justice diference between the scenarios are the diferent material attitudes that are expressed in them, ranking from open hostility to mere neglect. According to these material attitudes, unjustly treated persons are assigned diferential moral status by institutions: in scenario 1, they are regarded as of such inferior moral worth that their disadvantage may be openly and publicly mandated by the law and enforced; in scenario 2, they are regarded still as inferior enough to openly deny them any assistance should they be discriminated against by other individuals. As opposed to that, in scenario 4, institutions do not ofcially condone the discrimination of sellers of foodstufs against these people, such that a refusal to sell V could not, for example, regularly take place in the immediate presence of a police ofcer (imagine she is next in the queue in the supermarket); but they do not regard them as important enough to spend signifcant resources on coming to their help by, for example, actively searching out such cases of discrimination. Te diferent scenarios hence present diferent implicit institutional judgements about the worth of the disadvantaged people, and the claim is that, other things being equal, our diferent intuitive reactions to the diferent scenarios rest on our perception of these judgements. In each such case, social hierarchies are instantiated or made possible by such implicit judgements of worth; and this is what the relational egalitarian primarily objects to. Tis expressive interpretation of the relational objection fts well with the claims of past and present egalitarian movements, such as movements for the equality of women, of people of diferent sexual orientations, and of ethnic minorities (see 1.4 in ­chapter 1). Such movements generally demand

40  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

treatment that afrms their equal moral standing. What they are afer is the confrmation that the people they represent are not, by virtue of belonging to a group such as women or gays, of inferior moral worth, and, accordingly, they demand state action that makes this clear. Te task for a justice-​based conception of relational equality, to be worked out in the next chapters, is then to deliver a framework that makes sense of, and orders, such demands, and can thus function as a principled basis for expressive assessment, in order to substantiate (and possibly also revise) such preliminary intuitive orderings and assessments as the one undertaken here. But before we move further, it is necessary to take several steps back again. Te expressive perspective on the justice of institutions requires further explanation, along two dimensions: First, does it really make sense, both conceptually and morally, to ascribe attitudes to institutions? Which formal conditions do institutions have to fulfl in order to be appropriate targets for such ascriptions? And, why, in justice assessment, should we (also) focus on their attitudes, and not (only) on the attitudes of individuals that uphold and sustain them? Focusing on institutions might seem fetishistic. Second, even if expressive assessment of institutions is meaningful and irreducible to expressive assessment of the actions of individuals, why can distributive egalitarianism not account for these attitudes; and why can it not be amended, or modifed, so as to account for them? Only if these questions can be satisfactorily answered, is expressive assessment of institutions meaningful and morally warranted; only then can we proceed to the question of which implications it has for the substantive content of relational egalitarian social justice. Te following subsections thus tackle these questions in turn.

2.6.1.  Institutional versus Individual Attitudes First, then, to the question of whether institutions can express attitudes in their actions, and why, from the point of view of justice, we should ascribe ultimate importance to them, and not, or not only, to individual attitudes towards each other as expressed in the creation and upholding of institutions. Regarding the frst of these questions—​whether institutions are formally of the right kind to be carriers of attitudes—​I mainly refer the reader to recent scholarship on collective action and group, or institutional, minds. A  detailed argument against skeptical methodological individualists, who sustain that there can be no such thing as a group agent, is beyond the scope of this

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  41 chapter, so I assume that the following is a defensible interpretation of group agency:  In collective intentional action, individual actors properly regard themselves as engaged in a common enterprise, and this common enterprise generates claims on each of them to behave in certain ways in support of the common enterprise. Together, they act in ways that enable the collective agent in question to form its own beliefs and intentions, and to act on these. Tese beliefs and intentions in turn are subject to moral assessment: just as in the case of individual action, acting with the right attitude towards the object of one’s actions means taking the right considerations as reasons for actions regarding that object.42 Te formation of group intentions may sometimes happen in a spontaneous manner; for example, in the case where a collection of bystanders to an accident instantly organizes itself in support of the victim by implementing a division of labour, in common knowledge that this is what they are doing, as part of the group action ‘assisting the victim’: one calls an ambulance, another one gives frst help on the spot, a third warns oncoming trafc, and so on.43 But the distinctness of group beliefs and intentions is even clearer in the case that is more relevant for this chapter: institutions. Institutions are non-​spontaneous group agents: they form beliefs and intentions on the basis of a constitution of decision-​rules adopted and upheld by (a sufcient portion of) the involved individuals. Research by Pettit and List shows that a rational reduction of collective beliefs and intentions of a functioning group agent to the beliefs and intentions of the individuals involved in the group’s decisions is not possible even in relatively simple cases: depending on the constitution of the institution in question, it may adopt beliefs and intentions that are not shared by the majority of the involved individuals; or even beliefs and intentions that are not shared by any of them.44 For institutions to count as distinct agents, all that is needed is a constitution that allows it to form its own beliefs and intentions, and to display consistency in acting on these over time; such as a legal court does when it develops and upholds a particular doctrine for a feld of law, as its composition changes. If it is true that institutions can fulfl these conditions, then such institutions can also express normative attitudes in their actions; they do not need to be able to have 42 See Anderson and Pildes 2000, p. 1510. 43 See Gilbert 1989, pp. 139f. 44 See List and Pettit 2002, 2006, and 2011. Te simplest case they use is that of a jury of three judges deciding law cases: majority voting on the verdict and majority voting on the premises of the verdict may generate diferent results. See also Pettit 2003, and 2007 for an account of the conditions that collective bodies have to fulfl in order to be ft to be held morally responsible.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

42  Liberal Relational Egalitarianism the mental states that ofen accompany attitude expression by natural agents, such as feelings of contempt, disgust, hate, and so on. Returning now to the V case, one more thing needs to be noted: institutional agency as such does not sufce to make sense of the possibility of assigning a scalar justice ranking to the respective scenarios. Tis is because there are diferent agents involved in the diferent scenarios: for example, in scenario 1, the perpetrators of injustice are lawmakers and the police; in scenario 2, the same two plus private sellers of foodstufs; and in scenario 3, it is the economic order of a society. Since a relational view focuses on the justice and injustice of actions, which is, inter alia, dependent on such situational factors as the responsibilities and capacities of the respective agents, a variety of involved actors may complicate comparative justice assessment, or even make it impossible. Accordingly, in order to preserve the possibility of a clear ranking of the V scenarios, it is necessary to adopt a holistic stance: the primary subject of justice involved must be conceived of as capable of—​complex—​agency (see 1.4 in c­ hapter 1). To put it in Rawlsian terms, this subject may be called ‘the basic structure’, that is, ‘the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation’.45 Major institutions are ‘the political constitution and the principal economic and social arrangements’46, such as a market economy and the family, or at least some dimensions of it. Up to now, for reasons of simplicity, I have used ‘the state’ to refer to the agent whose causal role or attitude and its relevance to justice was discussed; in the case of a single society, as our primary focus, it is at least provisionally permissible to regard the state and the basic structure in the relevant sense as co-​extensive. Scenarios 1 to 5 may then express diferent attitudes of the basic structure: 1 expresses hostility, and 2 contempt, while 3, 4, and 5 may be scenarios of decreasing neglect; and, other things being equal, according to an intuitive preliminary assessment (see earlier discussion), treating a group of people under your power with hostility is more unjust than displaying contempt for them, which in turn is more unjust than neglecting them. A scalar justice ranking of the fve scenarios is thus possible if the basic structure in question can be regarded as one institutional agent, unifed by one constitution 45 Rawls 1999a, p. 6, and for institutional holism, p. 50: ‘a social system may be unjust even though none of its institutions are unjust taken separately; the injustice is a consequence of how they are combined together.’ 46 Ibid., p. 6.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  43 enabling it to act consistently—​at least in the sense that this constitution reliably connects diferent institutional agents under it, and enables these to coordinate their actions efectively. Tis leads to the second question mentioned earlier: even if it is possible to regard institutions as agents expressing attitudes in their actions, why should we focus on their attitudes, and not only on the ones expressed by the individuals that sustain them? If, for example, we are worried about the contempt expressed by institutions for some individuals in their openly discriminatory treatment of them, are we not really only worried about the contempt that the people that cooperate in upholding the institutions thereby express? Isn’t it fetishistic to focus on the institutions themselves? In a trivial sense, the answer to this question is ‘yes’. Every institutional action must be embodied in some individual action; if no individual does anything, no institution does, either. But, in a more interesting sense, the answer to the question is ‘no’, as the preceding remarks have already made clear. As Pogge notes, individual attitudes ‘only make a diference to how blameworthy persons are for their role in imposing’47 the institutions in question. Anderson’s and Pildes’s example is the following: ‘lawmakers could pass a law that expresses contempt for blacks by denying them the right to vote even if none of the lawmakers personally feel contempt for blacks, and all are merely pandering to their white constituents’.48 Te institutional action in question, because of its injustice, should alarm everybody upholding the institutional scheme of which the parliament in question is part into counteraction, and remedial duties fall, in the frst instance, most clearly on the lawmakers themselves, both because their contribution to the injustice was greatest and clearest, and because they are in the best position to reverse it. But these remedial duties follow from their responsibility for the institutional attitude expressed, not from the identity of their own individual attitudes with it. To make the point even clearer, consider the case from the point of those who are treated unjustly. Being treated in certain ways by others triggers ‘reactive attitudes’;49 in the case of injustice, resentment and indignation. In the preceding case, the appropriate object of these reactive attitudes is, in the frst instance, parliament, and the overall basic structure into which it is embedded, and which makes it an efective institution. We might say



47 Pogge 2003, p. 157.

48 Anderson and Pildes 2000, p. 1508.

49 For the locus classicus, see Strawson 1974.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

44  Liberal Relational Egalitarianism that ‘everybody with their citizen hat on’ is, to some extent, an appropriate target of resentment and indignation. But little is gained by this. For example, it is not clear to which extent a lawmaker who opposed the law in question is an appropriate target. As a general rule, it does not make sense to charge the people who generally uphold an institution with having the same attitudes that the institution expresses in its specifc actions, just because their upholding it allows for them. Cases of straightforward alignment of institutional and individual attitudes are, of course, possible:  take the case of a society whose basic structure is controlled by a dominant group of like-​minded people, who merely use it to express their pre-​existing attitudes towards other individuals and groups. However, what makes this case so straightforward is a prior structural injustice: the dominatory nature of that basic structure—​the unchecked opportunity of the group in question to use it for their private purposes. One fnal clarifcation is needed, before proceeding to the question of why exactly distributive views cannot account for institutional attitudes: the previous considerations do not rule out that the attitudes expressed by individuals are also an object of concern, in addition to collective ones, nor are they meant to. A liberal view will merely insist that standards of appropriateness for individuals are, in many cases, diferent from those for institutions:  it will ofen not subject individuals to as stringent demands as institutions, and make space for prioritizing individual relations, even where these do not have an egalitarian character, insofar as they remain within certain protective bounds (see 4.3.2 in c­ hapter 4), and do not endanger egalitarian institutions and societal norms of non-​domination (4.5 in c­ hapter 4), or corroborate pervasive hierarchies of social status (6.3 and 6.4 in c­ hapter 6). As opposed to that, more radical egalitarian views may recognize that institutional settings raise special issues, but will demand that people display similarly stringent concern for social equality in their private lives (see ­chapter 5, 5.3, and c­ hapter 6, 6.5).

2.6.2.  Why Distributivism Cannot Account for Expressive Attitudes If the argument of the previous subsection is sound, it lays to rest any worries that concern about institutional attitudes is mysterious, or fetishistic, in comparison with concern about individual ones. But the central point of the

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  45 argument of this section is still missing. It seems that a distributivist may agree on Dworkin’s formula, quoted in section 2.2, that the state has to express an attitude of ‘equal concern’ for those subject to its power, but insist that it does so best precisely by bringing about an egalitarian distribution. So, frst, why is it that distributive views cannot account properly for the justice relevance of institutional attitudes? And second, even if distributive views of a traditional kind—​focusing, for example, on the distribution of opportunity for welfare, or resources—​indeed cannot, is it possible to amend them so that they can? In order to answer these questions, it is necessary to explain in more detail what is involved in expressing an attitude in one’s actions. In the course of this explanation, some connections between the relational objection to distributivism and the hoary doctrine of double efect will become evident. Te expressive interpretation of institutional action delivers an account that displays some important similarities with that doctrine, according to which the quality of intentions expressed in actions can make a diference to the justice of these actions, even when they produce identical outcomes.50 Te attitude expressed by an action is the right one if the agent in question acted on the right reasons. Actions express intentions, and hence have a meaning; interpreting the meaning of an action is the same as assessing the attitude it expresses. Te meaning of an action is not just a matter of what the agent in question means to express with her action, but also of how those who are subject to the action may reasonably understand it. Here, expressive assessment diverges signifcantly from the traditional doctrine of double efect. In attitude assessment, it is necessary to look both at the motivations and purposes of the agents, and the way the action can reasonably be understood by those afected—​and is in fact understood.51 If the action in question is, for example, the institutional action of parliament passing a law, sources that expressive attachment may draw on are (a) the reasons given for the law by parliament,52 and (b) the way the law has been understood by those subject to it. Needed is then an interpretive assessment of both: whether the reasons 50 Te following account of attitude assessment is heavily indebted to Anderson and Pildes 2000, pp. 1511f. 51 If a disrespectful action is indeed understood to be disrespectful by the afected, that makes it, other things being equal, worse; see Voigt 2018, pp. 451f. 52 Tese difer, in the frst instance, from the private reasons that any individual lawmaker may have had for voting for the law in question—​see the previous subsection. Such individual reasons may be drawn in at a later stage of interpretation; for example, when it becomes evident that the ofcial reasons do not sufce to motivate the law in question.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

46  Liberal Relational Egalitarianism given adequately support the law in question and whether the way it has been understood is reasonable. In order for the expressive interpretation of the relational objection to distributivism to be successful, it must not be possible to account for this dimension of meaning as just another distributive outcome. Tree examples may serve to clarify what is at stake in expressive assessment of institutional action, and to elucidate the challenges it poses to distributivism. First, take the example of group inequalities. Here the relational claim goes as follows: other things being equal, it is more unjust if, due to some state measure, among a group of disadvantaged people—​say, they lack adequate access to the healthcare system—​there is a signifcant number of members of a pre-​identifable group, such as women, or blacks, than if the disadvantaged group consists solely of random people—​men, women, whites, blacks, Asians, etc.—​not unifed by any common ascriptive characteristic.53 Te problem is not restricted to ethnic or gender groups; it is the same in the case of social class.54 Such state action gives rise to an additional, irreducible grievance on the part of the members of such groups, because it is reasonably taken as the expression of an objectionable attitude of institutions towards them: institutions do not, in their actions, regard the members of the group in question as equals to members of other groups. So, in the V example, under the assumption of an efective basic structure, the respective injustice of the diferent social scenarios 1–​4 would increase if we were to assume that the disadvantaged group in each scenario consisted of people unifed by a common characteristic other than the disadvantage. Note that for such a group grievance to arise, it is not necessary that the disadvantage in question really comes about, as a matter of fact. Even if it is prevented by some accident, it is enough that, at the decision-​making stage, the information that it would regularly ensue has been taken into account by the institutions, or should have been taken into account, given reasonable requirements regarding the collection and use of information. Tis is nothing else but to say that the action itself expresses the attitude, even if it should not produce the intended, or foreseen, result. Injustice varies with the way that the outcome fgured at the decision-​making stage: if the institution aimed at it, the 53 See Pogge 1995, pp. 251–​255. ‘Due to some state measure’ means that the disadvantage is not due to natural factors that institutions are not required to ofset (see 2.5), or to free choice of the members of that group, according to a suitable account of free choice that takes into account relevant factors such as, for example, the importance of pervasive peer-​group pressure. 54 Social class must then, of course, be understood not simply in terms of distributive position; see ­chapter 6, 6.2.2, on social esteem and class.

Distributive and Relational Equality  47

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

action is more unjust than if it merely knew about it, or should have known about it.55 A second example clarifes further how the injustice of institutional action may vary with the adoption of certain means towards intended outcomes. Consider the case of a government aiming to improve the ‘competitiveness and efciency of the country’s economy’. One policy choice towards this aim is not to intervene in the confict between owners of the means of production and trade unions: the owners then manage to strike a deal that allows them to worsen the contracts of low-​skilled workers, lowering their pay, and making their position more precarious. Te result is due to the present balance of power in this confict, which is entirely known to the government. Owners win by threatening with closing the factories in question, and moving production abroad; given that the global economic situation makes this threat credible, trade unions give in. Te second policy choice is to abolish the laws establishing collective contract negotiations for the workers in question. Such institutional protection withdrawn, they all individually agree to the worse new conditions imposed by owners. Ex hypothesi, the distributive outcome here is exactly equal; as is, let us suppose, the overall cost of both policies. Yet the second policy is more unjust than the frst.56 Tirdly, consider cases of collective communication of attitudes, as a subclass of cases of collective attitude expression. Anderson and Pildes mention the example of token compensation of (justly) expropriated owners, which serves to communicate the state’s awareness that it is imposing a considerable sacrifce on them.57 Or take cases of public apology, in which a state 55 Scanlon argues that the agent’s intentions in performing an action do not, in many cases, ultimately matter for its rightness or wrongness. What they matter for instead is the assessment of the agent’s moral performance in carrying out the action in question, as distinct from the question of its permissibility; Scanlon 2008, pp. 1–​87. As said, this chapter does not defend the doctrine of double efect in its traditional form, but focuses instead on the meanings that institutional actions express. Scanlon concedes that such meaning may, depending on the relationship between agent and ‘recipient’ of the action, infuence rightness in some cases (ibid., p. 79) and agrees that discrimination is such a case, because of the judgement of inferiority expressed by it ibid., pp. 69–​74. 56 What if the best interpretation of the frst scenario established that the state was motivated by the best of intentions—​say, it genuinely believed that non-​intervention aimed at enhancing economic effciency in this and similar cases would ultimately be in the service of all its citizens, and to a reasonably equal degree? Would the expressive view not have to admit that, in such cases, the state commits no injustice at all—​which seems hard to swallow for egalitarians? Te answer is: it depends. If the state decides not to intervene because all reasonably accessible evidence favours this course of action, it indeed does not express any disrespect. Some forms of global competition may, afer all, present states with tragic choices; on a relational view, they should not be charged with injustice if they choose the lesser evil (injustice may reside elsewhere, for example, in the global regulation of trade, or the lack of adequate such regulation; see the Conclusion). If, on the other hand, the state’s choice not to intervene is based on a false belief that it could have avoided, it is unjust. 57 Anderson and Pildes 2000, p. 1526.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

48  Liberal Relational Egalitarianism recognizes the grievances of those against whom it has committed injustices. Of course, such apologies mostly also have distributive consequences, such as compensation payments; ofen they have to, in order to count as sincere. But the apology itself serves as an invitation by the state to the aggrieved individuals or groups in question to regard their relationship as altered, by accepting it—​as it does in the case of apologies between individuals, which aim at dissolving reactive attitudes of resentment, and at reinstating attitudes of mutual goodwill. Tis point leads back to the one discussed at the end of the previous subsection: if we think that states can apologize meaningfully, it must also make sense to regard them as appropriate targets of reactive attitudes. In the case of communicative action, the role of intentions is even clearer:  knowledge that an apology is unintentional destroys its value for the addressee. Take the case of somebody who actually wants to insult a foreigner, because she thinks their collision on the street was his fault, but instead apologizes to him, since she confuses the respective expressions in his language. If these examples are sound, then how could distributive theories react to them? One possible strategy is, of course, simply to dig in one’s heels. Distributivists focusing on the traditional currencies of justice mentioned in section 2.2—​opportunity for welfare, resources, or a mixture of both—​could insist that, since the point of institutions is to produce a just outcome, any meaning that institutional action may have, or any attitude it may express, should be understood in these terms: there is simply no principled reason why, if an institution is in a position to foresee and avoid an unjust outcome, and yet does not avoid it, this should not be, from the point of justice, fully on a par with intending that result.58 Accordingly, any contrary intuitions we might have about the preceding examples must either be unreasonable, or ultimately rest on empirical beliefs. A distributivist might respond to the group inequalities case that, in many cases in the real world, common disadvantage in one dimension leads to common disadvantage in other dimensions as well. For example, if women earn less for doing the same jobs as men, they will presumably also have worse access to positions of infuence than men. So it might be said that our intuitive objection to fnding more people of one kind in one disadvantaged group merely refects an implicit assumption that there will be similar fndings for other dimensions of disadvantage as well,

58 See Enoch 2007.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive and Relational Equality  49 so that they are very likely to be distributively disadvantaged overall. Tis is largely true; disadvantages tend to ‘cluster’.59 Similarly, in the ‘no collective bargaining’ case, we might fnd it probable that a government withdrawing one kind of institutional protection for low-​skilled workers will also withdraw other kinds, so that more distributive disadvantage will ensue. Tese responses on the part of the distributivist are empirically sound, in the frst case, or at least superfcially plausible, in the second. But, to put it cautiously, it is at least an open question whether a plausible such empirical counterstrategy is available in every signifcant problem case. For example, it is hard to see how it could work in public apology cases. And it is a further open question whether the possibility of hypothetical counterexamples, as long as they are reasonably realistic, would not nevertheless raise problems for theories of justice that aim at delivering fundamental principles, as the distributive theories mentioned in section 2.2 do. Tere is no reason why cases in which overall distributive advantage is equal, as it is in all examples employed in the previous sections, cannot arise.60 A second strategy is to try to deal with expressive considerations by extending one’s metric of justice so as to include a special class of social, cultural, and psychological consequences of institutional action.61 For example, one might seek to incorporate purely psychological notions of self-​respect and self-​esteem, focusing on how positively or negatively people view themselves and their way of life as a consequence of state action. But this would be to incorporate expressive considerations in the wrong way. Legal discrimination against a group, for example, constitutes a particular kind of objectionable relation between the state—​or ‘us with our citizen hats on’—​and the members of that group: the assignment of inferior legal status, which expresses the inferiority of members of that group to those of other groups. Whether or not this discrimination then also becomes socially salient, for example, in everyday relations with other citizens, in a way that has harmful psychological consequences for the members of that group, this need not be the case for the laws in question to be unjust. Tey would not cease to be unjust even if their targets would experience a boost in self-​esteem, 59 Wolf and de-​Shalit 2007, chs. 5–​7. 60 If one added the—​plausible—​assumption that, in many cases, it cannot be known with any degree of precision to which distributive outcome exactly an institutional action, or policy, will lead (see March and Olsen 1989, ch. 2), this would further strengthen the case for demanding that institutions at least make clear that they have acted on the right kind of reasons, and hence expressed a just attitude—​but I do not rely on this further assumption here. 61 For some suggestions along these lines, see Adler 2000, pp. 1434f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

50  Liberal Relational Egalitarianism for example, due to their solidarity in indignation. Conversely, people who merely feel disadvantaged because of a just policy would suddenly have a legitimate grievance against the state.62 Just take the case of people objecting to laws instituting special protections for ethnic or sexual minorities because ‘nobody cares about ordinary people anymore’.63 A third, more promising strategy overcomes strict distributivism by including certain social and political relations in a list of goods that should be equalized as a matter of justice. Examples of such theories are capability views including the factor of efective access to social and political institutions, for example Nussbaum’s,64 and theories that seek to extend the concept of resources so as to include ‘relational resources’ more generally.65 Important examples of such views, and their problems, will be discussed in c­ hapter 5 (5.4). However, as regards the good of access to basic structural institutions specifcally, it is worth already highlighting a problem that such attempts encounter. From a distributive point of view, the question to ask is why this factor should fgure in a list of things that justice should be ultimately concerned with: why should it be an intrinsic problem of justice that some people lack efective access to social and political institutions—​why have such institutions ceased to play a merely instrumental role, that of enabling people to get what they are entitled to?66 One way to answer this question is to argue that access to political institutions is a constituent part of what is objectively good for people. Tat would be to commit to a conception of the human good that includes an element of civic perfectionism. But this is a very controversial view. People may well need meaningful relations with others in some social institutions, but why in political institutions wielding coercive power?67 On the expressive view, no such perfectionist claims are necessary to account for the intrinsic importance of access to political institutions: by granting such access, institutions are merely expressing proper respect for 62 Anderson and Pildes 2000, p. 1574. 63 Tis is not to say that concern for the right kind of self-​respect is not a very important social egalitarian consideration; see 6.3.2 in ­chapter 6. 64 Nussbaum 2000, pp. 78–​80. Wolf and de-​Shalit adopt and expand this list, 2007, p. 56. 65 Cordelli 2015a and b. 66 For a refreshingly die-​hard instrumental account of democracy, see Arneson 2004. See also Arneson 2000, p. 342: ‘[T]‌he quality of relationships is itself reasonably regarded as instrumental to well-​being, not morally important in itself.’ 67 Sen’s capability view regards political rights as non-​instrumentally valuable because of their constructive role in conceptualizing other basic capabilities; Sen 1999, pp. 153f. However, he seems to hold that they also have non-​constructive intrinsic value—​without giving an argument for this, apart from pointing out that people do in fact seem to value them; ibid., pp. 151f. Tis may well be because they recognize their instrumental and constructive roles, or understand that they express respect.

Distributive and Relational Equality  51 those subject to their power, whether such access is also intrinsically good for people in any more substantive sense, or not.68 To make the diference even clearer, consider once again the role of intentions: on the civic perfectionist view, in the—​somewhat outlandish—​ scenario of access to institutions being granted unintentionally, individuals would still have (a part of) their objective good secured. On the expressive view, the institutions’ failure to grant access intentionally would mean that they have failed to express proper respect. Te point of this view is that we do not only have to distribute properly, we also have to treat people properly while doing so: the latter considerations function as constraints on the former.69 Finally, it should be noted that this assessment does not depend on taking the views mentioned earlier as mandating that citizens indeed participate in political institutions. Tey generally do not, and restrict themselves to requiring opportunities to participate, just as the expressive view put forward here does. Te point is that a justifcation for regarding precisely these opportunities, as opposed to others, as intrinsically required by justice still needs to be given, and the expressive view can deliver this justifcation without resorting to the claim that having them is intrinsically good for people.

2.6.3.  Expressing Attitudes without Relations

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

To conclude, a completely diferent kind of objection needs to be defused. Tis is the objection that the meanings of certain actions might well matter in their own right, and that there might thus be moral requirements against

68 For a defence of luck egalitarianism that hovers between the second and the third strategy, see Barry 2006. Barry recommends that luck egalitarians switch to an objective account of individual well-​being, including factors such as efective access to social, political, and cultural institutions, because ‘oppression is a form of bad brute luck that lowers the well-​being of individuals’ (p. 93). Tis is the third strategy. Te statement that ‘[a]‌n objective account of the good life [ . . . ] is sensitive to the efects of social oppression’ (p. 96, my emphasis) suggests the second. 69 Dworkin has objected to a classifcation of his theory as purely distributive along similar lines: he argues that the consolidated version of ‘equality of resources’ ofers a multidimensional theory of equality as the best interpretation of what it means to treat people as equals; 2003, p. 190. Rights to democratic participation are of intrinsic importance to justice because they express respect for individuals as agents, and they might require tolerating deviations from distributive equality of resources; 2000, p. 187. I accept that the argument of this chapter applies straightforwardly only to the articles on ‘equality of welfare’ and ‘equality of resources’ (Dworkin 2000, chs. 1 and 2), which have been at the centre of the debate about the ‘currency of justice’, and that the later version of the theory might not fully correspond to the ideal type of a ‘purely recipient-​oriented’ view laid out earlier. However, it still ofers a valuable perspective for critical assessment of Dworkin’s consolidated theory; see the critique of Dworkin’s conception of political equality in c­ hapter 7, section 7.5.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

52  Liberal Relational Egalitarianism expressing certain such meanings, which cannot be incorporated into, and accounted for by, distributive theories qua distributive theories (but only combined with them)—​but that this has nothing to do with relational equality, or social relations in general. Tat is shown by the fact that it is possible, and objectionable, to act in a way whose meaning, on the best available interpretation, is that some people have inferior status—​even though one has no relations worth the name with them.70 For example, state ofcials in one country might make, in public capacity, dismissive remarks about the inhabitants of another country—​say, about their typical lack of honesty, or work ethics—​with whom, as a matter of fact, they have no interactions, and over whom they seem to have no power at all. And that does seem objectionable, to an extent that warrants some sanctions (such as, at least, strict, publicly expressed, censure by other state ofcials). Te objection is certainly right to point out that the existence of social relations is not a necessary condition for the possibility of expressing objectionable meanings. However, for there to be expressive constraints of the kind discussed previously, as a matter of justice—​as opposed to mere reasons, such as avoiding gratuitous ofense, not to—​it is important that there be relations, or at least non-​trivial opportunities for interaction. What is more, the more such relations are characterized by the presence of power, and especially asymmetrical power (see ­chapter 3), the stricter the constraints on the more powerful party become. In all examples used in this chapter, the institutions in question were assumed to have very considerable power over the individuals in question. In the preceding example, this is not given, but it is still state ofcials who express that citizens of another country are inferior. Tere is a stringent constraint against this precisely because there are international and global interactions, and there is, as a matter of fact, a state system regulating them, to some degree, at least, of which the ofcials in question are part. And it is of importance for the justifability of that state system that it regards, and treats, all individuals around the globe, in important respects, as equals. Against that, private citizens are, I submit, not subject to a comparable constraint, whether or not there is some moral reason to object to any similar remarks of theirs.71 70 Lippert-​Rasmussen 2018b, pp. 90f., p. 200, fn. 24. 71 Lippert-​Rasmussen’s own example is that of a state failing to accept responsibility for a ‘tragically “successful” genocide’ it committed in the past, so that it has no relations with victims or their descendants because none are lef; ibid. Tis is a problematic example, for three reasons. Te frst is that the state did have relations with the group in question: the problem is precisely that it terminated them in the worst of all possible ways. Tat makes for an intuitively salient diference to cases where

Distributive and Relational Equality  53 So the expressive perspective focusing on the meaning of actions is indeed particularly salient for justice in cases in which there are social relations. It is here that it gives rise to strong constraints.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

2.7.  Conclusion Te argument of this chapter has made clear that expressive considerations, focusing on the attitudes that institutions express in their treatment of individuals and groups, play an irreducible role for justice that is overlooked by purely recipient-​oriented views. Te next step is the following: in order for the expressive perspective to ground a viable alternative approach to justice, it must be possible to develop a conception of justice on its basis. Te principles of justice proposed by such conceptions cannot be as simple as those proposed by purely recipient-​oriented theories; but they must be able to give both reasonably general and precise guidelines for answering the question of when and how institutions express proper respect and concern in their treatment of individuals and groups. So, to conclude with a brief preview, here are three directions into which candidate conceptions could go. Te frst possibility is the radical social egalitarian view mentioned earlier (2.6.1):  according to such a view, individuals should display egalitarian attitudes towards each other in their actions across all areas of social life, and institutions should express equivalent collective attitudes. Views such as Tawney’s (1.2 in c­ hapter 1) are of this kind. A second possibility is a view that focuses on the communication of institutional attitudes to individuals and groups: on this view, institutions would have to communicate to individuals and groups equal respect and esteem for them and their ways of life, such as some conceptions of the ‘politics of identity’ demand. However, this view may be in danger of overestimating the importance of institutional communication. Communication is especially important to signal a reversal of past unjust action, such as in apology cases, no relations exist, or existed. Te second is that the action in question is, as a crime against humanity, of the kind where a failure to accept responsibility might taint present relations of that state with all other agents around. Te state might have a duty to redeem itself in front of all others, whose fulflment requires, at the very least, such acceptance, to repair its relations. Te third is that such a failure also especially impairs present relations with some specifcally situated other agents or groups, such as ethnic minorities within the state now, which might reasonably regard themselves as threatened by it, in the absence of credible assurance that it has now changed. So there are several particular, relation-​dependent, reasons in play in this example which are not in play in the example of state ofcials making ofensive remarks about foreigners.

54  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

or to accompany actions that impose sacrifces on individuals and groups, such as in expropriation cases. But if institutions express proper respect for people in their treatment of them, they should not also need to say this constantly; and if they do not, then communication to the contrary merely seems hypocritical. A third possibility is that of a conception of egalitarian relations as a matter of liberal social justice, according to which institutions treat individuals with the proper attitude when they ofer them equal structural protection against subjection to others and unjust treatment by them, acting alone or in concert: against ending up on the wrong end of inegalitarian relations, such as, in particular, domination, and against subjection to pervasive inegalitarian norms of social status. Tis is the alternative that the next chapters will pursue.

3 Liberal Non-​Domination 3.1. Introduction

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.1.1.  From the Expressive Perspective to Liberal Relational Equality, and Non-​Domination Te expressive concept of justice directs our attention to the way in which individuals are treated within social interactions and relations, as of independent importance for social justice, and the previous chapter has shown that it is applicable also to the way social and political institutions treat individuals in the process of structuring cooperation, and distributing its goods and bads. As shown, this is sometimes a matter of how institutions relate causally to goods shortfalls, where the natural-​social distinction is in play. Where it is not, it is a matter of the quality of the attitude that institutions express in their treatment of individuals and goods: of the social hierarchies it implies, makes possible, or reinforces. Certainly, the expressive perspective does not, by itself, strictly imply relational equality, or pin down the precise role of institutions for it. First, it needs to be combined with a principle of moral equality of persons and the principle that, therefore, they have to be treated with equal respect (2.2 in ­chapter 2). If there were good reasons to think that individuals are of fundamentally unequal moral worth, then the institutions of a society would have to express that in their treatment of them, by instantiating social and political hierarchies appropriate to that moral inequality. Powers, ofces, and social honours would have to be allocated so as to express this ranking of persons. Tat could, in principle, be compatible with aiming at forms of distributive equality for all, too, for example, in economic resources. We could insist that the appropriate rewards for superiority are only such powers and honours, not any further rewards, and that such a ‘benevolent’ hierarchy thus take all the measures it can to make sure that those subject to it fare equally well in

Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0003

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

56  Liberal Relational Egalitarianism dimensions not centrally connected to expressing such proper respect for diferential worth.1 If, however, individuals are moral equals, then institutions have to treat them as such, and a central requirement will be to make sure that they encounter each other in relations of appropriate social and political equality. Te question of which kinds of power relations between them are compatible with, ruled out, or called for by that ideal is then the frst question that needs to be answered, diferent from the frst question of purely recipient-​oriented views, which asks which distributive outcomes best express moral equality. Tis establishes the link between the expressive perspective and a general requirement of relational equality between individuals; however, by itself, it still tells us little about the content of these requirements. It does not answer the questions of how exactly power is to be dispersed and controlled in society, of which social relations are unjust, which permissible, and which positively called for by social justice, and of which roles both basic institutions and individuals have to play in securing the latter. Recall that the preceding chapter proceeded in an intuitive way. It proposed a preliminary way of ordering injustice assessments in order to tease out what is distinctive about the expressive perspective. To supply a principled basis for such assessments, we need a conception of relational equality as a matter of social justice. Working out the most important part of this conception, pertaining to power—​its requirements of non-​domination—​is the task of this and the next chapter. But arriving at such a principled basis frst requires the injection of more content, in the form of background conceptions of the person, and the nature of society. Interpretations of the attitudes expressed in actions can then be justifed in a more than merely intuitive way by tracing their connection to these conceptions. Te interpretations so arrived at—​for example, that diferent ways of withholding important benefts (recall the V scenarios) are diferentially unjust, depending on how disrespectful they are—​must be capable of being shared by the individuals subject to such actions; otherwise, claims about expression have no traction. For that to be the case, they must be, in the frst instance, accessible to them, based on such background conceptions of the person, 1 One might object that this is not fully possible—​that powers and honours will always give their bearers some opportunities to allocate to themselves more of other benefts, too. Tis is most likely true, but recognition of this fact would still commit this ‘benevolent’ hierarchy only to do all it can, within its design, to ward of such abuse, not to abolish ranks, because instituting them would be dictated by the dominant expressive requirement of acknowledging diferential worth.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  57 society, the institutions structuring the latter, and the kind of tasks one should expect them to fulfl. As such, background ideas for the undertaking of developing a conception of relational equality, this chapter proposes, in a Rawlsian vein, the ideas of individuals as free and equal members of society, endowed with certain moral powers; the idea of society as a scheme of cooperation; and the idea of a just society as a fair scheme of such cooperation. Te following sections will fesh these out further. Tese are not proposed as background ideas that are in fact already shared among the relevant individuals, or even a large proportion of them, within the kind of society in which we are primarily interested in: contemporary large-​scale societies characterized by an intricate division of labour, with a set of basic institutions organizing them efectively (see 1.4 in ­chapter 1). I am not sure if they are; but it is not necessary for our enterprise.2 Its point is not to analyse and systematize how certain ways of treating people, and relating to each other, are understood by people in a given social setting, or can be understood by them without challenging any more deeply seated convictions, but, in a more ambitious vein, how they should be understood, drawing on the best arguments about relational equality and social justice we can make, on the basis of these background ideas. For that enterprise to be viable, and to deliver a candidate conception of relational equality for people in such societies, it is thus enough that the background conceptions in question are accessible to, and potentially shareable by, such people. Tat I take to be the case—​that the idea that members of society are free and equal in virtue of possessing certain crucial moral capacities, and that society is an enterprise of reciprocal cooperation that should be structured fairly can be understood by pretty much all of them (whether they already sign up to it or not), and I will not argue for this claim. Nor will this chapter, or other chapters, deliver a direct argument for why the Rawlsian conceptions of person and society are preferable to others that might be similarly accessible, for example, through a comprehensive comparison, issuing in a full balance sheet of the advantages and disadvantages of competing conceptions. What they will do is deliver an indirect argument as to why these conceptions and a concern about egalitarian relations, and 2 Tus, it difers from Rawls’s project in Political Liberalism (1996) to the extent that that project rests on the claim that these background ideas are in fact shared within the public political culture of such societies. It also difers from that project insofar as it does not argue that the resulting conception of relational equality is, in the end, shareable even between adherents of very diferent comprehensive philosophical doctrines—​even though that would certainly be nice (see ­chapter 1, fn. 14).

58  Liberal Relational Egalitarianism equal non-​domination, in particular, mutually support each other, and why there is stronger such internal support on this view than on rival views, such as Pettit’s neo-​republicanism. In this sense, liberalism is assumed, and the main argument of this and the following chapters consists of showing how it can serve as the basis for a demanding and plausible justice-​based conception of relational equality.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.1.2.  Te Argument of the Chapter Tis chapter is structured as follows. Section 3.2 feshes out the background conceptions of person and society further, and motivates its focus on equal non-​domination. A natural frst candidate for relational equality would seem to be comprehensive equality of power within, and over, social and political arrangements; however, liberals have very strong reason to recognize that some inequalities of power serve valuable purposes for social cooperation. So it is better to focus on equal non-​domination, which requires equal protection against unequal and arbitrary power, leaving open, for the moment, whether equality of non-​domination itself requires a form of equality of power, and if so, of which kind. Te central section 3.3 develops a liberal conception of non-​domination. It spells out what is unjust about domination (3.3.1), and what a concern for non-​domination has to range over: which choices and claims ought to be protected against others’ arbitrary power (3.3.2). Tis liberal account of non-​domination is distinctive in injecting, in order to identify this range, considerable substance into the concept of domination—​substance that is derived from the background conceptions. Furthermore, it does not hold that concern for non-​domination exhausts social justice. Both claims set it at odds with prominent neo-​republican accounts of justice as non-​domination, with which section 3.4 will compare it, in order to motivate it further. Frank Lovett’s conception of social justice as non-​domination (3.4.1) requires only that power be efectively constrained, with no regard for the aims it serves. Tis must leave liberals about social justice unsatisfed. ‘Non-​domination as democracy’ (3.4.2) is a more promising contender; on this view, questions of non-​domination ultimately reduce to questions of political justice, and political justice must demand a thoroughgoing form of equality of political power proper to ensure that any relations of unequal power within everyday social cooperation are truly justifed. Tis view, however, not only gives insufcient guidance as to how such social

Liberal Non-Domination  59 cooperation ought to be organized; the form of political equality it must require is not attainable in complex societies characterized by an intricate division of social and political labour. Tis result thus both underscores the need for substantive requirements of non-​domination of the kind delivered by the liberal view, and further vindicates the caution expressed towards equality of power in section 3.2. Philip Pettit’s more recent proposal for a conception of social justice as non-​domination (3.4.3) recognizes the need for such substantive requirements, but is arbitrarily under-​demanding, in a way that should trouble liberals and republicans alike, and an important part of the reasons for this is to be found precisely in its eschewing background conceptions of the persons and society of the kind relied upon by the liberal view. Te analysis of neo-​republican views thus corroborates the case for the liberal, substantive conception of equal non-​domination, and lays the basis for further developing its demands of protection in the next chapter.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.2.  Equality of Power and Social Cooperation Let us then work out the background conceptions of the person and society, and the resulting role of social justice in more detail. Doing so will enable us to see why liberal relational equality cannot demand strict equality of power between individuals:  permitting only forms of cooperation that are consistent with such power equality would unduly endanger the possibility of forms of, and gains from, cooperation, which liberals have very strong reason to want. Tus, a focus on equality of non-​domination is more promising. Te task of a conception of social justice is to structure society so as to institute a fair scheme of reciprocal cooperation among free and equal individuals. Individuals are free and equal insofar as they possess two fundamental moral powers.3 Te frst is an efective sense of justice, according to which individuals are able to identify, and act with due regard to, others’ justifed claims arising out of social cooperation. Tis power includes a component of care: to have a complete sense of justice, individuals not only need to be able to understand a general conception of justice. Tey also need to be able to identify the claims and needs—​needs insofar as these are the basis of claims made within a context of cooperation—​of particular others with whom they stand in a relation of mutual dependency. And they need to be able not only 3 Rawls 1996, p. 19.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

60  Liberal Relational Egalitarianism to identify these claims, but also to perceive them as practically salient, and to be reliably motivated by them.4 Te second moral power is the capacity to develop, pursue, and revise a conception of the good. Individuals’ fundamental interests are, on this basis, to develop and pursue such a conception of the good in the most favourable possible conditions, consistent with appropriately favourable such conditions for all, and to be able to express their sense of justice—​which precisely enables, and obligates, them to identify, and honour, these conditions—​by bringing it to bear on the social and political arrangements they live under. Te idea that the role of social justice is to set up and regulate social cooperation so as to enable the development and exercise of everybody’s moral powers, and especially of the second moral power of developing, pursuing, and revising one’s own conception of the good, is a specifcally liberal ideal: it is a liberal interpretation of the task of social justice of enabling all members of society to be a full participant in it (see 1.2 in c­ hapter 1). On this liberal ideal of justice, structuring social cooperation fairly by protecting these fundamental interests for all is to express proper respect for its participants. Te task of a conception of relational equality, as a central building block of such a conception of justice, is that of structuring relations of power and status between individuals so as to express such respect. Of these two, relations of power are naturally the frst subject that has to be tackled: power to set the terms of cooperation and to enforce them, as well as power within cooperation conferred by whatever terms are currently in place. So how exactly do such relations of power have to be structured? It seems that if relational equality is our aim, then comprehensive equality of power among individuals should be the frst candidate for consideration. What could be more relationally egalitarian, and more accurately express respect for individuals as free and equal, than enabling all individuals to protect their interests on the basis of equal power over and within the social and political arrangements they live under, so that they have equal opportunity to infuence, by their own doing, the nature of their society? Such an ideal of equal power would require that individuals retain it on a continuous basis. It would, taken literally, require an invariant opportunity to exercise the same degree of infuence on the structure of society, and to that extent, to be extremely suspicious of any standing positions and ofces conferring superior 4 Tis is not to argue that the virtue of care can be entirely subsumed under justice, but merely to make clear that the diference between the two can be, and sometimes is, overstated.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  61 power and authority, in particular if they are constituted, and backed up, by a formal, coercive institutional structure regulating that society. One straightforward proposal for realizing this kind of ideal is, thus, cooperative anarchism, which demands the absence of such a formalized, coercive institutional structure, in favour of voluntary interaction and exchange between individuals on an equal footing—​stabilized merely by a system of social norms favouring such free exchange between equals.5 Now, comprehensive equality of power is, at least at frst glance, intuitively attractive, and therefore, so is cooperative anarchism. However, from the point of view of liberal relational egalitarianism, it encounters two serious difculties. Te frst is the familiar and general one that aiming for such equality of power may be self-​ defeating, and lead to great power inequalities. Tis danger is clear in the case of anarchism, where the absence of regulated coercive structures may enable some individuals, or, more likely, gangs, to seize disproportionate power over others. Warding of this danger will require stabilizing institutions, and these may involve at least some positions of superior power (such as guardians). If these considerations are sound, then not all inequality of power may be a net detractor from the ideal: some inequality may be necessary to prevent even greater inequality.6 At the very least, adherents of equality of power have to show how it can be stable without any such inequality. Te second difculty is more directly linked to liberalism, and questions whether equality of power is desirable at all: even if the ideal is not practically self-​defeating, it is very unlikely to be able to give liberals what they want. Liberals want, on grounds of enabling, and protecting, everybody’s second moral power, an extensive set of liberties which give individuals a wide and diverse array of opportunities to develop and pursue their own conception of the good. Tis may not be available insofar as maintaining equality of power might require tight social control of all by all, and a resulting high degree 5 For two anarchist proposals for social equality, see Kropotkin 1976 [1902] and 1906, and Bakunin 1996. For an earlier socialist view containing some anarchist elements, see Proudhon 2011 [1846]—​ although it is debatable whether these elements should be given much weight in Proudhon’s overall view. Neither of these three views fts the stylized anarchist view of equality of power proper, drawn on in the main text, perfectly. It is a view constructed for the purposes of contrast with liberal relational egalitarianism, in order to bring out the main questions the latter has to answer. 6 I thank Hillel Steiner for suggesting this way of putting the point. If we take as the starting point for attempts to implement power equality the basic structure of existing, large-​scale, and highly differentiated and economically developed societies embedded in a context of other such societies (see 1.4 in c­ hapter 1), it seems even clearer that, in this case, such attempts might lead to more power inequality. Tis would be both for reasons ‘from within’, as the break-​up of existing power structures would lead to diferent components of such a society struggling for power, and ‘from without’, as other states would seek to fll any resulting power vacuums through political and economic colonization.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

62  Liberal Relational Egalitarianism of conformity with reigning social standards. Furthermore, and more importantly still, liberals will want a degree of productive development which guarantees not only a sufcient level of basic goods for all, but also enables a suitably rich array of social practices, and diverse social groups and associations, within and through which a wide array of conceptions of the good can be pursued. Tese will very likely not be available. Such a level of material and cultural productivity arguably requires the capacity to organize an intricate division of societal labour. Such a division necessitates a complex material and legal infrastructure instituting many positions of diferential power and specialized knowledge, in both the political and the social and economic spheres (such as bureaucrats, judges, managers, experts in the educational sector, urban planners, and so on), in order to solve coordination problems stably, and render diverse material and cultural resources continuously accessible to all. For example, a society of subsistence farmers, or hunter-​gatherers, may well be socially egalitarian, but cannot yield the rich possibilities for living a life of your choosing that liberals prize (even if it should otherwise be liberal in its social mores, which it probably will not be).7 Admittedly, these considerations rely on some general empirical assumptions about the available possibilities of structuring large-​scale cooperation in production. We should be glad if they were proven wrong, as this would open up a richer feld of candidate conceptions of cooperation for in-​ depth investigation. But if they are right, liberal relational egalitarians will have to accept the existence of at least some standing positions of authority and power within social cooperation, and the fact that, though opportunity to attain them may be equal, not everybody might get to occupy them, and a fortiori not for the same amount of time during their life. We have to develop our conception with this in mind, and look for a justice-​based conception of relational equality ftting for a society able to fulfl the preconditions of liberalism. Tis does not yet rule out that such diferential positions can only be justifed if there is indeed, at some special level, equality of power proper, so that, say, all power inequality within social cooperation is ofset by full equality of political power over its terms. We will later encounter conceptions of non-​ domination which precisely lead to this demand (3.4.2); and the liberal view, though having to stop short of it, must itself assign a privileged place to a demanding form of equality of political opportunity (3.3.2; 7.3 in c­ hapter 7). 7 Anderson 2017b, p. 50; see also Anderson 2012, p. 51.

Liberal Non-Domination  63

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Nor is it to deny that all relations that instantiate some inequality of power, wherever it is to be found, not only have to be justifed, but may, even if they are, have an aspect that is regrettable. But regret is cheap. Te important point is that not all such inequalities can be incompatible with liberal relational equality itself; they cannot be if they are necessary for functioning social cooperation, and can be constrained so as to serve justifed purposes, and appropriately controlled. Te alternative would be to develop a conception of relational equality as comprehensive power equality, without regard to constraints and imperatives following from a liberal conception of the purpose of social cooperation, in order to then balance it with the latter, to see how much of both we can plausibly have.8 Against that, the task of this and the following chapters is to develop a conception of relational equality that is, from the start, tailored to the latter. We have to focus on the question of which power inequalities liberals have particular reason to worry about, and what is specifcally unjust about them.9 Liberal relational egalitarianism thus asks which positions of unequal power, if any, there have to be, and which choices, interests, and claims of individuals they may, and which they must not, range over, and how that power must be circumscribed and checked so as not to be arbitrary. Tese are the main guiding questions for a conception of non-​domination. Accordingly, we have to turn to explaining the procedural and substantive aspects of the concept of non-​domination, and to examining how these have to be understood within a liberal framework.

3.3.  Te Injustice of Domination Domination is subjection to arbitrary power; you are dominated if you are caught in a social relation with another agent who is capable of exercising that power in a way that imposes signifcant costs on you, without adequate checks or control (especially by yourself). Tis is a broad way of describing domination; we will narrow the concept down, identify diferent variants of it, and select among them, as we proceed. ‘Domination’ is a term with clear 8 Tis would be a ‘pluralist’ strategy (2.4 in c­ hapter 2). It will be taken up again in c­ hapter 5 (5.3). 9 Does this mean that this conception is liberal frst, and relational egalitarian only second? Tis is a possible description of it, as the liberal conceptions of the person, society, and the role of justice, are, as seen, taken to be fundamental, and assumed. On the rival, pluralist view, we do not know what is frst and what is second; it is all a matter of balancing in given cases.

64  Liberal Relational Egalitarianism negative connotations. It seems difcult to declare oneself in favour of domination, intuitively perhaps just as difcult as declaring oneself in favour of injustice. Domination gives reason for complaint.10 For our purposes, we need to know what the specifc injustice of dominatory relationships consists of (3.3.1). Tis informs how the range of concept has to be construed: which kinds of individual choices, interests, or claims must be protected from it (3.3.2). Te resulting liberal conception of non-​domination difers in important respects from prominent conceptions of non-​domination put forward by contemporary neo-​republicans, who have done most to revive and fesh out the idea: it is a substantive conception, whose role for social justice is, at the same time, more circumscribed than on such accounts. Te subsequent section (3.4) further substantiates the case for the liberal conception by way of comparison with these.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.3.1.  Domination as Relational Inequality According to Philip Pettit’s seminal defnition of domination, one agent dominates another ‘to the extent [ . . . ] that they have the capacity to interfere on an arbitrary basis in certain choices that that the other is in a position to make.’11 Tis refers not only to the choice-​set the other person faces at any given moment, but also to the choices that she could make in the future, if she wanted. It thus includes power to structure other people’s choice sets, from the outset. Generally, dominatory power has to satisfy two conditions: the frst, as said, is an arbitrariness condition: interference is not subject to suitable direction, checks, or control. Tis is a procedural condition.12 Te second is that the power of interference in question must range over choices that are of some importance for the life of the chooser; its exercise must be able to afect her negatively to a signifcant extent.13 Tis is a substantive condition. In the 10 McCammon 2015, p. 1029. 11 Pettit 1999a, p. 52. 12 In earlier work, Pettit defnes arbitrariness as not being ‘forced to track’ relevant interests of the chooser; ibid., p. 55. In later work, he drops the concept of arbitrariness: domination there consists of the capacity for interference not being controlled by the interferee; Pettit 2012, p.  58. Other prominent accounts defne arbitrary power as power ‘not externally constrained by efective rules, procedures, or goals that are common knowledge to all persons or groups concerned’; Lovett 2010, p. 111 (see 3.4.1); or wielded in ‘deliberative isolation’ (McCammon 2015, p. 1046), that is, without accountability—​specifcally to those afected (see 3.4.2). 13 Pettit, as seen (fn. 12), cashes this out with the help of the chooser’s interests; Shapiro with ‘basic interests’ (‘security, nutrition, health, and education’); 2012, 293–​335, p. 294. Lovett demands that

Liberal Non-Domination  65 following subsection (3.3.2), we will specify these conditions further for the liberal view of social justice. In the recent neo-​republican literature, non-​domination has been put forward as a conception of freedom that can serve as the ‘paramount value’ of political morality.14 As such, it is meant to be central to both political justice, or legitimacy, and to social justice, where the former concerns relations of power within political decision-​making, and the latter relations between individuals in social life more generally, outside of political decision-​ making.15 In both domains, domination has to be minimized,16 or individuals must be protected against it up to certain thresholds, which specify when protection is intense enough to count as satisfying the requirements of social justice, or legitimacy.17 Tese requirements of intensity spell out how non-​domination is a modal notion:  it demands robust protection against others’ power in a wide range of possible scenarios, and is especially directed against the need to rely on others’ contingent goodwill, or favour, in any of those scenarios—​in the running of institutions of political decision-​making as much as in the pursuit of one’s personal projects and aims. Tis is a nice, neat picture, and the prospect of a single value unifying political morality and issuing similar demands in all its domains is certainly philosophically attractive. However, we should resist its temptation, and pause to refect further on the specifc evil of domination. In particular, we need to

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

the dominated be ‘dependent’ on the relationship with the dominator ‘to some degree’; 2010, p. 96. McCammon proposes to generalize over these (and other) defnitions by requiring the dominator to have ‘impositional power’; 2015, pp. 1037f. 14 Lovett and Pettit 2009, p. 17. While neo-​republicans do not deny that there are other issues of importance to political morality, they holds that these will, in efect, be taken care of if non-​domination is taken care of, so that they do not need to fgure in frst principles. Tis claim is present in Pettit’s work from the early phases. In Republicanism he argues that it is a value ‘with a distinctive claim to the role of yardstick for our institutions’; Pettit 1999a, p. 80; see also ibid., ch. 5. Pettit’s construal of non-​domination, and of its role for justice, has undergone several variations over the course of his works, but this claim remains central to this work as a whole. See also Pettit 2012, 2014, p. xix, and 2017. 15 Republicans sometimes phrase the former as a concern with avoiding imperium—​public power must not be arbitrary, and properly controlled by those subject to it—​and the latter as a concern with avoiding dominium—​there ought to be no masters and subordinates in social relations in everyday life; Pettit 1999a, p. 13; and see Pettit 2012 chs. 2 and 3 for a corresponding distinction between political legitimacy and social justice. For liberals, both concerns are part of a wider concept of social justice. Tis terminological diference is of no consequence for the argument of this and the following chapter; liberals and republicans disagree about the priority relation between these two components (see 3.4.2; 7.2 in c­ hapter 7). 16 Pettit 1999a; Lovett and Pettit 2009, p. 19; Lovett 2010. 17 In Pettit’s case, the ‘tough-​luck’ test demands equality for republican political justice (see 3.4.2), while the ‘eyeball test’ sets a sufciency threshold for matters of republican social justice (see 3.4.3); Pettit 2012, pp. 84, 176.

66  Liberal Relational Egalitarianism refect on the kind of relational inequality which domination constitutes, and the complaint it gives rise to on that basis. One feature that is prominent in the republican tradition as much as in its contemporary revival is a focus on asymmetrical power of interference: it is one side that has the power to arbitrarily interfere. Tis is explicit only in some defnitions of domination,18 while not being fully explicit in others;19 but it is not only an explicit, but a central feature of all neo-​republican accounts of the reasons for being so prominently concerned with arbitrary power. Republicans focus on unequal, unchecked power relations because of ‘the evil of being subject to a master’20; ‘mastery, dictatorship, [and] patriarchy’21 are the paradigmatic relations which they object to. More positively,

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

[b]‌ eing a free person became synonymous with being sufciently empowered to stand on equal terms with others, as a citizen among citizens [ . . . ]. Te idea that citizens could enjoy this equal standing in their society, and not have to hang on the benevolence of their betters, became the signature theme in the long and powerful tradition of republican thought.22

Coupling this with the specifcally late modern core idea of social justice that it is the state’s task to guarantee for all citizens the background conditions to fully participate in society (1.2 in c­ hapter 1), this gives rise to ‘the idea [ . . . ] that the state could provide for all citizens in such a measure that they would each be able to walk tall, live without shame or indignity and look another in the eye without reason for fear or deference’.23 At the same time, contemporary neo-​republicans accept, as liberals do (3.2), that inequality of power is not necessarily unjust: it can be justifed.24 But if it has to exist, then we have all the more the reason to make sure that it cannot be arbitrarily exercised—​hence the focus on domination, as a specifc

18 Lovett 2010, p. 82. 19 Pettit’s diferent defnitions. 20 Pettit 2012, p. 2. 21 McCammon 2015, p. 1030. Objections to ‘patriarchy’, specifcally, are, of course, a contemporary, neo-​republican addition to the tradition. 22 Pettit 2012, p. 2. 23 Pettit 2012, p. 3. Te last clause is the ‘eyeball test’. 24 See, for example, Lovett 2010, pp. 83f. Pettit, in one of the many contrasts between liberal and republican freedom which he draws, argues that ‘on the liberal understanding, asymmetries in interpersonal power are not in themselves objectionable’ (2012, p. 11), which would seem to suggest that, on his republican understanding, they are. But in his constructive discussions of non-​domination he only ever objects to ‘arbitrary’ or ‘uncontrolled’ asymmetrical power.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  67 kind of power hierarchy that is to be ruled out.25 Tus, if non-​domination is freedom from power hierarchy, specifcally, then we can draw three important conclusions for the place it should enjoy within a liberal conception of social justice: one positive, one negative, and one mixed. Te frst, positive, one is that we should share the republican conviction that domination is a particularly urgent concern, a particularly objectionable form of inegalitarian relation. If individuals are to be free and equal participants in social cooperation, then none should be able to boss others around without due regard for their interests, and/​or being accountable to them. All power inequality needs justifcation, and this also commits us to making sure that whatever power inequalities are allowed to stand are tightly reined in and controlled, so as not to be able to spill beyond the domain where they are justifed. Leaving some individuals at the mercy of others expresses disrespect for them; it is a particularly conspicuous and objectionable form of disregard for their equal standing.26 Tis also helps to explain why, for freedom from domination, the central concern is not whether the arbitrary power in question is indeed probable to be exercised, or not.27 One reason why this is so is that inferiors might render interference more unlikely by cosying up to their masters. It would seem odd to regard this as a gain in freedom. But even if we leave this case aside—​let us assume that the master in question is stably benevolent, just not disposed to interfere, and let us concede, as we should, that in one, straightforwardly liberal, sense, there is a gain in freedom here—​one is nevertheless at her mercy, while she is not at one’s own, and this is still an insult to equal standing. How exactly non-​ domination between individuals has to be specifed for liberal purposes, and what it ranges over, will be the subject of the next subsection.

25 In Kantian accounts of non-​domination, or independence, on the other hand, asymmetrical power is not essential; the problem here is dependence on any non-​‘omnilateral’ will, whether it be the will of a more powerful agent, or not; Kant 1996 [1797], p. 48. For a recent exploration, see Ripstein 2009. For criticism, see Kolodny 2019. 26 Recall the V scenarios 2 and 4 in 2.3 (­chapter 2). 27 Tere is a lively debate about whether adherents of a traditional liberal concept of freedom as non-​interference can satisfactorily account for republican concerns about domination by explicitly incorporating the probability of interference into their freedom assessments; see, for the position that they can, Goodin and Jackson 2007; Carter 1999, pp. 237–​245; and Kramer 2003, ch. 2; and for the position that they cannot, Pettit 2008. As regards the specifc evil of domination as power hierarchy, I side with the latter; but it is not the only kind of freedom which matters (see what follows in the main text). Against that, List and Valentini, 2016, pp. 1053f., propose a unifed analysis of freedom as modally robust non-​interference, within or outside power hierarchy; however, that analysis is meant to be descriptive, and thus neutral on the question of whether such lack of modal robustness is always an evil, or gives rise to any complaint.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

68  Liberal Relational Egalitarianism Te second conclusion is, however, negative: on the liberal conception, these considerations do not compel us in any way to restrict concern of social justice to the elimination, or reduction, of dominatory relations, understood as such asymmetrical power relations between agents. For liberals, there is no reason to do so—​to assign non-​domination the position of exclusive master value. Individuals ought to cooperate under the constraint of fairly enabling the development and exercise of everybody’s moral powers, so that all can live the life they want to live. Tat would, rather straightforwardly, seem to require more than not being subject to others’ superior and arbitrary power. It entitles individuals to certain things, in that sense, by themselves—​such as access to the vital vitamin V (2.3)—​and not only because lacking them might, or even likely will, make them objectionably dependent on the whims of others. Tat is only one kind of injustice, albeit, as just shown, a particularly pressing one. Similarly, it also requires that we are kept from interfering with each other in certain domains even if our power to do so should be exactly equal, and exactly equally (un-​)checked. For example, we should want the state to make sure that all of us buy only safe cars, and maintain them properly, so as to minimize dangers of injuries to others, and not only to make sure that the bigger and faster cars that the richer can buy do not pose particular dangers for less well-​armed participants in trafc. Of course, we could re-​describe the more egalitarian scenario of mutual endangerment in trafc as one of ‘mutual domination’;28 and, insofar, as neo-​republicans want to insist on non-​domination as the exclusive master value of political morality, there is considerable pressure on them to do so. However, in that case, the intuitive connection between domination and hierarchy is lost. Consequently, there is no argument any longer as to why we should care about modal robustness of non-​interference, in such a scenario, as its absence does not express any inequality of standing. Liberals should rest content when they have made such interference appropriately unlikely (as unlikely as it is reasonable to wish, whatever that degree of probability is).29 Ensuring reasonable certainty of non-​interference is not the same as ensuring non-​domination, even though the two concerns will very ofen overlap signifcantly. Tis is not to rule out that concern about non-​domination might turn out to be very demanding; we will see in the next chapter (4.3 and 4.5) that it is. 28 See Schmidt 2018. 29 Schmidt concedes that traditional liberal concern with rendering interference appropriately unlikely may, in substance, fully accommodate problems of ‘mutual domination’; ibid., p. 194.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  69 Neither is it to rule out that there might be, in many cases, non-​domination-​ based rationales for requirements, which, on a liberal view, can also be grounded diferently, and that, sometimes, these requirements could even be extensionally equivalent. Te point is that there is no reason to restrict ourselves to these rationales, on the liberal picture. Nor is it to rule out that the concept of domination could perhaps be extended in various ways so as to increase its catch of intuitively unjust social scenarios, as in the trafc case described earlier. Here, the point is simply that, on a liberal view, there is no reason to embark on such a theoretical enterprise of enlargement,30 while, from the point of view of relational egalitarianism, we have positive reasons not to do so, because it risks obscuring the distinctive evil at stake. Te third, ‘mixed’, conclusion is that this account of the specifc evil of relational inequality which is at the heart of concern about domination should leave us somewhat puzzled about our relation with the state, and the demands applying to this relation. Historically, republicans have been primarily concerned with some individuals, such as monarchs, or groups, such as the aristocracy or undisciplined mobs, seizing political power and dominating others politically. Insofar as the driving concern here is that some individuals, by capturing institutions, or through membership in particularly powerful groups, obtain asymmetrical and arbitrary power over others, these cases present no problem. Tey fall straightforwardly under the relational egalitarian rationale for objecting to domination. In the next chapter (4.2), we will investigate more closely how exactly participation in a powerful group agent can put an individual agent in an objectionable position of superior power over another, even if she does not wield that power all on her own. However, contemporary neo-​republicans extend concern about political non-​domination to concern about domination by the state as such, as a particular collective, or group, agent.31 It is that which is puzzling, for relational egalitarians. Te reason for this is that, on the face of it, it makes no sense to demand that individuals and the state (or the ‘basic structure’; see 2.6.1 in ­chapter 2) encounter each other on an equal footing, grant each other equal social standing, or enjoy relations of equality in any other way—​at least, if we are 30 Sometimes, neo-​republicans present their concern about free persons as those who are not unilaterally dependent on their ‘betters’ (see fn. 22), more modestly, as a heuristic that is useful for arriving at a full account of the freedoms that all should have; Pettit 2012, p. 82. Te point just made disputes that this heuristic is useful across the board. 31 Pettit 1999a and 2012.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

70  Liberal Relational Egalitarianism afer any reasonably literal sense of equality.32 Te state has to be at the service of, and responsive to, its citizens in various ways, but even when it is, that does not make citizens and the state equals. As noted in c­ hapter 2, fulflling the standards of justice is the purpose of existence of the state (recall the high injustice ranking of V scenario 1, where the state acts with direct hostility towards some of its citizens); and we should make sure that the state not only gives individuals what they are due, but that they get robust rights to hold it to account and demand their due where the state falls short—​paradigmatically, in the form of both democratic rights and access to an independent judiciary able to call other agencies of the state to order. Furthermore, states not only have to make sure that their citizens encounter each other on a footing of equality, but have to treat them as equally possessing the capacities needed to live lives of their own (the moral powers), and we will see in the next chapter (4.3.2) how exactly such basic requirements of respect constrain how the state may protect individuals against domination by others. Finally, democratic rights are needed not only to make sure that the state as such is responsive to its citizens; a specifc conception of equality of such rights (­chapter 7) is similarly required on grounds of such equal standing of individuals. When all these requirements are fulflled, the state does what it has to do, and insofar as the justice of a state obliges individuals to adhere to it and the rules it makes, we can also say that the ensuing relation between state and citizen is one of just reciprocity. But none of this means that, literally, there is any form of equality between them, nor does it make any clearer what it might mean to demand it. Consequently, the question of what, if anything, is unjust, if the state as such—​as opposed to given state ofcials, or particular agencies or sub-​ institutions of the state—​has dominatory power over its citizens needs a separate answer. Are there good reasons to demand, with neo-​republicans, modal robustness here, too—​should states as such not be capable of arbitrary, or uncontrolled, interference? Is that even possible, for a state that is powerful enough to guarantee justice in the relations between individuals in social life?33 Or is it more reasonable to rest content if the state as such is, for the right kind of reasons, very unlikely to interfere in the wrong kind of matters: if it is non-​accidentally ‘benevolently disposed’ towards us, because it is democratic and responsive, and diferent state institutions check

32 Kolodny 2019, p. 112.

33 For negative answers to both questions, see again Kolodny 2019, pp. 99f, 111.

Liberal Non-Domination  71 and balance each other efectively? Whatever the answer to these questions, we can put them aside for our purposes. We can focus, in what follows, on the requirements of protection against domination among individuals, as a key requirement of social justice, as this is the case where the rationale of ensuring equality of standing applies in a literal manner.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.3.2.  Te Range of Non-​Domination Having thus restricted our focus, and accounted for specifc injustice of domination, we need to pin down its range, or domain: which of an agent’s choices, interests, or claims have to be under asymmetrical, arbitrary power, to trigger this kind of complaint? If I badly want access to a swimming pool, you own the only one around, and are not dependent on any contribution of mine for maintaining it, this gives you some power over me that we might call arbitrary:  your bargaining position over my conditions of access will be much better than mine. Still, it does not seem that this is domination in any sense that triggers moral concern, and gives reason for complaint.34 Similarly, your superior tennis skills might give you the unchecked ability to thwart my attempts at successful returns of your service at any time of your choosing. But it is clear that there is no complaint here. Teorists who aim at a free-​standing conception of domination, not tied to specifc background conceptions of the person and society like the ones assumed here, seek to make do with an account of that domain that is slim and non-​moralized. In this vein, Lovett proposes that it is enough if the dominated is ‘dependent on this social relationship [the one between the dominator and himself] to some degree’.35 As this condition might be over-​inclusive, and might capture scenarios like the pool case, McCammon proposes to narrow it down by demanding that dominators possess signifcant ‘impositional power’,36 which is paradigmatically given when they can make the ‘costs to noncooperation higher than the costs of cooperation for a very wide range of forms of cooperation’.37 Tat would be given, in a modifcation of the example made earlier, if my love of pools is pathological, a severe form of addiction, and I would therefore even agree to become your

34 Te example is adapted from McCammon 2015, p. 1034. 35 Lovett 2010, p. 112.

36 McCammon 2015, p. 1041. 37 Ibid, p. 1040.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

72  Liberal Relational Egalitarianism servant in order to gain access.38 Pettit, in earlier work, demands that the dominator be able to compromise the interests of the dominated, as they themselves would see them.39 Such slim, non-​moralized determinations are obviously attractive, because they promise to be able to explain what is objectionable about arbitrary, asymmetrical power as such, and thus, in that sense, to get considerable theoretical leverage out of the concept of domination itself. However, for precisely this reason, they cannot deliver a solid enough basis for a conception of social justice. What it is, in substance, that individuals have unequal power over is important for the question of whether that power is unjust, and, if so, what precisely is unjust about it. Tis is clearest in examples where somebody’s capacity to arbitrarily interfere with somebody else’s choices ranges over an advantage that may itself be unfair, or unjust. Take this one: we have run a company together, but have fallen out, and are about to dissolve it. I know that you know, and possess documentation showing, that I have been very sloppy with my tax returns over the last years, including perhaps signifcant tax evasion. Being able to threaten to tip of authorities about this puts you into a position where you could demand a very lopsided dissolution agreement, if you wanted to. Does my dependence on your will raise urgent justice concern of the sort discussed in the previous subsection? We can presumably all agree that it would not be just for you to use your knowledge for private gain. But that would be an objection to the use of power for a certain, objectionable, end, not an objection to your having that power. It is the latter—​capacity to interfere—​which is distinctive of, and therefore mostly taken to be sufcient for, the evil of domination. Furthermore, we can probably also agree that the best resolution here would be for me to end your power over me myself, by coming clean on my tax practices, including owning up to evasion, should any have taken place. But it is in my interest not to do this, and to avoid possible sanctions, just as it is in my interest to escape your lopsided dissolution proposal.40 38 Ibid., p. 1035. 39 Pettit 1999a, p. 68. 40 On Pettit’s view, it is not clear whether, in this example, my interest in avoiding sanctions counts. In some places, he insists that, for the case of domination among privates, the interests in questions are non-​moralized. Tese interests do have to be ‘avowable’ in the sense of being ‘avowal-​ready’, but that is to be understood merely as asking whether the interferee is ‘disposed to avow’ them, as a matter of fact; Pettit 2006, p. 276. In other places, he demands that the interests in question be generalizable, and hence publicly acceptable, because shareable by others; Pettit 1999a, p. 55; or holds that proposals for what should count as common interests have to rest on ‘cooperatively admissible considerations’; Pettit 2001, p. 156. Te latter is supposed to be the right conception of interests to

Liberal Non-Domination  73

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Free-​standing accounts of the substance of domination thus seem to apply.41 One might try to dig in one’s heels, insisting that, in cases such as this, you really should not be able to impose deals favourable to yourself, such as the lopsided dissolution proposal, not only that you should not in fact do so—​that this kind of leverage simply should not exist, even if it is brought about by a fault of the other party. But, apart from the question of which set of requirements could possibly ensure this without unduly stifing the possibilities of exchange among individuals, it seems free-​standing conceptions of domination would still apply, because you could still make me cooperate in many diferent ways, even if none is directly for your personal gain. You could make me donate most of my share of the company’s capital to an NGO of your choosing, for example. Other examples of the domain of domination are less clear-​cut, but similarly demonstrate the case for relying on thicker notions of relevant choices, or interests, over which the power at issue must range in order to constitute domination. Take the example of a trade union which is in a position to threaten a company known to guarantee its employees only the lowest wage and safety standards that are at all reconcilable with current law with imposing a strike. Te strike will signifcantly indent the company’s profts, and thus harm the owners’ and managers’ interests. Te trade union is not forced to consult the owners and managers, or forced to take their interests into account in any other way. Other things being equal, this would seem to be a case of domination, on non-​moralized accounts. Is it objectionable; is it unjust? It looks like we need to know more about the case. On the liberal be used in democratic procedures aiming at non-​dominatory laws, while the former is supposed to apply to private relations; Pettit 2006, pp. 279f. In the preceding example, we might think that even the non-​moralized interpretation already rules out my interest: people who are sloppy with their tax reports tend not to be disposed to avow their interest in avoiding sanctions for this. However, under repressive sexual and gender regimes, many LGBT people are also, as a matter of fact, not disposed to avow their interest in sexual self-​ determination, so understanding the test in this way leads to absurd conclusions. 41 Other views of non-​domination demand that arbitrary power must range over ‘basic interests’, such as ‘security, nutrition, health, and education’; Shapiro 2012, p.  294; or ‘basic capabilities’; Laborde 2010, pp. 53f. Tis yields a straightforward connection between domination and injustice, but is too restrictive; there is no reason why justice should be concerned only with basic interests or capabilities. Shapiro regards non-​domination over basic interests as the bedrock of justice, which leaves open the possibility that full justice goes further, but has independent arguments for restricting theorising domination to this bedrock; see Shapiro 2016, ­chapter 1. Laborde proposes that, in matters other than ‘basic capabilities’, opportunities for participating in egalitarian democratic procedures determining their content constitute non-​domination; Laborde 2010, pp. 51f. To this extent, she adheres to ‘non-​domination as democracy’ (3.4.2).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

74  Liberal Relational Egalitarianism account, the question of which ends, if any, such power serves, directly enters into domination assessment:  in the example at hand, is the trade union’s power useful for securing the standing of workers as free and equal (to managers) within the company, and any other possibly justifed claims of theirs, such as a distributive claim to a fair reward for their work—​is it perhaps necessary to secure these? Tis account operates with a substantive baseline for determining which kinds of claims non-​domination, as a demand of social justice, should range over. According to this conception, then, a more powerful agent does not dominate another (a) if her superior power is constrained so as to make sure that she has to honour the claims of justice of the less powerful, where these are fnal. Where they are not fnal, she does not dominate (b)  if her superior power is so constrained as to make sure that she has to take the other party’s prima facie relevant claims of justice arising out of social cooperation into account, or if the procedure which gave her that superior power in fact had to take them into account, and is repeatable and revisable. Paradigmatically, this requires democratic procedures of decision-​making in which the weaker party has an equal say. On the liberal view, the question of the domain of non-​domination—​ claims of justice, including prima facie claims—​and the question of what makes asymmetrical power arbitrary, and how it has to be constrained to cease to be so, are thus tightly connected:  the more substantive the view in question, the tighter the connection. Tis liberal conception of non-​ domination is thus indebted to Pettit’s original account of arbitrariness, which defned a capacity for arbitrary interference as a capacity for interference that is not forced to track the relevant interests of the interferee, and uses claims of liberal social justice to fll out the notion of ‘relevant interests’.42 Te two questions of what these claims are claims to, and how strong and entrenched protection for them has to be, are to be decided by taking recourse to the fundamental ideas underlying the conception: the idea of society as a fair scheme of reciprocal cooperation between individuals understood as

42 See fn. 12. As noted, Pettit has moved away from that conception. Tat arbitrariness can have a substantive meaning—​of not being forced to produce the right decision, or at least not to get it too wrong—​is precisely one of the reasons why, in recent works, he seeks to defne domination entirely without that concept, using ‘uncontrolled interference’ instead; 2012, p. 58. However, this does not eliminate the question of the appropriate substance for just, non-​dominatory relations; on Pettit’s recent account of neo-​republican social justice, it simply resurfaces in the question what control amounts to, and what it should range over; see 3.4.3.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  75 free and equal, and the liberal idea of the person as endowed with the two moral powers, spelling out this freedom and equality.43 What non-​domination amounts to can thus be illustrated with the help of a spectrum of claims. At the substantive end of the spectrum, the relevant claims of justice are not prima facie, but fnal, and being forced to take them into account simply means being forced to honour them. Take a case of ofcial ethnic discrimination, where state ofcials can bar members of a certain group, in virtue of their ethnic identity, from accessing public services, or attaining important social positions; or cases where owners of the means of production can dictate conditions of employment so as to make sure that even those who work at the level of what counts as full-​time employment in a given society get just enough to meet their basic needs, and scrape by, while they themselves reap lavish rewards (to make things as straightforward as possible, let us explicitly add that there is no way in which this inequality could possibly be construed as efcient). Both these scenarios are entirely incompatible with the idea of reciprocal cooperation among free individuals of equal standing; they are not justifable within a liberal egalitarian framework. Ofcials and employers dominate in them, because of substantive arbitrariness. So this would be the case even if their powers had been granted in an otherwise perfect democratic process; we will need to inquire into the complex relationship between such substantive requirements and requirements of political non-​domination and equality separately.44 A focus on non-​domination at this fully substantive end of the spectrum amounts to a focus on the modal nature of procedural protection, on the question of how frmly potential dominators are to be constrained not to produce the wrong results (to in fact discriminate, or pay breadline wages). However, we should not infer from this that protection against domination should necessarily be more intense the more frmly a case is located at this substantive end of the spectrum. Tis is a matter of the importance of the claims at hand, measured by reference to the moral powers and the idea of fair cooperation, not of

43 See Richardson 2002, pp. 37, 47, for the view that liberal republican democratic procedures have to be informed by these background ideas. For important diferences to Richardson’s account of non-​ domination, see fn. 67 in this chapter, and 4.2 in c­ hapter 4. See also Tomas 2017, pp. 15–​17. 44 In ­chapter 7. Conversely, in a social relation that is fully justly regulated, and gives the less powerful party the right kind of rights against the more powerful (and robustly so), the more powerful party does not dominate the less powerful one even if the latter did not have the required participatory opportunities in the political process yielding the regulations. Having been denied that opportunity is then an independent political injustice, and likely a case of political domination, as we will see in that chapter.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

76  Liberal Relational Egalitarianism fnality,45 and the next chapter will develop diferent requirements of intensity based on such importance. In the two examples made, high protection is needed because they concern central questions of cooperation. From the substantive end of the spectrum, let us move to the area of taking prima facie claims into account, where the account treads closer to more procedural, constitutively democratic accounts (3.4.2). Over this part of the spectrum, it is both possible not to be dominated even though a decision mechanism may, post factum, turn out to give the wrong kinds of powers to the more powerful, and to be dominated even though such a mechanism produces what in the end may turn out to be the right result, in terms of constraints on the latter. Te former is the case if the procedure did take into account all accessible information about prima facie relevant claims at the time of the decision, and respected all rights to participation in it; the latter is the case if the mechanism did not take it into account, or did not respect all such rights.46 Procedural arbitrariness so defned is sufcient for domination. Te claims which non-​dominatory procedures have to track thus do not have to have fnal moral status. Whether they overcome the threshold of relevance—​of prima facie plausibility, or reasonableness—​depends on their having an appropriate connection to the basic liberal ideas of the person and society underlying the conception of social justice, as well as by general features of rationality, such as consistency and coherence. Very ofen, claims arising out of social cooperation will be reasonably contestable, with regard to their correct formulation, their correct weighing against other, and their precise specifcation in concrete social contexts. Take the following example: to what extent should employers be forced to have to disregard any possible disabilities of applicants in job selection? Should they be forced to work from a fxed list of which disabilities, if any, render doing the job in question impossible, or should they retain some scope for individual judgement? Answers to questions such as these will depend, inter alia, on how we conceptualize disability, as well as broader ideas of fairness in cooperation and the resulting burdens associated with diferent roles in the productive process. Tey will thus depend, to some extent, on substantive considerations of social justice; but it is very unlikely that any independently identifable such considerations 45 If you have the power to decide all by yourself whether you will give me back the fve pounds which I lent to you, this is not an instance of domination against which I should be entitled to very intensive protection—​even though my claim to restitution may well be absolutely fnal. 46 Tat the available information might change is one of the reasons why non-​domination always requires standing rights of correction and appeal. Tis is standard fare on all conceptions of non-​domination.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  77 can settle the answers all by themselves, without participatory, democratic procedures involving the afected parties. Furthermore, the substantive claims of justice in question, fnal or prima facie, may well ofen be claims to distributive shares of relevant goods, but they need not always be. Openness to all prima facie relevant claims in the process of determining powers over each other thus embodies a particularly important form of recognition for people’s standing as free and equal cooperators. Tat one has been treated in the right way in the production of an outcome can compensate for the fact that, in some cases, individually compelling substantive reasons for disagreeing with that outcome may remain. On the other hand, the conception implies that where a claim does not even enjoy any prima facie standing, it does not matter for non-​domination which kind of procedure, if any, is followed in deciding over it. Take, again, my pressing a claim to be allowed a say equal to yours on the conditions of my access to your pool, on grounds of my special love of pools. In fact, if institutional resources are scarce, as they always are in the real world, making procedures over-​inclusive of potential claims may bring about domination elsewhere, as guaranteeing a hearing to the outlandish claims of some will divert such resources away from truly relevant claims of others, and their appropriately intensive protection. Finally, the fact that domination is tied to at least prima facie relevant claims of justice makes it unnecessary to demand any kind of particular actions, or intentions, on the part of the agent who is in a dominatory position in order for a complaint to arise (see 3.3.1). Te complaint is, in the frst instance, about somebody being in that position, not about what, if anything, they make of it (which may, of course, increase overall injustice).47 To conclude, then: this subsection has delivered the ‘grammar’ of liberal non-​domination, by spelling out how we ought to determine the range of claims and interests over which protection against domination is required. Filling out this grammar completely would be too much to ask; we do not need an exhaustive list of claims of justice, prima facie or fnal, for this purpose, and diferent liberal relational egalitarians can disagree at least about some of them. However, to determine the required intensity of protection against domination, we do need to identify, and defend, substantive 47 Against that, conceptions of non-​domination operating without a substantive baseline are under pressure to demand that more powerful parties actually do something to exploit their position, or to entrench it further, in order to render credible that it gives rise to complaint—​for example, in the pool case, your granting me access to the pool in order to cultivate my dependence on you; see MacCammon 2015, p. 1041 fn. 39.

78  Liberal Relational Egalitarianism requirements, and this is the task upon which the next chapter will embark, focusing especially on protection of a liberal conception of the basic liberties, where such intensity must be highest. However, before getting to these, we need to compare the liberal conception of domination in more depth to rival neo-​republican accounts of justice as non-​domination, in order to confrm that it is on the right track.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.4.  Republican Conceptions of Justice as Non-​Domination Such neo-​republicans about non-​domination and justice are likely to fnd the liberal account somewhat defating. Tis is so because it is, as seen, pruned back, compared to their accounts, in two signifcant, and related, ways. First, it does not conceive of non-​domination as the linchpin of justice, but, more modestly, as merely one—​though particularly urgent—​demand, and it recognizes other, independent, relational egalitarian demands, too, pertaining to social status (which we will develop from ­chapter 5 onwards). Second, it operates with a substantive baseline of the concerns non-​ domination has to range over, and identifes these with the help of relatively thick background conceptions of society, social cooperation, and the moral powers of persons. Because of this, we need to trace connections to these conceptions to confrm that somebody facing unequal and unconstrained power is indeed dominated, in the sense that gives rise to a complaint of justice, and the shape that domination may take can vary quite signifcantly, from more substantive to more procedural, as seen—​from insufciently constrained power over fnal claims to being denied an appropriate say in the determination of others. Accordingly, remedies may also vary. Neo-​ republicans want domination to play a more ambitious role for justice; on their view, a conception of domination should not need to be so reliant on such background conceptions, should identify a more uniform feature of all dominating relations, and should have a better claim to exhaust the space of justice—​or at least the space that theories of justice need to cover. Not all republican conceptions share all of these ambitions to the same extent, but all those analysed in this section regard at least some of them as essential. As we will see, all of them fail. Tey do so in diferent ways, but the liberal view avoids, or solves, the core problem of each. Tis corroborates the case for it.

Liberal Non-Domination  79

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.4.1.  Non-​Domination as Efective Constraint Frank Lovett’s account of justice as non-​domination is the most ambitious in these respects, insofar as it seeks to build a general theory of social justice on a notion of non-​domination that is maximally thin and procedural. As seen, in terms of the range of domination, Lovett merely demands that the dominated be dependent on the dominatory relation in question to some extent, so that exit would be costly; in terms of defning arbitrariness, Lovett argues that social power is arbitrary ‘if it is not externally constrained by effective rules, procedures, or goals that are common knowledge to all persons or groups concerned’.48 Social justice then requires minimization of domination, so understood. Tere are no substantive restrictions on what the goals, or aims of the said rules and procedures curbing social power, can be.49 Te problem with this account is not only that it would seem to require protection of individuals with regard to advantages, or choices, over which they might not merit any (see 3.3.2). Tere is another fundamental problem for this account. It implies, as Lovett himself notes, that, for example, procedures instituting a rigorous system of legal discrimination against certain groups, on the basis of their ethnic identity, do not amount to domination, and therefore not to injustice, as long as they are indeed rigorous—​as long as they do not deliver opportunities for further, unchecked abuse of their victims by those in positions of superior power.50 It is important to dissect what exactly is the problem here: it is not that, on it, the state ofcials enforcing the discrimination might not be dominating their victims by discriminating against them. On the relational egalitarian account of the injustice of domination (3.3.1), it is not implausible to hold that, should they really be fully constrained (if this is possible at all), and do not have any choice in the matter, they are not actually dominating: the victims are then not, in the relevant sense, at their mercy. Te real dominators might sit somewhere else; it might be those who made the law in question. Te problem is simply that Lovett’s account cannot account properly for the injustice of the rigorous 48 Lovett 2010, p. 111. 49 One advantage of such a ‘descriptive’ defnition of domination is supposed to be that it ‘quickly hands things over from the political philosopher to the social scientist’, who can then measure non-​ domination; Lovett 2010, p. 164. 50 Ibid., p. 118f. Most probably, they will bring about such opportunities; the point is that only this seems to be unjust, on Lovett’s view. For similar objections, see Arnold and Harris 2017, pp. 58f, and p. 61 n. 2; however, they wrongly interpret Lovettian non-​domination as requiring the minimization of all discretion in any, including trivial, scenarios where ‘no one has any real interest at stake’, ibid., p. 60. Tat is not the case, because in such scenarios, nobody is dependent on the relation in question.

80  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

discrimination itself, whether or not there are any dominators around (which, most likely, there will be).51 Similarly, take the case of the managers dictating poverty wages discussed earlier. Let us assume that the markets in question are so competitive that any manager who ofers more thereby immediately puts her frm out of business. Ten, it becomes more plausible to hold that, to that extent, the managers are not themselves dominating. But, whether or not there are dominators elsewhere, this wage system brings about massive distributive unfairness among social cooperators; a framework for social justice like the one adopted here can account for the injustice of this also independently of questions of domination, while Lovett’s account cannot. Lovett does recognize a value of ‘fairness in the distribution of goods’ as a value external to social justice as non-​domination, which may have to be, in instances, traded of against the latter.52 But this strategy leads not only to a severely revisionist conception of justice—​systematic discrimination and massive distributive unfairness count as paradigm instances of social injustice on most people’s views, and do so in themselves, and not only because they engender vulnerability to further domination (which they typically do). It is also in evident tension with neo-​republicans’ own claims about the paramountcy of non-​domination (see 3.3.1).53 Lovett’s conception of justice as non-​domination, understood in a fully procedural way, thus seems to capture mostly one aspect of justice:  certainty of expectations, as perhaps the most crucial formal feature of the rule of law;54 and closely related to what Rawls calls ‘formal justice’.55 Tis is not enough to base a conception of social justice on. 51 Take V scenario 1 (­chapter 2, 2.3): Tis scenario does not lose its high injustice rank even if state ofcials are fully constrained to enforce V deprivation. It is sufciently accounted for by the state’s hostility to the victims, whatever else is going on in it. Even though injustice increases further if the rule in question was set up by a gang of unconstrained strongmen at the helms of power, this is not needed to rank it highly. 52 Lovett 2010, p. 187f. 53 As well as with the paramountcy of justice itself, which not only most liberal theories of justice accept (see 2.4 in c­ hapter 2), but many neo-​republicans as well, such as Shapiro, Laborde, and Pettit (in recent works, see 2012 and 2014): republicans accept the paramountcy of justice as long as requirements for political decision-​making are understood as requirements of justice, and regarded as prior to requirements for relations in general social life (see fn. 15 and the following subsection in this chapter, and 7.2 in c­ hapter 7). 54 For a substantive understanding of the rule of law, which regards it as instantiating a form of social equality, see Gowder 2016. For Lovett’s own understanding of the rule of law and its relation to his view of social justice as non-​domination, see Lovett 2016a. 55 Rawls defnes this as ‘the impartial and consistent administration of laws and institutions’, in short ‘obedience to system’; Rawls 1999a, p. 51, referring to Perelman 1963. Tat ‘conformity to rules’ is another plausible sense of ‘non-​arbitrariness’ is a further reason for Pettit to do without the concept

Liberal Non-Domination  81

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

3.4.2.  Non-​Domination as Democracy Other republicans, however, will argue that the main problem with Lovett’s account is not that it fails to give the idea of social justice as non-​domination appropriate, sufciently substantive, content, but that it mischaracterizes the relationship between justice and non-​domination in a diferent way. Perhaps what is fundamentally unjust about relations such as the ones of ofcial state discrimination, and imposition of poverty employment terms, is simply that the more powerful are not accountable to the ones at the weaker end of the relation. Whether or not the ofcials or managers in question really count as dominators if they have no choice but to impose what they impose, the real problem is that those in power—​whoever they are—​should not be able to proceed without the less powerful party having an appropriate say on the terms of the relationship.56 As the less powerful in such relationships will, typically, not be the only ones with a relevant interest in it—​in the manager-​employee case, consumers, shareholders, and taxpayers at large are signifcantly affected by the structure of employment relations; in the discrimination case, everybody under the power of ofcials has an interest in non-​arbitrary treatment by them—​this requirement of a say quickly scales up, on such views, to the society in question: it generally demands non-​domination as democracy. Such constitutively democratic views are the most forceful free-​standing accounts of non-​domination. On such views, questions of the justice of particular relations between individuals in social life (dominium) are entirely subordinated, or reduced, to questions of public decision-​making and political justice (imperium). Tey also promise to have their own answer to the baseline problem (3.3.2): where that baseline should be could itself be decided in society-​wide procedures of democratic deliberation. Non-​domination, on these views, is present if these procedures are in place, and are followed. Within these procedures, other matters relevant for the justice of a society, such as proper remuneration for work, can be raised by everyone, and decided upon. Even if these procedures may sometimes yield what seems to be the wrong result—​whether in terms of power relations between individuals in social life outside those procedures, or on other matters—​non-​domination is still present on this view, as long as they have of arbitrariness in his latest attempts to defne non-​domination; Pettit 2012, p. 58 (see fn. 42 in this chapter).

56 MacCammon 2015, 1044, 1048f.

82  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

indeed been followed. And it always remains open to run the procedures again, perhaps attaining results that are more satisfactory to more individuals. In this sense, this view seems to hold out the promise of being able to relegate theorizing about such substantive matters to a subordinate level.57 Since, as seen, there cannot be fully individualized control over matters of structuring cooperation which are relevant for many members of society, what all individuals can reasonably demand, accordingly, is not such individual control over everything that afects, or subjects, them, but an appropriate share of collective control, in conjunction with the share of everybody else. Terefore, such views give rise to a demanding requirement of substantive political equality: a requirement of an equal share of control, understood as opportunity for political infuence. Tis is because it seems plausible that others who control a given procedure more than you do thereby dominate you. Such a view of political equality as collective control has been recently put forward by Philip Pettit.58 However, Pettit couples it with a distinct conception of republican social justice as non-​domination between individuals in social life (dominium). On his most recent, considered, view it is thus possible, as it is on the liberal view, that political non-​domination is fully in place, but there is still domination in other areas of social life.59 Te shortcomings of his conception of non-​domination in social life will be analysed in the next subsection. A pure example of ‘non-​domination as democracy’ is one in which questions of domination between individuals in social life reduce, at the 57 A possible misunderstanding should be eliminated at the outset: the focus on procedures does not imply that it could ever be enough to merely institute some kind of legal procedure. People need real opportunities—​including resources, education, and time—​to take part in them. In this sense, the democratic view is substantive, too. 58 Pettit 2012, pp. 168 and 169: ‘[T]‌here cannot be equally shared control without equally shared infuence’. Tis is the basis of his ‘tough-​luck test’: ‘[L]egitimacy will be adequately ensured . . . to the extent that you and your fellows have good grounds to think that any unwelcome results of public decision-​making are just tough luck’; p. 177; and: ‘Insofar as the tough-​luck test is satisfed, the will displayed by the state is under an efcacious form of control that you share equally with others in imposing’; p. 178 (my emphasis). Pettit uses ‘political legitimacy’ in Pettit 2012 (throughout), and ‘political justice’ in Pettit 2015a; as noted (fn. 15), this terminological diference is irrelevant. 59 Te view that non-​domination is fully constituted by democratic procedures may, against that, have been Pettit’s earlier view, depending on whether the kind of ‘relevant’ interests over which arbitrary power must not range are to be entirely determined by democratic procedures, as common interests; see fn. 40 on diferent notions of interest employed by Pettit. Tis claim that relevant interests have to be defned by democratic procedures was present already in Pettit’s Republicanism (1999a), and has become more prominent in subsequent writings; see Pettit 1999b, pp. 172f., and 2001, pp. 138f., 156–​168—​though without emphasis on political equality. Pettit’s earlier work also argued that non-​domination ought to be maximized; for criticism that the relationship between the maximizing and the democratic components is unclear, see again McMahon 2005, and Pettit’s reply (Pettit 2006). Te maximizing view will be analysed separately in the next chapter (4.4).

Liberal Non-Domination  83 fundamental level, to questions of political non-​domination, in the broadest possible sense. Te clearest and most forceful statement of this kind of view has been provided by Rainer Forst. According to Forst, domination is any kind of arbitrary rule without proper justifcation and without proper possibilities for justifcation in place, and the [ . . . ] claim to non-​domination is the claim to be respected as a justifcatory equal.60

Because the ‘frst question of justice is the question of power’61, the primary task of justice is to eradicate domination. It is the task of ‘fundamental justice’:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Whereas the task of fundamental justice is to construct a basic structure of justifcation, the task of full justice is to construct a justifed basic structure. Te former is necessary in order to pursue the latter, that is, a ‘putting-​ into-​efect’ of justifcation through constructive, discursive democratic procedures in which the ‘justifcatory power’ is distributed as evenly as possible among the citizens. [ . . . ] Based on a moral right to justifcation, arguments are presented for the basic structure in which those who are part of it have real opportunities to codetermine the institutions of this structure in a reciprocal and general manner. Fundamental justice guarantees all citizens an efective status ‘as equals’ in this sense.62

Te distinction between fundamental and full justice makes clear that the former is primary, and that relational equality is to be taken care of, in the frst instance, by fundamental justice. Te term ‘full justice’ acknowledges that, within the relevant procedures, citizens are to phrase their arguments about the substantive implications of diferent possible confgurations of the basic structure—​for example, about the distributions of relevant social goods produced—​in justice-​relevant terms (e.g., by making arguments based on an idea of fairness), and not, or not merely, in terms of self-​interest. Tus, participants ought to be committed to determining what counts as a relevant interest within the procedure with the help of a justice baseline (3.3.2); the

60 Forst 2014, p. 211. 61 Forst 2012, p. 195. 62 Forst 2013, p. 160 (emphases in the original); see also 2012, p. 196. ‘Justifcatory power’ is not to be understood in merely hypothetical terms; see Forst 2013, p. 159: ‘[A]‌just social order is one to which free and equal persons could give their assent—​not just their counterfactual assent but assent based on institutionalized justifcation procedures.’

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

84  Liberal Relational Egalitarianism task of defning it is shifed onto them. For theories of justice, it is then merely a secondary task, if at all, to make substantive proposals for ‘full justice’. Now, one objection that one may raise against this proposal is that this kind of political equality, no matter how demanding it is, is little consolation if the substantive outcome of the required procedures may still be cases of vastly asymmetrical power in social relations. It could be an outcome of such procedures, for example, that workplace managers end up with legally sanctioned, comprehensive directive power over workers, including over their remuneration and their powers to structure their own working lives, can hire and fre them at will, and perhaps even have far-​reaching control over the opinions they are permitted to utter at work. Such cases, on this view, cannot be cases of domination—​of the kind that gives rise to complaint.63 And that seems a problem. However, ‘non-​domination as democracy’ has an efective retort: if such laws indeed emerge from egalitarian democratic procedures that are, ex hypothesi, perfectly set up, then we should rather take this to indicate that the substantive policy area is an area of reasonable disagreement. Substantive political equality is compensation for such workplace ‘domination’, if there is such disagreement.64 Democratic decisions are reversible, and every worker has the basic right to retain everything which is needed to participate as an equal in these, which puts efective external constraints on how bad such workplace ‘domination’ can be, and guarantees her a chance to convince her fellows to change it. Tis seems a powerful reply; its point is that the constant guarantee of equal power over the basic structure is all the consolation one can reasonably want, especially if reasonable disagreement is pervasive. However, this reply is nevertheless too quick, for two reasons. Te frst concerns the shape that even the most demanding form of political equality can take, in complex societies; the second concerns the extent of power that the political system as a whole can have over the rest of the basic structure, in such societies, at any given time. Both will be importantly limited; because of this, ‘non-​domination as democracy’ comes up short. To the frst point, then:  political equality will remain insufcient compensation for inequalities in broader social life if it cannot demand perfect 63 See McMahon 2005, pp. 81f. 64 McMahon (ibid.) argues against Pettit’s earlier view that, if the relevant common interests for non-​domination are those that survive democratic procedures, then this view fails to yield any determinate policy implications as soon as there is reasonable disagreement about these interests (this is somewhat sofened in McMahon 2007). Te response denies that this is a fundamental problem if there really is political equality of the required kind.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  85 equality of political power; and there is very good reason to think that it cannot. In highly complex, productive societies characterized by an intricate division of labour—​which liberals have reason to want (3.2)—​there will have to be some relatively stable positons of authority and power within the basic structure. And even if there should be a suitable way of distinguishing, and keeping apart in practice, social power within the basic structure, in its present confguration, from political power to shape diferent confgurations of it, that still does not solve the problem. Tis is for the following reason: positions within the political system in charge of shaping the structure will have to be diferentiated, as well, and cannot, in large-​scale, complex societies, be fully equally occupied by all over the course of their lives. In short, there will be positions such as that of ministerial bureaucrats, judges, and, probably also MPs, and these will, on all reasonable interpretations of political power, enjoy more of it.65 An appropriate conception of political equality (to be developed further in ­chapter 7), then requires that a demanding form of equality of opportunity to attain those positions has to be in place, at least at crucial points in time in individual lives, that all inequalities of political power instituted through such positions is justifed at the bar of justice, and that all enjoy substantively equal opportunity to infuence the outcome of periodical selection procedures (such as elections). But that is not equivalent to fully diachronic equality of power proper. Only equality of political power proper, however, would seem full compensation for any intuitively problematic inequalities in broader social life.66 If this is so, theorists of justice should not restrict themselves to making proposals about political justice. Even under full political justice, of the kind 65 Such power will not only be mechanical, merely executive: these ofcials will also enjoy more ‘justifcatory power’, as expertise and experience on the job will normally increase that, as well. 66 It is important to emphasize that this objection to ‘non-​domination as democracy’ does not hinge on imputing to its adherents any kind of narrow account of equality of political power, which applies only to a very limited set of formal political institutions. Forst insists that the proper aim of a democratic conception of just non-​domination is an account of the latter not only for such political institutions, but also for a ‘comprehensive’ basic structure, ‘from the family to the workplace’ (2013, p. 164), and ‘[t]‌hat the question of power is the frst question of justice means that the sites of justice are to be sought wherever the central justifcations for a society’s basic structure, which determine social life in its entirety, have to be provided’ (2012, p. 196). Leaving aside the empirical question of whether it is indeed the justifcations of the basic structure which determine ‘social life in its entirety’, the point of the objection is that, for at least some of these sites, theories of just non-​domination should focus on providing the content of the justifcations. Taken literally, a view focusing on providing egalitarian structures for determining the justifcations in all relevant sites of justice demands equality of power to shape, and change, all of them, including the central mechanism in charge of coordinating them—​political institutions making the law. Tis is what does not work.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

86  Liberal Relational Egalitarianism that can be had, not everybody will be constantly able to bring their own substantive views to the fore with equal force. Teorizing the questions of what counts as substantive injustice, of which of these injustices are most urgent, and of how diferent ones may hang together, then remains an important task, in order to ofer direction and guidance to political institutions. Making substantive proposals for the content of social (or ‘full’) justice is thus not a secondary task. In response, one could seek to bolster the case for ‘non-​domination as democracy’ by demanding that political procedures be constructed on the basis of more substantive background conceptions about the nature of social cooperation, people’s standing as free and equal within them, and their higher-​order interests, such as adequate development and exercise of the moral powers. Tis will increase the chance that participants are directed at the right kinds of questions for their agenda. Once one does that, however, participants not only lose some fundamental justifcatory power over these background conceptions themselves; some substantive requirements of social justice will also emerge, as independent ones which democratic decision-​making just has to respect (3.3.2).67 Te second reason is that the political system in charge of shaping the basic structure of society will, in complex societies with a diferentiated division of labour, not have complete control over that shape, in the sense of being able to change about it whatever those in control of the political system want, at any given time. Setting apart the question of political power inequality between individuals—​between individuals alive and participating in society together at a given moment—​what can be changed at any given time are, almost always, only some elements of the basic structure, not all together. Others have to remain constant, perhaps being taken over by tradition and habit, perhaps awaiting their turn. What is more, ambitious reforms of the basic structure always carry risks, and major changes, once undertaken, may not be easily reversible, and may heavily constrain the direction and possibility of future changes.

67 As noted earlier (fn. 43), Richardson’s account of non-​domination within democratic procedures argues that these must be informed by the liberal ideals of the person as free and equal, and procedural fairness; Richardson 2002 and 2006. Te view developed here is indebted to Richardson, but recognizes that reliance on these ideals gives rise to independent substantive requirements of non-​ domination in social life, too—​while Richardson restricts the scope of domination to relations of authority within political decision-​making; see 4.2 in ­chapter 4. See Tomas 2017, pp. 16f, who argues that designing democratic fora deciding over common interests already requires recourse to these interests themselves.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  87 Tis second point is a commonplace of political science; but it implies that there are clear limits to the possibility of democratically experimenting with the shape of the basic structure itself.68 Rare moments of revolution and radical political innovation apart, most of the time, it may only be possible to do so with single policies, or very limited sets of policies. 69 However, it seems ‘non-​domination as democracy’ would have to require more than that to hold out the promise to be able to do without a substantive conception of social justice in broader social life, in the name of letting the people themselves determine its content. If the possibility of undertaking ambitious reform is not always present, and if the reforms we do undertake, whether successful or not, themselves heavily constrain what can be done in the future, then this is another reason for political theory delivering substantive proposals for just policies: we ought to make sure, to the extent that we can, that we have our substantive priorities right—​or at least that we have delivered some guidance for the process of sorting them out—​when the relevant moment comes, because it might not come again. It is also a reason not to be too cavalier about reasonable disagreement, and its extent, and to seek to give clear guidance on what can count as reasonable, within the theoretical framework on ofer. To be sure, there is a risk of theorists and philosophers seeking to usurp the authority to dictate to others what is reasonable; but the previous discussion shows that there are real risks connected to being very permissive about it, too. Tus, ‘non-​domination as democracy’ is certainly a powerful, and appealingly simple, view. But it cannot do all it promises to do. We retain good reasons for substantive theorizing of justice and non-​domination of the kind undertaken by the liberal view, and for not regarding it as in any way a subordinate task—​even in the face of apparently reasonable disagreement.

3.4.3.  Pettit’s Account of Social Justice as Non-​Domination In recent works, Philip Pettit has recognized the need for a specifcally neo-​ republican conception of social justice as non-​domination regulating the 68 For a Deweyan conception of democratic experimentalism, see Anderson 2006. 69 An example of such a non-​revolutionary but innovative moment is the birth of welfare state institutions and policies. For an infuential argument that diferent initial confgurations of these in diferent countries, due to the diferent kinds of class coalitions supporting them, played, and continue to play, a very large role in determining their subsequent diferent trajectories, see Esping-​ Andersen 1990, chs. 4 and 5.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

88  Liberal Relational Egalitarianism relations between individuals in general social life, as distinct from an account of political justice, and has provided one.70 We have seen in the preceding subsection that this is a good idea. What, then, are the remaining key diferences to the present account? We have already noted (3.3.1) that one key diference between a republican and a liberal account of social justice is that the latter does not regard the demands of non-​domination as exhaustive of its requirements, whereas the former does, and we will pursue the implications of this broader scope for relational egalitarianism further in subsequent chapters. Here, it is more instructive, and useful, to demonstrate how Pettit’s account of social justice as non-​domination fails to correctly identify, and object to, at least some important cases of domination itself, because it is not based on appropriate background conceptions of the person and society. Te task descriptions for the Pettitian and the liberal relational egalitarian conception of social justice are very similar. Pettitian social justice requires that state institutions be committed to expressive egalitarianism:71 institutions must honour individuals’ fundamentally equal moral worth, and thus treat those under their power as equals. On a neo-​republican view, they do this by honouring their equal status as free persons, which requires respecting and promoting their equal freedom to choose, which, in turn, requires entrenching these freedoms against others’ powers. Tus, so far, the question of exhaustiveness apart, the two views are in agreement. Pettit argues that justice among individuals in general social life requires giving people both the resources and the protections against others needed so as to be free to choose within the parameters of basic liberties. Basic liberties are all those that are (a) co-​exercisable and (b) co-​satisfable.72 For (a), people must be able to exercise the said liberties at the same time; for (b), people must be equally able to enjoy the exercise of basic liberties, that is, to derive welfare from it. What counts as welfare is to be judged ‘by received social criteria’.73 When setting up a list of basic liberties, we should aim at the most general formulation of co-​exercisable and co-​satisfable liberties, and the one that allows individuals to have the most far-​reaching efects, as intended by them, on other agents and the world at large (‘distality’).74 70 Pettit 2012, ch. 2, and 2014. Earlier accounts of the role of non-​domination as the master value of political morality did not connect it explicitly to justice, in any of its dimensions. 71 Ibid., p. 79. 72 Ibid., p. 93. 73 Ibid., p. 98. 74 Ibid., p. 102.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  89 Since many diferent combinations of co-​exercisable and co-​satisfable basic liberties are possible, we need guidance on the question of which combination to pick, or which set of combinations to regard as viable contenders (among which we should then choose in democratic procedures). Furthermore, the question of intensity of protection against others’ potentially superior power needs to be answered. We need to know how frmly each of the resulting basic liberties is to be entrenched—​the degree of robustness with which it is to be protected. Pettit proposes the same solution to both tasks: the ‘eyeball test’. We determine the range of choices as well as the degree of resourcing and protection by asking what people need on these fronts to be able to ‘look others in the eye without reason for the fear or deference that a power of interference might inspire’.75 When the eyeball test is satisfed, people enjoy a set of basic rights that safeguard their not being under ‘alien control’.76 Pettit thinks that this yields a ‘liberal republicanism’77, but precisely one that avoids reliance on more determinate background ideas, such as about the moral powers of persons.78 However, this thinness comes at the cost of two problems: Pettit’s view of social justice is neither very liberal nor very republican, or if it is so, then only contingently. Together, these problems show that reliance on thicker background conceptions of the person and society is necessary to yield both reasonably determinate and defensible guidelines for how to entrench demands of non-​domination between individuals in social life; and they suggest that a liberal view is particularly well-​placed to answer the questions of why both an extensive range of resourced and protected choices is needed, and why the degree of protection should be high (though it should be able to vary over diferent areas; see 4.3 in ­chapter 4). Te root of both problems is that, because Pettit’s view seeks to avoid reliance on such independent substantive ideas in its grounding, it gives too much room to de facto prevailing social standards, which come in as external variables. Tey enter in two diferent places, corresponding to the two problems: the frst is that ‘received social criteria’ and ‘cultural expectations’79 determine which exercises of liberty can count as co-​satisfable. Tis concerns the extent of protected choices. Te second is that the ‘eyeball test’ makes the intensity of protection depend on 75 Ibid., p. 84. 76 Note thus that the absence of ‘alien control’ here is not defned by having a say, but by simply having full individual control. 77 Ibid., p. 11 fn. 8. 78 Ibid., p. 94. 79 Ibid., p. 106.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

90  Liberal Relational Egalitarianism when prevailing social and cultural norms judge fear to be an appropriate reaction to the constellation of powers in the relations in question, as a opposed to a personal emotional overreaction: it demands that people be able to look each other in the eye ‘in the absence of what would count, even by the most demanding standards of their society, as mere timidity or cowardice’.80 Demands for more robust protection cannot be based on the latter. With regard to the frst issue, Pettit notes, by way of examples, that different societies have wide leeway in determining when ‘ofensive speech is damaging and should be restricted’, and, on the issue of gun ownership, in setting up the ‘background tests that gun owners should pass’.81 In both cases, it seems clear that a liberal view has to demand more checks on such leeway. Giving cultural norms on the possibly damaging nature of ofensive speech a free pass likely means licensing a majority’s power to curb speech that its members regard as damaging, such as, perhaps, speech that challenges the precepts of a dominant religion—​a historical paradigm case for freedom of speech. It is a staple of all liberal views of freedom expression that individuals in search of their own good must be able to voice views which are not already guaranteed to be acceptably non-​ofensive to possible interlocutors. With regard to gun ownership and control, lax testing means a higher chance of more guns around, and more guns around tend to lead to more violent deaths.82 Avoidable violent deaths are not good for people’s moral powers, and it is not clear what comparable interest (‘welfare’) might stand on the other side. Tis liberal objection is, however, not particularly deep. It merely casts doubts on Pettit’s assertion that his view of non-​domination and a foundationally liberal one, which ties non-​domination to thicker moral notions such as the moral powers, do not come apart in substance.83 Te second issue—​that the degree of protection against domination, too, is determined by reference to variable cultural standards—​cuts deeper. It raises doubts that Pettit’s view is a recognizably republican view. ‘Recognisably republican’ 80 Ibid., p. 84. 81 Ibid., pp. 106–​107f. 82 Tis seems to be the case even in places where prevailing social and cultural norms heavily emphasize the value of avoiding violent confrontations, such as Switzerland; see http://​www.npr.org/​ 2013/​03/​19/​174758723/​facing-​switzerland-​gun-​culture (last accessed on 13 November 2020). Widespread private ownership of guns once had defensible, distinctly republican rationales in both the US and Switzerland: enabling citizens to defend their freedom against oppressive governments (US, especially), and against invasion by foreign forces (both). But these rationales seem to have lost much of their force at the current stage of development of military technology, and state capacity. 83 Pettit 2012, p. 58, fn. 34, with reference to Richardson 2002.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Liberal Non-Domination  91 means, straightforwardly, that it ought to be in accordance with salient republican convictions about the objectionable nature of uncontrolled asymmetrical power, as a danger to individuals’ status as free and equal members of society (3.3.1). Contrary to Pettit’s claims, the eyeball test does not do enough to satisfy these, either. It might sanction very low levels of protection, because ‘people are liable to vary across societies in the diferent levels of vulnerability to which they have become inured’.84 Take, again, a case of inequality in the world of work: an economic regime in which employers can boss around, and fre, employees at will. To isolate what is problematic, assume that background social guarantees of meeting basic needs are in place, and that discrimination between diferent groups of employees is efectively ruled out. It is not clear why all possible construals of co-​exercisable and co-​satisfable equal basic liberties, on Pettit’s model, should rule out such a regime. And it is not outlandish to suppose that people, or many people at least, can be socialized into such a system so as to accept it without fear—​or at least so as to accept that they have no reason for fear, and consequently regard any fear of theirs as simply a personal defciency. Tey can be taught that ‘life is tough’, and that, even though it may not be great to be at the receiving side of this arrangement, everybody has some opportunity to make it to the other side. So that is what one should work towards, accept any present hardship, and encounter those who momentarily have these powers without deference or fear, for your chance may come, and the tables may turn in the future. Social norms may confer stigma on those who crack up, and become fearful, labelling them ‘losers’ who just are not tough enough: ‘mere cowards’.85 What other grounds there could be, on the eyeball test, to conclude that they nevertheless have ‘reason for [ . . . ] fear or deference’ is not clear.86 Pettit gives

84 Pettit 2012, p. 85. 85 Similarities to the present US employment regime are intended (even though it does not guarantee the satisfaction of basic needs for all). Pettitian republicanism is able to object to a scenario in which de facto absence of fear is based on a demonstrably false social narrative, insofar as it is a democratic position incorporating transparency and publicity conditions. In this example, this could be a sustaining myth of equal substantive opportunity to become a boss. Many studies confrm low social mobility in the US compared to other afuent liberal democracies, such as the Scandinavian countries; see, for example, OECD, ‘United States: Tackling High Inequalities’, p. 6, https://​www.oecd.org/​ unitedstates/​Tackling-​high-​inequalities.pdf (last accessed 13 November 2020). Tis is why the example requires merely some, not equal, opportunity. 86 Both McCammon’s and Lovett’s accounts can easily identify such grounds, in this example. Bosses have ‘impositional power’, and many of their actions are not ‘externally constrained by efective rules’.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

92  Liberal Relational Egalitarianism none.87 How exactly workplace relations have to be structured instead is, of course, a topic for deeper investigation. Here, the point is simply that power inequalities such as the ones in the example look like domination. At the very least, they stand in need of a stringent justifcation in terms of competing values and considerations, because, insofar as the tables do not turn for everybody, people do encounter each other in continuous and entrenched positions of uncontrolled asymmetrical power, and that would seem, on the face of it, incompatible with republican equality. 88 Te example shows that both the range and intensity of protection need to be more frmly anchored in background conceptions of the person and society. Te liberal view has no difculty whatsoever with scrutinizing the preceding example further. Te very frst question it will ask (see 3.3.2) is which reasons individuals who rightly regard each other as free and equal participants in a fair scheme of social cooperation could have to accept such an arrangement of dependence on the goodwill of others, in an area of activity—​paid work—​that is very signifcant for most people, and not contingently so, as it covers a great part of overall societal cooperation. It is, however, important not to overstate the reach of this argument. Other conceptions of the person and society may also be able to account for why people should be interested in a high degree of protection. What the argument shows is (a) that some such conceptions have to be plugged in, and (b) that the liberal conceptions of the person and society look well-​ placed to account for both far-​reaching and intensive protection against domination. Teir suitability for an egalitarian, demanding account of non-​ domination is thus an indirect, internal argument for these conceptions, 87 It could of course be that, as a matter of fact, such a system does not bring about fearlessness and absence of deference to the degree required by the eyeball test, and that countervailing cultural norms depriving people of reasons for fear cannot be stably maintained. But it seems odd, for an avowedly egalitarian republican, to say that, absent this problem, there would be no other problem here. It could also be that, under such a system, individuals’ struggles to overcome dispositions towards fear or deference, even where successful, occasion higher levels of stress, tenseness, and aggression. However, the eyeball test is not sensitive to these. 88 Tis is somewhat ironic, as Pettit criticizes Rawls’s conception of social justice for its supposed neglect of the modal requirements of robust protection against domination; 2012, p. 108. A similar complaint of contingency and insufcient demandingness cannot be made against Pettit’s view of political justice based on the tough-​luck test; see fn. 58. Te tough-​luck test is so demanding that very few (if any!) states may be legitimate; see Patton 2015 p. 666. Pettit accepts this; Pettit 2015b, p. 691. He is uninterested in what reigning social norms in diferent contexts may say about political legitimacy. Perhaps Pettit’s considered view is that a demandingly egalitarian form of political justice compensates for lacunae of protection lef by his conception of social justice; we have explored and rejected this line of argument in the preceding subsection.

Liberal Non-Domination  93

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

aimed at those concerned about the evils of inegalitarian social relations, or tempted by the positive aspects of egalitarian ones. It suggests that liberalism can be a home for these convictions, and, surprisingly, a better one than some important neo-​republican accounts. Which guidelines exactly this conception yields for the intensity of protection over diferent kinds of claims and choices is one of the issues to which the next chapter will turn.

4 Te Demands of Liberal Non-​Domination

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.1.  Introduction Chapter 3 has developed a liberal relational egalitarian account of domination as social injustice, and outlined its key advantages over its most important neo-​republican competitors. However, in order to be able to pin down, in more concrete terms, what liberal non-​domination demands, we need to specify the relation of domination further, work out key requirements of intensity of protection against domination over diferent substantive matters, and give an indication of the diferent kinds of policy measures through which they are to be implemented. Tese are the tasks for this chapter. It is structured as follows. Section 4.2 argues that, while the specifc injustice of domination resides in arbitrary power inequality between individuals, we should take a wide view of that inequality: it is neither restricted to relations of de facto authority, nor to relations of power inequality which are a matter of public, or common, knowledge. Attempts to further widen the scope of domination to include ‘structural domination’, as a relation which need not hold between individual agents, are, however, misguided, precisely because they miss this distinct wrong-​making feature. Section 4.3 develops diferent requirements of intensity of protection against domination, specifying how safe the diferent kinds of choices and claims which are the substance of the conception have to be made against others’ superior, and arbitrary, power. It also identifes a basic constraint of respectful protection that liberal non-​domination has to fulfl. Liberal non-​domination thus requires equal, respectful protection at various thresholds: while basic liberties call for the highest possible equal protection, in other matters of individual choice, a trade-​of between intensity of protection and extensive choice has to be struck. Section 4.4 rejects an alternative, simpler account of intensive protection, put forward by Pettit in his earlier works, according to which no such salient trade-​of is likely to arise, so that we can always safely maximize equal intensity. Section 4.5 leads over from the principled demands of non-​ domination to the range of its policy implications. It shows that this range Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0004

The Demands of Liberal Non-Domination  95 is wide, and goes considerably beyond demanding the right kind of formal, legal, institutions, extending to shaping appropriate informal social norms for everyday behaviour. It argues that this need not worry liberals. To wrap up, and link the conception of liberal non-​domination developed in this and the previous chapter back to the argument for the expressive perspective and against purely recipient-​oriented views of social justice made in ­chapter 2, section 4.6 investigates whether liberal non-​domination can itself be cashed out as such a purely recipient-​oriented conception. It argues that many of its demands can certainly be formulated in distributive terms, and indeed that they sometimes should be, for practical purposes—​but that this does not impugn in any way its distinctness from distributive conceptions of justice of the kind discussed in that chapter, nor its distinct connection to the expressive perspective.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.2.  Domination, Narrow and Wide Tis section clarifes some respects in which our conception of domination should be wide, and some in which it should not be. It should be wide insofar as it should require neither de facto authority of the more powerful party, nor common knowledge of its superior, unconstrained power among the parties (4.2.1); these features are not needed to make such power unjust. It should be narrow insofar as it should remain strictly agential, and focus on superior, arbitrary power of individual agents over others (4.2.2): As section 3.3.1 of ­chapter 3 argued, the special force of the domination complaint ultimately rests on the inequality of standing between such agents that domination expresses. However, such inequality is present also when the superior party participates in a group agent, or collective, which wields arbitrary power over the weaker party, whereas that weaker party does not so participate, or, if she does, has less power within it. Te superior party does not need to have the capacity for arbitrary interference all by herself, as it were. Social structures can constitute, or contribute to, such capacities in various ways, and therefore warrant condemnation; however, there is no ‘structural domination’, where a structure takes the place of a dominating agent; such ‘domination’ cannot give rise to the same kind of complaint.

96  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.2.1.  De Facto Authority and Common Knowledge Domination is not confned exclusively to those relations where the party with superior power can claim some form of authority over the less powerful one.1 Tis is the correct conception for analysing and assessing possibilities of domination in, inter alia, procedures of political decision-​making, and in the powers it confers in turn. Such decision-​making claims authority over those subject to its power, in the form of law, and such authority, where efective, limits the freedom of these subjects beyond the constraints imposed by ‘naked’ enforcement power. Under a functioning legal system, with the requisite social norms of general law-​abidance in place, individuals are expected to conform to laws because they accept them as sources of obligations. Tey should not regard them as merely a kind of brute force potentially standing in the way of achieving their goals and intentions. Accordingly, they will face social sanctions in the case of disobedience already for a failure to recognize such authority, not only for the (perhaps) bad outcomes that their failure to comply produces in particular instances. Generally, relations of authoritative power are, of course, of particular salience for relational egalitarians, whether they be formal legal relations or not. Positions of authority are needed to stably realize efective social cooperation (3.2 in c­ hapter 3). Authority depends on recognition by others, and so recognized individuals command, by that token, and to that extent, higher social status than these others, or at least have a privileged opportunity to acquire it. Such status is a problem for non-​domination where it can be used to infuence others even in matters that extend beyond the scope of the position of authority in question (assuming its existence is justifed). Tis is thus a further reason for why relational egalitarians need to have a particular focus on sharp circumscriptions of positions of power, and why this focus should include the social norms sustaining them. To take an outlandish example, imagine that judges could use their status of respectability to dictate fashion. To take a less outlandish example, consider the media airtime that managers and businessmen and -​women who are deemed successful sometimes get, in capitalist countries, also for instructing others on how to be successful in life more generally. Tis gives them an opportunity to contribute to shaping social norms reaching far beyond their own ambit of functional competence.2 1 As maintained by Richardson 2002, pp. 26f. 2 For an example, see Sandberg 2013. Perhaps, however, hardly anybody is listening, or listeners are not taking them seriously. Tat would, for liberals, most probably be a welcome result.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  97 Chapter 6 will analyse the exact connection between status and domination further, based on a fne-​grained account of social esteem-​based status norms, and the problems they pose for relational equality also beyond domination. However, none of this means that questions of domination are confned to relations of authority; it is not clear what reasons there could be for such a restriction, if non-​domination is to be a general requirement of social justice. A gangster boss in your neighbourhood dominates you if she can have you beaten up at will, whether or not she claims, and is widely regarded as having, some form of authority over you (as traditional Mafa bosses ofen were, and sometimes perhaps still are, in their territory). Te presence of unjustifed de facto authority merely makes the relation in question, other things being equal, more unjust, because it subjects the will of the dominated to that of the dominator in a more intimate fashion. Neither do relations of domination have to involve common knowledge among the parties, so that both the dominator and the dominated know about the asymmetrical, uncontrolled power, and know that the other knows. Common knowledge is, in the republican tradition, generally regarded as a necessary feature of domination, and this is largely maintained in contemporary neo-​republicanism. Pettit’s eyeball test (3.4.3) seems inapplicable without it: fear and deference are reactions to knowing about uncontrolled power.3 However, clearly such power can exist without the dominated knowing about it, or at least not about its true extent and form. Take, as an example, secret Mafa Cupola meetings, where spoils and territories are distributed among clans, without anybody else, including lowlier Mafa members, knowing about it. Other examples of political and economic manipulation can be easily found. Take oligopolistic collusion:  the largest players in a given market secretly cooperate to set prices. It would seem that the cartel dominates price-​takers.4 Tere is thus, again, no reason to exclude such cases from the scope of our conception of domination. To the contrary, given that non-​domination is a kind of freedom—​freedom from the unconstrained will of more powerful others—​there is a strong case for including them. It would be very odd to say 3 Lovett’s account also requires common knowledge; see Lovett 2010, p. 111. As the eyeball test checks for reasons for fear and deference, it could perhaps be counterfactually extended, asking whether one would have reasons for fear and deference if one knew about the power. 4 Such collusion can be so secret that the players do not even talk about it among themselves; sometimes, the intention to collude can simply be read of others’ behavior. Pettit 1999a, p. 60, advances a broader account than in later works, and correctly classifes cases of unconstrained opportunity for ‘backroom manipulation’ as domination.

98  Liberal Relational Egalitarianism that my subjection only starts at the point where I become aware that another agent has the capacity to interfere arbitrarily with relevant choices of mine. It would leave us unable to explain the status of these choices before I became aware of my subjection, even though they had already been at her mercy. Of course, common awareness of that capacity will enhance the dominator’s capacity to interfere with other choices of mine, as well, as she can now use that common awareness of her power over the frst set of choices to issue credible threats against me. But this is a diferent, further, problem. However, the common knowledge condition does correctly indicate that open relations of domination have a particular quality.5 Tis quality makes them, other things being equal, more unjust than ‘secret’ domination. While being somebody else’s unwitting dupe is objectively bad, whether or not one comes to know of it at some later point, common knowledge enables domination’s attendant psychological evils of fear, deference, and servility, which the eyeball test seizes upon, and which can be harmful to self-​ respect.6 And its particular quality is not only a matter of such psychological consequences, nor of their connection to the dominator’s extended threat power. It can be captured well in terms of what that the respective relation itself expresses about the unequal standing of its parties, and its acceptability. Te secrecy of manipulation is ofen a tribute to its unacceptability.7 Open domination fouts justice publicly, and is thus more disrespectful of its victims.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.2.2.  Agents, Norms, and Structures Dominatory relations can, of course, be constituted, and propped up, by social structures in diferent ways. Tis should, however, not lead us to conclude that, in such cases, it can also be the structures that dominate. Tat would either be conceptually confused, or lead to moral dilution—​a failure to recognize the specifc evil of domination—​or both. If the analysis of the specifc injustice of domination delivered in the preceding chapter (3.3.1) 5 McBride 2015, p. 363. 6 See 6.3.2 in ­chapter 6. A relevant portion of the attendant evils of common knowledge can be given also in cases where the dominated views her superior’s power as a good thing; McBride 2015, p. 364. Even if there should be no fear here, there might be deference and servility, and thus likely defective self-​respect. 7 ‘Hypocrisy is the tribute that vice pays to virtue’, says La Rochefoucauld. Silence and stealth are other forms of tribute.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  99 has force, we have strong reasons to maintain (a)  that social norms, and structures, as opposed to agents, cannot themselves dominate, and (b) that those agents who uphold norms that facilitate, or enable, domination do not thereby necessarily dominate. In an important range of cases, domination is enabled by a structure of customary social norms, which sit somewhere between the ‘naked’ power of the gangster boss mentioned earlier, and fully formalized, in particular legal, authority. Such social norms might, for example, decree that husbands can divorce wives very easily, while wives cannot do the same, even if this is not enshrined in law. Te norms might, for example, exert pressure on the weaker party to formally consent to divorce, making it legally valid, and to give ofcial reasons for it which might not correspond at all to the real ones. Te primary task of just non-​domination is to upend such norms. If the shape and reach of these norms is not, however, presently under the control of particular powerful agents, there is only one instance of domination here:  husbands dominate wives. Nothing is gained, and clarity is lost, by saying that the social norms themselves dominate. Tey are not agents; they are norms that facilitate, or enable, domination. Domination is ‘structural’ in the sense that a domination complaint is triggered already by an agent occupying a position of asymmetrical, arbitrary power, without having to do anything, as shown in the preceding chapter (3.3.2). But it requires that dominators be agents. Giving up on the agency requirement would change the nature and signifcance of the ‘asymmetry’-​condition beyond recognition, and would mean losing track of the specifc evil of domination. In order to be problematic, asymmetry must hold between parties who could, in principle, be symmetrically related—​so that asymmetry requires justifcation at the bar of justice (3.2). Tat is not the case for our relations to social structures. To maintain that social structures have ‘asymmetrical power’ over us would be merely another way of expressing the truism that social structures determine a large part of our individual lives, while we, individually, do not determine them to the same extent. Of course, such norms and structures only exist insofar as they are sustained by the continuous conforming actions of a multitude of individual agents (including expressing, in various ways, disapprobation of those who do not). Tese agents thereby also facilitate, or enable, domination. But these agents are, in turn, not necessarily themselves dominators, because they might not enjoy superior power over the formation and the maintenance of these norms, or over the person whose domination by somebody else their

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

100  Liberal Relational Egalitarianism actions facilitate. Tey may even be co-​opted victims. Some superior agential power there has to be. We should, however, not hold that the evil of inequality of standing characteristic of domination is present only where an individual, natural agent wields arbitrary power over another all by herself, without having to coordinate with others. Tis would make the catch of the concept too low, and would especially be inadequate for dealing with arbitrary power inequalities in more complex social settings, which, in contemporary societies, are the normal case. For the evil to be present, it is enough if an individual agent is part of a group agent, and has some infuence on its actions, while the dominated party is not, or, if it is, has less power over the agent—​or, of a collective which, even though it may lack the formal constitution of decision-​making and execution that characterizes group agents proper (see 2.6.2), can act effectively on the basis of coordination among members. Tus, in a changed version of the gang example, where the gang is a looser, non-​hierarchical group of violence-​minded individuals, you may be dominated, in the strict sense, by the gang, but that would institute an objectionable relational inequality not so much with the group itself, but with all of its members who have some kind of say over its actions. Tere will be many difcult cases, and perhaps even ineradicable grey areas, where we cannot pin down with certainty where agency, especially of a collective nature, starts, or where we have only structures and norms disadvantaging some individuals, but without there being dominators.8 But the possibility of such cases does not render the distinction between agents having power over you and structures and norms enabling such power unintelligible.9 Tis distinction is of crucial signifcance for understanding the specifc, and specifcally objectionable, nature of domination.

8 Take the example of involuntary and inefcient unemployment: labour markets do not clear, and employers consequently enjoy disproportionate power over terms of employment. Tis may be due to employers acting, in some way, in concert in order to safeguard the power of each over her employees (at the expense of short-​term proft), or due to many employers simply following norms of (over-​)caution in taking on employees, without consulting, or coordinating, with each other. In such cases, it is hard to say where collective agency starts, and what exactly to require for its presence. 9 Pettit agrees that domination has to be agential; what he does not agree with is that the specifc evil of domination is always inequality between individuals. However, he muddies the waters somewhat by distinguishing between ‘direct’ and ‘indirect’, or ‘structural’, domination; 2012, p. 63; 2014, p.  53; where the former apparently refers to dominators using their asymmetrical, uncontrolled power to interfere with the dominated, and the latter to their merely being in a position of such power, enabled by social norms and structures. Both are equally cases of domination; the diference is simply that in the former kind of scenario dominators actually interfere. Both may be enabled by social structures and norms, or by ‘naked’ power, or by a mixture of both.

The Demands of Liberal Non-Domination  101

4.3.  Te Intensity of Non-​Domination

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.3.1. Diferent Tresholds With the diferent possible shapes that relations of domination can take more frmly in view, we can turn to the question of the intensity of protection against it, over the range of relevant claims of justice, staked out in 3.3.2. Te analysis of Pettit’s conception of social justice as non-​domination in everyday social life based on the eyeball test (3.4.3) has shown that the particular threshold of protection demanded by this test is too dependent on contingent cultural norms, and may well turn out to be too low, in salient cases. Against that, on non-​ domination as democracy (3.4.2), the level of protection in social relations is simply a matter for deliberation and decision among equals (as are all other substantive matters, too). So how, on the liberal, substantive view is the required degree of protection to be determined instead? How high protection has to be depends on the subject matter: on the centrality of the claims in question. Tis centrality has to be assessed by reference to the conceptions of the person and society underlying the liberal conception of social justice, giving rise to a scale of required intensity of protection. Te more central the claim in question is for individuals’ two moral powers—​capacity for a sense of justice, and capacity to develop and pursue a conception of the good—​the more intense protection has to be. For illustration, let us return to an example discussed in the preceding chapter (3.4.3): protection of freedom of speech is crucial for both moral powers. It gives individuals the opportunity to discuss matters for political decision-​making with others freely before decisions are taken (frst moral power), and the opportunity to fnd out, in free critical exchange with others, what should matter most to them personally, and measure this against others’ conceptions of the good (second moral power). Accordingly, being dependent on others’ goodwill for engaging in activities such as these ofends centrally against the liberal ideas of freedom and equality, based on these moral powers. Parallel arguments can be made for other core freedoms, such as freedom of conscience, association, and movement. In this way, we arrive at a principle of equal basic liberties as those liberties which require most intense protection. Basic liberties on this model are thus not simply those which happen to be co-​exercisable and co-​satisfable, as on Pettit’s model, but those which

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

102  Liberal Relational Egalitarianism are most central to the moral powers, as on Rawls’s conception.10 And they do not exhaust the range of protection required by justice, but indicate the area where protection has to be highest, at an equal level for all: where trade-​ ofs with other claims and their protection are not permissible. An objection to this line of thought might appear to be that almost any kind of protected and resourced choice can be construed as making some kind of contribution to the development and exercise of the moral powers. Following through on this thought would empty the idea of a special class of basic liberties warranting special protection of any meaning, and would lead back to Pettit’s fatter model of the basic liberties, or something akin to it. We must thus hold that the basic liberties proper have a special, more intimate connection to the moral powers. Tis is explained well by van Platz: ‘[A]‌basic right is a liberty necessary for the adequate development and full exercise of the moral powers.’11 It is clear that some liberties, such as the liberty to speak, are so necessary, while others, such as the liberty to own frearms, are not; and it is plausible to hold that a given range of choices can be sufcient to fully exercise the moral powers, with additional choices not making that exercise any fuller. Over these basic liberties, protection must thus be as intense as possible, and provided continuously over the course of individuals’ lives; this is what it means to guarantee them. One might object that it is always possible to protect them more fully, so that this frst requirement, in practice, would exhaust all the eforts that social justice can demand. However, this is not the case: given how basicness has to be understood, as just laid out, the demands of the basic liberties are, in principle, satiable (even if they have perhaps never been fully met anywhere, so far). And it seems possible to guarantee them fully for all, and have resources lef to spend—​at least this will be so in productive societies characterized by an advanced and intricate division of labour (3.2, ­chapter  3). It is important to recall that protection against domination is protection against asymmetrical power. It does not require that anybody, anywhere, must be entirely deprived of any capacity to interfere in others’ basic liberties. Social life is such that this cannot be guaranteed without itself ofending against these liberties.12 But protection against



10 Rawls 1996, pp. 202, 319, 333.

11 Von Platz 2014, p. 26 (emphases in the original). See also Costa 2009. 12 Such as, in particular, privacy: freedom from surveillance.

The Demands of Liberal Non-Domination  103

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

domination over the liberties most tightly linked to personhood and moral agency has to be delivered in full, continuously, and to all.13 Matters become more complicated when we move down the scale of required protection. Tere are many kinds of claims which are not, or, at any rate, only more tenuously connected to the moral powers. Over these claims, a liberal view of non-​domination may only require a lower threshold of protection, guaranteed for all. Over this part of the range, it is thus possible that, as a matter of fact, some will fnd themselves efectively in a position that is safer from domination over some claims than others; and this is not an injustice, if all are nevertheless protected up to the threshold in question. Why is that? Such possible inequalities cannot simply be outside the purview of liberal relational egalitarianism. Te status of individuals as free and equal in cooperation requires that any possible inequalities in power be given justifcation (3.2, ­chapter  3), and this requirement extends to intensity of protection. Tere are three liberal justifcations for permitting their possibility. Te frst is that cranking up the level of protection to maximum on all claims of justice might well leave us short of resources better invested in extension of choice; and extending choice might sometimes itself create new opportunities for domination. Te second is that individuals have a claim to freely engage in personal relationships, particularly of an intimate sort, which are characterized by a high degree of dependency, and consequently by heightened risks 13 Tere is a technical complication in the formulation of the demands of this principle. Te view of the basic liberties and their protection developed here requires, as Pettit’s does, not just the absence of prohibitions on, and of direct interferences of others with, the exercise of basic liberties, as does Rawls’s view, but also that one is efectively enabled to make use of them, and not dependent on the mercy of another for that use. Imagine that the resources you need to be able to efectively make use of basic liberties—​to be well-​nourished enough, for example, to have some efective opportunity to speak and debate with others about questions of justice and the personal good—​are granted exclusively by another person, for whom you work, and who can withdraw them at will, at any time of her choosing. Imagine further that there is no alternative way to sustain yourself. Ten this is not only a case of more ‘general’ economic domination, but makes you dominated over the basic liberties, as well. I see no plausible way around this extension, as taking the elevated status of the basic liberties seriously must mean caring also about the substantive, protected opportunity to engage in the choices covered by them, at the same level for all. However, it has to be noted that Rawls similarly incorporates a ‘social minimum providing for the basic needs of all citizens’ (1996, p. 228) as a ‘constitutional essential’ into his theory, thus giving it, in practice, a status similar to, if not prior to, that of the principle of equal basic liberties, more narrowly understood. If basic needs are understood by reference to the moral powers, and not by reference to some other, less demanding (perhaps ‘physiological’) standard of basic functioning—​as they should be—​this signifcantly reduces the diference between the two views. See Arnold 2018 for a related argument that, within Rawls’s conception, all material preconditions necessary for the moral powers ought to enjoy the same priority as the basic liberties, because they matter for the same reason.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

104  Liberal Relational Egalitarianism of domination. Te third is that eforts to eradicate domination by focusing on the victim’s side, on eliminating personal defciencies and weaknesses that may render a person especially vulnerable to domination, carry risks of disrespecting that person, in a way that itself endangers that person’s equal standing. Let us take the frst two in turn. Te particular problems raised by the third one will be addressed in the following subsection. To the frst problem, then:  protection, resources, and extent of choice. Liberals want to make a large range of choices and options available to individuals, to give them the greatest opportunity possible to develop and maintain their own conceptions of the good. Assigning intensity of protection against domination lexical priority, as a matter of principle, over programs of choice extension, runs counter to this ambition. Take the example of publicly run, rights-​based, egalitarian education and healthcare systems. To be sure, that these systems reduce domination is one very important rationale for them: they do so, compared to other alternatives (such as a laissez-​faire market for these goods), when they are set up properly: administered impartially, on the basis of individual rights to access, which ensure that individuals are not at the mercy of the public agency, and its ofcials, responsible for running them.14 Still, within these systems, we may ofen face a choice between minimizing domination for all and extending choice for all. For example, we may face a choice between reducing domination in the existing healthcare system further, by more tightly controlling health staf and its attendant bureaucracy, or investing into developing, and making available, new health technology with the prospect of extending the life span of people aficted by certain conditions. Tis will be costly, and setting up a scheme administering these new technologies might also itself introduce new risks of domination, not present under the old system; for example, in the decisions by health staf on which individuals will get priority access to the new treatments. As the technology in question is new, we simply might not have the information (yet) to control these decisions reasonably fully. Acquiring the information will be costly; and keeping healthcare professionals from using the technology, and making decisions on a less than fully controllable basis in the meantime, may cost lives. Now, as long as an adequate threshold of protection is guaranteed, this new scheme should not simply be out of bounds for relational egalitarians on 14 See Marshall 1950 for the locus classicus on welfare state provision as fulflling social rights of citizenship, not as a charitable handout, or as conditional on specifc levels of personal economic performance and contribution.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  105 grounds of non-​domination, because of the real benefts it brings, in terms of enhancing, or sustaining, individual capabilities. A liberal theory of social justice must be sensitive to the possibility of trade-​ofs between degrees of intensity of protection—​the modal dimension of non-​domination—​and increased choice and capabilities.15 Where exactly that threshold should be, in cases such as these, is difcult, and ofen perhaps impossible, to determine in advance. According to the model of non-​domination as taking appropriately into account all prima facie relevant claims of justice developed in ­chapter 3, we should require that a decision about an appropriate threshold of protection be made in egalitarian procedures of democratic deliberation, identifying and weighing all such relevant claims based on the liberal background conceptions of the person and society. Te second problem is the problem of intimate personal relationships. Intimacy implies heightened dependency on each other, and this dependency engenders risks of domination. Take the example of marriage: this is a social institution which binds two people together and allows them, typically on the basis of their emotional attachment to each other, to undertake commitments to each other—​such as those pertaining to bringing up children—​which are recognized as binding not only by them, but by the rest of society, as well. Obviously, marriage gives spouses increased opportunities to make each other’s lives hell: they might spoil each other’s projects, overburden each other with expectations, and so on. As long as they have these opportunities symmetrically, and reciprocally, this is not domination (3.3.1). However, and equally obviously, domination may arise in such a relationship, especially as it develops over time. Tis could happen simply because one spouse loves the other more, and accordingly cannot help but give the other’s claims and needs much more weight than their own, with the other perhaps exploiting this, and threatening withdrawal of love in the case of frustration of her wishes. Still, ruling out marriage—​or similar intimate relationships16—​on grounds of risks of domination would clearly be

15 For a similar argument against the priority of intensity of protection, see List 2006, pp. 217f. Pettit (1999a) argues, against that—​unsuccessfully, as the next section will show—​that we can, and should, always assign such priority. 16 For a liberal egalitarian argument against marriage, understood as a holistic arrangement, and for more partial, type-​specifed personal arrangements, see Chambers 2017. Te argument applies, mutatis mutandis, to such more type-​specifed relationships, as well, even though their restricted scope also restricts opportunities for domination.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

106  Liberal Relational Egalitarianism excessive. Liberals must respect individuals’ choices to go for intimate personal relationships, even if risky, as expressions of their moral powers.17 Tis does not mean that a liberal view of non-​domination demands no protection. It demands, frstly, that such life choices do not arise against a background of already structurally unequal risk of domination. In our example, this background could consist of gender norms, such as norms teaching girls that catering to the whims of men is a woman’s true love. It could also consist of the world of paid work being structured so as to typically remunerate male work more highly, and to consequently pile pressure on women to leave most of the earning to men, and to concentrate on the more domestic aspects of family life—​with one’s market-​based earning power then, in turn, also being seen as a reason for having more of a say in family decisions. Secondly, it demands a level of ongoing protection within the relationship itself; for example, in the case of one partner taking over the lion’s share of care work for household and dependants, split pay-​checks.18 Tis is in addition to the protection of the basic liberties, which always has to be continuously, and fully, guaranteed (4.3.1). Tirdly, it demands opportunities to exit relationships that have gone wrong, and to revert to the full level of protection that was present ex ante—​in the case of marriage, these might be divorce provisions that render legally void any spouse’s waiving of claims to fnancial support undertaken in the course of the relationship. Still, all this does not mean that such relationships do not carry risks of domination not present outside them. It cannot, as such risks are the fipside of dependency, which, in turn, is a necessary condition of intimacy for many personal relationships. One might maintain that this already followed, rather straightforwardly, from the scope of non-​domination being restricted to prima facie relevant claims of justice. Tere just is no claim to be kept from falling in love with, and binding oneself to, the wrong person, not even a prima facie one; this is a matter for oneself, and perhaps one’s family and friends. And there is a claim to the freedom to make one’s own relationship errors, and this claim does not stop at the possibility of some personal relational inequality.19 Tis is true, but the discussion has shown there is also 17 Tis is consistent with holding that taking advantage of such risks, even where they are not ruled out by social justice, constitutes a kind of personal injustice. A just person does not make use of opportunities for domination that a just liberal society has to leave open. 18 Okin 1989, pp. 180–​182; Anderson 1999, pp. 323f. 19 To be more precise, a liberal view requires not only the freedom for people to make errors, where they would be, in hindsight, and upon refection, disposed to acknowledge these, but also freedom for them to accept that some domination in a personal relationship is present, in the sense of

The Demands of Liberal Non-Domination  107 a claim to not being systematically pressurized to so fall in love, or to have relationship inequality and exploitation declared socially normal—​as well as claims to some level of continuous protection within a relationship going badly, and to opportunities for exit.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.3.2. Respectful Protection Te third problem of possible overprotection against domination raises a particular, and foundational, issue, and therefore merits separate treatment. It arises from the observation that domination can be facilitated by certain personal characteristics of the victim. Tere are many such characteristics that make it easier for others to establish dominatory relationships: for example, gullibility, excessive propensity to fear, incapacity to make up one’s own mind in the face of external infuence, and incapacity to stick to decisions in the face of such infuence. Such characteristics also make manipulation easier. ‘Pushovers’, who drop any decisions and projects at the frst hint of opposition by others, have it coming, we might say. With all that proper legal procedures, and general social norms (see 4.5.1), can to do reduce domination, some level of personal resilience is also required. Tis consideration is somewhat linked to the well-​known republican topos of a vigilant citizenry, but it applies not only to political relations, and the control of political decision-​making, but also to social life more generally. It may thus prompt the thought that proper protection for all against domination may require screening individuals for particularly heightened levels of vulnerability in the senses just described, and ofering, or perhaps even mandating, remedial eforts. Imagine, for example, mandatory ‘consumer tests’ checking for diferential levels of competence, and immunity to manipulation, in market interactions. Tese kinds of measures are problematic from a liberal point of view not only because they are intrusive. Intrusiveness is a reason for objecting which is similar to the second rationale against overprotection just discussed: creating appropriate space for personal relationships. Tey are, in an even more basic sense, in danger of being disrespectful of individual agency—​in a sense

preferring this to losing the relationship altogether (‘What can I do? I just love her so much’), as long as it remains within the protective bounds established previously.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

108  Liberal Relational Egalitarianism that may undermine the very idea of the basic freedom and equality underlying liberal conceptions of social justice. Why is this? Te domination-​enabling weaknesses listed earlier are all, to some degree, connected to general levels of moral competence and agency. Tey register in the two moral powers:  gullibility and lack of capacity to make up one’s own mind and resist others’ unwanted attempts at persuasion lower one’s capacity for an efective sense of justice as well as one’s capacity to develop and maintain one’s own conception of the good. However, possession of the two moral powers are the grounds for regarding people as free and equal in social cooperation. So, if inequalities in these moral capacities exist, and if people’s standing tracks their possession, the possibility of variation would seem to endanger this basic equality. And clearly, they do exist. Many of us will know someone about whose defciencies in these quarters they are convinced (that person may be oneself). In response to such worries, liberals ofen hold that that moral agency—​the possession of the two powers—​is a ‘range property’:20 people stand as moral equals as soon as they make it into a range of possession of moral powers above some minimum threshold, with variations above that threshold being of no concern.21 But why would that be so? We need a reason for accepting the range property view, keeping us from inquiring into the extent to which diferent people in fact possess these moral powers, and regarding such diferences as salient for justice. In Ian Carter’s words, this reason is that all moral agents are, qua moral agents, entitled to basic ‘opacity respect’.22 Once individuals have made it past the minimum of moral capacities into the realm of moral agency, they have a right to preservation of their ‘outward dignity’:23 to exercise their agency free from others inquiring into, and possibly exposing, their level of it. It would thus be, in this sense, fundamentally disrespectful for agents of social justice to undertake any assessments of moral qualities that would allow them to rank individuals on a scale of moral competence (degree of possession of moral powers, in our Rawlsian case).24 People have a right to be shielded against this particular form of inquiry into

20 Rawls 1999a, p. 444. 21 Rawls 1999a, p. 443; Rawls 1996, pp. 109, 302. 22 Carter 2011, p. 553. 23 Ibid., p. 555. 24 It is important that this competence scale must be absolute, not relative to other people—​it could measure, for example, how far each person falls short of perfection. Prohibiting measuring in comparison to others does not contribute anything to grounding basic equality; it merely restates a commitment to it. Tanks to Ian Carter for instructive discussion of this point.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  109 their moral competence, given its essential role for personhood. And because they have this right, we have to treat them as moral equals even if there will be diferences in moral competence. It is difcult to pin down when exactly opacity respect is violated in this basic equality-​undermining sense. An uncompromising position holds that as soon as agents of justice assess any quality whose possession will, other things being equal, have an impact on one’s moral capacity score, this is the case. A more moderate position objects that this is not plausible: individuals vary along very many diferent characteristics which have some impact on the moral powers (capacity for care being an example, see 3.2 in ­chapter 3). One’s weakness of will when it comes to carrying out decisions for one’s own personal life may well be counterbalanced by sensitivity to others’ claims, which increases one’s capacity for an efective sense of justice. Single-​ dimensional assessments do not allow sound inferences to people’s overall moral capacity, and only the latter are the problem. Ruling out any possibly pertinent assessments is therefore overly cautious, and might be overly restrictive in practical terms, especially when it comes to delivering policies for people with special needs, such as those who need professional help in overcoming an addiction. Deciding this issue would need further argument.25 Te salient point here is that it is not difcult to imagine measures aiming at full interpersonal equality of protection against domination which breach this constraint of respect on any interpretation of it, and thus end up undermining the very equality that protection against domination was meant to express.26 If this argument is sound, respectful protection is a fundamental requirement of a liberal conception of non-​ domination. Furthermore, this requirement, in whichever interpretation, does not rule out general, untargeted measures aimed at enhancing personal resilience. In fact, creating such resilience is part of the purpose of a liberal republican education (see 4.5.1); but there is no reason why assistance should not also be ofered to adults, as long as it is generic, and anonymous.

25 Te more permissive position will have to deliver an indication of the threshold that cumulative assessments must not cross, or at least general guidelines on how to go about fxing such a threshold in particular cases. Te former position is Carter’s view. For instructive analysis, see Floris 2019, chs. 3 and 4. 26 We will see in the next chapter (5.4.3) that relation-​sensitive distributive views of equality, which aim at equalizing people’s ‘relational resources’—​their opportunities to engage in important and fulflling personal relations—​are particularly prone to falling foul of this constraint, on any reasonable interpretation of it.

110  Liberal Relational Egalitarianism

4.3.3.  Te Priority of Non-​Domination How intensive protection against domination has to be thus varies with the subject matter, and it has to honour constraints of respect. However, once the required degrees of protection are fxed, the resulting requirements enjoy priority over other egalitarian demands of social justice, not pertaining to domination, such as requirements to impede the formation of inegalitarian norms of social status (for reasons other than preventing domination; 6.4 in ­chapter 6), and egalitarian requirements of ‘pure’ distributive justice among participants in social cooperation, without consequences for either domination or social status (8.3 in c­ hapter 8). Tis follows from the argument of the previous chapter: to be subjected to the superior, arbitrary power of other agents denies one’s equal standing more directly, and forcefully, than, for example, merely not receiving, without adequate justifcation, a beneft which they enjoy. As subsequent chapters will show, these other egalitarian requirements are unlikely to come into sharp confict with non-​domination requirements, and likely to mutually support each other; but it is nevertheless important to appreciate the priority of non-​domination in principle.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.4.  Why Not Just Maximize (Equal) Non-​Domination? Te protective demands of liberal non-​domination are thus complex, and require not only compliance with basic requirements of respect, but, in areas other than the basic liberties, negotiation of trade-​ofs between extent and intensity of protection of choices leading to various thresholds of protection. Tey cannot be reduced to a simple formula. At this point, one may thus wonder whether more determinate and simpler requirements of protection cannot be had, yielding more straightforward action-​guidance. According to Philip Pettit’s earlier conception of neo-​republicanism, they can be; non-​domination should be maximized. In order to assess the specifc problems of his maximization proposal, we need to assume, for the sake of the argument, agreement on how the substance over which non-​domination ranges is to be determined. Pettit’s earlier conception of arbitrary power, to which the liberal conception is indebted, defned it as a capacity for interference not being forced to track the interferee’s relevant interests;27 let us

27 See c ­ hapter 3, fn. 12.

The Demands of Liberal Non-Domination  111

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

thus simply assume agreement about how such interests are to be identifed.28 Te next step is to make clear what exactly about non-​domination is to be maximized: if no choice is made between the extent of protected choices connected to these interests, and the intensity of protection over a given set of choices that are so connected, then the maximization requirement is obviously indeterminate. Pettit proposes to give priority to maximizing the intensity of non-​domination,29 thus rendering the view determinate.30 Te argument of the previous section suggests the main problem with the maximizing view: it is that it is too determinate—​and thus too constraining. It ends up inevitably attaching too little importance to the extent of choice and opportunities that individuals have.31 However, Pettit is alive to this problem, so it is instructive to see how exactly the various resolutions of it proposed by him fail. Tis corroborates the case for the more nuanced liberal solution. It is important that the resulting objection is not that maximizing intensity of protection may give rise to unjustifable inequalities of such protection. It is not some version of the hoary objection against consequentialism that it does not ‘take seriously the distinction between persons’,32 because it may licence sacrifcing the protection of some to the better protection of others. Tis is not the case, for two reasons. First, Pettit rightly points out that it is empirically unlikely that maximizing intensity of non-​domination could be achieved by protecting some much more than others. Even better protection for the already well-​protected will likely mean that they will generally enjoy more power than the worse protected, which will, in turn, make it more likely that they dominate the latter.33 Tis is not only a plausible assumption; it also

28 As noted in the previous chapter (fn. 59), Pettit’s earlier work also relies, at least in important parts, on the constitutively democratic strategy for identifying relevant interests and determining what counts as non-​domination over them, and has been criticized by McMahon (2005) for the resulting unclarity of the relation between the democratic and the maximizing element. Treating the problems of maximization and democratization separately might encourage the response that combining them could compensate for the problems of each. Tat cannot be ruled out. However, in the absence of any proposal to this extent, combining the two strategies would merely seem to combine their respective problems. 29 For the early Pettit, domination compromises freedom, whereas (non-​arbitrary) restrictions of choice merely condition it, and hence only ‘ofend in a secondary way against domination’; 1998, p. 95. Later work draws a distinction between ‘invasion’ and ‘vitiation’ of choice; Pettit 2012, pp. 37–​ 44: invasion directly subjects the chooser to another will, while vitiation consists of other hindrances of choice (such as lacking the resources needed for it). 30 Pettit 1999a, p. 106. 31 Constraints of basic respect (4.3.2) could, against that, be accepted by it. 32 Rawls 1999a, p. 24. 33 Pettit 1999a, p. 115.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

112  Liberal Relational Egalitarianism illustrates that non-​domination, because of its specifc characteristics, is a good that does seem, initially at least, particularly suitable for a maximization proposal. So this proposal ought to be assessed on its own terms, setting aside more generic objections to maximizing. Second, Pettit emphasizes that, should there nevertheless be cases in which maximizing non-​domination would allow such inequalities of protection, the maximandum of the view can be changed, at the level of principle, to ‘equally intense non-​domination’.34 Tis would do little, if any, damage to the central ambition of delivering a determinate view, and make the view more clearly compatible, at the fundamental level, with the expressive egalitarianism that is at the heart of both liberal relational egalitarianism and Pettit’s own later conceptions of republican social and political justice analysed in the previous chapter. Instead, the objection is based specifcally on how maximizing (equal) intensity of non-​domination fails to display adequate concern for the extent of choice. It can be illustrated in two steps, using two sets of cases that were discussed in the last section—​intimate relationships and resource-​intensive extensions of choice—​and demonstrating for each how Pettit’s attempts to make sufcient room for the extent of choice within the maximizing framework fail. As to the frst case, we have seen that intimate relationships, such as marriage, inevitably engender risks of domination. Tus, from the point of view of maximizing neo-​republicanism, such risks ought to be minimized. But that risks unacceptably constraining the possibility of such relationships. Tis is not the case because neo-​republicans cannot be in favour of marriage, or other intimate relationships. Pettit plausibly argues that it would simply be impossible to safeguard, by way of state regulation and sanctions, against all risks of mutual dependency degenerating into dominatory forms of asymmetrical dependency, and that any attempts to do so will most likely be counterproductive. It would be a prime example of an invasive state assuming too much power, which will turn it (or its ofcials) into a dominating force. Te maximizing view is naturally sensitive to, and indeed designed to account for, the trade-​of between risks of domination by the state, or groups and individuals within the state (imperium), and risks of domination in general social life (dominium).35 It allows us to weigh the advantages of the state regulating the relations between individuals in social life with the help of law and state

34 Ibid., p. 117. 35 Ibid., p. 105.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  113 policing, and control, against the dangers of an overbearing state arrogating itself, or its ofcials, too much power, which cannot be efectively constrained and controlled anymore.36 Given the extent of state power involved in comprehensive regulation and policing of intimate relationships, it is plausible to suppose that the impact on the overall domination score (which we are to minimize) would be enormous. So, republicans should instead encourage a general morality of trust, in personal relationships as much as in less intimate social and political relations, which recognizes that trusting somebody expresses a judgement of virtue about the trustee. Tis, in turn, creates a powerful incentive for the trustee to indeed act in a trustworthy manner, and reap the social rewards of being generally regarded as virtuous.37 Tis is plausible, too, and congenial to the argument for the necessity for social norms shoring up non-​domination that will be made in 4.5.1. However, it still does not do away with the objection. Maximizing republicans must remain committed to the highest degree of institutional protection in intimate relationships which is at all feasible without giving the state, or its ofcials, dominatory power. But it is not, on the face of it, unreasonable to hold that people—​some people, at least—​may have an adequate chance to experience the goods of living together on intimate terms, such as love, the development and pursuit of common projects shared without constant reservations, and the capacity to sacrifce individual interests narrowly conceived, only if the social and political framework they are living under leaves open some more opportunities for domination than the unavoidable minimum. For that, the maximizing view cannot account. But the interests at stake in such relationships are very important. We have seen in the preceding section how the liberal view deals with the problem: it proposes a combination of eradicating unequal background vulnerability to domination (due, for example, to gender norms) with constant in-​relationship protection, while explicitly accepting a trade-​of at the level of principle. Te maximizing view, on the other hand, by denying the trade-​of, in efect turns into a kind of republican perfectionism, in which the requirement of non-​domination

36 Lovett notes that Pettit’s later twin accounts of social justice as sufcient social non-​domination, measured by the eyeball test, and political justice as equal non-​domination within decision-​making (Pettit 2012) seek to do away with this account of an inescapable trade-​of; Lovett 2016b, p. 691. In fact, the maximizing view avoids the problems of contingency and insufcient demandingness which plague the eyeball test (3.4.3), because, within it, risks of political domination set an internal cut-​of point for protective measures by the state regulating social life. 37 Pettit 1999a, pp. 265–​270.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

114  Liberal Relational Egalitarianism takes on the character of a mandatory aim for personal lives, at the possible expense of personal attachments and relations.38 One might seek to resist this argument by objecting that it is, on the face of it, unclear whether maximizing neo-​republicanism, even if implemented fully, would not, in the end, leave sufcient opportunities for intimate relationships intact—​even if it does jettison liberal (neutralist) justifcatory ambitions. We can concede that it is hard to know how practically salient this problem is. Even if it should not be very salient, the objection still clarifes, and underscores, the diference of principle between the liberal view and maximizing neo-​republicanism: that the former takes into account such central personal projects at the level of principle is a theoretical virtue of it, no matter how exactly the facts pan out. However, partisans of maximizing can plausibly retort at least that, if the problem is not practically salient, then their view does retain the advantage of giving clear action-​guidance on the basis of an appealingly simple principle at no, or little, cost in terms of undesirable consequences. In fact, the defciencies of the maximizing view can be brought out more forcefully, in practical terms, by focusing on the second case mentioned earlier: the case of resource-​intensive extensions of choice and life opportunities, such as those brought about by extending the welfare state through a state-​run healthcare system with comprehensive coverage. Pettit argues that we can safely focus on maximizing the intensity of non-​domination frst, before turning to extending choice, because ‘the cause of more efective protection would not generally make substantial demands on the resources that expansion [of choice] would require’.39 Tis assumption relies on the idea that instituting efective protection against domination is generally done via ensuring the rule of law, as well as via the presence of an appropriate republican ethos, and implementing such protective laws, and encouraging such an ethos, are not so costly that they will not leave sufcient resources for expanding choice.

38 Lovett and Whitfeld argue that there is no way in which republicans concerned with the maximization of non-​domination can lay claim to neutrality, because it is not possible to make a case for non-​domination as a primary good with ‘some degree of priority over [ . . . ] other primary goods’, without relying on premises about its pre-​eminent desirability that will be controversial, and cannot be justifed to everybody—​in particular to those who might not object to domination because their conception of the good might even require being dominated; Lovett and Whitfeld 2016, p. 131. For Pettit’s avowed resistance to perfectionist ideals, see 1999a, p. 291; for his liberal ambitions, see 3.4.3. For a neo-​republican position that explicitly espouses ‘quasi-​perfectionism’, see Maynor 2003, especially p. 87: ‘[W]‌here [ . . . ] liberal neutrality stops at regulating how individuals and groups behave, the republican state continues to challenge how individuals or groups cast their ends’. 39 Pettit 1999a, p. 106.

The Demands of Liberal Non-Domination  115

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Here, too, he thus denies the trade-​of which motivated the proposal of differentiated thresholds of protection made in the previous section. However, the assumption is very doubtful, for two reasons. First, implementing non-​domination via the rule of law might well require a lot of resources, especially in enforcement.40 We cannot know in advance how expensive such protection would be. In cases of very deprived areas, extensive ghettoization, and ‘lawless’ areas, for example, whose residents are very vulnerable to domination by criminal gangs (as well as by the police), it may very well require very expensive programs of urban restructuring and integration, the provision of vastly improved educational and other opportunities for inhabitants, and so on.41 Secondly, the assumption fails to register that the expansion of choice might itself bring about some domination. Tat, too, was a point made in the earlier discussion of healthcare provision: some expansions of choice and opportunity may well be accompanied by some domination, for example, by doctors, where the appropriate allocation of likely life-​prolonging but experimental treatments cannot be fully controlled. It is arbitrary to assume that such cases will not arise, so we need a framework able to recognize the genuine interests and claims sometimes speaking against maximal non-​domination. Not having one will, in these cases, certainly lead to undesirable policies. Tere is thus no reason to suppose that the overshooting tendencies of the maximizing view can be neutralized with the help of empirical assumptions of the kind made by Pettit. Tere is no acceptable shortcut to a simpler, general principle governing the intensity of non-​domination.

4.5. Realizing Liberal Non-​Domination Tis concludes the analysis of the principled requirements of liberal non-​domination, and of how these difer from those put forward by neo-​ republicans. To give a fuller picture of its demands, we should adumbrate the 40 Neo-​ republicans could take recourse to the further assumption, mentioned earlier, that enhancing the enforcement power of the state runs up against the requirement to avoid domination by it (and/​or its ofcials). However, we could then invest further resources into controlling such enforcement. 41 Insofar as such protection is necessary to safeguard the basic liberties of inhabitants, the liberal view agrees that maximizing it enjoys priority; see 4.3.1. Te point is that it is not generally true that protection is not expensive, and that this is a reason against maximizing intensity of protection tout court, as opposed to maximizing it over a restricted area of elevated importance.

116  Liberal Relational Egalitarianism range of policy and institutional measures it may require—​before the next chapters will work out further relational egalitarian requirements, pertaining to status and self-​respect, and Part II the implications of liberal relational egalitarianism as a whole for political and distributive justice. Tis section argues that this range is wide, and includes, in particular, a focus on shaping appropriate informal social norms (4.5.1), without there being a reason to fear that this wide mandate will run afoul of any liberal strictures (4.5.2), neither in ideal theory, nor under circumstances of injustice (4.5.3).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.5.1.  Te Variety of Policies of Non-​Domination It is natural to think of the primary task of non-​domination measures as that of designing appropriate legal procedures ruling out the existence of arbitrary power. Paradigm examples are the strong procedural protections for suspects in the criminal law system. Another judiciary example would be provisions to protect individuals from being covered in lawsuits by richer people, or companies (for example, because they have publicly said something to the disliking of the latter). But while such procedural protection is certainly a very important part of such eforts, they cannot exhaust them. For once, the law itself needs democratic input by citizens, also on a substantive conception of domination. However, even perfectly crafed, just law with impeccable democratic credentials cannot, by, itself, bring about appropriately full non-​domination. Te law cannot regulate everything; even the best of laws will leave room for discretion. Te venerable ideal that traditional republican political philosophy has described as the ‘rule of law, not of men’ should be approached as far as possible, but it cannot ever be fully attained.42 Non-​dominatory laws therefore need to be sustained by the right kinds of social norms and individual attitudes, to make sure that discretion is exercised in the right way: they need to be supported by a general disposition on the part of individuals to exercise power over each other, where they have the opportunity to do so, with self-​restraint and attention to others’ claims (including prima facie claims, see 3.3.2); this, we can say, is the spirit, or ethos, of non-​domination. Accordingly, relational egalitarians must also draw on 42 A striking example of this ideal is Montesquieu’s account of the role of judges: ‘[T]‌he judges of the nation are [ . . . ] only the mouth that pronounces the words of the law, inanimate beings who can moderate neither its force nor its rigor’; 1989 [1748], Book 11, ch. 6, p. 163 (my emphasis). Tis is a powerful image for the lack of discretion that Montesquieu believed to be possible, and advocated.

The Demands of Liberal Non-Domination  117

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

socio-​psychological strategies concerned with creating and sustaining these dispositions, supporting and bringing about the right kind of social norms that are active in everyday social life, and combating the wrong ones. For example, such norms play an important role in structuring workplace relations, which can never be fully legally codifed. Tey play a similarly, if not more, important role also for structuring typical interactions between diferent types of people:  men and women; people of diferent sexual orientations, ethnic minorities and the majority, ethnic minorities amongst themselves, and so on. A particularly salient case here is that of upbringing: what kind of attitudes towards others are you likely to develop if you are, for example, the son of a sexist couple, who pay more attention to your needs and development than to your sister’s? And how, at all, will these difer, if this kind of preferential treatment is very widespread, and approved by general social norms? Relational egalitarians have to consider the socio-​ psychological impact of such social forms, most importantly in the formative stage, but also at later stages, on both one’s capacity to treat others with respect, and one’s capacity to respect oneself, and resist others’ unjust power. In this sense, what is needed, as Schefer puts it, is a serious psychology of egalitarianism. Such a psychology would include, for example, an account of the motivational structures and resources that egalitarian institutions could be expected to engage, a demonstration of how egalitarian norms would support the reactive attitudes and emotions that are an important part of human relationships, and a description of the psychological processes by which egalitarian social forms would sustain individuals’ self-​respect and their sense of themselves as free and efective agents.43

Furthermore, distributive strategies also play an important role in the non-​domination toolkit. Chapter  2 argued that relational egalitarianism seeks to account for the difering signifcance to justice of diferent modes of treatment also where distributive outcomes are identical, which purely recipient-​oriented cannot capture. However, distributions matter also because distributive inequality can, and is in fact likely to, engender domination, on its own, or through its impact on social norms (see 8.4 in c­ hapter 8); this is one of the reasons why it has to be limited, and a particularly urgent

43 Schefer 2005, p. 24.

118  Liberal Relational Egalitarianism one. What is more, we need some limits on background distributive inequality already in order to make sure that people have an appropriately functioning sense of justice (3.2 in c­ hapter 3)—​one that is not impaired by social circumstances of segregation and radical diference of life conditions. Tat is, it is needed for the enterprise of social justice to get of the ground from an epistemological point of view: to put us in the condition of being reliably able to identify and understand each other’s claims of justice properly. Given how diferent everyday life experience tends to be for the better of and worse of in very unequal societies, the former may lack the capacity to sympathize with the latter, not only in the sense of not being disposed to act (at least also) in their interests, but in the sense of not understanding these interests, to start with: of not being capable of putting oneself in the other’s shoes.44 Consequently, under conditions of increasing economic inequality, elite decision-​makers will be increasingly unlikely to understand what is important in the lives of those below them; formal procedures, however well designed, cannot compensate for such lacks of basic understanding.45 Tere is, thus, a wide variety of diferent dominatory relationships which require responses by diferent policy and institutional means.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

4.5.2.  An Ethos of Non-​Domination However, the fact that non-​domination yields such a wide policy mandate might be thought to give rise to an internal objection: that such a wide mandate cannot be reconciled with the ambition of our conception to remain faithful to fundamental liberal commitments. In particular, the mandate to shape and control general social norms, as laid out earlier, might give rise to this suspicion: the view might seem to require just too much intervention in individual lives, in the specifc sense of not incorporating a principled enough restraint on the forms that intervention can take. Te answer to this objection should be given in two steps, the frst focusing on the possible content of such social norms, and the second focusing on 44 For accounts of the problem of sympathy between rich and poor in Adam Smith’s work, see Fleischacker 2004, pp. 66f.; and Rasmussen 2016. 45 Anderson 2007, pp.  598–​606, 616. Such epistemological problems can, of course, arise not only due to economic inequality. Tey arise also where social norms dictate the lesser standing of members of some groups (ofen based on ascriptive identities), so that their testimony is given less credibility, and put through distorting lenses imposing the categories of understanding of the socially privileged on it—​instead of engaging in faithful, and charitable, interpretation; see Fricker 2007.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  119 their form: on the medium of social norms, as such. On the frst point: the content of the social norms called for must correspond to the core ideas animating liberal relational egalitarianism. Tis chapter has argued that protection against domination, at diferent thresholds, according to the subject matter at hand, is called for in the name of limiting our pursuit of our own good so as to make sure that others do not fnd themselves at the wrong end of dominatory relations. It is a conception spelling out constraints following from the ideals of the person as free and equal and of society as a fair scheme of social cooperation among such persons. It is not an ideal assuming a mandate to tell individuals how to live so as to realize their personal good. Recall the argument of subsection 4.3.2, which laid out how individuals must remain free to enter into personal relationships which expose them to heightened risks of domination, as long as suitable background protections are in place, and how concern with the extent of choice similarly gives reason to adopt a lower than maximal threshold of protection in areas that do not concern the basic liberties. It is in this sense, that the conception aims to stay neutral between diferent conceptions of the good.46 Whatever social norms it mandates must be sensitive to, and express, this fundamental commitment. However, this merely clears away a possible misunderstanding of the focus of the objection. Regarding the second point, one might still worry that the focus on social norms as such implies that it will not be possible to guarantee that counter-​domination policies will stay within such principled limits, and demand a further restriction, which states that the eforts of non-​domination be located frmly in ofcial, formal institutions, and that individual duties of social justice be correspondingly limited to support for these institutions. Individuals must not be required, the thus sharpened objection holds, to abide by, and help shape, appropriate informal social norms, and have a claim of justice not to be subject to the informal sanctions their violation would prompt (such as social censure, shaming, or perhaps even some forms of shunning). Tese would constitute individual eforts that must not be demanded of them, and liabilities they cannot be asked to incur. Tis would be to insist on a particular kind of ‘basic structure restriction’ for demands of justice; a kind that restricts them to formal—​paradigmatically legal—​institutions.47



46 See Rawls 1996, pp. 191f., and 1.2 in c ­ hapter 1, especially fn. 14. 47 For criticism of such a position, see Cohen 1997.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

120  Liberal Relational Egalitarianism In response to this worry, it is not necessary to take a stand on the general need for, and merits of, a basic structure restriction. Tis is because, even if there should be independent reasons to call for such a restriction,48 on any plausible relational egalitarian view, at least some important and reasonably general informal social norms must come out as part of such a basic structure, so that a narrow, formal construal of it is ruled out. Tere are neither strong reasons to regard it as desirable, nor good enough reasons to think that it can work. On the one hand, domination, as asymmetrical capacity for arbitrary interference, is a structural phenomenon. On purely distributive views of the demands of justice (­chapter 2), it always seems possible to undertake individual eforts to get closer to the required egalitarian distributive pattern in one’s preferred currency, such as opportunity for welfare, or resources; for example, through private donations to disadvantaged individuals. Tis may seem to give rise to very open-​ended obligations, and thus make a restriction to formal institutions look, to that extent, desirable.49 But individuals cannot maintain non-​domination, or eradicate domination—​not even as far their own power positions are concerned—​by isolated individual eforts. If they fnd themselves in a dominatory position, they should abstain from exercising their power in harmful ways, and they should give the dominated party assurance that they will continue to do (see 4.5.3). But that is not the same as thing as eliminating the capacity itself; it is second-​best. Proper anti-​ domination eforts must always have a structural target. Te real question is how much efort individuals can be asked to expend on such structural measures, together with others, and which kinds of measures are called for. On the other hand, limiting anti-​domination eforts to formal institutions is unduly restrictive for exactly the reasons previously mentioned: even the best of laws needs not only to be supported, but also to be complemented by corresponding individual dispositions, which are nurtured by the right kind of social norms.50 Not only is it too much to ask of institutions to police all relations, where people are unwilling to abide by the law’s terms. Tis

48 For defenses of diferent kinds of basic structure restrictions, on diferent grounds, see Estlund 1998; Williams 1998; Pogge 2000; and Schefer 2006. 49 Initially, the main focus of the basic structure debate was in fact more narrowly confned to the question of the proper addressees of the distributive requirements issued by Rawls’s diference principle. With regard to social justice in general, Rawls consistently stresses that the basic structure is its ‘frst’, or ‘primary’, subject, not that it is its exclusive subject. 50 See Pettit 1999a, ch. 8 (and 4.4. in this chapter), on republican social norms, especially of deserved trust.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  121 observation does not cut against the formal restriction. Its supporters are entitled to insist that what is called for is general compliance with commands of institutions, not just compliance where one can be sure that, in cases of contravention, they will step in coercively. Te deeper point is, once more, that it is not possible to fully legalize the right kind of non-​dominatory relations. People have to internalize an ethos of non-​domination, beyond sticking merely to the letter of the law, however comprehensive it may be. What thus needs to be decided is rather the right balance of measures between formal institutional interventions and eforts to create conducive social norms. However, it would be too much to ask for a general principle determining that balance; it ought to be struck on the merits of the particular case at hand. Take the contrasting examples of combating racism, and raising funds from individuals for the pursuit of justice by institutions. It would not be promising to hold that the right way of combating racist attitudes is to rely exclusively on formal institutions issuing legal commands. Te right way to combat racism is arguably to seek to bring about social norms that enable people to identify, and sanction, racist behaviour also informally, including inviting individuals to call each other out in cases of contravention.51 Ofen formal institutional measures will precisely contribute to bringing about such social norms where they are as of yet lacking in civil society. Generally, a liberal position has to rely on a degree of optimism, according to which it is possible, to a signifcant extent, to change social norms, at least in the long term, by changing formal institutions and policies.52 But this does not change the point that such social norms are the best bulwark against racism: the most direct one, and the one that can identify and sanction kinds of racism that the heavy-​handed machinery of the law just could not process adequately. And it is hard to see, in this case, how their existence, as such, could violate other liberal requirements. In the case of ensuring fnancial support for state policies, against that, it makes more sense to restrict eforts to a formal taxation system, whose rules are publicly promulgated and formally checkable, as is individuals’

51 A  note of caution:  putting stress on ‘calling out’ others invites individuals to hide their true views, and resent sanctioners; at least some positive stimuli, rewarding individuals for overcoming racist convictions, especially of an implicit, and unconscious, kind, will also be called for. 52 See, for example, Rothstein 1998 on the positive efect of unconditional welfare policies on social attitudes towards the recipients of welfare benefts. Tis kind of optimism is bolstered by the observation that inegalitarian social norms, such as gender norms, are ofen lef-​overs of historically inegalitarian formal institutions, such as patriarchal family law; see Ronzoni 2008.

122  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

compliance with them.53 Difuse social norms demanding appropriate contributions are arguably not efective enough at assuring individuals that others comply, and reciprocity is present. Eforts to try to make them more efective, by inciting individuals both to apply very demanding ideas of what may be appropriate, and, in case of presumed contravention, to be very trigger-​happy in terms of calling out and shaming others, might, on the other hand, easily lead to excessive intrusiveness, and dangers of persecution, with little opportunity for defence. Tis is because it is difcult, if not impossible, in principle, to calculate the exactly just amount of individual fnancial contributions to be asked of every citizen. Tere is arguably some fuzziness in the requirement itself, not present in the injunction against racism, so offcial promulgation of the level currently deemed just is comparatively more important.54 However, even in the tax case, there may be some appropriate corresponding informal norms: social norms sanctioning those who are legally ascertained signifcant tax-​evaders, on top of legal sanctions they incur, instead of responding with tolerance, or even sympathy, to their attempt to ‘look out for themselves’, might very well be a very good idea, from the point of view of social justice—​as long as the informal sanctions are proportionate. So, whether or not the requirements of non-​domination are, in a technical sense, restricted to a basic structure, they require a set of informal social norms, which, taken together, form an ethos of non-​domination. Such an ethos is not, in virtue of being an ethos, in danger of burdening individuals—​ morally, informationally, or in any other way—​so as to impugn its liberal credentials.55

4.5.3.  Responding to Arbitrary Power Similar considerations apply to the question of which eforts can be demanded by individuals in order to eradicate domination. In cases where individuals fnd themselves in a position of arbitrary power overs others, they must abstain from exercising it, and give corresponding assurance to the 53 For the importance of publicity, see Williams 1998. Williams, however, points out that the demands of an ethos may be sufciently public, if the social norms constituting it are specifed well enough, and compliance is appropriately checkable, ibid., pp. 242f.; he thus puts forward a wide conception of the basic structure which might be appropriate for our purposes. 54 Tis is not, of course, to say that identifying racism is always easy, in practice. 55 For congenial arguments about a liberal ethos, see Tomas 2017, chs. 3 and 5; and Titelbaum 2008.

The Demands of Liberal Non-Domination  123

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

weaker party. If they suspect their power is arbitrary, they must make eforts to ascertain whether it is, in cooperation with those they have power over, as well as by pursuing more formal avenues. Tey must not go ahead pressing their own interest in the hope of getting away with it, and be merely disposed to accept possible ofcial remedial action. If their own position is not at issue, and there is no question of them dominating, individuals must support institutional eforts to eradicate non-​domination, and it might well be that these eforts have to go beyond what their fair share of support would be under conditions of justice, and include having to take over also a fair share of the additional cost of remedying injustice.56 But there is, again, no principled reason to restrict requirements to these two avenues of individual eforts and support for formal institutions, and to exclude more informal social measures: depending on circumstances, they may also be asked to contribute to developing informal norms that help disable dominatory ofcial institutions, by circumventing and undermining them, at the price of incurring some sanctions (even though such a strategy can only be second-​best, unless it manages, in efect, to dismantle said institutions entirely). Limits on the possible reach of requirements to combat injustice can thus only be derived from the core background conceptions underlying liberal non-​domination. For example, individuals plausibly cannot be asked, as a matter of justice, to increase their support for measures to eradicate domination, of whichever sort, to an extent that deprives them of any opportunity to engage in central personal projects. Doing so would be supererogatory. But whether an efort is supererogatory or not does not hinge on its form.

4.6.  Conclusion: Distributing Non-​Domination? Te core features of the liberal conception of non-​domination, as a central part of a conception of relational equality required by social justice, are now in place. A society which protects its members against domination in the ways specifed goes a long way towards expressing their freedom and equality in social cooperation. A good way to conclude the investigation into the shape, and demands, of liberal non-​domination is to double-​check where the resulting conception stands in relation to the distributive egalitarian

56 For a contrary view, see Murphy 2000.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

124  Liberal Relational Egalitarianism views discussed in c­ hapter 2, and to examine whether, and, if so, in which respects, it could be classifed as itself a distributive view. One possible objection coming from such more traditional distributive quarters takes the form of an attempted friendly takeover: it is that the liberal conception of non-​domination is not actually very diferent from familiar conceptions of liberal egalitarian distributive justice, and will, in fact, be parasitic upon one such conception, precisely because non-​domination as defned here has an irreducible substantive component, and such distributive conceptions supply the substance in question: the claims of justice non-​ domination has to range over. Te reply to this observation is that the conception is not parasitic on any such traditional distributive theory. First, it certainly presupposes liberal background conceptions of the person and society; but it does not presuppose any specifc distributive theory, nor that all the claims of justice in question are essentially distributive, in whatever currency one may propose.57 Because of this, it can go together with diferent views of what individuals’ prima facie claims of justice are, and gives guidance on the question of what non-​domination requires, especially in democratic terms, where such claims confict. Second, even where such claims are indeed fnal, the conception still adds, by focusing on the modal nature of protection against domination, an important diagnostic dimension for injustice. It is better able than distributive conceptions to make sense of an important class of injustices, and their gravity: it is able to make sense of the idea that it can be a greater injustice to be subject to arbitrariness in the processing of one’s claims even where the substance of the claim is less important, or even where it is ultimately unjustifed, than it is to have a substantively more important claim disregarded in a procedure that, however, scores better on non-​arbitrariness (see 3.3.2).58 Furthermore, the conception is particularly well placed to explain why people need robust rights against particular, potentially more powerful, others to the things justice entitles them to, as these empower them to afrm their equal standing and ward of possible encroachment by these others, and give them control over their own afairs (including the possibility of choosing not to press one’s claims). 57 For the claims of distributive justice issuing from liberal relational egalitarianism, see ­chapters 8 and 9. 58 We might say that this corresponds, roughly, to a distinction between injustice as error, and injustice as insult.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

The Demands of Liberal Non-Domination  125 Another possible objection is that it may be possible to ‘distributivize’ the view, that is, to formulate its central requirements in distributive language: people are entitled to get equal amounts of protection against domination, at varying levels. Tis is then the currency of justice: the answer the view gives to the question of ‘equality (or sufciency, or priority, or something else) of what?’. Te answer to this is to point out that this may be the case, but that it is not really an objection. First, insofar as we do not merely want to explain, in qualitative terms, the kind of injustice somebody may be subject to, but to measure justice and injustice in diferent cases, and compare them to each other by assigning justice/​injustice scores, we will indeed ofen have to fnd ways to formulate the demands of the conception in such terms, and these inevitably embody a ‘purely recipient-​oriented’ perspective. However, given the complexity of the measure needed, which has to spell out not just the importance of the substantive claims at hand, but also the required level of protection, and has to assign weights to these two dimensions, this will be a very difcult task. It is likely that any measures will have to be tailored to specifc policy and institutional tasks, and will thus be, to some extent, rough-​and-​ ready—​that constructing a global, comprehensive such measure applying to all tasks is not possible. Second, whether or not it is possible, a deeper response to this objection is to point out that it is important to keep the stages of explanation and justifcation of a conception, and of its operationalization for, and application to, particular tasks, distinct. Te former is about pinning down the central aims of a conception, and explaining why these, and not others, are central. For example, if we took the aim of the conception to be that people must get equality in some currency, and took non-​domination to be that currency, we would fnd it more difcult to see that some ways of protecting people against domination are ruled out because they would constitute disrespectful treatment, perhaps even of a basic equality-​undermining kind (see 4.3.2). A distributive formula attracts our attention away from the quality of treatment, and thus away from the meaning it might express. Such requirements of respect are, on such views, if they appear, likely to be mischaracterized as external constraints based on diferent values, not as internal parts of the conception motivated by the same basic concerns. In section 4.2, we have also noted two further expressively important dimensions of domination—​ whether domination rests on de facto authority, and whether it is a matter of public, or common, knowledge. Te

126  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

same holds for other considerations which were identifed in ­chapter 2, but which the conception of non-​domination developed in this chapter has not yet accounted for: this may, for example, pertain to cases in which the same degree of required protection against domination is withheld from two different, but equally large, groups of individuals—​one consisting of random members of society, the other prevalently of members of one specifc ethnic group, or of women. We may well be convinced that the latter scenario is more unjust.59 Te requirements of protection against non-​domination spelled out in this and the preceding chapter are more readily understandable as (part of) an answer to the question ‘How do we have to treat each other in order to respect our status as free and equal?’, which then also directs us to such other cases, not as an answer to the question ‘What is it that people, as equals, should get equal amounts of?’. Te former is the pertinent question for the task of justifcation, focusing our attention on the specifc injustice of domination, and to how the demands of non-​domination have to be construed in response. A diferent, and further question is, of course, whether there are other types of egalitarian relations, beyond non-​domination, whose presence social justice requires, and, if so, whether it does so because they give rise to specifc relational goods which can, and ought to, be included in egalitarian distributive metrics, because they make a distinctive contribution to the overall good of individuals, or to opportunities for it (as purely recipient-​ oriented views demand). Tis question will be among the ones tackled in the next two chapters.

59 It might, again, be technically possible to capture such constraints—​both regarding generally respectful protection, and regarding group composition—​within a suitably sophisticated metric. For specifc policy tasks, this may sometimes be a helpful operationalization; but it would contribute nothing to justifying them, and would, in that sense, be ‘gimmicky’ (Pogge 1995, p. 254, with reference to Nozick 1974, p. 157 n.).

5 Relational Equality beyond Non-​Domination

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

5.1.  Introduction Te previous two chapters have developed a liberal conception of non-​ domination. Tis conception is distinctive insofar as it conceives of domination as necessarily involving insufciently constrained power inequality between individual agents over a substantively determined range of actions and choices. It is this feature which makes it a particularly urgent concern of social justice, because of the hierarchy such unregulated power inequality instantiates, and the inequality of standing between agents it thus expresses. It determines the required range and intensity of non-​domination by recourse to the liberal background conceptions of persons as free and equal in social cooperation, and the idea of a just society as a fair scheme of such cooperation, as the sources of the claims of justice, of a fnal or prima facie nature, over which non-​domination has to operate. And it has to abide by specifc constraints of basic respect in its attempts to implement it. Te natural next question to ask is what else, if anything, apart from equal non-​domination, liberal relational egalitarianism should demand. Tis and the following chapter aim at answering this question, before the second part of the book develops the implications of its demands for political and distributive justice, including health and healthcare. Tis chapter surveys and analyses which other demands there may be, and the shortcomings of existing attempts to deal with them. On its basis, the next chapter delivers a liberal account of the injustice of inegalitarian norms of social status, and of the place and reach of requirements to further and uphold individuals’ self-​ respect as equals. We have already seen that, on a liberal conception of social justice, there is, generally, no reason to restrict the demands of justice to non-​domination, as neo-​republicans want to do. But our specifc focus is on relational equality, and our ambition is to develop a conception which explains and justifes why Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0005

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

128  Liberal Relational Egalitarianism demands of relational equality are demands of liberal social justice, which shape they take, and why they have a strong claim to be the most demandingly and stringently egalitarian elements of such a liberal conception. So our focus has to be twofold. On the one hand, we have to ask which demands of relational equality there may plausibly be, other than non-​domination: the abolition of hierarchies of social status, an egalitarian allocation of recognition and social esteem (where this might require more than abolishing hierarchies), the furthering and maintenance of individuals’ self-​respect as equals, enhancing relations of community, and civic friendship, enhancing personal relationships of trust and friendship? On the other hand, we have to ask which of these can, and ought to be, conceived of as matters of liberal egalitarian social justice—​how they are to be incorporated into it in a way that is consistent with, and supported, by the background ideas on which it rests, so that they can justifably command the special force that requirements of justice are widely thought to have (2.4 in c­ hapter 2). To prepare this task, this chapter analyses, and rejects, two prominent and seemingly promising strategies for accounting for a wide array of requirements of social equality: the pluralist strategy, and the strategy of incorporating relational goods into egalitarian metrics of distributive justice, which we already encountered at the end of c­ hapter 2 (2.6.2). It is structured as follows:  section 5.2 illustrates the need to extend requirements of relational equality by demonstrating the insufciency of the exclusive focus on non-​domination that is characteristic of neo-​ republicanism. To this end, it briefy returns to Pettit’s conception of social justice as non-​domination, revolving around the ‘eyeball test’, as a good example for this problem. It shows that, in addition to being under-​demanding regarding non-​domination itself (3.4.3 in c­ hapter 3), it arbitrarily excludes other relational egalitarian concerns. Section 5.3 tackles a family of views which we can call pluralist social egalitarianism. Tese views conceive of the requirements of social equality as, at least partly, based on a free-​standing value alongside justice. Such a free-​ standing value seems particularly well placed to explain why social equality gives rise to a wide variety of distinct demands. However, pluralist views either fail to make a convincing case for such a distinct value, insofar as they misconstrue the distinction between social justice and social equality; or they rest on an account of the value of social equality which gives us hardly any mandate to shape society in conformity with its demands. Nevertheless, one of their central ambitions, that of accounting for the evils of diferential

Relational Equality beyond Non-Domination  129

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

social status, and of treating each other as unequal even where domination is not at stake, needs to be kept, and c­ hapter 6 will show how it should be fulflled. Section 5.4 analyses several members of the nascent family of relation-​ sensitive distributive theories of justice mentioned at the end of c­ hapter 2. Tese seek to incorporate the goods constituted by egalitarian relations into the metric of egalitarian distributive justice, because of the distinctive contribution to the individual good, or to opportunities for it, which they are supposed to make. Tis strategy, if successful, promises to be able both to account for a wide variety of social egalitarian concerns, and to explain why they are matters of social justice. Tis would thus answer the objection to purely recipient-​oriented theories developed in that chapter, and avoid the difculties encountered by pluralist social egalitarianism. However, different such views encounter diferent problems: they either miss the specifc goods such relations might constitute or misconstrue their value; or they cannot lead to a form of relational equality, because aiming at equality in goods such as relations of trust and friendship runs up against independent requirements of basic respect (4.3.2 in ­chapter  4). So, once more, liberals have to look for other ways to account for requirements of social equality beyond non-​domination. Furthermore, some pluralist and relation-​sensitive distributive approaches also aim to account for non-​domination, so demonstrating their shortcomings further corroborates the case for the liberal conception of non-​domination. However, the main focus of the chapter is on how to extend relational equality, and how not to.

5.2.  Te Insufciency of Non-​Domination A focus on non-​domination as the master value of political morality is central to the neo-​republican project, at least as Pettit, and others such as Lovett, conceive of it (­chapter  3, 3.3.1; 3.4). Te previous two chapters discussed and criticized several prominent neo-​republican proposals for justice as non-​domination, some of which also aimed at yielding specifc guidelines for the relations between individuals in social life (dominium), outside political decision-​making. Philip Pettit’s recent proposal for just non-​domination between individuals came closest to the liberal view. It explicitly claims liberal credentials, but ends up being under-​demanding, or, if appropriately demanding, then only contingently so—​on both liberal and republican counts.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

130  Liberal Relational Egalitarianism It is a particular kind of sufciency view of social justice as non-​domination:1 sufciency is reached when people are able to look each other in the eye. More protection against domination is not needed. It is not egalitarian, as cultural and other social norms may allow for reaching this point in the face of otherwise signifcantly unequal protection against domination. However, Pettit’s conception has another very important shortcoming, from the perspective of expressive egalitarianism, avowedly shared by Pettit, which requires that agents of justice, such as basic institutions, express in their treatment of those under their power, within the scheme of cooperation they organize, their standing as free and equal. It fails to include matters other than non-​domination within the scope of social justice, and does not worry, once non-​domination is guaranteed, about the fairness of societal arrangements in other respects, without justifying this exclusion.2 Te question of whether individuals enjoy equal, or not too unequal, opportunities to live their lives as they see ft does not register as a question of social justice in its own right3—​whether or not the obstacles to disparities of opportunity are in some specifc way socially created (see 2.5.2 in ­chapter 2) or not. Pettit’s view thus seems to see no problem at all with a social order that puts considerably more difculties in one’s individual path to what counts as success, by her own lights, than in that of another individual—​as long as the former enjoys resourced and protected basic liberties that fulfl the ‘eyeball test’. No justifcation for this is apparently needed, at the bar of justice. An example for this is educational opportunity, where the question of equal opportunities, in whatever dimension we may deem relevant, is not even mentioned.4 It is not necessary to validate any particular conception of equality of opportunity to see the general problem. However, relational egalitarians will also worry, more specifcally, that social relations other than domination may present their own, distinct problems of hierarchy and subordination. Entrenched inequalities of social status, for example, might create obstacles for self-​respect, or represent particularly objectionable ways of depriving individuals of social opportunities. People looking down on, or up at, others according to entrenched social norms and patterns might thwart desirable relations of equality, and the 1 Lovett 2016b, p. 692. 2 Pettit extends the concept of domination further than the liberal view does, covering also cases where no power inequality is at stake (3.3.1 in c­ hapter 3); but what follows remains outside. 3 See Gilabert 2015, pp. 647f. 4 See Pettit 2012, p. 111.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational Equality beyond Non-Domination  131 wrong of such inequalities might not always be traceable to distinct agents having arbitrary power over others.5 If so, these considerations will also call for their own responses. It would be merely fortuitous if implementing sufcient non-​domination, as a matter of fact, happened to solve any such problems, too, whatever they might turn out to be. Pettit’s restrictions of social justice thus seem arbitrary; and no argument for them is presented.6 He merely asserts that views which are more demandingly egalitarian, and issue requirements of social justice beyond assuring sufcient non-​domination over basic liberties according to the ‘eyeball test’, ‘seem like moral fantasies’.7 We have seen that, in order to avoid the problems of contingency and insufcient demandingness besetting Pettit’s conception of social justice as non-​domination in social life, we should take recourse to thicker background ideas of the person, their moral powers, and social cooperation. Once one does that, however, the additional problems just mentioned become even more pressing. As soon as one focuses on spelling out people’s equal standing in cooperative relations so understood, it is even less clear why we should restrict ourselves to questions of non-​domination, and not allow other substantive requirements, as well; in particular, as doing so is entirely consistent with holding that non-​domination is an especially urgent demand (4.3.3 in c­ hapter 4). If we are looking for substantive social justice, why not go all the way? Among the diferent conceptions of non-​domination analysed before, ‘non-​domination as democracy’ seems, ultimately, strongest also with regard to being able to account for other requirements, or at least for justifying that they are indeed secondary. If all had political equality of the kind that this view calls for, the fact that, from such a position, they could continuously bring their own substantive views to the fore, on matters such as opportunity, social status, or any other, would increase the chances of them indeed being sufciently addressed in political decisions—​and count as adequate compensation if they were not. Furthermore, we might surmise, with some 5 See Laborde and Garrau 2015, pp. 61f., for an analysis of status norms stigmatizing the poor, and an argument that these cannot be fully accounted for as matters of interpersonal domination. 6 Tat combating domination will go some way towards combating other relational inequalities is certainly a plausible conjecture, but not such an argument. Tat requiring political equality is sufcient for addressing them would be one, but it is not successful; see c­ hapter 3, fn. 88, and what follows in the main text. 7 Pettit 2012 , p. 126; he applies this label both to luck egalitarian views (see also ibid., pp. 79, 92), which focus on social and natural inequality alike, and to more restricted views, which focus on social processes, and do not require any form of strict equality, such as Rawls’s view, and the diference principle.

132  Liberal Relational Egalitarianism plausibility, that a society that is so thoroughly politically egalitarian would not be characterized by any kind of very serious relational inequality in any other dimension, anyway. However, we have seen that the kind of full and egalitarian democratic control over the basic structure of society on which this view ultimately trades—​diachronic equality of political power proper among all—​can, in complex societies with an intricate division of labour, not be upheld. Tis then, turns out to be an argument both for adopting a substantive conception of non-​domination instead, as we did, and for taking seriously the need to extend relational egalitarianism beyond non-​domination.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

5.3.  Pluralist Social Egalitarianism How might one go about such an extension? A very important family of theories within the broad social, or relational, egalitarian camp approaches the demands of social equality from a direction that is diametrically opposed to that of the neo-​republican family. It does not call for the reduction of central social egalitarian demands to one master value, or principle, such as non-​ domination, but for recognition of a plurality of fundamental values, beside justice, in order to account for the fullness of such demands. Tis is the family of pluralist social egalitarian views: according to such views, at least some relational egalitarian requirements are not grounded in social justice, but in a diferent, free-​standing value. Social and political relations in which participants treat, and regard, each other as equals, instantiate a kind of value for these participants, for the community of which they are members, or impersonally, which is not linked, or at least cannot be reduced, to that of justice. In this respect, these views thus hark back to an older tradition of egalitarian theorizing, which considerably precedes theorizing about social justice—​found, for example, in the works of the Levellers, and Diggers, and arguably also Rousseau (1.2 in ­chapter 1). However, the relevant views in contemporary political theory do not put forward an ideal of social equality as a comprehensive yardstick for how we should organize societal life. Tey all accept that the demands of social equality are merely those of one value among others, and have to be viewed together with, and sometimes in competition with, those of these other values, such as justice—​hence the label ‘pluralist’ social egalitarianism (PSE). If PSE is the right approach towards relational equality, its demands will stretch beyond non-​domination; but our

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational Equality beyond Non-Domination  133 main ambition, to account for them as requirements of social justice, has to be given up. To qualify as a pluralist social egalitarian in the sense used in this chapter, it is enough to be committed to the view that (a) there is at least one important social egalitarian rationale that social justice cannot account for, and which therefore calls for reliance on a value of social equality distinct from justice, and (b)  that the requirements of social equality have to be balanced with those of other values—​including justice, should they confict—​at the level of principle. Tis is a broad and minimal defnition; PSE denotes a family of social egalitarian views, whose members can be more or less pluralist, according to the degree of possible confict which they identify between justice and those social egalitarian rationales that lie outside it, and according to the degree to which the requirements of social equality are thought to go beyond those of justice.8 Of course, in some instances, diferences to justice-​based views could turn out to be mostly, or even entirely, terminological: it obviously depends on what is meant by ‘justice’. Tere are, however, clear contrasts between such positions and the liberal justice-​based conception of relational equality developed so far, which aims at spelling out fair terms of cooperation between free and equal individuals, and aims to account for specifc egalitarian relations, such as that of non-​domination, as a central part of these terms. So we should analyse these contrasts, with a view to identifying, and comparing, the advantages and disadvantages of both types of position. One particularly important apparent advantage of the pluralist view is precisely that, because of its focus on the intrinsic value of treating each other as equals, it easily includes social norms of esteem and status within its purview, and militates for egalitarian such norms. Such norms do seem a natural concern for social egalitarians, one that should be quite close to the heart of any position that deserves that name. But it is not clear how, if at all, such concern can ft under a liberal conception of social justice. One might think that one straightforward contrast that speaks for the justice-​based position is that the pluralist position, by defnition, does not, and cannot, lay claim to any kind of principled priority of its demands over those of other values. Nor can it put forward any clear priority ordering 8 Pluralist positions are advanced by Cohen 2009; Mason 2012 and 2015; Miller 1995 and 1998; O’Neill 2008; and Wolf 1998. Views discussing a plurality of egalitarian demands without committing to pluralist social egalitarianism in the sense of this chapter include Fourie 2012; Hausman and Waldren 2011; and Scanlon 2018.

134  Liberal Relational Egalitarianism between its own demands. Te liberal justice-​based view aspires, against that, to such priority over other values (2.4 in c­ hapter 2); and it also delivers internal rankings between its own demands (see c­ hapter 4, 4.3.3; c­ hapter 6, 6.4; ­chapter 7, 7.3 and 7.4; ­chapter 8, 8.4; ­chapter 9, 9.3). But this is not so much an objection to pluralism as merely a restatement of its central feature, and of the main formal contrast to justice as conceived of here. If there are important, and desirable, features of social egalitarian relations that cannot be accounted for properly within a justice framework, then so much the worse for the latter, it would seem. Te theoretical price of having to resort to determining the weight of the demands of social equality in a more case by case fashion (2.4 in ­chapter 2), including their weight vis à vis considerations of justice, as the case may be, is surely one worth paying for getting the substance of social equality right. So we need to inquire into which substantive rationales there may be for conceiving of (at least some of) the demands of social equality as operating beyond the realm of justice, and into what such demands may be.

5.3.1.  Distributive Justice versus Social Equality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te most straightforward such rationale would be that justice ranges only over the distribution of certain goods, and has nothing to say about social equality. Tis is the view of David Miller: [T]‌here are two diferent kinds of valuable equality, one connected with justice, and the other standing independently of it. Equality of the frst kind is distributive in nature. It specifes that benefts of a certain kind [ . . . ] should be distributed equally, because justice requires this. Te second kind of equality is not in this sense distributive. It does not specify directly any distribution of rights or resources. Instead it signifes a social ideal, the ideal of a society in which people regard and treat one another as equals, in other words a society that is not marked by status divisions such that one can place diferent people in hierarchically ranked categories, in different classes for instance. We can call this second kind of equality equality of status, or simply social equality.9 9 Miller 1998, p. 23, emphases in the original. Miller’s view is based on, and further develops, Michael Walzer’s account of ‘complex equality’; Walzer 1983. Walzer’s view is not included in this discussion of pluralist social egalitarianism for the following reasons: First, Walzer holds that social

Relational Equality beyond Non-Domination  135

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

On this view, it is also clear how justice and social equality could confict: justice could require a distribution that directly ofends against social equality, by instantiating such a hierarchy, or, somewhat more indirectly, by mandating distributions that create risks of such a hierarchy emerging; risks of a kind, or magnitude, which social egalitarians will fnd intolerable.10 However, this rationale relies on a concept of justice which we have very good reason to reject. Tat was one of the main points of the preceding chapters. If social justice is about structuring cooperation between individuals conceived of as free and equal fairly, then one very prominent demand on such fair terms of cooperation is that they make sure that, at the very least, these individuals do not encounter each other in relations of domination—​ they do no treat each as other as equals if they deny each other the required protection, and do not regard each other as equals if they fail to recognize goods, such as education, healthcare, money, and others, should be distributed according to their social meaning: ‘All distributions are just or unjust relative to the social meanings of the goods at stake’; ibid., p. 9. Hence, health care should be distributed according to need; ibid., p. 91; education—​at least above a certain minimum level—​according to talent and willingness to learn and study; ibid, p. 202; and so on. On the basis of this, Walzer defnes complex equality as follows: In formal terms, complex equality means that no citizen’s standing in one sphere or with regard to one social good can be undercut by his standing in some other sphere, with regard to some other good. Tus, citizen X may be chosen over citizen Y for political offce, and then the two of them will be unequal in the sphere of politics. But they will not be unequal generally so long as X’s ofce gives him no advantages over Y in any other sphere—​superior medical care, access to better schools for his children, entrepreneurial opportunities, and so on. (ibid., p. 19). Te frst reason to exclude this theory from the present discussion is thus that it is not clear how it is a theory of equality: it might well be that all social goods are distributed according to their meaning, and spillovers between goods (‘dominance’; see ibid., pp. 10f.) do not take place—​granting for the sake of argument that we know what this means, and that it is also feasible—​but nothing resembling any kind of equality between persons ensues. It is more plausible to understand Walzer as claiming that, if the distribution of goods is properly kept separate according to their spheres, it turns out that we cannot rank people because the fact that diferent goods have diferent meanings implies that holdings across spheres cannot be compared—​they are incommensurable (see, for this interpretation, Miller 1990, p. 96). But while it is true that diferent individuals can, under such conditions, not be said to be unequal overall, it does not follow that they are therefore equal; what follows is simply that their overall positions cannot be compared. In order to claim that some form of overall equality does indeed hold, diferent individuals must be equal in some foundational sphere: Miller argues that this sphere is the sphere of equality of citizenship, which is implicit, and lef underdeveloped, in Walzer’s account of complex equality. Citizenship equality, if understood not simply as formal equality, but broadly enough, instantiates equality of status; Miller 1990, pp. 97–​99. Te second reason is, as the preceding quotations make clear, that Walzer conceives of his view of complex equality as a conception of distributive justice, so precisely not as a free-​standing conception of social equality. 10 For an example, see Miller 1995, p. 204 (repeated with slight variations in Miller 1999, p. 48): A boy wins all competitions for honours in his school. Tese are set up fairly, and proceed against a background of equal opportunity, so justice, according to Miller, is done. But the fact that he wins all of them instantiates a clear, and fairly comprehensive, hierarchy between him and the others; and this is a problem for social equality.

136  Liberal Relational Egalitarianism their importance.11 So this way of drawing the distinction between justice and free-​standing social equality fails, because at least relations of power are frmly within the scope of justice.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

5.3.2.  Social Equality beyond Justice (I): Special Goods However, this is only one kind of social relation; what about others? A more promising rationale for conceiving of social equality and justice as at least partly distinct is that social equality goes beyond justice. Teir demands may ofen overlap, but social equality comprises surplus demands that cannot be accommodated by justice. Tis is in fact the construal of the relationship between the two values which most adherents of PSE subscribe to. Social equality is not only ofended by domination, but by a host of other social phenomena. People can fail to treat, and regard, others as equals, even if they do not dominate them. Tey can, for example, judge others to be inferior, and therefore sneer at them, or shun them. Similarly, there are a host of different positive qualities that social relations can display which are intuitively connected to social equality, yet cannot be reduced to, or might not even involve, power equality, or protection against abuse of superior power. Positive accounts of such relations variously describe them as ‘healthy fraternal relations’12, or as instantiating ‘the good of equal membership of a social and political community’13 (over and above the good of fulflling each other’s claims of justice), or ‘general social friendship’.14 It is important to note, that, on some conceptions of such egalitarian relations, for there to be objectionable social inequality, it is not even necessary for any party to them to regard the other party as unequal in any way, and behave accordingly. It is enough, for example, for there to be material disparity in people’s life situations, which, even if it should be licensed by justice, nevertheless makes it so that they cannot communicate with ease, and, to that extent, cannot share each other’s fate. Here is an example by G. A. Cohen:

11 In other writings, Miller recognizes that social justice is not exclusively distributive, and possesses an irreducibly procedural dimension; 1999, ch. 5; but he does not draw a connection between just procedures and social equality. 12 O’Neill 2008, p. 126. 13 Mason 2012, p. 2. 14 Cohen 2009, p. 52.

Relational Equality beyond Non-Domination  137

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

You have to take the crowded bus every day, whereas I pass you by in my comfortable car. One day, however, I must take the bus, because my wife needs the car. I can reasonably complain about that to a fellow car-​driver, but not to you. I can’t say to you: ‘It’s awful that I have to take the bus today.’ Tere is a lack of community between us of just the sort that naturally obtains between me and the fellow car-​driver.15

Other ideals of social equality are not as wide as that. Tere are, of course, as many diferent conceptions of egalitarian relations as there are authors writing on them. At this point, it is more important to focus on what underpins such ideals—​how exactly their value is to be spelled out, what makes it plausible to think that that value stands apart from justice. One fairly common claim is that egalitarian relations of the right kind are an important good for individuals, as they allow them to have the right kind of self-​respect.16 People engaged in them will rightly come to believe, and internalize, that they are each other’s equals, not only in some abstract sense, but in the concrete social and political contexts they fnd themselves in; and being so engaged will also make these convictions secure. Such relations may also instantiate other, related, goods, which are clearly valuable for individual well-​being, such as the good of enhanced mutual trust even among people who do not know each other intimately—​which is a key component of ‘general social friendship’, or community. However, the problem with this understanding of the rationale of social equality as a distinct ideal is that it is, again, not clear why it necessitates moving beyond justice. If we are dealing with goods that are valuable for individuals, there seems to be no reason why such goods could not, in principle, be incorporated into the metric of justice. And that is the case also within theories focusing on social cooperation, and the burdens and benefts it gives rise to. Enhanced self-​respect and trust may well be benefts which people engaged in the right kind of cooperative relations enjoy, and have to go without in the wrong kind of relations. Insofar as these goods are crucial, 15 Ibid, p. 36. Inegalitarian behaviour would be involved if, in the example, I did complain to you in that way—​in that case, I would not have ‘checked my privilege’, and would not have given any attention to your less privileged situation. But Cohen’s point is that even if I do not commit this error, the ideal of community is already thwarted, because the disparity in our situation obstructs the chance to attain it. As noted in the preceding chapter (4.5.1), excessive disparities in material life circumstances will also present obstacles to people acquiring and maintaining a functioning sense of justice; however, the problem for Cohenite community kicks in much earlier, as the example demonstrates. 16 O’Neill 2008, p. 126; Miller 1998, p. 30.

138  Liberal Relational Egalitarianism and potentially non-​substitutable, for individuals, what they necessitate is a move beyond narrow metrics focusing, for example, purely on economic resources; but that can be understood precisely as a call to devise a broader metric for justice (2.6.2 in c­ hapter 2), not as a call to move beyond it.17 Various such enlarged metrics of justice have been proposed in the recent literature, and merit deeper analysis. We will see in the next section that they encounter their own problems; based on this analysis, the role and requirements of self-​respect on the liberal view will be tackled in the next chapter (6.3.2). Te intermediate conclusion for social equality here is simply that this interpretation of its rationale does not deliver good reasons to think that its requirements have to operate beyond the realm of justice.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

5.3.3.  Social Equality beyond Justice (II): Impersonal Value Tis leaves the third, and most promising, rationale for their free-​standing nature: egalitarian relations instantiate a kind value that is independent of, or at least not fully reducible to, the value they have for the individuals engaged in them. Claims of this kind are perhaps not as frequent among pluralist social egalitarians as those about the specifc personal goods connected to social equality, but they are not uncommon. Miller holds that the requirements of social equality are not ‘centrally about unfairness in distribution; the claim is that our society should not be like that.’18 Tis would seem to imply precisely that they are not reducible to individual claims to certain benefts, or the avoidance of certain burdens, but that a socially egalitarian society impersonally instantiates a particular kind of value.19 Te most explicit statement 17 White raises a similar objection to Miller’s account of social equality; 2007, pp.  18f. Wolf ’s writings on social equality have developed in exactly this way. Earlier writings accept, at least for the sake of argument, that egalitarian justice is narrowly distributive, and only ranges over fair opportunities of the kind that traditional luck egalitarian views are concerned with; Wolf 1998, p. 101. Tis makes reliance on a broader, not entirely justice-​based ‘egalitarian ethos’ necessary, to incorporate the demands of self-​respect. Later writings operate, against that, with an enlarged capability metric of justice, which includes social relations, and/​or their psychological efects on individuals; Wolf and de-​Shalit 2007, p. 2; Wolf 2009, pp. 121f, and 2010, p. 347. 18 Miller 1998, p. 24, emphases in the original. 19 For this interpretation of Miller’s view, see Elford 2012. Elford 2017 proposes this as the distinctive feature of relational egalitarianism tout court. Schefer objects to such construals that it is misguided to suppose ‘that the only alternative [to a narrowly distributive, individualist conception of equality] is a communitarian defence [for social equality] according to which equality is good for society as a whole because it is a condition of the right kind of relations among its members’; Schefer 2005, p. 21. He regards this as a false dichotomy, and holds that accounting for the demands of social equality does not necessitate ‘a departure from basic liberal principles’; ibid. Making good on this liberal promise is precisely our task; it requires taking seriously the strengths of the impersonal view.

Relational Equality beyond Non-Domination  139 of this view is to be found in O’Neill, who entertains the suggestion that egalitarian relations ‘have a basic moral signifcance that is not exhausted by their value for any particular individual’: they are ‘impersonally valuable.’20 Andrew Mason advances a similar line of argument about the surplus requirements of social equality. He devises the following example: Local Shop. A man refuses to shop at his local store because it is run by a member of a particular ethnic minority. Instead, he walks to a store that is farther away and more expensive because it is run by someone from his own ethnic group. It is not simply that he is opposed to the immigration policies that have led to a rise in the numbers of people from this ethnic minority; it is because he thinks that members of this group are inferior because they generally lack various qualities, such as honesty and integrity, that are widespread among members of his own ethnic group, so he does not want to give them his custom.21

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Mason argues that there is something morally problematic about this case even if justice is not at stake. By that, he means that it is problematic even if such behaviour does not have any kind of negative impact on the shopkeeper discriminated against, because it causes no psychological harm (let us say the shopkeeper never even fnds out), or any deprivation in terms of other goods to which he might conceivably have some kind of claim (let us say the shopkeeper reliably has more potential customers than he could possibly handle, anyway).22 Such cases nevertheless involve ‘a failure to treat others as equals’, and his contention is that what makes such a failure morally problematic is the fact that it involves withholding or denying the equal recognition respect to which they are all morally entitled [ . . . ].23 20 O’Neill 2008, p. 146. He wants to leave open the possibility that the diverse requirements of social equality might all ft under a conception of ‘all-​things-​considered social justice’; ibid., p. 144; without, however, explaining how and why justice would demand the promotion of impersonal value. 21 Mason 2015 , p. 132. 22 Ibid., p.  142. Mason also allows that domination is another, separate rationale of justice; he defnes it, however, as ‘being subjected to the arbitrary exercise of power’ (ibid., my emphasis), which does not capture the specifc injustice of domination as an unchecked capacity for interference (3.3.1 in ­chapter 3). 23 Ibid., p. 145. To be precise, Mason employs a tripartite distinction between duties of justice, duties of equal citizenship, which are violated when the person treated as unequal is also a member of the same social and political community, and duties to treat others as equals simpliciter. So a pure case of the third category, failure to fulfl a duty to treat others as equals simpliciter, might be one in which the prejudiced shopper behaves in a similar way while on a tourist trip to a remote country. For

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

140  Liberal Relational Egalitarianism Such recognition respect is withheld when agents fail to give due consideration to the intrinsic value of others as persons—​which is the case here, insofar as the shopper simply allows his pre-​formed opinions about members of the group in question free reign.24 Now, the language of ‘entitlement’ employed by Mason is somewhat problematic, as it seems to suggest that the shopkeeper has some kind of directed claim against the shopper not to behave in that way (including, perhaps, a right to remonstrate in case of contravention). Tis might, intuitively, push the case closer towards justice than Mason thinks it is: its central feature was supposed to be, as seen, that the shopper violates a requirement of social equality even if nothing for the shopkeeper is at stake. But whether or not the shopkeeper has any such claim, it seems right to say that the shopper behaves in a way which social egalitarians ought to fnd objectionable, even if no specifc burdens or benefts for the other party are at stake. Tis is why Mason’s free-​ standing requirements of social equality are close to those advanced by Miller and O’Neill on the basis of the impersonal value of relations of social equality. Tis rationale, if defensible, thus promises to account for the free-​standing nature of at least some requirements of social equality. But is it defensible? One objection one might advance against such impersonal value and consequent requirements of social equality is that they are simply ‘mysterious’.25 But this does not seem to be a good objection. It is intelligible that a relation that is generally good, or bad, for its participants is also good, or bad, in some impersonal way—​and that it is good or bad in that latter way even if in a specifc instance, such as in Mason’s example, it is not good or bad for some of its participants. Take, as an analogous case of a good relation, that of a caring relationship.26 Tis is not only typically immensely valuable for the person cared for, and valuable for the carer, as well, for example, in terms of their moral

the distinction between the frst two categories, see Mason 2012, ch. 2. Its exact nature is of no importance for the argument of this subsection. 24 Mason 2015, p. 139f. For the concept of recognition respect, see Darwall 1977. 25 Wolf and de-​Shalit 2007, p.  6; Tomlin 2014, p.  175. Te objection motivates Wolf and de-​ Shalit’s reliance on an enlarged capability metric incorporating (opportunities for) desirable social relations (see fn. 17). Tomlin does not endorse the objection, but merely points out that some social egalitarians have faulted so-​called telic distributive egalitarians, who hold that equal distributions have impersonal value even when they are not required by justice, for putting forward mysterious claims (see O’Neill 2008, p. 140; the label ‘telic’ stems from Parft 1991). His point is that it would be somewhat ironic if they held, in the end, a similar view about the value of egalitarian relations. 26 Elford 2017, pp. 84f.

Relational Equality beyond Non-Domination  141

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

development.27 Precisely because that is so, it is not difcult to see that it also, as such, instantiates a particular kind of value: a value which we could, for example, cash out as the approval that an impartial spectator would give, and experience, when contemplating it, in addition to the vicarious, empathetic happiness felt specifcally for the parties involved.28 Nor would it necessarily be a good objection against pluralist social egalitarianism, including accounts of its impersonal value, to argue that these positions involve perfectionist claims about the individual or social good.29 To be sure, claims about the personal good based, for example, on such far-​ reaching visions of community as Cohen’s, would be perfectionist if they were put forward as part of a conception of justice whose aim is to regulate society, and with whose demands individuals can therefore be made to comply. Whatever the personal good of community based on shared modes of transport, liberals will baulk at the suggestion that people ought to be made to enjoy it for its own sake, at the expense of other goods they might personally value much more highly (such as comfort).30 On a liberal view of justice, these are not the kind of claims that we can base requirements to structure society on, including structuring basic, and coercive institutions, requirements for education, and requirements to bring about the right kinds of social norms (4.5 in c­ hapter  4). In particular, basing such far-​reaching requirements (also) on impersonal goods, would, from such a liberal point of view, be a rather clear-​cut case of oppression. It would be a case of requiring people to maintain and support a certain set of societal arrangements (also) for reasons that are not connected to their personal good, and thus go beyond

27 At least if the carer is not inescapably locked into this relationship, and disadvantaged because of it; see Kittay 1996. 28 To me, this does, contra Tomlin (fn. 25), seem less mysterious than supposing that such a spectator would feel similar approval for a mere pattern of distributive equality among people, who are perhaps entirely unrelated, without any connection to the question of the goodness, or justice, of its causes (such as a scheme of social cooperation). However, that is not an argument; those with telic egalitarian intuitions will suppose the spectator to have such positive feelings. Note also that, if objectionable mysteriousness is present already when something is required even if is not specifcally good for anyone, then the expressive perspective argued for in ­chapter 2 is also mysterious. What is mysterious is not (or not only) the deontic requirement to respect the freedom and equality of social cooperators, but the claim that certain ways of failing to fulfl claims of justice are in themselves more unjust than others even if the same claims are denied to the same extent, and at the same cost. 29 Contra Schemmel 2015a, pp. 163f. 30 And Cohen might well, too. Tis is not to deny that shared modes of transport may be instrumentally valuable for generating a sense of common interests, which is useful for a functioning sense of justice (see fn.15 in this chapter), and that this may constitute a pro tanto reason for incentivizing their use, and disincentivizing other modes, on top of environmental and efciency reasons.

142  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

the role of liberal justice to fairly reconcile diferent claims based on their own conceptions of it. However, that need not worry pluralist social egalitarians very much; at least not the ones who stress that social justice and social equality overlap, but that the latter incorporates surplus requirements outside the former. Tey could agree that considerations of the good of social equality are in themselves proper bases for these kinds of requirements only insofar as they can be accounted for by justice, and that justice is liberal and should avoid perfectionism—​including the injunction that impersonal considerations can never be such bases.31 Tis would be a restricted pluralist position, which holds that, should justice and social equality ever confict, the latter cannot outweigh the former—​but it can outweigh the demands of other social values, such as efciency. Any such surplus considerations of social equality could then mainly ground requirements for individual actions, including attempts at persuasion of non-​egalitarians. And, insofar as justice leaves space for collective, democratic self-​determination in the pursuit of projects which are optional from its point of view—​which it should (7.3.1 in ­chapter 7)—​pluralist social egalitarians could also seek to feed their views into such democratic processes, in order to obtain more social equality, consistent with justice, in fair competition with those with less socially egalitarian views. Tis kind of modesty is difcult to sustain only for those pluralist social egalitarians who conceive of the claims of justice and of social equality as entirely distinct, and susceptible to signifcant confict in their implications, such as Miller. But that was, as seen, the least convincing account of the distinction between the two values, to start with.32 31 Hausman and Waldren 2011 argue that, whatever the distinction, if any, between justice and social equality, it is a criterion of adequacy for conceptions of social equality that they can deliver principled arguments as to how much their requirements matter vis à vis other values (p. 571), and that anchoring them in fundamental interests such as those based on the Rawlsian moral powers is a promising strategy (p. 584). Tey should thus welcome the liberal conception of relational equality. 32 In fact, Miller’s view about the relationship between distributive justice and social equality is somewhat unclear, and does not uniquely support the ‘confict’ interpretation. On the one hand, he stresses their distinctness, and that social equality has its own distributive requirements (Miller 1998, p. 33): those needed to discourage the emergence of hierarchy. Tis speaks for there being conficts. On the other, he also holds that ‘social equality may help shape other principles of distributive justice which are not internally egalitarian’ (ibid., p. 34). Tis is unclear, and could be interpreted in two ways. Either distributive justice can itself accommodate such shapings. Tis interpretation fts well with the quote, but in this case it is not clear why justice and social equality should, as such, be distinct values. It rather seems that any confict would have to be between conceptions of justice which accommodate social egalitarian rationales and those which do not; and Miller argues that we should opt for the former (this would then, on his understanding of social equality, imply that justice mandates the pursuit of impersonal values, which will trouble liberals). Or, we can fulfl the demands of distributive justice, and, once that is done, go on to shape the distributions in play according to social equality. Ten the two may be distinct, but there is no confict; we have two sets of requirements

Relational Equality beyond Non-Domination  143

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

5.3.4.  Lessons from Pluralist Social Egalitarianism So, to wrap up, there are two conclusions we should draw from this discussion of pluralist social egalitarianism. Te frst is that there is indeed a strong case for extending the requirements of relational equality beyond non-​domination. It is plausible not only that some special relational goods, such as self-​respect, and their connection to egalitarian relations, need to be accounted for. It is also plausible that cases such as the one put forward by Mason should trouble us, and should do so in themselves, and not only when, and because, they bring about further disadvantages for those treated as inferior (which they regularly do). Tis is so, at the very least, because cases of treating others as inferior based on social markers such as ethnic group membership, as in Mason’s example, link up closely to collective, widespread, social norms propagating stereotypes, inviting discrimination, and meting out unequal social status, which we clearly do recognize as problematic, in our societies. It would be disappointing for social egalitarians to be able to object to them only because of their efects. To this extent, this thus confrms the power and attraction of pluralist social egalitarian views, because they can naturally, and directly, account for them. Te second conclusion is, however, that it would also be disappointing if social egalitarian objections to such cases could only generate the very restricted mandate for societal action that such pluralist views yield—​once we hold a commitment to liberal justice constant. Terefore, while this is not, strictly speaking, an objection against such views, we can say that the more such social egalitarian rationales we can bring under liberal justice, the better—​because it opens up such a wider mandate for collectively shaping the right kind of social norms (which we already possess for social norms of non-​domination; 4.5 in c­ hapter 4). So we have a reason to seek to extend the liberal view to account for such cases of treatment as inferior, and the social norms regularly informing, and mandating, such treatments, not only because of their efect in given cases. Tis is what the next chapter will do; it will develop a fuller account of norms of social status and of the diferent ways in which they can be unjust, which can account for problematic cases of the kind singled out by Mason.

on which to base societal arrangements (at least one of which mandates the pursuit of impersonal values, which will, again, trouble liberals, even if it is not called ‘justice’).

144  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

5.4.  Relation-​Sensitive Metrics First, however, we need to turn to the enlarged relation-​sensitive distributive egalitarian views mentioned already in ­chapter 2, and pointed to in the preceding section, which seek to account for relations of equality, of non-​ domination and beyond, because of their distinctive contribution to the individual good, or to opportunities for it. We need to see how far these may plausibly reach, in terms of both the kinds of relational goods they can hope to cover, and the kind of concern for such goods which they can hope to justify: can that be a recognizably egalitarian concern? As seen, the main attraction of such relation-​sensitive metrics is that, if sound, they ofer a relatively simple and straightforward, ‘non-​mysterious’ way to account for egalitarian social relations as demands of justice. Furthermore, as also seen, the more relational egalitarian demands can be accounted for in that way, the less pressing is the need to rely on impersonal requirements of social equality beyond justice. Tere are, broadly speaking, two diferent kinds of attempts to account for egalitarian social relations as goods to be distributed. According to the frst, more general approach, concern for egalitarian social relations can always be translated, without signifcant moral remainder, into concern for egalitarian distributions of corresponding social goods (5.4.1). Tis approach is either merely formal, and therefore uninformative; or it fails both to account for deontic reasons to care about egalitarian relations, and to characterize properly what is specifcally good for individuals about them, where they might not (only) be required for such deontic reasons. Te second kind of attempt is substantive. It identifes specifc relational egalitarian goods, and seeks to justify their inclusion into an egalitarian metric of justice on the basis of the distinctive contributions they make to the individual good. Tis seems more promising, and can be attempted both for overall societal relations of power and status equality, and solidarity (5.4.2), and for more intimate, personal relations of trust and friendship (5.4.3). However, both attempts fail, for diferent reasons. Te frst fails to fully explain what is distinctively good about society-​wide egalitarian relations, and is liable to mischaracterize the wrong of being at the wrong end of inegalitarian ones. Te second leads to proposals which can only demand equality in such personal relational goods at the price of fundamentally disrespecting the individuals at whose good it aims. Tis is too high a price. Te analysis of these attempts thus confrms the soundness of the liberal approach pursued

Relational Equality beyond Non-Domination  145 so far, which is grounded in respect for individuals’ standing as free and equals in social cooperation, not in the contribution that egalitarian relations might make to the individual good, and informs how it has to be extended to matters of social status and self-​respect in the next chapter.

5.4.1.  Translating Relations into Distributions To the frst kind of attempt, then: it would seem that the most straightforward way to account for the demands of relational equality within distributive metrics is to insert an overall social good of relational equality into such a metric. Tis is what Lippert-​Rasmussen has in mind when he proposes a good of equal ‘social standing’:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Consider the good of having social standing that is equal to that of another. Mary and John have equal amounts of this good, ceteris paribus, if, and only if, they relate to one another as equals. Mary has more of this good than John has, ceteris paribus, if, and only if, John relates to Mary as a superior individual to an inferior individual, and vice versa.33

Tis good of social standing could also be split up into diferent sub-​goods, of power, of status, and so on, which, together, account for relational equality overall. According to this proposal, relating to each as other as equals can thus be translated into having equal amounts of corresponding social goods, to which each relevant social relation makes a specifc contribution—​and this is supposed to account for relational egalitarian concerns: I challenge anyone to explain the morally relevant diference between rejecting hierarchical social relations [ . . . ] and an unequal distribution of whatever good hierarchical social relations consist of.34

If this approach is sound, then whatever conception of egalitarian relations social egalitarians might come up with will be able to be spelled out in distributive terms. Distributive equality then subsumes relational equality, 33 Lippert-​Rasmussen 2015, p. 195. 34 Lippert-​Rasmussen, 2018b, p. 200. See also Lippert-​Rasmussen 2018a, p. 94, which adds: ‘[T]‌o be at the lower end of social hierarchy simply is to have less social standing than others have’ (emphasis in the original).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

146  Liberal Relational Egalitarianism and presumed disputes between distributive egalitarians and relational egalitarians can be reformulated as disputes between diferent kinds of distributive egalitarians about which goods should appear, or be more prominent than others, on our metric of egalitarian justice. Te shortcomings of this approach can be spelled out in two steps. In the frst step, we should distinguish merely formal reductions of relational egalitarian demands to distributive ones, which do not take into account the reasons for caring about the former, from more substantive ones, which do take them into account. It is the second which we should be interested in, for the purposes of this section. Te second step consists of pointing out that, insofar as we might aim for relations of equality because of the goods which they instantiate for individuals, this approach risks severely mischaracterizing these goods. To the frst step, then: insofar as reducing relational to distributive demands is possible whatever the reason for caring about egalitarian relations, such reductions are uninformative. Strictly speaking, this has already been shown in ­chapter 4, with regard to non-​domination. Te liberal approach demands equal non-​domination because permitting domination violates individuals’ standing as equals in social cooperation, and thus fails to express appropriate respect for them. It is a deontic, expressive approach. Tat does not rule out that its demands can be formulated in distributive terms, as claims to equal amounts of non-​domination, to diferent thresholds depending on the subject matter—​indeed, sometimes, depending on practical purposes, it might be a good idea to formulate its demands in this way (4.6 in ­chapter 4). And there is, per se, nothing wrong with describing what people have a right to, on this approach, as the ‘good’ of non-​domination, even if the reason for which this good is demanded is that it honours people’s status as free and equal, not that it is good for them in any other way. However, if this turns liberal non-​domination into a distributive egalitarian view, then distributive egalitarianism becomes uninterestingly capacious as a category for theories of justice, precisely because the views now falling under it might give completely diferent types of reasons for their demands.35 As seen, what we should instead be interested in here is whether relations of equality might be required because of the contribution they make to the individual good. Tat is the question we arrived at the end of ­chapter 2, 35 Lippert-​Rasmussen recognizes deontic rationales for relational equality, and indeed holds that these might capture best ‘what is intuitively attractive about relational egalitarianism’; 2018b, p. 174. Tis speaks for this attempt at reduction aiming to be a formal one.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational Equality beyond Non-Domination  147 because this is typically the case for the kinds of goods that fgure on egalitarian metrics of justice, be they Rawlsian primary goods, resources, capabilities, or opportunities for welfare. And it is also that question which we need to tackle in order to see whether relational egalitarianism can, and ought to be, extended beyond non-​domination, and, if so, how. Such a contribution to the individual good need not take the form of relations of equality being intrinsically good for individuals. Tat is not the case for all metrics. Nobody thinks, for example, that resources are intrinsically good for people. However, we should ask whether the goods in question make a distinctive contribution—​one that is not, or not easily, substitutable by other goods. Tis kind of alternative explanation of relational egalitarian demands cannot be given for liberal non-​domination. It is because it honours people’s status as free and equal in social cooperation that protection against domination ought to enjoy priority over other concerns (­chapter 4, 4.3.3), and is itself constrained by considerations of respect (4.3.2); and that kinds of domination which display special qualities, such as being based on de facto authority, and public (4.2.1), are more unjust than others lacking these qualities, even if the domination in question is otherwise of equal extent, and might have identical consequences for individuals’ lives, in terms of other goods. Domination with such special qualities might have a bigger overall impact on individuals’ good then domination without—​or it might not. Tis clears the deck for the second step of spelling out the shortcomings of the general reductive approach. Even if we hold, as other kinds of social egalitarians do, that the reason for caring about (some) relations of equality is that they constitute, or distinctively contribute to, the individual good, Lippert-​Rasmussen’s kind of reductive approach risks missing exactly what that distinctive contribution might be. Let us illustrate this with the help of a diferent good which may be part of ‘social standing’, recognition. Surely, receiving recognition is, other things being equal, good for people. In this vein, Lippert-​Rasmussen suggests that concern about the good of recognition relationships is ultimately concern about equal distributions of recognition: While recognition may be a feature of social relations between individuals, some individuals may enjoy more recognition than others and, for that reason, [distributive] egalitarians could sensibly object to unequal distributions of recognition [ . . . ].36 36 Lippert-​Rasmussen 2015, p. 195. Te original quotation focuses on luck egalitarians, specifcally, not all distributive egalitarians. Te former could object, at least, to unequal distributions of

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

148  Liberal Relational Egalitarianism Does this capture what might be distinctive, and distinctively valuable, about recognition relationships? To fnd out, we need to inquire more into the quality of the relations in which recognition is displayed, or not—​as seen, for example, in Mason’s case of a failure to show ‘recognition respect’ discussed in the preceding section.37 Having equal amounts of recognition is compatible with all parties having very little of it. Tat may be the case also in two-​person cases such as the one of John and Mary quoted earlier. If John and Mary equally despise each other, and deny each other recognition on the basis of their prejudices about each other, then, from the point of view of distributive equality of recognition, they would, other things being equal, seem to be doing fne. Of course, all things considered, distributive egalitarians will prefer people to have equal amounts of recognition at a higher rather than lower absolute level; but that consideration has to be captured with the help of a diferent principle (such as ‘other things being equal, maximize recognition’). Similarly, if Boris is either deeply admired by people, or utterly despised, while Lydia is engaged exclusively in relations of mutual esteem, it might be the case that, in one straightforward sense, they have equal amounts of recognition overall. Many social egalitarians will insist that this picture does not capture the distinctive good at stake in recognition relationships. From the point of view of the quality of relations, Boris and Lydia are engaged in diferent types of such relations, while John and Mary would be merely treating each other as inferior in a way that balances out overall. Social egalitarians might well accept that this is better than having only one party despise the other;38 and they will agree that being admired is, other things being equal, good for people (at least if it is for good reasons). But John and Mary might not only not enjoy any overall recognition, they also have none of the distinctive good that an egalitarian relation of recognition instantiates; and neither does Boris, while Lydia does:

recognition ‘that do not refect diferential exercise of responsibility’; ibid. Te next subsection will focus on a conception of luck egalitarianism which makes a case for relational equality even in the face of diferential such individual responsibility. 37 Recognition of the kind which Lippert-​Rasmussen has in in mind must be a personal good, while I have suggested earlier that Mason’s ‘recognition respect’ (also) instantiates impersonal value (for example, even if the recipient is entirely unaware of it, and will remain so); that diference is of no relevance here. 38 Lippert-​Rasmussen 2018b, p. 199.

Relational Equality beyond Non-Domination  149

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

An egalitarian relationship is one in which the parties have certain attitudes, motives, and dispositions with respect to one another. [ . . . ] And the point is not that these attitudes, motives and dispositions must be distributed equally between the parties. Admittedly, the relationship will not have an egalitarian character if one of the parties exhibits the relevant attitudes and dispositions and the other does not. Te attitudes and dispositions must hold reciprocally. But neither will the relationship have an egalitarian character if the parties possess those attitudes and dispositions to an equal but low degree.39

Tis characterization of what is distinctive about relationships of equality is general; it could apply to all kinds of potentially valuable relations, between fellow citizens, within and across diferent social groups, and also in more intimate settings, such as in friendships. One might seek to resist this analysis by restricting one’s focus strictly to the behaviour displayed by parties to a given social relation, and hold that attitudes and dispositions, while perhaps valuable in their own right, are just a diferent matter. John and Mary might, for example, be recognizing each other in this sense simply by not obstructing each other’s pursuit of their projects, or by furthering these projects in some way, but without doing so out of recognition of their worth for the other, but simply out of self-​interested, strategic reasons. Lippert-​Rasmussen proposes an analogous example of a peaceful relation: ‘Te criteria for relations being peaceful do not include anything about the dispositions of the involved parties’.40 Social relations are peaceful simply when they are characterized by the absence of attacks. However, that just amounts to cutting out what is distinctive about the relations in question, and focusing on a diferent good: absence of attacks. For that good, it does seem plausible to hold that it ought to be distributed equally, other things being equal, as a matter of justice—​but it could be realized in all kinds of ways, including making sure that people do not have much to do with each other. Against that, a relation in which the parties actually have peaceful attitudes towards each other—​they recognize the value of peace, also for the other party, and act, or refrain to act, because of this recognition—​may very well have distinctive value for the parties (and perhaps also impersonally; 5.3.3). Te specifc good in question could be that

39 Schefer 2015, p. 31.

40 Lippert-​Rasmussen 2018b, p. 204.

150  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

of being at peace with one’s fellows, in addition to enjoying freedom from attack—​perhaps because it makes individuals feel at home in their social world, as part of a community. Similarly, the good of Cohenite community mentioned in the previous section, according to which members of community should share each other’s fate, is not yet fulflled when material circumstances are such that, as a matter of fact, they tend to lead lives characterized by the same kinds of challenges, but rather when they know this to be the case, are disposed to reciprocally afrm it in social interactions, and build on that common recognition within them. For these kinds of goods, one thus has to accept the constitutive role of attitudes. Te relevant attitudes have to be present (relatively) equally among parties to the relevant relation, and present to the right degree, and only relations that exhibit both features make contributions to these goods.41 So, accounting for these goods within distributive metrics has to involve looking at, and measuring, these attitudes. For people to have overall equality in such distinctively relational egalitarian goods, they have to enjoy an equal number of such attitude-​rich relations, or equal opportunity to attain them; or sets of such relations, or opportunities for them, which are overall of equal signifcance for their lives. 41 One might be tempted to simplify matters by requiring only (rough) equality of relevant attitudes for a relation to make a contribution to the corresponding relationship good, such as egalitarian trust—​with only the size of the contribution depending on the absolute level of attitudes displayed. However, by now, it should be clear how this still risks missing the distinct value of the relationship in question. A number of relations that are doing very well on these attitudes plus a number that do very badly might not, taken together, be as good as the same number of relationships with sufcient presence of attitudes in all of them—​even if the overall amount of attitudes present should be exactly the same, and their distribution among diferent persons should also add up to the same, and that distribution should be perfectly equal (because all relations exhibit equality of attitudes). Here is an illustration with four persons, A, B, C, and D, and, consequently, six relations; the numbers indicate the degree to which the relevant attitude is present. A–​B

4–​4

2–​2

A–​C

4–​4

8–​8

A–​D

4–​4

2–​2

B–​C

4–​4

2–​2

B–​D

4–​4

8–​8

C–​D

4–​4

2–​2

Total attitude points

48 (12 each)

48 (12 each)

In both cases, there are 48 attitude points to go around; each individual gets 12; and all individual relations exhibit equality of attitudes. Now assume the sufciency threshold for a relation displaying the required quality—​for example, trust, not mistrust—​is 3–​3. Ten we have to discard all 2–​2 relations.

Relational Equality beyond Non-Domination  151 It could, of course, be that there is a strong case for including some such egalitarian relationship goods on egalitarian metrics of distributive justice. But in order to see whether there is, we need to look beyond more formal approaches, at substantive views which deliver an analysis of specifc relationships, of which contribution exactly to the individual good they might make, and of what role, if any, attitudes play for them—​and argue for an egalitarian distribution of corresponding goods on that basis. Tis is what the next two proposals do.

5.4.2.  Luck Egalitarianism and the Individual Good of Overall Social Equality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te frst proposal takes up the challenge of showing how society-​wide relations of equality make an important contribution to the individual good; and one that luck egalitarianism, too, should be sensitive to. It thus aims to account for the goods instantiated by these without resorting to the impersonal value approach discussed in 5.3.3. Tis view is Anca Gheaus’s: [ . . . ] DRGs [democratic relational goods] can plausibly be seen to contribute signifcantly to individuals’ resources, welfare, advantage or capabilities. Equality of status, power and solidarity are intrinsically good, and it is unlikely that they can be substituted by, for instance, material resources. If living in a society shaped by the ideal of democratic equality is, other things being equal, signifcantly better for each individual than living in a society that lacks democratic equality, then DRGs are plausible distribuenda of justice.42

As the quotation makes clear, the goods in question are not power and status simpliciter, but equality of power of status, and prevailing egalitarian relations of solidarity. Tis is an ingenuous proposal, for two reasons. Te frst is that this view does not demand that all individuals have particularly fulflling, or intense, attitude-​rich egalitarian relations with all of their co-​citizens, throughout, or with a high number of them (whatever that number is). It only demands that societal life, overall, is dominantly characterized by such relations. Tis might be the case even where most people have no relations

42 Gheaus 2018, p. 59 (italics in the original).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

152  Liberal Relational Egalitarianism with co-​citizens worth speaking of, and no attitudes towards them, as long as the ones which they do have possess that egalitarian character. We have seen in the preceding section that demanding positive relations of community as a matter of distributive justice is perfectionist; to this extent, Gheaus’s view might be able to circumvent perfectionism.43 Te second reason is that this approach, if sound, gives luck egalitarians strong internal reasons to care about such overall relations of equality, too. A  problem for luck egalitarianism is that, insofar as it requires that individuals bear the consequences of their responsible choices, sometimes such consequences may consist of exposing them to domination—​for example, where forfeiture of claims to assistance based on responsibility for one’s situation puts one at the mercy of others in the job market, or leads to losing the resources one needs to participate as an equal in politics. Or such consequences could consist of dramatic losses of status and recognition, which social egalitarians will fnd intolerable, even if they agree that it is acceptable, or perhaps even required, that the individuals in question lose some economic resources due to their imprudence.44 On a luck egalitarian view, compensation of the irresponsible cannot be demanded from others because that would be unfair to them. Now, the virtue of Gheaus’s approach—​if sound—​is that it shows how, in case of non-​compensation and ensuing relational risks, these others also lose out, for reasons beyond their control, in terms of their own enjoyment of the good of collective equality of power and status, and solidarity. So, in order to keep them from losing these goods, compensation of the irresponsible is called for. Tere is thus a countervailing reason within luck egalitarianism itself to ward of relational inequality, a reason that is not excluded by the requirement not to unfairly privilege the irresponsible, because that is not the reason for which relational risks are to be forestalled. However, the view runs up against two serious objections. Te frst is that this account of the supposed individual good of relational equality really is objectionably mysterious (5.3.3); the second is that, even if it were sound, it would still fail to do justice to the specifc claims of those on the wrong end of inegalitarian relations. 43 Among the relations Gheaus mentions, those of solidarity plausibly require reciprocal positive attitudes; she does not specify whether, on her view, such attitudes have to be part of egalitarian relations of power and status, too. 44 Objections of this kind, raised especially by Anderson (1999, pp. 295f.) and Schefer (2005, p. 15), fueled much of the initial stage of the debate between relational and distributive egalitarians, and are ofen summarized as the ‘harshness objection’; see Voigt 2007.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational Equality beyond Non-Domination  153 As to the frst objection: the view centres on the value of collective relations of equality of power and status, and solidarity, for individuals, but it remains unclear what is intrinsically, or otherwise distinctively, good about these relations, especially for those who would otherwise be on their upper end. To incorporate the collective goods in question into a metric of justice whose demands we ought to realize in society, it not only has to be shown that such relations have some unique goods in store also for those who would otherwise enjoy more power and status. Tat they may well have; for example, the social norms sustaining hierarchical relations can lock the privileged into a position of superiority. Tis not only hampers opportunities for easy-​going intercourse with those classed below them, but arguably also requires the privileged to behave in ways that are in fact recognized as superior by those around them, according to the demands of the norms in question45—​which can be a source of intense stress. However, this is not enough. It would also have to be shown that such goods are goods that everybody, on balance, ought to want, and ought, if not to assign priority over, than at least accept to trade of, to some extent, against other goods on the metric, such as economic resources, or other opportunities for self-​advancement—​as opposed to having the opportunity to forgo some of the latter goods in favour of egalitarian relations in their own lives, if this is what they wish for. Tat opportunity is already guaranteed by liberal views. What is more, strictly speaking, Gheaus’s view requires showing even more than that, because it is the fact that society is characterized by such egalitarian relations that is supposed to be intrinsically good for everybody, not (just) the egalitarian relations in which one is oneself involved. It may well be that a society shaped so as to instantiate ‘democratic relational goods’ scores, on the whole, much higher in terms of a host of other goods that all individuals have good reason to want, and whose presence on a metric of justice is therefore reasonably uncontroversial, such as enhanced security against crime, and a lower risk of incurring physical and mental health problems.46 But this only concerns the instrumental benefts of equality, and is diferent from claiming that equality of power and status intrinsically (or otherwise distinctively) makes everybody’s life better. Te

45 Fourie 2012, pp. 119f. 46 For some evidence, see Wilkinson and Pickett 2010. Wilkinson and Pickett do not distinguish very sharply between distributive equality of income and wealth and social equality, because they hold them to be not only empirically tightly correlated, but also mutually causal.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

154  Liberal Relational Egalitarianism view thus invites the suspicion that it simply translates claims about the impersonal good of egalitarian relationships into claims about the personal good. It does not justify the latter. At best, it is incomplete, and needs further development. Perhaps, for example, there is a specifc, and particularly desirable, kind of self-​respect which the overall presence of societal equality enables, or facilitates—​and its connection to social equality is close enough to efectively disarm the kind of mere instrumentality objection just made.47 We will see in the next chapter, however, that concern for self-​respect does, at least on a liberal conception of its role, not justify a blanket requirement of overall social equality. Te second problem is that, even if a case for the intrinsic goodness for individuals of societal equality of power and status can be made, this would still fail to account for the special harms sufered by those deprived of power and status, and their consequently special, and stringent, remedial claims. As a stand-​alone view of the goods in question, it would imply that both dominators, and entirely uninvolved third parties (insofar as they are members of the same society, whatever that is), have exactly the same kind of claim to the removal of domination as the dominated.48 Tis seems absurd. Being dominated is a grave ofence against one’s equal standing, which grounds a claim to removal of a certain priority (4.3.3 in c­ hapter 4). Even if something is to be said about the intrinsic goodness for individuals of a collective state of equal non-​domination, its absence is less troublesome, from the point of view of a dominated individual, than being dominated herself.49 So Gheaus’s view does not hold much promise for social egalitarians. Te general implication of this analysis is that any kind of view that seeks 47 Another, related possibility is that only societies characterized by prevailing relations of equality can generate a general sense of safety, or trust, in social interactions, and that this sense of safety and trust is valuable in its own right, on top of the good efects of equality on specifc other goods such as lack of crime, or good health. Te contribution that social equality would then make would still be instrumental—​to such a sense of safety—​but it would be a distinctive contribution to this distinctive good. I thank Lisa Fuller for this suggestion. It would need to be backed up by empirical research showing that hierarchical societies reliably fail to generate such a sense of safety and trust, due to their hierarchical nature. 48 Equal non-​domination is not the same as equality of power (­chapter 3, 3.2), but that is of no concern here, as Gheaus includes both. 49 One might seek to object that the question of the stringency of claims of justice just is a diferent matter; on Gheaus’s view, justice need not give rise to action-​guiding claims at all; see Gheaus 2013. However, this is not relevant for the point that being dominated is worse than not enjoying a societal state of non-​domination. Perhaps grave instances of domination elsewhere in society detract more from one’s individual good than being personally very slightly dominated over a much less important matter; but extent and intensity of domination should be held constant.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational Equality beyond Non-Domination  155 to incorporate collective egalitarian relations into metrics of the individual good will encounter its own version of each of these two objections.50 In conclusion, it is worth pointing out that, even if Gheaus’s conception of the value of egalitarian relations were sound, it would, within luck egalitarianism, only lead to a requirement to balance (2.4 in c­ hapter 2) reasons against compensating, and safeguarding, those who forfeit claims of fairness because of the exercise of individual choice against reasons for maintaining societal equality of power and status. Te view would thus pull into both dimensions at the same time. Whether the resulting balance would be close to a recognizably social egalitarian view is an open question. However, that would, by itself, not be a knock-​down objection. Relational egalitarian views also have to deal with considerations of individual responsibility, and it is possible that, on this issue, there is more convergence between luck egalitarians and relational egalitarians than the sometimes overly polemical focus of the debate suggests.51 On the one hand, we can say that the liberal relational egalitarian view is itself, at least in part, responsibility-​ and choice-​sensitive, because it leaves it open to individuals to pursue personal projects which, once entered, ofer only a lower threshold of protection against domination (4.3.1 in c­ hapter 4). To be sure, it demands, in these cases, that there is a tight connection between the nature of the project (intimacy, or extension of choice sets), and the lower threshold, that an array of background protections remains in place, and that individuals have exit options. So it is still a far cry from more unrestricted luck egalitarian positions. However, on the other hand, recent work on responsibility-​sensitivity stresses that, while luck egalitarianism must demand that individuals be able to infuence their (distributive) fate through individual choice, it is a diferent question how severe the consequences may be.52 An answer to it cannot be deduced from

50 A diferent kind of attempted reconciliation of relational and distributive views construes the latter as seeking to determine when a relation is just—​namely, when the parties’ distributive entitlements are met, so that there is no unjustifed distributive inequality—​and the former as explicating what a just relation consists of—​namely, of relating to each other on terms all could accept, or none could reject, precisely because such entitlements are met; see Quong 2018, p. 17. Realizing that one stands in such a relation, and being disposed to afrm it, might then be particularly valuable for individuals; Moles and Parr 2018, p. 7. On these views, relational egalitarian considerations thus do not inform the content of social justice. Tey cannot capture the distinct injustice of domination and status hierarchy. 51 See Lippert-​Rasmussen 2018a, p.  101, who rightly points out that relational egalitarians do not typically maintain that responsibility for choice plays no role whatsoever for justice, and also 2018b, ch. 8, for an exploration of some substantive overlaps between important luck and relational egalitarians positions. 52 Stemplowska 2015.

156  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

the luck egalitarian master principle of responsibility-​sensitivity; it requires an additional ‘principle of stakes’ determining the range of these possible consequences.53 So, even leaving Gheaus’s attempt aside, luck egalitarians could insist that the payof of bad choices must not include relational inequality, or at least restrict the risk of the latter very signifcantly. Tis is not to say that a happy marriage between luck egalitarianism and relational egalitarianism is guaranteed. Even if a large extent of substantive rapprochement can be achieved, luck egalitarians will insist that relational equality merely injects part of the substance that their framework requires, while relational egalitarians will regard their framework as indicating where considerations of individual choice and responsibility can enter, in the frst place. Having a fully worked-​ out liberal conception of relational equality at hand will make future comparison and confrontation more productive—​as would having a more fully developed view of what the ‘stakes’ of responsible choice might be, and how one could go about fxing them within a luck egalitarian framework. Te more determinate, and important, conclusion we can draw is that Gheaus’s proposal fails to capture the central relational egalitarian concerns which have been addressed in previous chapters, and would need a more convincing account of the distinctive value of a collective state of relational equality for individuals. For relations of power and domination, a deontic, expressive view does much better. It remains to be seen how it can apply to matters of status and self-​respect.

5.4.3. Relations of Trust and Friendship as Liberal Primary Goods Perhaps a better case can be made for the distinctive value for individuals of positive egalitarian relations of a more personal sort, and perhaps such a case can be appropriately connected to wider social and political relations, and can be upheld without inappropriately mandating the pursuit of such relations as intrinsically good. Tis is what Chiara Cordelli sets out to do. She proposes to incorporate concern for personal relationships of trust and friendship into a liberal metric of Rawlsian primary goods.54 Tis metric has

53 Olsaretti 2009, p. 167. 54 Cordelli 2015a and b.

Relational Equality beyond Non-Domination  157 to be extended to such ‘relational resources’, insofar as being able to draw on (some) trust and friendship is, for most people, an important precondition for success in developing and pursuing their conceptions of the good—​ whatever these may ultimately be about. Terefore, she proposes a Rawlsian principle of fair equality of opportunity for such relational resources. All institutions and norms which have a bearing on people’s access to these relations are thus within the purview of justice; they constitute the ‘relational distributive structure’.55 Relations of non-​domination come out, on this view, as an important enabling factor for developing the ‘skills’ needed to enter into, and maintain, relations of trust and friendship:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

[L]‌ack of relational talents that is socially produced includes the lack of self-​confdence and social skills due to growing up within authoritarian, abusive, or segregated relations, whether within the family, neighborhoods, schools, or civil society associations. In this respect, fair equality of relational opportunity requires, as its own precondition, relational equality understood as the absence of social relations of domination, segregation, and hierarchy.56

However, the view is not restricted to merely demanding the absence of such relations because of their negative impact on relational skills and talents, but also requires tackling other sources of deprivation of these skills, as well as their positive cultivation, and real opportunities for their exercise. Tis is a promising proposal; its claim to liberal credentials is plausible insofar as it explicitly abstains from putting forward the case for relations of trust and friendship in virtue of their intrinsic value. Whatever individuals reasonably regard as of such intrinsic value for their own life, for most, if not all, such relations will play at least some role in upholding that conception of the good. However, the view encounters the following problem. Either it does not go far enough to constitute a recognizably egalitarian conception of relational opportunities, thus deserting the value of equality. Or, if it goes further, it ends up fundamentally disrespecting the individuals about whose relational disadvantage it is concerned, and is therefore in danger undermining

55 Cordelli 2015a, pp. 89, 101. 56 Cordelli 2015a, p. 107. General relations of community and civic friendship may thus enter insofar as they facilitate individual friendship and relations of trust.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

158  Liberal Relational Egalitarianism individuals’ basic moral equality (­chapter 4, 4.3.2), based on possession of the two Rawlsian moral powers. Te objection is not that the view will require of some individuals what cannot, or must not, be demanded of them, as a matter of justice: that they enter into relations of trust and friendship with those lacking these goods, even if they are unwilling, or even if they have good reason not to do so—​let us say, because the so disadvantaged really would not make good friends, or really cannot be trusted, for whatever reason (as long as the unwilling are not themselves responsible for this defect).57 Perhaps this objection can be overcome. Cordelli is sensitive to it and proposes to focus on ‘shaping patterns of socialisation in a way that (roughly) equalises, ex ante [ . . . ] opportunities’58 to attain the goods in question. Let us grant, for the sake of argument, that it is indeed possible to sustain such a requirement while avoiding both that (a) its demands fall on particular individuals in a problematic way, and (b) it remains too shapeless and untargeted to be able to count as a demand of justice, on a standard account of justice yielding requirements for action, including coercive action (which Cordelli wants to maintain).59 An important objection still remains, which focuses entirely on the recipient side of the requirement. Cordelli is alive to problems on this side, too. She notes that collecting the relevant information about people’s relational skills and opportunities may lead to ‘excessive intrusion’ into personal lives.60 Because of this, as well as because of the dangers of making inappropriate demands on those on the supply side noted earlier, she proposes both to require only opportunities for relational resources, not their guaranteed provision, and to interpret fair equality of opportunity as merely ‘requiring an equal opportunity to access an overall adequate level of each relational resource.’61 So the requirement of fair opportunity is already scaled down considerably. Now, the objection is not that this surviving requirement cannot count as egalitarian in any signifcant way, and is better described as a sufciency requirement. Let us again grant that it could count as signifcantly egalitarian,

57 For this line of objection, see Valentini 2016, pp. 60f. 58 Cordelli 2015b, p. 692. 59 Cordelli’s aim is ‘a political theory of distributive justice’; Cordelli 2015a, p. 88, fn. 6. I assume this is (part of) what she means by ‘political’. 60 Ibid., p. 99. She also argues that it may be impossible; ibid. Tis is less convincing. Psychological tests ascertaining people’s relational situation, including their ‘relational skills’, may well be available; or, if they should not yet be, it is not clear why developing them should not be realistically possible. 61 Ibid., p. 108 (emphasis in the original).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational Equality beyond Non-Domination  159 could it be upheld. Te objection is that it cannot be upheld because there are additional reasons, stronger than those focusing on mere ‘intrusiveness’, which militate against it—​and do so precisely in some cases where it might seem most needed. Tese are reasons of basic respect: being serious about seeking to guarantee equal opportunity for relations of trust and friendship would not merely require intrusion into personal lives, in order to ascertain people’s overall relational situation, which liberals will fnd problematic. It would, other things being equal, also seem to require assessing the level of relational skills that the relationally disadvantaged can muster: ofen, such disadvantage will be connected to such lack of skill. But this would be fundamentally disrespectful, because it would amount to assessing moral competence. Friendship and trust come with moral requirements, of, inter alia, reliability, being able to understand the particular interests of one’s friends, and others whose trust one seeks, and to act on these—​and, to that extent, of being able to abstain from the pursuit of mere self-​interest, and to overcome weakness of will and indiference. All these capacities are intricately linked to persons’ two moral powers, of a sense of justice, and of a capacity to develop, and pursue, a conception of the good. Accordingly, a lack of capacities needed for proper trust and friendship will, ofen at least, be linked to defciencies in those moral powers. However, since individuals are fundamentally entitled to Carterian ‘opacity respect’ (4.3.2 in ­chapter 4), agents of justice, such as those upholding, and being in charge of improving, the ‘relational distributive structure’, have to abstain from inquiry into their level of possession of the moral powers—​as long as they cross a minimum threshold. And since these powers are the basis for individuals’ standing as equals, violations of such respect would, as argued in the preceding chapter, endanger this standing (although this is not the reason why it is required—​ that would, as seen, be circular). It is hard to see how such opacity respect can be maintained while seeking to equalize opportunities for intimate relationships with signifcant moral dimensions, such as trust and friendship. And it is also signifcantly harder to see how that can be done than it is in the case of equalizing protection against domination. While some possible such protections might involve such violations—​such as those that would require comprehensive assessments of intelligence, gullibility, and strength of will, to try to equip people against possible manipulators—​protection may also be achieved in other, more impersonal ways (though perhaps not at maximum level). Tis is not possible to the same extent for friendship and personal trust, because of their intimate

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

160  Liberal Relational Egalitarianism and personal nature.62 For example, those whose susceptibility to manipulation gives rise to heightened risks of domination will beneft from general, non-​targeted information about what are sure signs of manipulative relations, and advice on how to avoid them. But avoiding relations is not a way to fnd friends, or trust. Accordingly, if we retreat to analogous impersonal, and anonymous, measures to improve people’s friendship and trust opportunities—​as we should, in the name of opacity respect—​this will leave many personal defciencies and problems standing in the way of friendship and trust untouched. Accordingly, the surviving requirements of the ‘relational resources’ view are even less egalitarian than Cordelli maintains.63 Tis is not to say that these resulting requirements should be rejected. Provided that the objections from the ‘supply side’ mentioned earlier can be met, we should endorse them; but those looking for relational equality will end up disappointed by them. And while it is true that domination and social hierarchy generally impede opportunities for trust and friendship,64 and that this is a further reason for objecting to them, unequal protection against domination is unjust even where it does not obstruct these opportunities—​ simply because it expresses disrespect for people’s equal standing in social cooperation. Te general implication of this analysis is that any view which is similar to Cordelli’s view, insofar as it aims at (opportunities for) personally signifcant relationships, with their own moral requirements, because of their contribution to the individual good, will encounter a version of this problem. It will fnd it difcult to navigate between respect and equality. However, as Cordelli’s analysis has shown, it is these relationships—​as opposed to wider, more society-​wide ones—​which will enter most directly, either intrinsically, 62 It was lef open in the preceding chapter whether opacity respect is violated as soon as any partial assessment of capacities that will have some impact on overall competence is made—​as Carter maintains (see ch. 4, fn. 25)—​or whether only those assessments are ruled out which allow reasonable inferences to the exact level of overall competence. Tis question does not need to be resolved here, as assessments of competence for friendship and trustworthiness pull towards the latter alternative, in any case. 63 Tis is not so, however, in the practically very important case of friendship-​and trust-​enhancing education of children, where opacity respect does not yet apply, or at least not fully. 64 Arneson rightly points out that overall inequalities of power and status between parties need not always impede the quality of their friendship; 2010, pp. 42f. However, not only does the presence of such inequalities reduce opportunities to fnd friends across society; friendship requires some rough equality of positive attitudes, and reciprocity. If an aristocrat regarded her higher rank as a justifcation for imposing greater demands on her commoner friend within their friendship—​say, she cannot be expected to keep their appointments, while the commoner can be—​that would impede its quality, even if the commoner accepted these demands.

Relational Equality beyond Non-Domination  161 or distinctively in other ways, into a great variety of individual conceptions of the good, for diferent reasons, and thus have the strongest claim to fgure prominently within a metric of distributive justice.

5.5.  Conclusion

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te objections raised to diferent kinds of relation-​sensitive accounts of egalitarian distributive justice do not conclusively establish that no promising such account can be found. Tis kind of conclusion cannot follow from analysing existing proposals. However, it has clarifed important, and distinct, challenges which each type of incorporation proposal has to overcome. An account which managed to overcome them would certainly not be unwelcome; the case for relational equality could then be overdetermined, or good-​centred and deontic, expressive approaches might cover diferent types of egalitarian relations. In the absence of such an account, we have all the more reason to look for ways of extending the expressive approach to account for the demands of self-​respect, and social status, as matters of justice. So it is to this task that we must now turn.

6 Social Status, Self-​Respect, and Opportunity

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.1.  Introduction Concern about unequal power and domination does not exhaust relational egalitarianism. Concern about treating each other as equals in everyday social interactions, beyond ensuring non-​domination, is another long-​standing demand within the social egalitarian tradition, and we have seen in the last chapter that it is plausible. Terefore, no view that fails to account for what is wrong with failing to treat each other as equals even where domination is not at stake, and to specify when that is so, can count as an adequate theoretical embodiment and elaboration of the political and social concerns that animate social egalitarians. However, we still have to see how this concern can be incorporated into a liberal conception of social justice; and incorporated not only because of its bad, or unjust, efects. Tis chapter proceeds as follows. First, the dimensions in which individuals can objectionably treat each other as unequals need to be spelled out with more precision. Section 6.2 focuses on the dimension of diferential social esteem, and consequent social norms of social status, as this is the dimension that poses the deepest theoretical problems for social egalitarianism. It develops an account of how individual and collective esteem judgements are expressed in social interaction, and how these are corroborated and reproduced so as to give rise to norms of social status. Tis focus on norms is necessary because their collective, self-​reproducing features pose problems for social equality which are not posed by individual judgements, and acts, of (dis-​)esteem, as the course of the argument will show. In order to show that such norms, and individual actions conforming to them, are both a genuine and independent social egalitarian concern, and one that can be brought under justice, it is necessary to show that the injustice of inegalitarian such norms cannot be reduced to their efects, but that the way in which they bring about these efects matters, independently. Tis Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0006

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  163 is similar to how diferent ways for institutions to bring about the same distributive efects matter independently for justice, as shown in c­ hapter 2. Section 6.3 prepares this argument by analysing diferent efects of these norms. Subsection 6.3.1 explains how social status norms can engender domination, deepening the analysis delivered in ­chapter 4, while subsection 6.3.2 focuses on their impact on self-​respect and self-​esteem. It argues that such impact is unjust, and that it is at stake in many of the more pernicious relational inequalities we are familiar with in our societies; however, liberals should be committed to a conception of self-​respect on which unequal social status as such need not threaten self-​respect, and indeed sometimes should not. Subsection 6.3.3 shows that inegalitarian status norms also curtail opportunities, even where they do not engender domination, or are properly regarded as threats to self-​respect. Section 6.4 then argues that such opportunity loss due to status norms can be unjust even where losing the same opportunities due to other causes, such as social processes which are not coordinated on the basis of judgements of esteem, or natural events, is not. Te reason for this is that a denial of signifcant opportunities due to norm-​coordinated negative esteem judgements by others impairs individuals’ standing as free and equal in society in a manner that is absent when they are lost in other ways. Tis yields both a genuinely social egalitarian rationale for combating them, and one that can be brought under justice, where status norm-​induced loss of opportunity is sufciently signifcant. Section 6.5 wraps up the argument by applying the conclusions reached so far to individual inegalitarian actions. If a norm is unjust, all norm-​following actions are unproblematically unjust, because of their meaning, independently of the particular efects they might produce in any given instance. However, the fipside of this argument is that some inegalitarian actions that are not connected to norms in this way are not a problem of justice. Tey may be morally objectionable, but no more, or diferently so, than actions failing to instantiate the full range of other-​regarding moral virtues, whatever these may be. Te argument thus shows that the most important and plausible concerns of the pluralist social egalitarian positions analysed in the preceding chapter, which aim to go beyond justice, can be accounted for by justice, and that the remainder should not exercise us, within political philosophy. Or, if it should, then at least not in any way that is diferent from how all action which is less than morally perfect should exercise us.

164  Liberal Relational Egalitarianism

6.2.  Social Status Norms

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.2.1.  Respect, Esteem, and Status In unearthing, and motivating, the problem of individuals treating each other as unequal even in the absence of inter-​individual power inequality and domination, the previous chapter has used diferent concepts—​of recognition, respect, esteem, and status—​quite freely. Tis was sufcient to motivate the problem, but in order to solve it, these concepts need to be sharpened. Respect and esteem are ofen distinguished in the following way: respect is taken to apply to fundamental moral considerations, and especially considerations of justice, such as people’s basic moral standing, the capacities that ground it—​possession of the two moral powers, on a Rawlsian view—​and consequent rights and duties, especially the ones most tightly linked to these capacities. Tis is how it has been used in previous chapters (­chapter 2, 2.6; ­chapter 4, 4.3.2.). Esteem, against that, is commonly taken to refer to individuals’ appraisal of particular capacities, skills, and traits of others, including not necessarily morally relevant ones, and, consequently typically allows of more gradation than the former.1 ‘Recognition’ is ofen used as an umbrella concept covering both these dimensions.2 Te dimension of esteem is the salient one for this chapter. Failure to treat others with the kinds of respect individuated in previous chapters is unjust.3 Esteem poses, against that, the more difcult problems, as diferential esteeming is a necessary fact of social life, always in play when we value some things—​some projects, traits, products—​more than others, and, therefore, at least to that extent, also their bearers and producers.4 So the question is: how could esteeming or disesteeming something be unjust, if not for mere instrumental reasons, unless it happened to produce results condemned by independent principles of justice? Te best way to start the investigation of this

1 For a nuanced discussion of the difculty of sustaining a distinction between respect and esteem in important practical cases, see Fourie 2015, pp. 96f. 2 See Honneth 1995 and McBride 2013. Similarly, Anderson uses ‘recognition’ for both the dimensions of ‘standing’ (see fn. 5) and esteem; 2012, p. 48. 3 Te only problem here is that of what to think about individuals who fail to regard others with such respect—​who do not have mental states and attitudes of respect—​but nevertheless act towards them as if they did; see 6.2.3 for the problem of hypocrisy in expressions of respect and esteem. 4 Authors discussed in the previous chapter might not agree that the most difcult cases of individuals treating each other as unequal in the absence of domination, or other unjust efects, play out in the dimension of esteem, as it is understood here. I take it that Mason (5.3.3) would not; we will be in a better position to assess this disagreement at the end of this chapter.

Social Status, Self-Respect, and Opportunity  165 question is by putting in place an account of norms of social esteem, focusing on how they build on, and, in turn, reinforce individual judgements of esteem. Tis is so not only because, as a matter of fact, many people’s relevant esteem judgements and consequent actions take their orientation from such norms, much of the time. It is also, as the argument will show, because it is the norms themselves that should be regarded as the primary subject of justice assessment. Te injustice of individual inegalitarian actions, if any, depend on their relation to them (6.5). Tese norms of esteem are norms of social status. Social norms can range over many diferent matters; not all are centrally about according social esteem or disesteem. Similarly, ‘status’ can be used in very many diferent ways, and ‘status inequality’ can denominate virtually any inequality in social position between diferent individuals which social egalitarians may worry about.5 So it is important to be precise about the specifc meaning of ‘social status’, and to outline the possible content, scope, and mechanisms of self-​ reproduction of norms determining it, in some detail before moving on to the substantive argument as to when they are unjust, and why (6.3 and 6.4).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.2.2.  Esteem Judgements and Actions 6.2.2.1. Esteem and Social Class Humans are judging creatures, and, very ofen, how they judge others is heavily infuenced by how third parties, such as their peers, do. Our social status is a function of the esteem, or disesteem, we receive from others, for our identities, personality traits, social appearance, possessions, skills, 5 Anderson distinguishes between three diferent kinds of hierarchies which social egalitarians are typically opposed to: hierarchies of authority, esteem, and standing; Anderson 2012, p. 41. ‘Authority’ mainly refers to socially recognized power, and is covered by requirements of non-​domination—​ which, however, extend to cases of power inequality not mediated by authority, too (4.2.1). ‘Esteem’ occupies us in this chapter. ‘Standing’, against that, is a wider term which refers to ‘the interests of those occupying superior social positions [being] given special weight in the deliberations of others and in the normal [ . . . ] operation of social institutions’, which leads to ‘those of higher rank [enjoying] greater rights, privileges, opportunities, or benefts than their social inferiors’; ibid. p. 43. Tus, for example, diferent institutional attitudes expressed towards individuals, as in the V example of ­chapter 2, make for diferential standing, whether or not diferential authority or esteem of these individuals are at play. Epistemic privilege (see ­chapter 4, 4.5.1, fn. 45), which allows some to feed their own formulation of their interests into deliberation, while others do not have this opportunity, is another case of superior standing. Tese three ranking systems could all easily be formulated as ranging over diferent kinds of ‘status’. For reasons of conceptual hygiene, it might then seem tempting to drop the concept of status altogether, but, in ordinary language, ‘social status’ is ofen used, more distinctively, to refer to the norm-​based command of others’ esteem.

166  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

talents, projects, pursuits, and achievements. In modern societies, talents usable in occupational pursuits, and achievements based on these, for example, in recognized professions, are a particularly important object of others’ esteem judgements. Ofen, when we refer to social status, we primarily, or even exclusively, refer to these. Typically, occupations requiring a set of complex, not easily attainable skills, which are exercised in activities generally regarded to be socially valuable—​such as the traditional professions of physician or judge—​command, other things being equal, higher status than others. However, our conception of social status is purposely as wide as possible, and not restricted to such specifc esteem-​markers. Consequently, it is also not restricted to issues of social class. Class distinctions in contemporary societies with market economies tend to be based on levels of education and type of occupational activity, and ofen include consequent command of economic resources. Tey are a particularly important example of status distinctions, but not the only one. Because of this, it is also not necessary to commit to any particular account of social class, whether largely based on education and occupation or not.6 Only those defnitions on which receipt of social esteem makes no diference to one’s class membership are incompatible with the present analysis. Any possible social esteem-​marker is thus included, and the connection to the question of how exactly to defne social class can be largely lef open. 6.2.2.2. (Dis-​)Esteeming, and Acting Accordingly Esteem judgements have to have some consequences for action in order to lead to diferential social status, but the set of actions which qualify is, again, wide. It is not necessary that those who do not possess the esteem-​markers, or possess the disesteem-​markers, in question meet with a reaction on the part of judgers whose main purpose it is to express a social ranking, such as scorn, shunning, visible disgust, or contemptuous smiles, in the negative case, and explicit praise, perhaps expressed in a comparative manner, in the 6 One prominent such account, to which I am largely sympathetic, is the account of habitus developed in Bourdieu 1984, which focuses on the possession, and continuous enactment, of a set of aesthetic norms expressing esteem for some ways of life over others, whose role is thus to distinguish those who fulfl their demands as superior to those who do not. Tose of superior social class use their greater educational and economic resources (‘capitals’) to enact status norms which, if successful, achieve dominance: those of inferior class struggle to fulfl their demands, and have to defne whatever norms of their own they bring into existence against the background set by the former. Tis hierarchical function, and its mechanism of enactment, need not, however, be fully transparent to, and approved by, these followers, as long as they display the ‘right’ kind of reaction to those judged below, such as disgust, or avoidance; see what follows in the main text.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  167 positive case. Esteem-​guided treatment of others can be a straightforward, and genuine, expression of what the judger regards as important, valuable, or desirable, and to which she hence orients her actions. She thereby gives others incentives to do likewise, in order to attain her esteem, and (inadvertently) penalizes those who do not want to, or cannot, live up to them—​but all this can happen without any intention whatsoever of instantiating, and communicating, a consequent social ranking among people, and the inferiority of those who fall short.7 Tus, according (dis-​)esteem to others, and acting accordingly, precisely because these already occupy a superior (or inferior) social position, or possess characteristics that the judger thinks will enable them to attain (or condemn them to) such a position, is certainly a frequent social phenomenon, and a pernicious one, from the point of view of social egalitarianism. And ofen it may take place because individuals, taking their cue from hierarchically-​minded others, fear that they themselves will lose esteem if they do not profess, and display, similar standards. However, social status is not merely a matter of what we may call ‘snobbish’ tastes and their expression, nor of herd behaviour dictated by fear for one’s own status. It is important to clear away these intuitively most problematic cases to see that the problem of social status strikes deeper. It is enough, in terms of (dis)esteem-​ expressing actions, that those who are judged against a certain esteem standard can reasonably infer that they are, in efect, regarded as inferior, or would be, if the judger gave it any thought: even though everybody at a social dinner is perfectly polite to each other, the host assigns the place of honour to those whom she regards as of highest social achievements, while you happen to get the bottom end of the table (the one closest to the lavatories). Or, to take a diferent, and more consequential, example, think about a society in which middle-​class parents take great care to select the playmates of their children, and discourage them—​perhaps not explicitly, but via their choice and scheduling powers over the activities of their children—​from playing with children from poorer, or less educated, backgrounds.8 It is hard to deny that children from poorer backgrounds (and their parents) are, in such cases, treated in ways that are apt to make clear to them that they are regarded as socially inferior. But it clearly need not be the point of this treatment to express their inferiority, let alone to cement it. Parents may simply think that 7 See Scanlon 2018, p. 35. 8 For an account of how childhood experiences difer across social class in the US, see Lareau 2011.

168  Liberal Relational Egalitarianism playing with children from more well-​placed backgrounds will better enable their children to develop skills that they will need in later life in order to succeed according to prevailing social standards. Tey could, to an extent, even be right about this. And they could act in this way while happily paying all the taxes that justice requires them to pay for beneft of the worse of (see ­chapter 8,  8.3). Our conception of social status and status-​induced actions is thus as wide as possible not only with regard to the objects judged—​(dis-​)esteem markers—​but also with regard to their expression in action; as long as there is some such expression accessible to others. If these judgements, and their expression, track widespread norm of social status, reinforce them in turn, and these norms are in fact unjust, according to the substantive argument yet to be made, then taking on the content—​the meaning—​of these norms and attitudes by erasing or changing them must be the primary remedy.9 Merely trying to block, or rein in, by external means, the actions they normally dictate, or their efects, can only be second-​best.10 To the norm-​generating and -​upholding features of esteem judgements, we now turn.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.2.3. Status Norms 6.2.3.1. Sanctions and Tird Parties For analyses of social status, it is essential to examine the action consequences of judgements of esteem and disesteem also towards third parties, who might judge in the same way, or fail to do so. Tis dimension is an important determinant of the strength of the social norms governing those judgements. At the strong end, judgers explicitly sanction third parties for failing to judge the same way, and reward them—​with praise and/​or other benefts—​ for conforming to the same standards. Clear examples of the former are verbal aggression towards non-​ conformers, shunning, and attempts to convince others to adopt the same sanctions towards them.11 Sanctions can take subtler forms, as well. Take the example of a group of sexist men making repeated derisive jokes not only about one of their members who, 9 See Lessig 1996. 10 To be sure, sometimes such second-​best strategies can still be best, all things considered, for liberal social justice; see ­chapter 8, 8.5, on the interplay between distributive inequality and social status norms. 11 I take it that physical aggression is ruled out by independent requirements of justice, whether or not it is dictated by a status norm, as are certain types of verbal aggression, such as defamation.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  169 in their view, helps his partner too much with their baby, but also about the only other member expressing support for him. At the weaker end, there is less policing. Judgers simply adjust their own actions towards a third-​ party non-​conformer, without sanctioning her directly, or trying to induce others to adjust their actions towards them in similar ways. For example, X’s knowledge that Y holds diferent standards of esteem might induce her not to engage in conversation with X, at least about subjects relevant to these standards, but to look for other conformers instead. Tis might be harmless in some instances, but harmful in others, for example, where relevant conversations inform a group decision, say, in the process of hiring somebody for a job. Standards-​induced propensity to engage in conversations with some rather than others can de facto bring about a coalition in favour of one candidate over the other—​even if such coalition-​forming was not intentional, but came about simply because everybody, individually, avoided talking to the non-​conformer, so that her diferent views in favour of another candidate did not get a hearing.12 Tere are many grey areas and difcult cases. For example, we can imagine cases of individuals who are very confdent about their own judgements of esteem and disesteem, and do not hesitate to express them towards the persons judged, but entirely abstain from sanctioning any third parties in any direct or indirect way. Tey may even welcome the fact that others hold diferent standards, prefer social contact with them over contact with the like-​minded, and relish debate with them. We could call them, somewhat uncatchily, perfect third-​party liberals. In their case, it becomes more difcult to speak of their following social norms of esteem, based on their standards—​even if they should apply these standards very regularly and frequently. In the limiting case, individuals may not hold any expectations of conformity towards third parties, and they may not even be disposed to accept that certain of their own actions towards people showing, or lacking, certain characteristics, are expression of esteem or disesteem, to start with—​as opposed to mere habits. Tey may even disavow these actions (perhaps pleading that they lack the strength of will to change them). To the extent that they are sincere, there

12 Formalizing such processes can go some way towards eliminating such problems, but not all the way—​in this example, we would have to think of a formal procedure which entirely cancelled any efects of informal conversations. But this is not only very hard, or perhaps impossible; trying to achieve this result may also have undesirable other efects (for example, on a relaxed and generally trustful work environment).

170  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

is no norm among them—​even if, again, these habits should happen to be very widespread.13 For a norm to be in place, its followers must be committed to the validity of the standards that are its content, with at least some generality. Te role of social sanctions against those judged, and third-​parties, is to reinforce these standards. 6.2.3.2. Te Scope of Norms Clearly, however, such standards need not be so general as to apply to all members of the society in question, but can apply only to a subgroup—​such as men, or women, people in a certain profession, or of a certain regional origin, within a certain peer-​group, and so on. Scope restriction can make social norms both more harmless, and more harmful, for social equality than perfectly general social norms dictating the application of equivalent standards to all. An example of the former case is a democratic social norm demanding, on pain of severe social criticism, that ofce holders display particularly afable social behaviour (while tolerating some rudeness and aloofness in others). An example of the latter is a social norm penalizing women for not shying away from confict, and from insisting on their interests, as they see them—​they are being branded as ‘pushy’ or ‘bossy’, while men behaving similarly get away with it, or are perhaps even praised as ‘determined’ and ‘decisive’. Tere are many more group-​specifc expectations we can think of. It is not necessary to be able to draw the line between social norms and other practices and habits with confdence in all cases, in order to agree to the general point that the strength of a social status norm increases with the number of people who follow it, and the force of the sanctions, also towards third-​party non-​conformers, which these followers enact. Increasing the scope of followers thus increases the strength and salience of a norm, while 13 Tis is not to claim that such habits cannot be unjust. But they will be unjust only to the extent that they bring about outcomes which are already independently condemned by justice. One complication is worth noting: people who profess that some behaviour is merely an instantiation of a habit, or based on entirely subjective preferences, without any esteem judgements involved, may well be wrong about this. Unconscious factors can play a large role in many cases, and individuals do not enjoy full interpretative sovereignty over their own actions. You may tell yourself that it is just a harmless fact about you that you tend to enjoy conversations with people with an English accent more than conversations with people with American accents, because you just like the sound of the former better, and that you therefore seek out the English for conversation; but perhaps you actually think the English tend to be more intelligent (you just do not think you think it). Tis line of thought need not lead to the conclusion that there is no such thing as a mere, ‘innocent’, habit or preference.

Social Status, Self-Respect, and Opportunity  171

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

increasing object scope (entire society versus subgroup) need not. With this account of norms of social esteem and consequent allocations of social status in hand, we can start to turn to the question of when they are unjust, and why. 6.2.3.3. Regarding, Treating, and Hypocrisy Before doing this, however, it is necessary to tie up one more loose end. Te analysis undertaken so far insists that what matters exclusively is esteem-​ induced, and norm-​reinforced, action towards others. Tis would seem to imply that where one’s actual convictions and beliefs diverge from the norms one professes, through one’s actions, to follow, that diference is immaterial. Tat might seem puzzling. And this kind of puzzle does not seem to arise only in the case of status norms; so it is worth briefy refecting in a more general manner on the relationship between regarding others as equals, and treating them as equals, and how the distinction between these enters into a liberal relational egalitarian conception of social justice. It is, in fact, a general problem, applying to matters of more fundamental respect, as well. On the one hand, it seems problematic for justice to demand not only that people treat others in the required way, but that, in doing so, they also have the appropriate mental states and attitudes. On the other hand, it seems plausible to hold that somebody who ostensibly goes through the motions of treating others as equals, while in fact despising them, falls fouls of some social egalitarian standard (whether or not the requirement of treatment that is in question is particularly important, or not): he is a hypocrite. From the expressive perspective (­chapter  2), it follows that diferences in intention can make a diference to the justice of an action in the absence of diferential outcomes, and thus that a hypocritical action can be less just than a properly motivated one even where outcomes are identical. As seen there, intentions are not the only thing that matters for the meaning of an action, but they do matter. To that extent, the relevant diference between liberal relational egalitarianism and other social egalitarian positions, such as pluralist ones, is merely that the right attitude toward others is one based on respect for their freedom and equality in social cooperation, and need not in any way be based on the good of egalitarians relations, beyond that. However, where intentions really are private—​as they are in cases of skilful hypocrisy—​we should not hold that hypocrites are open to any kind of sanctions, even of an informal, social nature, only for not having had the right intentions (say, facts about their real intentions somehow come to light aferwards).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

172  Liberal Relational Egalitarianism Tis does not mean that aiming at the right attitudes in other ways is not a proper aim for liberal justice. Te general connection between regarding and treating others as equals is rather straightforward: regarding others as equals is the modally most efective way of ensuring that one indeed treats them so. Te right kinds of attitudes are the best (because most robust) guarantee for the right kind of relations. People with the right attitude have the most reliable disposition to treat others as equals, independently of contextual factors which might induce them to show the right kind of behaviour, but for other reasons (such as fear of sanctions, ofcial or informal). So, to that extent, measures aiming at the presence of the right kind of attitudes within society are not only permitted, but will ofen be required. And it is perfectly appropriate for individuals to not only feel suspicion towards, but express guardedness in their relations with, those of whom they have reason to think that they merely pretend to have the right attitude—​even if they have not acted unjustly (yet). Status norms are, then, special only because they operate by directly prescribing certain attitudes: they tell individuals what to think, and how to judge, not merely how to act. A challenge to this way of accounting for the signifcance of attitudes is that of the perfect hypocrite. Tis is somebody who has the wrong attitudes—​ let us say she regards everybody around her as of lesser value, and feels contempt for them—​but who never acts on these, and instead always displays the appropriate kind of egalitarian behaviour, whether in the domains of respect or esteem. Such a person would then be beyond reproach, from the point of view of liberal justice, in the sense of being secure from any kind of sanctions by others, despite not being a social egalitarian: as seen, her relations with others would still be contaminated by her defcient attitudes. Can it be the right result that justice lets such a person get of scot-​free? And, even if it is the right result, from the point of view of justice, does this not cast doubt on the idea, central to this book, that justice can account, in a sufciently comprehensive manner, for the concerns that animate social egalitarians? Tere are two responses to this objection. Te frst challenges the idea that somebody whose attitudes, ex hypothesi, never infuence her actions in any way actually has these attitudes, in the frst place—​instead of merely thinking that she has them, but being wrong about herself. Tis response would catapult us into disputes about the connection between attitudes and actions which belong to the philosophy of mind and action. Te second response concedes, more modestly, that the perfect hypocrite does pose somewhat of a problem, but a much smaller one than it would be to regard attitudes as

Social Status, Self-Respect, and Opportunity  173 fully within the scope of justice independently of their expression in action. On this view, it may be a bullet that liberal justice has to bite, but not a very hard one, given the somewhat contrived nature of the example. Afer all, we think that what is special about hypocrisy, as opposed to other cases of dishonesty, is that hypocrites would change their behaviour if displaying their real attitudes and beliefs became socially acceptable: what motivates them is the desire to avoid social criticism, or to reap praise. But that just means that their dispositions do not have the modal robustness required. Somebody who reliably exhibited the dishonest behaviour in question even if social circumstances did allow for cost-​free honesty would not ordinarily be considered a hypocrite, but merely a bizarrely dishonest person, whose aims and motivations, if any, might remain a puzzle. Tus, even though, as far as liberal social justice is concerned, hypocrites are beyond reproach, in the sense outlined, the normally tight connection between attitudes and actions gives a mandate for sociopsychological strategies (4.5.1 in c­ hapter 4) encouraging those attitudes most conducive to relational equality.14 And this must be all the more so for the case of measures countering esteem-​based status norms, because their main point is to prescribe evaluative attitudes, and reactions ensuring their own reproduction—​ if they are, in efect, unjust, because of their efects, or because of the way they bring these about.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.3.  Te Diverse Efects of Status Norms A natural starting place for the inquiry into the injustice of status norms are the efects of such norms on other justice-​relevant goods and concerns. While receipt of esteem is, other things being equal, a good thing in itself, demanding equality of esteem for its own sake as a matter of justice makes no sense, even as a pro tanto requirement. We all believe not only that some things merit esteem while others do not—​being trustworthy and reliable merits esteem while being selfsh does not; producing things of social value merits esteem, producing useless things does not—​but that one should receive more esteem the more one possesses such positive traits and puts them

14 To be sure, that mandate must not run afoul of other liberal structures, such as especially of freedom of thought, and expression. As ‘hypocrisy is the tribute that vice pays to virtue’ (ch. 4, fn. 7), liberal justice should content itself if it reliably extracts the tribute.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

174  Liberal Relational Egalitarianism into action, or produces things of value. Gradation is of the essence, for esteem. For the same reason, it would make no sense to require, somewhat less demandingly, that the reigning set of social norms accord social esteem equally, while individuals remain free to accord diferential esteem beyond these norms, according to their own lights. If there are good reasons to think that some things merit more esteem than others, then it seems there can be no objection to corresponding social norms. To the contrary, having such norms is then valuable for all of us, as they enhance our capacity to orient ourselves in our social world, by indicating which traits, skills, and projects are worth having, cultivating, and pursuing. Tere is somewhat better reason to be suspicious of directly inegalitarian norms, whose main purpose it is to instantiate or reinforce an esteem hierarchy between individuals (6.2.2)—​and which thus make sure, rather than merely leaving open the possibility, that some individuals will end up with less esteem than others. Competitive norms, which decree that there have to be winners and losers in some social activity, and that only the winners get esteem, are of this kind. It is not necessary to endorse a positive requirement of equality of esteem to be suspicious of norms that make it impossible. Perhaps, then, a case not to have at least these kinds of norms can be made on the grounds of the value of esteem alone—​a requirement against zero-​ sum esteem competitions. Tis, however, could only be a very weak requirement. Such norms might, in principle, be justifable on the basis of positive efects, at the bar of justice—​say, spurring some esteem competition leads to greater productivity all round, whose fruits are then distributed to the disadvantaged. What is more, it might already count as such a justifcation if a signifcant number of people want to have such competitions, because they value the chance of winning sufciently to accept the risk of penalties for losing. Tis kind of freedom matters for justice, too. Simply ruling it out would amount to putting forward a kind of social egalitarian perfectionism. So, to arrive at frmer requirements governing social status norms, social egalitarians have to look to their further possible negative efects on individuals. Tese can be efects on power and domination (6.3.1), self-​respect and self-​esteem (6.3.2), and opportunity (6.3.3). Ultimately, however, for norms of social status to represent a genuinely independent concern for social egalitarianism, they must be capable of being unjust also in cases where the efects they produce are not already condemned by independent requirements of justice, or, where they are so condemned, the fact that they are caused by social status norms must aggravate the injustice in question. Tere must

Social Status, Self-Respect, and Opportunity  175 be something specifcally unjust in the mechanism through which such norms produce their efects. Tis argument will be made in the subsequent section (6.4).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.3.1.  Status Norms and Domination Status norms are unjust when they lead to the violation of requirements of non-​domination. Tis has been argued in a nutshell already in c­ hapter  4 (4.2), but we are now in a position to deliver a more complete analysis of the interplay between domination and social status. Status norms enable domination in, broadly speaking, two kinds of cases. Te frst case is the case of status attached to justifed positions of superior power. If things go well, these positions, and, therefore, their occupants, enjoy authority over others. Sociological authority—​being de facto recognized by others as rightly in power—​is an essential ingredient. It must be, for social cooperation to function well: efectively and stably. No cooperative social system can function well, and durably so, merely on the basis of coercive enforcement, or the threat of it. Tose who enjoy such authority can then try to exploit their elevated status by trying to extend it beyond the domain in which their authority was justifed, and success in this enterprise results in domination. In fact, having unchecked opportunity to try already constitutes domination. However, our subject in this chapter is esteem-​based status, and one might want to resist the inference from the possession of authority to the receipt of esteem. Te objection here is that others’ respect for authority, no matter how pronounced, need not translate into esteem. Indeed, it would seem irrational for those who accept the authority of an ofce, or position, to automatically accord esteem to its holder, on top; unmotivated at best, and, at worst, an instance of directly hierarchically-​minded herd behaviour (6.2.2). Tis objection seems right. In the frst instance, occupants of positions of superior power will, if the position is recognized as justifed, be able to draw on social norms commanding others’ acceptance of that power; but those norms can, and should, operate merely in the domain of authority, not in the domain of social esteem. If the power positions are justifed, it is misguided to object to the authority attached to them, but it should be confned to the narrowest functionally acceptable remit, to ward of risks of domination due to excess (sociological) authority, and people must generally display vigilance about

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

176  Liberal Relational Egalitarianism those remits. Tat then, fully explicates the point made in the penultimate chapter. It suggests that authority-​induced automatic esteemers of ofce holders are simply making a mistake. Tis mistake might be quite common, and it certainly merits deeper psychological investigation,15 but it does not raise a deeper theoretical problem for social egalitarianism. Tis objection is, technically, right about the lack of an intrinsic connection between the mere existence of a recognized ofce, or other power position, and esteem. However, it is nevertheless a little too quick. Te transition from authority to esteem is still, in many circumstances, quite smooth:  if the position in question is recognized as justifed, that will be so because it is regarded as fulflling a socially valuable role (judge, mayor, and so on). Accordingly, at least those power holders who wield their power reasonably well will command esteem for that exercise from others. Benign exercise of justifed authority must be a source of esteem, for social cooperation to function well: it must be recognized as doing something of social value well, and this recognition just is a kind of social esteem.16 If that is so, those with authority enjoy, other things being equal, more esteem than others, or at least have a substantive opportunity to acquire esteem which others lack; and they may well use that esteem to try to collect de facto recognized power over areas outside their remit, too.17 Somebody’s recognized excellence in one position of authority might, on the one hand, qualify them for other positions, too; but, on the other hand, it also turns them into a potential dominator.18 Tis is thus an esteem mechanism latching on to power positions which requires particular attention, because it is not just about mistakes due to particular, contingent, hierarchical, or ‘authoritarian’ psychological dispositions. To be sure, trying to counter it by automatically lowering the overall esteem received by the person in question would be absurd: why would their power and excellence in one dimension themselves be reasons to hold that they 15 For example, it may be a feature of an authoritarian personality type; see Adorno et al. 1950. 16 Te reverse of this is that, if there is a widespread perception that ofce holders exercise their power badly, resulting judgements of disesteem will be particularly forceful, and forcefully expressed—​at least in democracies, where such expression is allowed. While such vigilance is generally very healthy, it can backfre if it tips over into wholesale prejudice against all those in power, at the expense of proper inquiry into their actions. Hatred of ‘elites’, independently of what they actually do, is a problematic feature of populism. 17 Te transition from authority to esteem is insufciently explained, and thus vulnerable to the objection just defused in Schemmel 2015a, pp. 159f. 18 Countering this danger was arguably among the original motivations for the ancient Athenian practice of ostracism—​though in practice it has also been used merely as a tool in the struggle between two powerful leaders; see Pericles’ successful campaign to have Tucydides ostracized in 444–​442  bc.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  177 merit less esteem elsewhere? A spirit of vigilance including healthy opportunities to satirize, and make fun, of holders of authority, is a more appropriate strategy, coupled with stringent requirements of equality of opportunity for such positions (­chapter 7, 7.3; c­ hapter 8, 8.6.), and stringent limits on terms of ofce, within political institutions as well as regarding positions of social power more generally. Te second case is more straightforward. Tis is the case of esteem norms contributing to some ending up with informal positions of authority and power which simply should not exist. Because they should not exist, having them is domination. A case in point may be that of a company that seeks to make its hierarchies fatter, and appoints a commission to come up with proposals for doing so—​but on that committee, those with more seniority and experience then end up in a de facto dictatorial position, because these attributes command so much esteem that no one dares to challenge their proposals (or those who do are penalized by others for their lack of deference). In a case such as this, whatever esteem it is generally justifed to accord to experience, or professional success, it must not translate into any superior power and authority at all. Tis gives us a fuller picture of the connection between esteem-​based social status, power, and domination. Nevertheless, it would be disappointing if that was all there is, to norms of social status, for relational egalitarian social justice. Domination is, afer all, only one social egalitarian concern, and we have argued for a relatively narrow construal of it: it requires that there be an agent who has arbitrary capacity to interfere in the lives of others, and that that agent has more power over the agent liable to interference than vice versa (3.3.1 in ­chapter 3)—​or, at least, that some individuals have more infuence over the group agent who has the superior power to so interfere (4.2.2 in ­chapter 4). Tis narrowness played an essential role in justifying the elevated status of requirements of non-​domination within relational egalitarian justice (4.3.3 in ­chapter 4). While inegalitarian and exclusive esteem norms are very ofen connected to such power scenarios, especially in the case of structures of norms of gender, race, ethnicity, and sexuality, they need not always create, or prop up, defnite positions of superior power.19 Te ‘middle-​ classes’ do not seem to dominate the lower simply because they own fancy cars while the latter cannot aford them, and stand by on the pavement watching them whiz by (perhaps enviously); nor do they, at least on the face

19 See Anderson 2018, p. 93, for more examples of power and esteem hierarchies coming apart.

178  Liberal Relational Egalitarianism of it, dominate them by founding car owner clubs from which the latter are excluded. In cases such as these, other efects of status norms may be in play which merit investigation in their own right.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.3.2.  Self-​Respect and Self-​Esteem Self-​respect is another natural place to look for such efects. We have seen in the previous chapter (5.3.2) that many social egalitarians appeal to considerations of self-​respect to ground the special value of egalitarian relations. Insofar as self-​respect is a crucial personal good, social justice certainly demands that individuals enjoy a social and political environment that is appropriately favourable to its development and maintenance. Egalitarian relations might then, on this view, be an essential ‘social basis’ of self-​respect; inegalitarian norms of social status are unjust insofar as they endanger this environment. However, social egalitarians rarely specify what exactly self-​respect is, and does, for individuals, and what, therefore, its distinctive value is. Te social egalitarian literature, so far, proceeds largely in isolation from most of the philosophical literature on self-​respect. Closer inquiry into the question of which conception of self-​respect liberals have good reason to endorse reveals a more nuanced picture.20 On the one hand, certain pernicious inegalitarian relations, and ways of treating each other as unequal, should be classifed as threats to such self-​respect, and combated on this basis as a matter of justice. And insofar as self-​respect is indeed an essential personal good, this yields particularly stringent reasons to combat them. On the other hand, there is good reason to hold that norms of social status meting out diferential social esteem need not, as such, constitute threats to self-​respect. Tis is because holding the contrary assumes a conception of self-​respect which regards individuals as completely and continuously dependent on how they are respected and esteemed by others for orientation about their own worth,21 and the worth of their own projects and activities, and thus insufciently capable of being autonomous also in social circumstances that are less than 20 Te discussion in this subsection draws on Schemmel 2019. Schemmel 2011 seeks to base too comprehensive a set of relational egalitarian requirements, and in particular their distributive implications, on self-​respect, with insufcient analysis of that notion. Chapter 8 remedies this by deriving such implications from the full set of requirements argued for in c­ hapters 2–​7 instead. 21 Bird 2010, p. 17.

Social Status, Self-Respect, and Opportunity  179

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

perfectly favourable to them. Such a conception leads to implications that are seriously problematic for liberals. To that extent, a better solution for liberal justice is to seek to enable people’s capacity to attain and retain such orientation in a more robust fashion—​which leads to more of a focus on education and enabling resilience, rather than on norms of social status. Let us unpack this argument. 6.3.2.1. Two Kinds of Self-​Respect, and Teir Value Te literature on self-​respect standardly distinguishes between two kinds of self-​respect, which we can call standing self-​respect and standards self-​ respect.22 Te distinction between them roughly corresponds to the distinction between respect and esteem drawn previously (6.2.1). We can therefore also speak of ‘self-​respect’ and ‘self-​esteem’23, as long as it is kept in mind that, in ordinary language, the distinction between these two concepts is far from always clear. Tese two kinds of self-​respect also map well onto the two Rawlsian moral powers of a sense of justice, and the capacity to adopt, and carry out, a conception of the good (­chapter 3, 3.2). Standing self-​respect is a matter of moral convictions about oneself: for liberals, this is the conviction that one is a moral agent subject to the same basic moral demands as all other agents,24 and worthy of being accorded a set of rights that are equal to those of others. Standards self-​respect, on the other hand, is the conviction that one is capable of developing, and carrying out, on the basis of one’s talents and capacities, projects and activities that have value, and thus are worth carrying out. Standards self-​respect can refer to specifcally moral projects, but need not—​(perceived) moral demands can be of varying importance to the content of the standards one most holds oneself to in one’s life.25 22 Tese are intended to be the most neutral available terms for the distinction in play. Darwall 1977 proposes to distinguish between ‘recognition self-​respect’ and ‘appraisal self-​respect’; Bird 2010 between ‘entitlement self-​respect’ and ‘standards self-​respect’; and Kristjànnson between ‘Kantian’ and ‘Aristotelian’ self-​respect; 2010, ch. 7. Teir distinctions difer in some nuances from each other, and from the one drawn here, but not in ways which are relevant to the argument of this section. 23 As Honneth 1995 does. 24 For Kant, it is the conviction that one is, as a rational being, subject to the demands of the moral law, which one has to respect for its own sake, just as all other rational beings do; Kant 1996 [1797], p. 182 (Ak 6:429). 25 Rawls defnes self-​respect ‘as having two aspects. First of all, [ . . . ] it includes a person’s sense of his own value, his secure conviction that [ . . . ] his plan of life [ . . . ] is worth carrying out. And second, self-​respect implies a confdence in one’s ability [ . . . ] to fulfll one’s intentions’; Rawls 1999a, 386. Tis defnition is somewhat tilted towards standards self-​respect, as the emphasis on one’s (particular) plans and abilities shows, and Rawls tends to use ‘self-​respect’ and ‘self-​esteem’ interchangeably. Standing self-​respect is, however, a very important ingredient of his view, too. Tis becomes evident in other places where he refers to the importance of being publicly recognized by ‘just institutions’

180  Liberal Relational Egalitarianism Having these two kinds of self-​respect is evidently crucial for one’s personal good. As Rawls notes, if self-​respect is lacking completely, then

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

[n]‌othing may seem worth doing, or if some things have value for us, we lack the will to strive for them. All desire and activity becomes empty and vain, and we sink into apathy and cynicism.26

Tere can, of course, be defects of self-​respect which are not quite as catastrophic. But the quote illustrates well that its key function is one of practically efective self-​orientation: by assuring us of our worth—​both in terms of moral standing and in terms of our capacity to engage in esteem-​worthy activities—​it motivates us to act in ways that conform to the demands put on us in both dimensions. Functioning self-​respect assures us that we are efective authors of our own actions, that these are worth it, and that we are responsible for them; therefore, it plays a crucial role in enabling, and shoring up, personal autonomy.27 Successful self-​respecters display integrity in these two dimensions: their commitment to said demands is stable, and normally efective, and, in case of failure, they take responsibility for the failure, and actively seek to remedy it. Accordingly, functioning self-​respect also yields, or at least crucially corroborates, an expectation that others acknowledge our worth in their interactions with us—​and that we therefore need not, and indeed should not, approach them with deference and servility (even where they should hold justifed authority over us), or be despondent about the worth of our talents and activities, and envy others for theirs. Tese capacities and dispositions are crucial ingredients of successful relations to ourselves and to others, whatever our individual conception of the good. So there is a straightforward case for liberal justice requiring them to be protected, and their development appropriately enabled.28 (ibid., p. 477; my emphasis)—​not simply by institutions which happen to be favourable to one’s particular plans and abilities—​and stresses that a guarantee of equal political liberties, including a guarantee of their ‘fair value’, is an important social base of self-​respect; Rawls 1996, p. 318 fn. 29. Such political rights are a particularly important means for acknowledging one’s equal possession of the frst moral power, the sense of justice. 26 Rawls 1999a, p. 386. 27 See Govier 1993 and Benson 1994. 28 Accordingly, Cordelli lists the social bases of self-​respect among the ‘relational resources’; 2015a, pp. 96f. In line with the argument of the preceding two chapters, interventions to enhance and protect self-​respect must not themselves be disrespectful—​they have to abide by the same constraints of basic respect (­chapter 4, 4.3.2) which must govern the provision of opportunities for relations of trust and friendship (­chapter 5, 5.4.3).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  181 6.3.2.2. Treats to Self-​Respect, and Robustness Tere are indeed many ways in which inegalitarian relations can endanger the capacities constitutive of functioning self-​respect, and concern for self-​ respect is at play in many of the relations we object to most. Take domination: it is hard to see how someone who grew up, and is kept, in slavery, can develop and maintain the conviction that he is of equal worth to others, and morally entitled to the same basic rights, even though they have always been withheld from him, and he has no idea what it would be like to enjoy them. Similarly, pervasive inegalitarian norms of social status systematically withholding social esteem from a group of persons, or commanding disesteem for them, will, in many cases, severely impede the capacity of these individuals to arrive at the conviction that they do possess esteem-​worthy characteristics, and valuable projects. In some familiar cases, for example, of racism and sexism, such norms will also cross the line to violations of basic respect, insofar as they deny that members of the so devalued groups possess, to the required extent, the fundamental moral powers that ground their standing as free and equal. Tat is the case, for example, when women are, as they historically ofen were, systematically portrayed as lacking the strength of will, deliberative capacities, and public-​spiritedness needed to be an equal, and good, citizen. Furthermore, whether or not they cross that line, such norms can play a crucial role in propping up systems of domination (6.3.1). One particular important, and insidious, way of doing so is by successfully peddling sham justifcations for dominatory power inequalities—​justifcations that the subjected themselves internalize. Accordingly, protection against such inegalitarian relations constitutes one important subset of the social bases of self-​respect. Individuals’ opportunity to arrive at correct, and secure, convictions of their own worth must not be thwarted by subjecting them to relations which encourage them to conceive of themselves as inferior, consistently undermine the security of the conviction that they are equals in the relevant respects, and deny them access to other, better relations. Given the importance of self-​respect for enabling autonomy, this thus delivers particularly stringent reasons to protect people against such pernicious forms of domination and status inequality. However, for both the dimensions of standing and standards self-​respect, it is problematic to hold that all domination, and all norms of social esteem instituting some inequalities, are such threats to self-​respect—​and that self-​respect is thus a comprehensive rationale for objecting to all of them (in

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

182  Liberal Relational Egalitarianism addition to their being unjust for other reasons, where they are). Not all inegalitarian relations deprive those subject to them of the epistemic access to correct convictions about their own worth, and not all of them necessarily undermine the capacity to robustly maintain these convictions. For standing self-​respect, this is well-​recognized in the literature in moral philosophy. Standing self-​respect cannot deliver any requirements not only where the convictions constitutive of it are incorrect: where they rely on a conception of individual rights and duties that is misguided. For example, it would be absurd to hold that sexist men, whose standing self-​respect is, in part, grounded on the conviction that they are, as men, entitled to superior standing, have a self-​respect-​based complaint against the ongoing emancipation of women in their society—​even if these societal changes do in fact cause serious psychological disorientation about their worth (2.6.2 in ­chapter 2). More pertinently for our present purposes, it would also be inappropriate to hold that having been subjected to an injustice, of any kind, as such constitutes a reason to lose self-​respect. Tis is because holding it would, in efect, amount to conceding that individuals cannot be expected to have proper and full self-​respect unless circumstances are fully just. However, it is not merely an empirical contingency that some have robust self-​respect enabling them to withstand injustice, while others do not. It is constitutive of functioning standing self-​respect that individuals are robustly aware that injustices against them are not reasons to lose self-​respect; and not all injustices necessarily undermine such resilience. A person with self-​respect reacts with indignation to injustice, and protests against it, unless there are very good external reasons not to (say, they know the perpetrator will kill them if they do).29 Such protest afrms their worth.30 Part of the role of standing self-​respect is thus exactly to be able to deal with (at least some) injustice, and to direct people towards appropriate reactions to it. Take the example of inequality of power in the workplace between managers and workers (­chapter 3, 3.4): disenfranchised workers, who have to take up whatever employment they can get, on whatever terms, in order to fulfl their basic needs, are subject to a kind of domination that is so intense that it plausibly also endangers their social bases of self-​respect. Continuous complete dependence on the better-​of and more powerful for securing even the most basic goods makes it unduly hard to develop and

29 Hill 1991, p. 11.

30 It can thus be called for even if it has no chances of practical success; see Boxill 1978.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  183 maintain convictions of one’s own equal worth. But cases of workplace inequality, which, while guaranteeing a high level of security of needs fulflment, and protection against dismissal, nevertheless get the balance of directive power between managers and workers wrong, need not, as such, cross that line. Tere is no reason why workers would have to react to it by losing, or failing to develop, self-​respect. Tey should organize to change this balance, and functioning self-​respect will help them in that task. Tis may be diferent in cases where they are in the grip of a particular kind of capitalist ideology comprehensively devaluing the social contributions of workers, compared to those of entrepreneurs. If subjection to such an ideology is indeed comprehensive, and very hard to escape, this would have to be regarded as a threat to their self-​respect. Tis distinction between threats to standing self-​respect and ‘simple’ injustice does not fully correspond to a distinction between severe and less severe injustice, but the former do have to have the specifc quality or barring epistemic access to correct convictions, and/​or undermining their resilience.31 Similarly, in the case of standards self-​respect (or self-​esteem), a person with secure self-​esteem is able to tolerate the fact that not all comparisons with others’ talents and activities come out favourably towards themselves, and does not react with despondency, and envy, to such outcomes. We have encountered reasons for why not all social esteem can be equal. Tus, holding that injustice, and all inequality of social esteem, per se constitute threats to self-​respect which require social egalitarian counteraction amounts to relinquishing the expectation that people can have robust standing and standards self-​respect—​that is, self-​respect also under some adversity.32 It would also have the problematic implication that it turns envy, insofar as it is caused by, or constitutive of, low self-​esteem, into a reason for demanding more equality. Te frst problem is a problem for liberals insofar as accepting this kind of dependence of self-​respect on the continuous 31 Tere is an inevitable element of psychological contingency as to which arrangements should fall on either side of the line, and this could well vary with local context to some extent, as the requirements of Pettit’s ‘eyeball test’ for non-​domination do (­chapter 3, 3.4.3). Te argument of this section makes a case for insisting on such a line, for reasons internal to self-​respect; it cannot fully determine where exactly the line should be. 32 Expecting people to retain self-​respect has to be kept distinct from ‘holding people responsible’ for it, where this means that failure turns persons into legitimate targets of others’ criticism, or even attribution of blame. As self-​respect is a personal afair, nobody has, strictly speaking, the right to criticize others for failure. Criticism is appropriate only insofar as a failure to maintain self-​respect brings about a failure to abide by other moral requirements. For example, a person ‘sinking into apathy’ for lack of self-​respect may fail to fulfl his duty to support eforts to eradicate domination (­chapter 4, 4.5.3).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

184  Liberal Relational Egalitarianism receipt of respect and esteem by others amounts to treating individuals as ‘in every way [ . . . ] ‘other-​directed’,33 as not capable of sufcient self-​direction, and thus autonomy, unless external conditions are ideal for them. Te second is a problem for liberals insofar as it gives up on the key requirement that justifcations for both equality and inequality are to be given by way of independent, substantive reasons, not simply by reference to how people might be disposed to feel under them. Absent such reasons, proposing to withhold advantages from others merely on the basis of one’s feelings is disrespectful.34 It is crucial to note that this argument does not, as it were, try to wish away features of people’s emotional constitution that spell trouble for an ‘asocial’ conception of the self (which some liberals may have traditionally cherished), thereby ofoading the task of developing and maintaining robust self-​respect entirely onto individuals alone. It is not only hard to deny that it is better for individuals to have robust self-​respect of both kinds, if it can be had, because this renders them more securely capable of orienting themselves in their social world even where not all circumstances are favourable to them. Clearly, such a capacity for robustness also has its own social presuppositions; people do not automatically become robust self-​respecters. Te argument thus points to a second important subset of the social bases of self-​respect, which social and political arrangements have to aim at: these are the resources people need to arrive at correct convictions of their own worth, and to keep them robustly, even under adversity. Plausibly, the provision of this second set of social bases will especially require a sustained focus on an upbringing and education which enables children and adolescents to develop into resilient self-​respecters capable of dealing well with difculties, and of tolerating at least some social comparisons that are unfavourable to them (preferably by not even undertaking them).35 If this can be guaranteed, 33 Bloom 1975, p. 654. 34 See Anderson 2010b, pp. 8f., and 1999, p. 307, where she argues this against Dworkin’s proposal to determine just individual holdings by recourse to an envy-​test (Dworkin 2000, ch. 2). However, Dworkin actually relies on a diferent, more technical notion of envy. Her point applies better to a more everyday notion of envy like the one preoccupying Rawls, who speculates that certain degrees of distributive inequality may engender ‘excusable general envy’; 1999a, p. 468. What is missing in Rawls’s case is precisely an analysis of the kinds of threats to self-​respect that status norms can pose. 35 For some suggestions as to what this may require in practice, see Schemmel 2019, pp. 636, 640, 644. One may be tempted to argue that the importance of enabling robustness illustrates that it is misguided to go about the social basis of self-​respect in a distributive, purely recipient-​oriented fashion, seeking to engineer social contexts so as to make sure that individuals end up with equal amounts of self-​respect, rather than empowering them to take care of it themselves. However, it is perfectly sensible to maintain that the latter may be regarded as a matter of distribution—​just one of the distribution of the right kind of social bases: those that empower people in this fashion. A deeper distinction between relational and distributive views emerges once again in the next section, latching

Social Status, Self-Respect, and Opportunity  185

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

then unpalatable implications for others, regarding the need to compensate people for envy—​which could be achieved by simply levelling down the envied—​can be warded of. 6.3.2.3. Status Norms Revisited Let us summarize, then, what follows for the relationship between self-​ respect and inegalitarian norms of social status. We have seen that self-​respect, where it applies, grounds special requirements to ward of pernicious inegalitarian relations—​relations that deprive those at the wrong end of the capacity to develop and maintain self-​respect. But, in this chapter, we are mainly interested in reasons to extend requirements of justice to cover a larger array of social relations, in which diferential social esteem is at stake. Tere is good reason to doubt that self-​respect, if understood as the robust capacity outlined earlier, grounds any such blanket requirements on egalitarian social relations. What it does instead is ground other requirements to enable and encourage such robustness. Tus, it also fails to cover all that was important in pluralist social egalitarian objections to inegalitarian relations, and instances of treating each other as unequal (5.3). Tis is not to maintain that inegalitarian norms of social status are a problem for liberal justice only where they enable domination, and/​or tip over into matters of basic respect and self-​respect, and never on grounds of self-​esteem alone. Some inegalitarian norms of social status emphasizing the importance of certain esteem-​markers over others are still to be condemned on grounds of depriving individuals of the social bases of self-​esteem. Tis is so where such norms are so dominant that, as a matter of fact, they crowd out other opportunities to acquire social esteem.36 In capitalist market societies, this may especially be norms celebrating the command of large incomes, and wealth, and presenting these as the only social aims worth pursuing. Tis requirement thus points, in the frst instance, to instituting plural spheres of achievement, not to trying to make sure that people score equally highly in the ones currently socially emphasized, whatever these may be. But not all status norms are dominant in this way; and some fasten on esteem-​markers which we have good reasons to endorse—​such as moral and political qualities like reliability, trustworthiness, and vigilance, or on to the fundamental importance of the specifc mechanism through which (some) status norms bring about individual opportunity disadvantage (6.4).

36 Walzer 1983, pp. 10f.

186  Liberal Relational Egalitarianism performances, such as producing important social goods which beneft others. Tere is nothing in the latter kind of norms which necessarily implies that only particularly impressive performances should command esteem at all. Tey can be part of a broader system of norms stressing that any worthwhile contribution to society is a proper basis for self-​esteem—​and indeed also of standing self-​respect, insofar as individuals have both a right and a duty to participate in cooperation. Tis is a system of norms which we have good reason to encourage.37 It may, of course, be that even such a diferentiated system of norms does, as a matter of fact, cause damage to some people’s self-​esteem. Whether robust liberal self-​respect and self-​esteem of the kind outlined earlier can indeed be had for all is, in principle, subject to empirical investigation. It remains a possibility to change our conceptions of self-​respect and self-​esteem so as not to demand robustness as a constitutive feature of self-​respect. However, the preceding argument makes clear that this comes at the price of devaluing the currency of self-​respect, from a crucial capacity for self-​orientation in different, including challenging, social situations, to one ingredient of psychological well-​being among others, which may vary with one’s social situation, just as these others do.38 If the preceding argument is sound, this is a price that liberals should be wary of paying, because it severely endangers the internal connection between self-​respect and autonomy—​which accounted for its distinctive value. It is worth aiming at robust self-​respect frst.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.3.3.  Status and Opportunity Considerations of self-​respect thus do not lead to a blanket objection to all inegalitarian norms of social status. However, there may be further reasons of justice to object to such norms, and these may be, as it were, closer to home, in an important respect. Social esteem matters when it is expressed in actions (6.2); and these actions can have consequences other than 37 See Jütten 2017, and Stark 2012, who proposes, as the central feature of Rawls’s conception of self-​respect, the conviction that one’s contribution to social cooperation matters. Te corresponding duty of institutions of the basic structure is to secure this conviction. However, Stark also argues that having this conviction matters intrinsically, on Rawls’s view; ibid., p. 259. Whether or not this is the correct interpretation of Rawls’s view, it is not clear why the conviction should matter in itself, not only because it shores up agency in the way outlined earlier—​as opposed to it mattering intrinsically that everybody be appropriately respected and esteemed for the contributions that they in fact make. 38 Tis is how it is largely treated in epidemiological and socio-​psychological studies focusing on the consequences of inequality; for prominent examples, see Wilkinson and Pickett 2010 and 2018.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  187 domination or harm to self-​respect. Whether or not the child from a poorer family background who is regularly not selected as a playmate by the well-​ to-​do parents of his schoolmates grows up to have healthy self-​respect and self-​esteem regardless, such informal selection procedures may well have an important impact on her opportunities in later life. And whether or not, to take Adam Smith’s famous example, the labourer who lacks a ‘linen shirt’, and is thus ‘ashamed to appear in public’39, is so ashamed that we have to say that he is sufering a signifcant dent to his self-​esteem, his, by widely accepted standards, defective appearance may nevertheless foreclose important opportunities to him.40 To be sure, if he is denied, or in danger of being denied, a job solely on these grounds, he is subject to domination by the person or committee (as a group agent) in charge of hiring. But social esteem also structures many of the more ordinary of our social interactions with others, who need not have any kind of superior unchecked power. Social esteem opens many doors; disesteem closes them. Take the following example:  a collection of individuals—​say, the middle classes—​engage in ‘opportunity hoarding’, by favouring people who possess certain esteem-​markers, which they themselves possess—​a ‘decent’ public appearance, the ‘right’ accent and manner of speaking, and so on—​when establishing social networks and contacts, and passing on opportunities (for example, for internships, which are then used to establish further contacts).41 Tey need not in any way intentionally coordinate to do this; they need not be a dominating collective agent. And much of this can happen informally, before any kind of identifable decision situation emerges, in which one agent has superior power. Such status-​mediated opportunity loss can be unjust in its own right; explaining why is the task of the next section.

39 Smith 1976 [1776], Vol. 2, p. 399. 40 Scanlon 2018, pp. 29f. 41 See Wolf 2017, p.  176; for the classic discussion of opportunity hoarding, see Tilly 1998. Opportunity hoarding need not, in principle, be esteem-​mediated; it could be simply strategic action. Tis can push it closer to domination, insofar as, at least in cases with many agents, people will likely act strategically only if they have good reasons to think that others do, too—​so some intentional coordination is present (see what follows in the main text). However, an esteem-​based process has, for hoarders, the convenient feature of obscuring that what happens may be best described, from a more detached point of view, as strategic exclusion through collusion with similarly placed others. Tey can convince themselves that a denial of opportunities to somebody unlike them is really due to that person not meriting esteem, not by their mere desire to, for example, transmit privilege to their ofspring, and those like them. See fn. 6 on Bourdieusian habitus.

188  Liberal Relational Egalitarianism

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

6.4.  Te Injustice of Status-​Induced Opportunity Loss Inegalitarian norms of social status sometimes deprive individuals of opportunities to which they have an independently grounded claim of justice. Say, all others in society look down and you, thinking you do not merit any social esteem, and express that to you, by, for example, refusing to sell you any food (compare the V examples in ­chapter 2). Plausibly, you have a basic right to food, or at the very least to guaranteed substantive opportunities to acquire it; status norms, in this case, are just one way of having this right frustrated by others. Or, under a luck egalitarian principle of fairness, you have a claim not to be deprived of opportunities that others enjoy for reasons beyond your control, for which you cannot be held responsible. Again, status norms are just one of a whole array of possible causes of opportunity loss beyond individual control, which are all equally condemned by luck egalitarian fairness. For status-​induced opportunity loss to be a genuinely independent concern of relational egalitarian justice, something else must be true. Tere must be something specifcally unjust about losing opportunities in this way; concern for status must not be merely instrumental.42 So, whatever independent principles for opportunity distribution demand, losing some opportunities due to status must be unjust even if other ways of losing the same, or equivalent, opportunities are not; or it must aggravate the injustice in question. Take the following example: you intend to buy a house in a quarter of your city which you like. You have the money, and you can prove it; however, individual sellers in that quarter—​all comfortably middle-​class, and upper middle class—​and estate agents do not consider you as a possible buyer, because they think you do not ‘ft’. Perhaps they think so because of your appearance (you turn up in trainers to the house showing), your way of speaking, or your foreign background.43 Tey think that quarter should be reserved for other kinds of people, like them (or, in the case of estate agents, for their favourite type of clients). Tey do not intentionally coordinate in any way in your case, but proceed quite automatically, acting on 42 It would be somewhat strained to hold that the connections between diferential social status and domination and unequal self-​esteem analysed in previous subsections are merely instrumental, as they are more than empirically contingent causal relationships: there is a content match between status norms and opportunities for domination, and between esteem and self-​esteem. But they are nevertheless extrinsic connections. 43 It is important that they do not think you lack the money—​if that was the only relevant criterion, you would already be unjustly discriminated against by not being given the opportunity to demonstrate that you have it. Te question at stake is which impact on other selection criteria social status norms can, but must not, have.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  189 esteem norms which they all hold, and which their regular interactions with each other reinforce. Now, in this case, it is implausible to hold that you have any kind of independent claim of justice to acquire property in the quarter of your choosing, which is violated here; as opposed, perhaps, to a claim to enjoy efective opportunities to acquire a decent abode, wherever it may be (as long as the surrounding infrastructure is adequate, public goods are equally provided, and so on). So, there would be no problem of justice if people’s buying and selling preferences, impersonally coordinated by the market (and proceeding on the basis of an overall suitable distribution of the resources generated by social cooperation),44 worked out so as to efectively deny you the opportunity to acquire a house in that quarter. Too many people want to live there, so the prices are too high for you. It seems, however, plausible, to hold that this specifc, esteem-​mediated mechanism of exclusion is unjust, even if others may not be (and even if there is no domination, and your self-​esteem is robust, and unharmed). If this is right, then inegalitarian norms of social status can be independently unjust on opportunity grounds, and concern for social status constitutes a genuine rationale of relational egalitarian justice; one that can exist even in the face of disagreement about which distribution of opportunities would be otherwise just or unjust. Even if this is intuitive in this case, a principled reason is still needed to justify it, which will also give more general guidance on which kinds of cases to judge unjust. Just as with domination, we can make progress in uncovering such reasons by relating the mechanism in question to the liberal background ideas of fair social cooperation among free and equal individuals. Tat reason is not that the individuals in question, individually, judge you to be inferior. Te reason is the pervasive, collective esteem norm that holds in the case in question. Exclusion thus happens due to a norm-​coordinated process (even though that coordination can be, largely at least, unconscious, and unintentional), in which social sanctions are in play, also towards third parties: these corroborate the exclusion, and incentivize others to take part in it, so that, without counteraction, the process of exclusion is self-​sustaining. We can say that, through the norm in question, norm-​followers express a collective, objectionable inegalitarian attitude towards you (­chapter 2, 2.6); it is salient for justice because it has the efect of disadvantaging you, but the mere disadvantage might not itself be unjust, or equally unjust, were it not

44

See ­chapter 8.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

190  Liberal Relational Egalitarianism brought about in this way.45 What makes the process unjust is precisely the combination of the content of the social norms in question—​judging some to be inferior—​and the self-​sustaining collective mechanism of informal norm enforcement, which together produce disadvantage for those so judged. Both factors are individually necessary. Whether they are sufcient for injustice then depends on the kind of disadvantage produced (on which more later). Falling victim to this kind of collective mechanism—​a kind of ‘social esteem trap’—​is a risk that individuals who regard, and are committed to treating, each other as free and equal in cooperation should insure each other against. It is qualitatively diferent from merely individual cases of treating others as unequal. We have seen that esteeming some things more than others can be merely a straightforward expression of what individuals regard as worthwhile in their own lives. Tis is thus not an arbitrary distinction, nor one that merely relies on the estimate that, probably, the collective mechanism will produce worse outcomes than individual actions alone (which it, most ofen, will). It is also qualitatively diferent from other social processes which may produce equivalent results, but which lack problematic propositional content—​the judgement of inferiority in question—​such as constellations of mere self-​regarding preferences of others bringing about, through market coordination, prices that are too high for some. However, we still need to pin down with more precision what kind of efects the social norms in question must be liable to produce in order to be unjust. In the frst instance, social status norms that reliably disadvantage individuals in the dimension of other goods whose distribution is salient to justice will be candidates; goods such as income, and wealth, or desirable social positions, such as good jobs, as paradigm benefts of social cooperation. But the efects need not be restricted in this way; the mechanism can also play out in terms of ‘naked’ opportunities, with no further consequences for other goods attached—​as long as these opportunities are themselves of suffcient social signifcance. Tat is, having these opportunities, and making use of them, is an important part of many people’s conception of the good, and the surrounding norms put some pressure on all individuals to see it that way. Tese opportunities, too, are fruits of social cooperation.

45 Take the food case again: if all sellers of food refused it to you because they mistakenly thought they need it for themselves, that will be unjust, but, other things being equal, less unjust than if they refuse it to you because you are the kind of person who, according to reigning social norms, only needs to be served once all others have been.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  191 Take the example of a society in which participation in sporting activities is, as a matter of fact, very important to most people; consequently, accomplishments in sports are an important source of social esteem, and individuals derive signifcant personal satisfaction from participation. Say, however, reigning sports esteem norms narrowly fasten on the possession of very few characteristics, and disregard many others—​in particular, let us say, on physical force: you have to be able to deliver a tennis service, or smash a volleyball, at a certain top speed, to command esteem; to count as ‘really’ playing the game in question, or playing it suitably well. Clearly, this kind of norm privileges members of some groups, where these attributes are more frequently found, over those of others; most likely, men over women, in these cases. So, for women, it is signifcantly harder to participate in the relevant games, or perform in a manner that is regarded as adequate. Whether or not this unduly hampers their self-​esteem, this kind of distribution of opportunities to participate, and do well, is unjust, if sports really matter so much. However, having a society which values sporting activities very much is, other things being equal, perfectly permissible, from a liberal point of view. It is no way out to encourage female sports, anyway, just pretending the said norms did not exist; nor is it an appropriate solution to try to leave relevant esteem dimensions—​physical force—​in play, but to merely scale down what they require. Many esteem standards have their own logic of attraction, and competition is of the essence, for sports games, so this would just dull down the game, and make it less satisfying. Te solution is to broaden the norms so as to pluralize available esteem-​markers, putting more emphasis on others, such as technical skill (with some scaling down of existing ones precisely in order to make room for that plurality).46 Tis seems the right solution, for liberals; it is close to the general liberal strategy of pluralizing spheres of achievement (6.3.2). Of course, as noted, whether narrow esteem markers in sports really are a problem depends on the social signifcance that sports already have, in the society in question. Tis leads to the question of where, in the order of importance and stringency of the demands of liberal relational egalitarian justice, avoidance of status-​ induced opportunity loss, independently of other considerations (such as domination, and self-​respect), stands. It cannot, by itself, lay claim

46 Take the example of tennis: an emphasis on force tends to make rallies shorter, thus, other things being equal, reducing opportunities to display other forms of technical skill. And longer rallies can be much better in terms of dramatic value.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

192  Liberal Relational Egalitarianism to the kind of priority that requirements of non-​domination enjoy (4.3.3 in ­chapter 4), because domination institutes an inequality of standing between individuals through which the more powerful can harm the less powerful at will. Tat is not the case in the norm-​coordinated process of social disesteeming under discussion. Furthermore, because the opportunities foreclosed by such status norms vary in social signifcance, so does the weight of the corresponding requirements, if the stricter prohibitions against facilitating domination, or threatening self-​respect, are not triggered. Tere is thus no hard and fast solution to the question of how to decide trade-​ofs with other, potentially competing requirements of social justice. But the capacity of status-​induced opportunity loss to be unjust in its own right gives us reasons to seek to fulfl any such other requirements in a way that is not in confict with it, and a mandate to accordingly shape the social norms in question by collective eforts (e.g., tax-​fnanced ones).47 It also alerts us to the need to check that any penalties imposed by whatever status norms are entitled to survive are tightly circumscribed in scope and time, and leave those penalized with opportunities to acquire social esteem through diferent pursuits, as well as to try again and reacquire esteem in the same pursuits. Given that our starting point was the question of how salient forms of treating others as unequal which do not amount to domination, or produce other independently condemned results, can be brought under liberal justice (as opposed to merely constituting a kind of disvalue not condemned by it), this is a good result for social egalitarians. It opens up a mandate for counteracting such norms through collective measures.48 Tese will be of the kinds noted in ­chapter  4 (4.5.1):  socio-​psychological strategies, including education, awareness campaigns, and symbolic state action, such as ofcial pronouncements of esteem for formerly devalued groups; and distributive strategies eroding the background conditions for the formation of objectionable status norms.49 Te latter will be investigated further in Part

47 For example, even if some forms of status competition do well at enhancing productivity, and even if enhanced productivity is ofen desirable, we have to search for other ways to achieve whatever level of productivity can be reasonably required. For arguments that maximizing individual productivity is not required by liberal social justice, see Stanczyk, unpublished manuscript. 48 Tis means that, should there be good reasons to insist on a general basic structure restriction for liberal justice (4.5.2.), what counts as that basic structure has to be construed even more broadly than necessitated by the existence of informal domination-​enhancing mechanisms, to include such ‘simple’ opportunity-​denying or -​enhancing social status norms, too. As noted there, I take no stance on whether there are good reasons for a general basic structure restriction. 49 Strategies of formal procedural protection can, against that, only block, or reverse, some of the outcomes status norms generate, see fn. 12 in this chapter.

Social Status, Self-Respect, and Opportunity  193 II (­chapter 8). An important question for that investigation will be whether such norms can be erased or changed without measures exceeding what is permissible under liberal justice, and how, therefore, ‘frst-​best’ and ‘second-​ best’ strategies for combating social status norms of the kind which surround the possession of income, wealth, and desirable jobs and social positions will look like, and difer from each other.

6.5.  Conclusion: Treating Others as Inferior, With and Without Norms To conclude this chapter, we should double-​check whether the position arrived at accounts for all, or most, of the features we identifed, in the previous chapter, as intuitively salient in cases of individuals treating others as inferior which do not amount to domination. Tese cases exercised pluralist social egalitarians such as Mason, and led them to conclude that justice cannot account satisfactorily for the value of egalitarian relations. Take again Mason’s example of the Local Shop:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

A man refuses to shop at his local store because it is run by a member of a particular ethnic minority. Instead, he walks to a store that is farther away and more expensive because it is run by someone from his own ethnic group. [ . . . ] [H]‌e thinks that members of this group are inferior because they generally lack various qualities, such as honesty and integrity [ . . . ].50

Mason maintained, plausibly, that there is something objectionable about this scenario which social egalitarians ought to be able to account for, even if the man’s action has no efects for the shopkeeper that are salient for justice (and domination is not in play). It is nevertheless a failure to treat the shopkeeper as an equal. Specifcally, for Mason, it consists of ‘withholding or denying the equal recognition respect to which [ . . . ] are all morally entitled [ . . . ].’51 Recognition respect requires us to recognize the intrinsic value of others as persons, not to simply act on our prejudices about members of certain groups.



50 Mason 2015, p. 132. 51 Ibid., p. 145.

194  Liberal Relational Egalitarianism How does our analysis compare? Two points are particularly salient. First, it has shown that what matters, primarily, are not the efects produced by individual actions, but those produced by widespread social norms. If these are unjust, then individual actions conforming to the norms are unjust, whether or not any particular act of conformity produces any specifc effect. Te injustice consists of participation in norm-​coordinated exclusion; this is the objectionable meaning of the act in question.52 Tis seems to be in play in Mason’s example—​we are all familiar with exclusionary status norms targeting ethnic minorities, and many such norms tend to produce a great variety of unjust results, ofen going signifcantly beyond mere denial of social esteem. It is therefore plausible to contend that this familiarity structures the reader’s intuitive response to this particular example. In support of this contention, consider a diferent example:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Party invitation. A man decides not to invite his neighbour, with whom he is in polite, but superfcial, contact, to his dinner party. He does so because his neighbour is a PR consultant (while he is a civil servant), and he thinks—​based on sporadic past experience—​that PR consultants are less honest, and more prone to engage in self-​praise, than civil servants, and too much so to make good guests. Tus, he invites one more civil servant instead.’

Te party-​giver clearly fails to give the PR consultant due recognition respect of the kind stressed by Mason. He is acting on ill-​informed generalized judgements, and he does not give his neighbour a chance to prove him wrong. Perhaps this particular PR consultant is a beacon of honesty and modesty. Yet—​and this is the second point—​this example does not raise the social egalitarian concerns at play in Local Shop. Ideally, we should all give each other full recognition respect of this kind. Such recognition respect

52 In cases where such norms are present, it may sometimes happen that even expressions of entirely ‘innocent’ habits (see fn. 13) can take on an unjust meaning, which entitles the victims of the norms to take ofence. Say, you are nosy and have a penchant to ask everybody who does not have the local accent where they are from. If you ask somebody from a marginalized low-​status group, whose members ofen encounter others communicating to them, in some way or other, that they ‘really do not belong here’, they may be entitled to take ofence, and justifably interpret this as a ‘microaggression’; see McTernan 2018, pp. 278f (the example is taken, with adaptations, from Voigt 2018, p. 450). However, in these cases it is the exclusionary and denigrating social norms—​on which you did not act—​which are the primary problem. You should certainly be careful towards whom you express your nosiness, to avoid unintended ofence. But it would seem strange to maintain that demanding restraint from you contributes anything to combating this primary injustice.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Social Status, Self-Respect, and Opportunity  195 would pave the way for esteeming (or perhaps disesteeming, as the case may be) each other for what we truly are, for our personal attributes and characteristics, not for our group memberships. But such a wide conception of respect for the intrinsic value of persons ofers no viable method for sorting intuitively salient concerns of social equality from mere personal failures in other-​regarding virtues. What really matters, for social equality, are norms of social esteem disadvantaging some groups in the ways outlined in this chapter. Tus, if there were a pervasive norm judging all PR consultants to be dishonest, and penalizing those who sought relations with them by judging them equally dishonest, or complicit in the cause of dishonesty, and if such perceived dishonesty played out so as to deny PR consultants important social opportunities—​say, they never got invited to any parties at all, apart from their own, because of that norm—​then we should react to Party Invitation diferently. But this is just to say that what does the work in the example is the analysis of social esteem norms conducted earlier, not mere individual instances of regarding, and treating, each other as inferior. Social egalitarianism is not a position exhorting people not to make, and express, any negative, group-​based, judgements at all, at least on insufcient information, morally non-​ideal as such a tendency certainly is. Tus, not only does the justice-​based rationale for objecting to inegalitarian status norms capture the most important concerns of social egalitarianism. Pluralist social egalitarianism is in danger of being over-​inclusive, and thus of failing to account for what is specifc to them. And this danger is not exclusive to Mason’s account, but generally is present in views which focus on the goodness of egalitarian relations between individuals, on personal or impersonal grounds: they will have difculty to distinguish properly between the bad of norm-​mediated, collective processes of exclusion, and mere individual moral failures.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

PART II

IMPL IC ATIONS

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

7 Political Equality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

7.1.  Introduction Te task of Part II is to demonstrate that the fundamental requirements of liberal relational egalitarianism developed in Part I  have determinate and plausible implications for core domains of social justice; in particular, its task is to show that these implications amount to demanding far-​reaching forms of equality in these domains, and set stringent limits on any permissible inequalities, overall. Tese domains are those of political rights, distributive requirements pertaining to economic resources and opportunities to attain desirable social positions, and health. So what are the demands of justice-​based relational equality with regard to political equality, and what is their relationship to the demands of social justice regulating social relations in society at large, outside political institutions? On the one hand, it seems clear that political relations are a particularly important building block of relational equality; on the other hand, on a substantive view of social justice, the demands of relational equality cannot be reduced to political equality, however extensively construed. What should we say, for example, about a decision which is very democratically taken—​say, in a society-​wide referendum, enabling all to participate on an equal footing, on the basis of adequate deliberative preparation—​but which is substantively unjust, because it sets up important social relations on an hierarchical footing? Tink, for instance, about a referendum about marriage law whose result, for heterosexual marriages, is that men get considerably more rights than women over the shape of their marriage. We will certainly be unhappy with the result; but should we therefore constrain democratic decision-​making in a way that would avoid such results in the future? Tis is a question afer the priority of political equality. Tis chapter examines the grounds for, and the shape of, requirements of political equality within liberal relational egalitarianism, and the question of which kind of priority, if any, they should have over other requirements of social justice. It argues that there is a plurality of rationales for a demandingly egalitarian Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0007

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

200 Implications conception of democratic rights, akin in shape to Rawls’s conception of the ‘fair value’ of political liberties, interpreted as ‘fair equality of opportunity’ to participate efectively in politics.1 However, the case for it also depends on its conduciveness to producing substantively just results. Terefore, egalitarian political rights only enjoy a qualifed kind of priority; they are, in principle, open to some restriction. Te chapter goes on to develop a set of criteria for such restrictions, and shows that some, inherently restricted, deviations from political equality through special institutions, such as representation and constitutional courts, will sometimes be required. Te introduction of ground-​level political inequality among ordinary citizens is, against that, unlikely to be able to ever meet those criteria in salient real-​world circumstances, because alternative egalitarian strategies for improving the quality of decision-​making will regularly be available. Tus, any restrictions through special institutions need to be tightly circumscribed, and the demand for ground-​level political equality and its priority are, in practice, secure. Te chapter is structured as follows: section 7.2 takes on two arguments for a demandingly egalitarian conception of political rights with priority over other requirements of justice, and shows that they are either unavailable to the liberal relational egalitarian, or unsuccessful on their own terms. Te frst is the proposal of defning non-​domination by recourse to egalitarian democratic procedures (­chapter 3, 3.4.2). Tis is unavailable because liberal relational egalitarianism relies on a substantive conception of non-​domination. Te second draws, in an ecumenical spirit, on the prevailing practice of theorizing social justice, which rests on submitting candidate conceptions of social justice to democratic approval, and argues that this commits participants in it to giving democratic procedures priority over substantive requirements. Tis proposal fails in its ambition to establish a robust and far-​reaching priority of democracy, and, to the extent that it is correct, underdetermines the desirable shape of political rights. Te questions of their appropriate shape and priority thus need to be answered from within liberal relational egalitarianism. On this basis, the next two sections deliver the main positive argument of the chapter. Section 7.3 investigates the grounding of requirements of political equality in expressing equal respect for people’s sense of justice and in equal non-​domination, and the resulting general shape of these requirements. Section 7.4 then develops a set of explicit criteria for policies 1 Rawls 1999a, pp. 197–​199; 2001, p. 149.

Political Equality  201 and institutions restricting political equality, applies these to salient real-​ world circumstances, and demonstrates how ground-​level political inequality, in particular, can hardly ever be justifed by appeal to them. Section 7.5 concludes by analysing and rejecting Dworkin’s account of political rights. Tis account represents the most worked-​out and ambitious attempt within the camp of primarily distribution-​oriented theories of social justice to account for the importance of political equality, so it is instructive to see how it falls short: it delivers an under-​demanding account of political equality which fails to express appropriate respect for political agency. Tis strengthens the diagnosis of the shortcomings of primarily distribution-​ centred conceptions delivered in earlier chapters. Liberal relational egalitarianism thus yields a nuanced case for a demandingly egalitarian conception of political rights with a qualifed priority over other requirements of social justice; and this is an important advantage over both purely recipient-​oriented theories, which neglect the importance of the former, and over ‘purer’ theories of political equality, which neglect the importance of the latter.

7.2.  Two False Starts

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

7.2.1.  Non-​Domination as Democracy One attempt to justify a robustly egalitarian conception of political rights with priority over other requirements of justice argues that any inequality over power in decisions over the shape of social cooperation among free and equal members of society constitutes domination. Some can impose forms of cooperation on others, who lack symmetrical power. It is easy to see how this argument leads to a requirement of equal substantive opportunity to infuence political decisions: there is no reason to focus only on formally granted powers, such as equal fnal votes, and not to take informal powers into account, based on the possession of superior economic, or other social and cultural resources, such as elevated social status. Tese can all be used to arrive at, and maintain, dominatory positions. We have already encountered this argument: it is that of ‘non-​domination as democracy’ (­chapter 3, 3.4.2). According to this position, non-​domination is both the paramount, or even exclusive, requirement of social justice, and is constituted by democratic procedures. It is diferent from liberal relational egalitarianism:  the liberal, justice-​based conception of domination has a

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

202 Implications substantive element. On this conception, non-​domination is present when decision procedures reliably constrain decision-​makers to honour the claims of justice of those who are afected, and claims of justice are to be justifed by recourse to the background ideas of freedom and equality of co-​operators—​ their moral powers—​and fair cooperation. Now, where such claims are not fnal—​where they are prima facie, and subject to reasonable contestability—​this conception does overlap with ‘non-​ domination as democracy’ (­chapter 3, 3.3.2). Withholding equal participation rights here would amount to some simply substituting their judgement about a contestable matter for that of others; and that would be arbitrary. But this overlap between the two conceptions is far from complete. Tere are substantive requirements of non-​domination in general social life, determining its range and intensity, for example over the basic liberties, and applying to matters such as the structure of marriages, too (4.3 in ­chapter 4). It is not clear which complaint of domination a person might have who does not enjoy equal political decision-​making power over these, as long as her claims are indeed reliably correctly identifed and reliably protected over time within an alternative procedure—​should that be available.2 Any domination complaint here must be based on that other procedure not actually, as matter of fact, scoring as well on reliable identifcation and protection as a more democratic procedure, or exposing the persons concerned to domination in other, perhaps more indirect, ways. Tis is not to say that nothing else speaks in favour of political equality in such cases, but only that the possible grounds of such a complaint, and the resulting shape that political rights should take, need to be investigated further. Tis is also because non-​domination, albeit being of elevated importance, is not the only demand of relational egalitarian social justice. Other requirements of social justice might sometimes confict with political equality; or they might speak for more political equality, or its priority.3 Under liberal relational egalitarianism, the case for a demandingly egalitarian

2 Tis is, in turn, needs to be distinguished from the case in which political institutions incidentally make the right decision about the regulation of a social relation, without granting proper participation opportunities which increase the reliability of the procedure (say, the ‘experts’ just happened to have a good day); see ­chapter 3, fn. 44. Even if that decision is then kept in place robustly, there was political domination in its genesis, and there continues to be, if subsequent contestation opportunities are inadequate. 3 Tat is the case, as we will see, for the requirement of expressing equal respect for people’s sense of justice (7.3.1).

Political Equality  203 conception of political rights and their priority cannot be as straightforward as it is under ‘non-​domination as democracy’.

7.2.2.  Te Forbearance View: Practically Presupposing the Priority of Democracy

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te ‘forbearance view’ makes a more ecumenical case for substantive political equality and its priority: one that aims to extend to a large part of contemporary conceptions of social justice, including those primarily oriented towards identifying just distributions of resources, or other justice-​relevant goods, such as Dworkin’s, or Rawls’s. It has been put forward by Philip Pettit,4 and is based on the prevalent practice of theorizing justice itself, in contemporary political philosophy. Contemporary theories of justice are not addressed to a prince enjoying near-​absolute power over society, or a powerful clique, and press for their own implementation by whatever means, including unconstrained force or deceit. Tey present themselves as arguments addressed to an audience of fellow citizens, aimed at rationally convincing that audience. Tis is well expressed by Rawls: Political philosophy can only mean the tradition of political philosophy; and in a democracy this tradition is always the joint work of writers and their readers. Tis work is joint, since it is writers and readers together who produce and cherish works of political philosophy over time and it is always up to voters to decide whether to embody their ideas in basic institutions. [ . . . ] Tus, in a democracy, writers in political philosophy have no more authority than any other citizen, and should claim no more.5

Tus, even those theories who do propose substantive principles and policies for social life, as Rawls’s does, perceive of the theorist as a ‘democratic underlabourer’6 who delivers ammunition for public debate, and does not somehow seek to issue authoritative decrees about the content of social justice. Conceiving of the role of the theorist justice in these terms presupposes 4 Pettit 2015a. 5 Rawls 2007, p. 2. 6 Swif and White 2008, p. 54. For a convincing argument that accepting this role of ‘democratic underlabourer’ does not commit theorists to feeding only proposals into public debate which already cater to the audience’s existing views, see Baderin 2016.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

204 Implications respect for the audience as reasoners and deciders, including respect for their possibly diferent views. Now, Pettit argues, if we spell this out, we come to see that, whatever elaborate conceptions of—​social and political—​justice theorists put forward, they must be committed to a ‘basic, just process of decision-​making’ which should be used for accommodating diferences, including those between different conceptions of justice.7 And if this is the purpose of the process, then it enjoys priority over the substantive proposals between which it is supposed to arbitrate. Why should this basic process be egalitarian? As almost all contemporary theories of justice agree that societies must treat their members, in some recognizable sense, as equals (2.2 in ­chapter 2), this general equality requirement must hold for the process, too: nobody must be excluded from it, and its participants must be treated as equals within it. But that is just to say that the basic process must be recognizably democratic. Furthermore, the argument continues, it is not enough that participants have a chance of expressing democratic approval of proposals under discussion in a one-​of manner—​ sticking to the basic just political process, but only once. Tat might lead to signing away democratic rights, by, say, irrevocably empowering a ‘dynasty [ . . . ] or elite’8 to take political decisions in the future. Te basic process must be robustly in place, stably over time, and not in danger of being removed by anyone. Now, what better way could there be to achieve such security than by giving participants themselves the power to keep the process in place? In short, according to Pettit, the basic process requires not only democratic approval, but democratic control over both the decisions reached and the process used.9 It thus requires a continuous form of equal voice.10 What should we think about this argument? Te forbearance view is correct in its description of most current theoretical practice, in pointing out that this practice commits theorists to according democratic procedures an important role, whatever substantive views they put forward, and in arguing that this is how it should be. However, it does not establish, as it intends to do, that democratic procedures should enjoy a general and unrestricted kind of priority over other requirements of social justice. Nor does it establish which shape egalitarian political rights ought to have (which, admittedly, it does not

7 Pettit 2015a, p. 16. 8 Ibid., p. 27.

9 Ibid., pp. 26f.

10 Ibid., pp. 28, 31.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  205 intend to do). Both questions, of priority and shape, still need to be answered from within such substantive conceptions themselves. Take, frst, the priority claim:  substantive theories of justice, such as Rawls’s, come as packages of requirements and principles. Some elements of these packages will be very controversial, covering areas of substantial disagreement between diferent theories—​such as, in Rawls’s example, the diference principle. Indeed, developing and defending such controversial recommendations is a large part of the point of constructing theories of justice. It is thus unsurprising, and reasonable, that theorists put forward such elements as, at least implicitly, earmarked for democratic debate and approval. But this does not imply, nor does it commit theorists to holding in any other way, that egalitarian democratic procedures enjoy general priority over all substantive requirements of justice. Tere may be, and ofen are, some requirements that a theory regards as of elevated importance, of a kind that might even warrant restricting democratic procedures, were they to lead to their violation. Within Rawls’s theory, that is the case for the equal basic liberties (see 4.3.1 in c­ hapter 4). Rawls’s frst principle of justice, which enjoys priority over other requirements, covers both political liberties and other basic liberties, and accords them the same rank. Tese two elements of the frst principle can confict, without the principle itself giving a clear recommendation on how this confict is to be resolved. Te forbearance argument must claim that such a position is, in efect, incoherent. Te political liberties must have priority, because this is implied by the practice Rawls is engaged in: that of developing a theory addressed to a democratic audience. But this conclusion does not follow. Regarding oneself, generally, as a ‘democratic underlabourer’ does not commit one to accepting the priority of democracy across the board. To be sure, simply insisting that a substantive requirement is so important and uncontroversial that it warrants restrictions of democracy would amount to an attempt at arbitrary self-​arrogation of authority. But the theorist’s confdence in the requirement, and its status, may stem from a diferent procedure. Tis may be one of extensive dialogue with the ‘tradition of political philosophy’ (see the preceding Rawls quote), or with other contemporary political philosophers, or with other kinds of experts, in diferent disciplines, such as lawyers and constitutionalists. Or it may be one of extensive dialogue with the group of those who are regularly victims, or most at risk of being victims, of a kind of injustice which the theorist in question comes, through such dialogue, to regard as extremely severe, such as racial discrimination. More dialogue

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

206 Implications with this group comes at the expense of equal dialogue with other members of society, which would seem to be the fully democratic option. Or it might be any combination of these procedures. If that is so, it is not incoherent to argue that some procedure other than full democracy can then also to be used to safeguard that requirement. Or, at least, that it has to be combined with democracy—​which is enough to deny its priority. Second, regarding the shape of egalitarian political rights, political philosophers want to have their substantive proposals, or at least many of them (as just noted), tested and negotiated in a democratic setting, or at least should want this to happen—​but only in one which sufciently corresponds to the fundamental requirements and background commitments of their own theories. Even if theorists agree that a ‘basic just political process’ should be about delivering the best possible compromise between diferent views about justice, and that it requires a kind of substantive equality of opportunity for political infuence among all, this underdetermines the shape that the resulting political rights should take. Tat shape depends on the views about justice in question. For example, a liberal neutralist holding that democracy should be about deliberation about the fair terms of cooperation between free and equal cooperators will, and is within her rights to, reject a fundamentally antagonistic version of equality of political opportunity which allows participants in politics to push for the imposition of their own preferred conception of the good on others, with no regard for its justifability to them. Agreement on some features of a basic just political process is insufcient for determining a compromise conception of political equality in such cases of disagreement, because its shape cannot be deduced from the principle of moral equality and the requirement that society treat all as equals. Te ‘basic just political process’ only denotes some constraints that all conceptions of democracy have to respect. Tis much seems accepted by the forbearance view itself, so this observation is, unlike the preceding one about priority, not an objection to it. It merely demonstrates that the question of the appropriate shape of equal political rights admits of no clear-​cut ecumenical answer, either, and equally has to be answered from within diferent substantive theories themselves. Could the priority of political equality, or a more concrete conception of it, come out of the requirement that a basic just political process be robust, on which Pettit rightly insists? Unfortunately, that is not the case, either. If the preceding argument against the principled priority of democracy on practice-​based grounds of forbearance is sound, then, at least for some

Political Equality  207

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

substantive requirements of justice, instituting a special body with more power to decide on them is not ruled out in principle. Tat could be, for example, a constitutional court whose members make decisions about substantive justice—​such as on marriage law.11 And if that is so, why should it be the case that rendering such a body sufciently robust has to be achieved by way of all contributing equally, or having the opportunity to contribute equally, to its being upheld?12 Perhaps other strategies of achieving sufcient robustness are available—​perhaps they could even institute more robustness. What is needed is either an additional empirical argument that substantive political equality is uniquely well-​placed to deliver such robustness, or a normative argument that something is wrong with non-​egalitarian robustness. But that latter argument is likely already to object to substantive review by a constitutional court, in the frst place; and, if the preceding argument is sound, it is not available on a truly ecumenical basis.13 It has to rely on particular conceptions of political equality and its role. For example, under ‘non-​domination as democracy’, the conclusion that constitutional judges dominate us when they decide on substantive matters will be difcult to escape. None of this establishes that there are no stronger reasons for the priority of political equality than the forbearance view delivers, when put in its proper place. It only demonstrates that such reasons, as well as the resources needed for properly delineating the shape of equal political rights, have to come from within particular views and conceptions of justice and political equality. So let us now turn to identifying these within liberal relational egalitarianism.

11 Pettit does not pronounce on whether the forbearance argument is supposed to rule out constitutional courts with strong substantive reviewing powers, but a successful argument for the full priority of democracy would seem to imply that such substantive review is, at least, very problematic (while merely entrusting the court with invigilating powers over democratic procedures need not be). 12 In this case, decision-​making power over substantive matters is still, in the frst instance, unequal—​but the equal power of all to block deviations from that unequal procedure increases political equality. 13 In fact, Pettit only ofers, within the forbearance argument, a version of his ‘eyeball test’ for the required degree of robustness, and that test is not strictly egalitarian: a basic just political process is sufciently robust if it allows all under its power to look each other in the eye ‘without reason for fear or deference’; Pettit 2015a, p. 27. We have already seen (3.4.3 in c­ hapter 3) that the eyeball test is, in principle, compatible with quite a lot of power inequality. In the case of functioning constitutional courts, with a good track record, constrained by the constitution and principles of legal reasoning, and populated by well-​trained, conscientious, and experienced professionals, it seems plausible that citizens should have no fear of, but trust in, the judges. From within his own, less ecumenical, conception of non-​domination, Pettit employs the more strictly egalitarian ‘tough luck test’ for political justice; ibid., p. 29, fn. 17; see c­ hapter 3 in this book, fn. 58.

208 Implications

7.3.  Te Grounds and Shape of Political Equality 7.3.1.  Grounds

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te primary reasons for a demanding conception of political equality with some priority over other requirements of social justice are two: the frst is that egalitarian political procedures express equal and full respect for people’s frst moral power: their sense of justice. And the second is that, in many instances at least, they are the primary vehicles for instituting non-​domination.14 To the frst reason, then:  equal political rights make a unique, non-​ substitutable contribution to appropriately respecting persons as bearers of a sense of justice. Possession of such a sense of justice is one of the grounds of basic freedom and equality in social cooperation (3.2 in c­ hapter 3)—​in the sense that agents of justice have to treat cooperators as if they were equal in this dimension, and have to abstain from inquiring into, and acting on, diferences in such a possession above a threshold of moral competence (4.3.2 in ­chapter  4). What is special about political rights, compared to others, is precisely that cooperators making use of such rights can be recognized by others as having and exercising a sense of justice, not just as moral agents with their own views on how to lead their own lives.15 As Joshua Cohen puts it, [o]‌ther rights—​to liberty of conscience, or personal property, or privacy—​ do not have the same connection with the development and exercise of the sense of justice [ . . . ], nor is there the same showing of respect in acknowledging a right to sovereignty over one’s own afairs as there is in acknowledging that right as well as a right to an equal share of sovereignty in determining the conduct of common afairs.16

14 A third reason is equal social status (­chapter 6)—​being esteemed equally as (potential) political contributors. However, the frst two reasons—​basic respect, and non-​domination—​yield more stringent requirements (­chapter 4, 4.3; ­chapter 6, 6.4); and social status based on rights to political participation is empirically likely to move in lockstep with them. 15 For discussion of the sense of justice, see Rawls 1999, ch. 8, and, for later development, see Rawls 1996, pp. 19 and 77f. Te argument of this chapter is compatible with both these accounts. 16 Cohen 2002, p. 110. See Rawls 1996, p. 319: ‘By publicly afrming the basic liberties citizens in a well-​ordered society express their mutual respect for one another as reasonable and trustworthy’; the context of the quote makes clear that it refers to the political liberties in particular. Daniels highlights this ‘recognitional component’ of equal citizenship as a distinct egalitarian rationale of Rawls’s theory; Daniels 2002, p. 247; see also Gonzalez-​Ricoy and Queralt 2018, pp. 636, 638.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  209 Respect for people’s sense of justice is, of course, called for in other kinds of interactions and social relations, too. It is at stake in all social relations in which people are called upon to identify, and abide by, demands of justice, including personal relations. Such respect generally requires that people must, in the frst instance, be free to willingly comply with such demands, and that coercion or other sanctions must not be immediately and indiscriminately used, but only if there is good reason to hold that compliance will not be forthcoming. But respect for the sense of justice of participants in social cooperation cannot be full without giving them a share in collective power in societal decision-​making. It is a special kind of respect, tied to the fact that people can, and ought to, exercise their autonomy not just in self-​ regarding matters, but also through identifying and respecting the claims of others, in dialogue and debate with them. And societal decision-​making about common afairs is a privileged locus for putting it into action, because it settles the terms of social cooperation. For the same reason, it would also be misguided to regard equal political rights as merely making a very important contribution to communicating, across society, that people are, in a general sense, social equals. Tat could, in principle, also be done in other ways. We could, for example, write into all offcial documents that people are equals; or ofcials could be obligated to say it all the time. What communicates such equal standing best—​what the most efective symbols of social equality are—​is subject to some degree of historical and social contingency.17 But the argument is not about the general symbolic dimensions of political rights. It is about expressing proper respect for people’s sense of justice. We cannot express proper, and full, respect for each other’s sense of justice other than by granting each of us, reciprocally, the guaranteed, robust freedom to make use of it in those activities that matter most, from the point of view of justice.18 Tis is not contingent. And if people

17 Wall 2007, pp. 431f. 18 At this point, one may be tempted to question that there is a relevant diference between this argument and that of the forbearance view (7.2.2), that theorizing about justice aimed at democratic approval presupposes respect for people’s views about justice. However, the diference that the latter is based on the practice of theorizing, while the former is based on the content of the theory, is salient, because that content directly requires equal respect. Take a theory that, in line with the practice view, does aim at approval by a democratic audience, but also allows for making public judgements about diferential moral competence, requires diferential voting rights on this basis, and argues that this is consistent with regarding and treating people relevantly as equals overall. J. S. Mill’s view is of this kind; see Mill 1861, c­ hapter 8.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

210 Implications are equal in their possession of a sense of justice, those rights ought, in the frst instance, to be equal.19 However, the fip side of this of this argument is that decision-​making procedures can only lay claim to instantiating proper respect for people’s sense of justice if following them in fact reliably leads to the identifcation, and implementation, of just outcomes—​to fair adjudications of conficting claims. Granting people equal political opportunities expresses respect for political agency, in a general sense, but, as David Estlund puts it, ‘[t]‌he question is whether this strikes us as any valuable form of respect even if we purge the scene of all traces of epistemic value’.20 Conversely, expressing respect for people’s sense of justice is valuable, but instituting epistemically inapt procedures, which do not tend to correctly identify demands of justice, cannot be an instantiation of such respect.21 Furthermore, there is no other kind of respect which social cooperators rightly interested in fnding fair terms of cooperation should accord to a political procedure that is not sufciently conducive to this enterprise. On liberal relational egalitarianism, as a justice-​based view, there are no other grounds for demanding an egalitarian political procedure than its being required by respect for people’s sense of justice (if it is), or by other requirements of justice, such as non-​domination (see later discussion). If I am denied equality in political decision-​making because, demonstrably, there really is no other way of achieving just outcomes, I do not have a complaint based on my sense of justice. My claim to being able to express my sense of justice is one of being able to contribute to fnding and implementing justice for my fellows and myself, because this is what it obligates me to do, not of enjoying the greatest possible sway for my views about justice, whatever the ensuing result.22

19 Terefore, the symbolic signifcance of political rights supervenes on their non-​contingently expressing respect for some of our most important moral abilities, whether or not it rests on other, more contingent factors, too—​such as that we are arguably yet to see a society in which a politically privileged class does not use its superior power to secure (other) unfair advantages. 20 Estlund 2007, p. 95. On respect and deliberative democracy, see Gutmann and Tompson 2004, p. 22: ‘If deliberation tended to produce worse decisions than other processes in the long run, then it would not serve [an] expressive purpose. A process that generally produced bad outcomes would hardly express mutual respect.’ 21 Gutmann 2002, p. 179, discusses both the ‘expressive purpose of political freedoms’ to publicly signify equal standing within Rawls’s account of democracy, and their role as instrumental to justice, but does not connect these two dimensions. 22 Nor does a restriction of political equality necessarily imply that anybody’s sense of justice is viewed as inferior to that of anybody else. Tat depends on the reasons underlying the restriction; see 7.4. For a similar view—​that respect requires presumptively equal political rights because these enable people to discharge equally held duties of justice—​see Howard 2019.

Political Equality  211

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

In particular, the argument that power should be shared equally in decision-​making because, even if it should reliably lead to worse outcomes in terms of justice than some other procedure, it still constitutes a valuable way of living together as equals, is not available. It is only available to pluralist social egalitarians; and among these, only to those who accept a form of perfectionism. We have seen that commitment to perfectionism is not a necessary component of pluralist social egalitarianism (­chapter 5, 5.3.3). Plausible forms of PSE argue that egalitarian relations beyond justice institute distinct kinds of value, especially of an impersonal kind; but that can be argued without accepting that state power should be used to promote, or instantiate, them. Democratic pluralist social egalitarianism is, against that, necessarily perfectionist: institutions of political decision-​making issuing authoritative decrees, backed by the threat of coercion, are the paradigmatic form of state power. 23 So, respect for the sense of justice requires equal political rights if these are generally apt to correctly identify requirements of justice. Tey have to be instituted within a kind of procedure that is properly oriented towards identifying them—​not at, for example, satisfying the social preferences of as many people as possible. So, on views issuing substantive requirements of justice, such procedures must reliably respect and implement, and not violate, these. However, while this is a conditional argument, it is not a purely instrumental one. It is not indiferent between diferent ways of achieving equally just outcomes, but requires, on grounds of respect, selecting the 23 For such a pluralist argument that the value of relating as equals within democratic institutions is modeled on the kind of equality desirable for personal relationships such as friendship, and has to be traded of against the value of justice, where they come apart, see Viehof 2014, especially pp. 357 fn. 33, 361, 372 (however, for skepticism that reasons for power equality within such more intimate relations generalize to the case of ensuring overall societal equality of status, see Viehof 2019). Wilson similarly seems to endorse a pluralist position: he argues that political equality is required on grounds of ‘civic friendship’; 2019, p. 60; whose goods cannot be fully enjoyed if parties to the relation instrumentally focus on the results produced by it; p. 63; and which may sometimes have to be traded of against ‘other considerations of justice’; p. 49. However, he maintains not only, as the quote illustrates, that political equality is itself a requirement of justice, but also accepts that justice may be egalitarian overall and that political equality is one kind of equality among others required by it; p. 285. Perhaps conficts between diferent demands of that overarching ideal can then be internally resolved, stopping short of full pluralism. Important diferences to the liberal position, which militates against enshrining, within coercive political institutions, any result-​independent conception of civic friendship based on the relational goods it yields, remain, either way. Against that, the liberal view has no objection to people pursuing forms of democratic relations at sub-​state level which reliably and continuously come at the cost of other goods to which they have claims. Tey are free to waive these. For example, joining a Kibbutz might require forgoing recourse to external adjudicatory institutions, such as courts, in case of conficts with other members, because this may be irreconcilable with the particular Kibbutz ethos of egalitarian, non-​litigious self-​ sufciency. Tis is acceptable, as long as it does not violate basic liberties, and is revocable (4.3.1).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

212 Implications most inclusive and egalitarian procedure within the set of similarly well-​ performing procedures, because this is the one that ofers the fullest opportunities for everybody’s sense of justice to be heard, and to make a diference. If that is so, then the case for inclusivity and equality may still survive some divergence from substantive justice, afer all, at least for some time; and it will also dictate the manner in which such divergence is to be remedied. Te question as to the priority of equal political rights thus asks, on this view, how much divergence is allowed, for how long, and what shape remedies must take (7.4). Te second rationale, non-​domination, directly requires, as seen, demandingly egalitarian political rights to determine outcomes where claims of justice are reasonably contestable. Giving some greater power to determine the adjudication of such contested claims would be arbitrary. Furthermore, on a liberal view, justice will leave space for personal and collective projects to be furthered by means of democratic legislation, as long these stay within the bounds of its requirements. Such projects may be those of aiming at forms of social egalitarian relations which are not themselves required by justice (5.3.3 in ­chapter 5), but are permitted by it, and can be efectively furthered by law. For example, a society might want to exercise some centralized infuence in order to achieve the collectively most desired balance between optional social services—​such as ofering leisure activities for pensioners—​being provided through community volunteering, or through commercial, market-​ based supply. Te former alternative could greatly beneft from legislation giving, for example, generous tax-​fnanced subsidies, ofering locations and materials for free. Another important case is the familiar one of allocating funding between diferent cultural and sporting activities—​subsidies for the opera, or for football pitches. Such democratic freedom is not only consistent with, but itself required by liberal neutrality, properly understood. Ruling that those with projects beyond justice which necessitate collective activity and coordination can never enlist democratic means in their support would arbitrarily favour individualist conceptions of the good.24 And if these projects are indeed amenable to democratic pursuit, then everybody should have an equal opportunity to decide which ones to pursue, and how. Strictly 24 Rawls initially argued that adherents of such conceptions of the good must not use the legislative branch of government (even though an ‘exchange branch’ may help them to pool their resources); 1999a, pp. 248–​251, 291f. Gutmann 2002, pp. 193–​195, convincingly argues that this individualist restriction is no longer present in Political Liberalism (Rawls 1996).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  213 speaking, being denied equal opportunity to decide on such projects then constitutes domination, too—​even if justice itself has nothing to say about which projects should be pursued, and in which manner. However, as also seen, there is a third area, where substantive requirements of social justice are not reasonably contestable, such as requirements of maximally intense non-​domination over the basic liberties (4.3.1 in ­chapter 4). Tat is, they are not contestable in the sense that rejecting them also commits one to rejecting the framework on which they are based, because these are the requirements most tightly connected to the background conceptions of persons and their moral powers, and fair cooperation. Over this area it is possible that some procedure restricting equal political opportunities scores best. If it does so robustly, then it does not necessarily constitute political domination. It could, however, still be contingently dominatory, depending on how exactly it is structured and circumscribed. For example, even if a constitutional court with substantive reviewing powers brings about more substantive justice on the issues it is called upon to decide, perhaps judges get too much, or too indiscriminate, social status, which turns them into dominators on afairs not falling within their ambit (6.3.1 in c­ hapter 6). Just as in the case of democratic rights based on the sense of justice, the conclusion is thus that an egalitarian procedure for correctly identifying what non-​domination requires—​what exactly protection has to range over, and how intensive it has to be—​would be best, because it would ward of the risk of some such accompanying domination even in otherwise very reliable procedures. But we cannot conclude that the case for a demanding kind of political equality and its priority is always safe, and admits of no exceptions and limitations. Before inquiring into these, however, it is important to specify the content of equal political rights within liberal relational egalitarianism further.

7.3.2.  Shape Tis content is indicated by the respect-​and non-​ domination-​ based rationales for political equality. It is akin to Rawlsian ‘fair equality of political opportunity’. Te task of this subsection is to delineate it by distinguishing it from fully equal political power—​which, in line with c­ hapter 3 (3.2, 3.4.2), is to be understood as diachronic full equality of opportunity of infuence over the terms of social cooperation. On this basis, the next section will develop

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

214 Implications criteria for tackling specifc, and more problematic, cases of limitations and exceptions. As seen, only political equality in procedures which are both reasonably efective at regulating social cooperation and reasonably directed at justice expresses respect for participants’ sense of justice, and will reliably identify their claims of justice and the required level of protection against domination over them. Terefore, inequalities of political power which are necessitated by these aims are acceptable, in principle, and need to be taken into account already at the level of identifying the conception of equal political opportunity we have, in general, reason to value. Tere are two kinds of such inequalities. Te frst is inequality of power based on unequal capacity, or talent, for making good arguments which are capable of convincing one’s fellow co-​ deciders to adopt courses of actions conducive to justice.25 As long as all have equal opportunity to bring their views and opinions into the political process, the fact that some of these views attract more support than others at the voting stage cannot constitute a salient form of political inequality, if such support is based on good reasons.26 Tis point has become a staple in the literature on democratic theory. However, there is a further set of connected inequalities of political power which cannot be regarded as unjust, either, because they are necessitated by functional requirements on democratic procedures. Tis is the case at least in the circumstances of reasonably complex societies at an advanced level of development, characterized by an intricate division of labour. Tose are the circumstances we are interested in, not because of reasons of mere practicality, but because only these circumstances yield the rich array of life options which liberal egalitarians want (­chapter 3, 3.2). Limiting the permissible and required inequalities falling under this category entirely to the case of capacity for good argumentation likens democratic decision-​making too much

25 Brighouse 1996, p. 125. Te locus classicus is Habermas’s characterization of the ideal deliberative situation as one in which ‘no force but that of the better argument is exercised’; Habermas 1975, p. 108. 26 It is important to note that orientation of democratic procedures towards good arguments does not imply that they have to require entirely impartial reasoning about justice by all participants. Deliberative procedures are compatible with voicing self-​interested considerations, as long as these are constrained by a commitment to reciprocity among equals, and mutual justifcation; see Mansbridge et al. 2010, pp. 77f. Furthermore, as long as this commitment is fulflled, such procedures need not require particularly high levels of civility and argumentative sophistication, either—​which would give rise to worries of excluding the disadvantaged, as they are more likely to lack especially the latter; see Young 1996 and Sanders 1997.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  215 to philosophical discussions in the seminar room. Unlike these, democratic procedures have to issue in binding decisions which must be reached within reasonable time to solve existing social problems.27 Due to such unavoidable time constraints, participants in these procedures must ofen rely on a host of epistemic shortcuts in evaluating which political proposals merit support. Some of these shortcuts will be connected to the personal qualities of the people who advocate them, such as their political experience—​demonstrated, for example, in previous high-​quality contributions—​their reputation, in terms of intellectual capacities as well as integrity and honesty, and their education and credentials as an expert on certain specifc matters. Reliance on such shortcuts cannot be regarded as, in principle, giving rise to unjust inequalities in political power. Some degree of such reliance is unavoidable, and unexceptionable, if two conditions are fulflled. Te frst is that there is reason to trust that the general institutional context within which they take place functions properly—​that it, in fact, has a good track record at regulating societal cooperation justly. Te second is that the performance of those individuals who beneft from these shortcuts, in terms of power, is also individually checkable and subject to control by others, so that they can, and regularly do, lose their extra infuence if their record deteriorates. If expertise and experience can no longer be efectively challenged on any given matter, the experienced and the experts have turned into dominators. However, crucially, even if inequalities in political power connected to the proven capacity to make good contributions, interpreted so as to allow reliance on the aforementioned epistemic shortcuts in assessing contributors’ credentials, are not themselves unjust, concern for basic freedom and equality yields compensatory requirements. It must be made sure that the background social processes through which some people come to enjoy such superior political talents, skills, and experience to start with, are tightly controlled, and subject to a demanding conception of equality of opportunity. Tis conception has to have a particular focus on education, as well as on the entry conditions to jobs and activities in which political skill is developed. For example, internships with political parties and institutions may be one important such entry route. If these are unpaid, they will be accessible mainly

27 Tis is not to deny that it is a virtue of democracy that policies can be debated forever, but only to stress that at least some decisions will have to be taken in the meantime.

216 Implications to those who come from a relatively privileged background, which is able to compensate for such lack of pay. Te second set of generally permissible restrictions of equal political power concerns the formal mechanisms instituting a deliberative division of labour. We have seen that a complex social division of labour necessitates a political division of labour with some stable ofces, which not all will occupy to an equal extent over the course of their lives, or will have an equal chance of occupying, at all points in their lives (3.4.2). Such a division of labour needs to encompass not only the execution of political decisions, but also the deliberation leading up to them. Political representation is a key example of it.28 In contemporary, complex societies, ordinary citizens do not have the time to acquire, and process, all information necessary to decide on all issues in detail. Tis grounds a case for a deliberative division of labour between citizens and representatives. Te former ought to decide on the general agenda—​the issues whose resolution justice requires, in their view, including relations of temporal priority among them—​and to give indications in which way these issues ought to be resolved, while representatives are charged with formulating informed policy agendas for their concrete resolution. Te point is well put by Christiano:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Without the division of labor that the representative system supplies, citizens would be capable of doing a lot less than they would with such a division of labor. Hence, citizens would have much less power, and thus power would inevitably be transferred to other sectors of society.29

Allowing decision-​making power to be ceded to such ‘other sectors of society’—​for example, the corporate economy—​would leave citizens open to greater efective inequality of power, and domination, than under the

28 Tere are mechanisms other than representation to achieve a deliberative division of labour, such as the institution of randomly selected deliberative mini-​publics for specifc issues, and these may sometimes be epistemically superior; see Benson 2018, chs. 2, 3. Tis subsection focuses on representation as the practically most important and widespread such mechanism, but its argument applies, mutatis mutandis, to other mechanisms appointing special deliberators, too. 29 Christiano 2008. Brighouse 1997, p. 168, argues that representation might in fact serve equal availability of political infuence because it improves accountability: in large-​scale direct democracy, it is hard to monitor whether, and how many, decision-​makers act on purely selfsh reasons, or on the basis of afective ties, whereas public accountability mechanisms make the scrutiny of representatives’ behaviour easier. While this is plausible, it leaves open the possibility that direct democracies could develop a decentralized system of accountability, through strong social norms efectively combating selfshness and favouritism. Te argument based on enhanced overall power strikes deeper.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  217 representative system.30 Accordingly, since there is arguably no democratic alternative without special deliberators that would be comparably efective at regulating all societal matters that are in need of legislation, the proper conception of political equality must make room for such a mechanism. Tis is not to say that representation (or its alternative) is not actually itself a deviation from full equality of power over the terms of cooperation. Tat would arguably only be the case if representatives were mere delegates, whose mandate for decisions is fully circumscribed by their constituents.31 For example, they could be merely ‘vessels’ for aggregating and sorting their constituents’ views about justice, and executing the result, with the form and timing of such execution entirely determined by their mandate. But that is not what representation is needed for, if the preceding argument is correct. It is needed for processing complex information, and taking decisions on its basis, in response to pressing social problems. Even if citizens set the agenda, as they should, representatives will inevitably enjoy some discretion in how they go about fulflling it. Even if they always remain accountable to their constituents, which they should, the latter will not be able to challenge each and every judgement made in pursuit of their task—​in particular if it is made in collective deliberation and negotiation with other representatives, so that the resulting group judgement may not be reducible to the individual judgements of its members (2.6.1 in c­ hapter 2). If that is so, they have more power. And, in its basic shape, that problem will be the same for any kind of special deliberator, representative or not. Accordingly, while it is, in principle, unexceptionable that there be representation (or some functionally equivalent division of deliberative labour), the focus of our conception of political equality has to be on adequately circumscribing the powers of representative ofce, in terms of time and content, and accompanying it with instruments of direct democracy. And it has to be, once more, on equality of opportunity to attain such ofce. So, liberal relational egalitarianism has to require a demandingly egalitarian conception of political opportunity, such as Rawls’s conception of the ‘fair value’ of political liberties.32 According to this conception, 30 Even if, implausibly, power was not transferred to other sectors of society, the alternative would be not to have efective regulation of terms of cooperation, with accordingly random outcomes—​ this would be preferable in terms of avoiding, or reducing, domination, but would still frustrate cooperators’ fundamental interest in efective fair terms. 31 See Kolodny 2014, pp. 317f, who holds that a reasonably full such circumscription is possible for political representation without unduly endangering its function. 32 Rawls 1999a, pp. 197–​199.

218 Implications the worth of the political liberties to all citizens, whatever their economic or social position, must be sufciently equal in the sense that all have a fair opportunity to hold public ofce and to afect the outcome of elections, and the like. Tis idea of fair opportunity parallels that of fair equality of opportunity in the second principle.33

Tis is a stringent requirement, because of the importance of expressing respect for people’s frst moral power, and because non-​domination is a particularly weighty concern. For example, it enjoys priority over ‘simple’ distributive requirements connected to fair cooperation (­chapter 8, 8.3, 8.5), or requirements to prevent the problematic forms of inequality of social status analysed in ­chapter 6 (even if, in practice, it is very unlikely to come into confict with these, anyway).

7.4.  Restricting Political Equality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

7.4.1. Te Criteria Equal respect for everybody’s sense of justice and non-​domination thus ground a case for the fair value of political liberties, understood as fair equality of political opportunity along the lines laid out earlier. But this is only a rough and general conception of such equality of opportunity. We still need to specify the criteria that govern the permissibility of any specifc inequality of political power further, in order to be able to adequately address problematic cases, and select between diferent concrete policies and institutional set-​ups, fxing, for example, the exact balance between representative and direct democratic elements. Tese criteria are two—​one epistemic, one normative. Te epistemic criterion demands that a restriction of political equality—​ such as the institution of a constitutional court with reviewing powers over

33 Rawls, 2001, p. 149. See also Rawls 1999a, p. 199. Te general idea of fair equality of opportunity is that persons of equal initial talent, and of the same willingness to use it, have equal opportunity to attain desirable social positions, regardless of their class background (see c­ hapter 2, 2.5.1, and ­chapter 8, 8.5). However, there are important diferences between general fair equality of opportunity and its political variant. Te latter must require a sizeable level of diachronic opportunity for political infuence for all, while the former can be less diachronic, as long as selection for positions is fair and all have, at a relevant point in time, substantive opportunities to acquire the required skills; Scanlon 2018, pp. 77f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  219 substantive matters of justice, or specifc deviations from ground-​level equal political opportunity—​(a) demonstrably leads, over time, to better results in terms of substantive justice, such as the safeguarding of liberal basic rights, and (b) that alternative egalitarian measures to improve the substantive quality of decisions, such as measures at generally enhancing the political competence of ordinary citizens, are highly unlikely to achieve similar results. Tis high threshold of evidence is required by the expressive rationale of respect—​it demands erring on the side of trust in ordinary citizens and their sense of justice, where no such evidence exists. In particular, one-​of bad decisions by the demos do not already constitute such evidence, and do not warrant, for example, moving away from more direct democratic instruments, such as referenda, to stronger decision rights for representatives, or judges, or any other special deliberator. Just as in a society transitioning to representative democracy, representatives frst have to learn their trade, so a demos not used to referenda has to learn, collectively, how to make proper use of them. Tis takes time.34 Te normative criterion applies to questions of transition from a substantively defective political system to a superior one. If a less egalitarian system fulfls the epistemic criterion, we should implement it; if it does not, we should not. But a society might face a decision between two prospectively equally well-​performing decision-​making systems, one more egalitarian than the other, where it is likely that the less egalitarian one—​such as a constitutional court with substantive reviewing powers—​would get to better substantive results more quickly. In this case, the normative criterion demands accepting some substantive injustice for some time while moving to the egalitarian system. Te reason is simply, again, that the latter will more fully instantiate equal respect in the future, and this licences accepting some costs now. Tis is acceptable, because it is not a gamble: ex hypothesi, we have sufcient reason to hold that the egalitarian system will eventually perform adequately.35 Tere is no algorithm for calculating how much transitional 34 Dworkin, for example, rejects referenda on the basis of the ‘plausible assumption that elected ofcials, rather than popular assemblies, are better able to protect individual rights from dangerous swings in political opinion’; 2011, p. 394. Tis is an example of a kind of reasoning that is insufcient to justify political inequalities. It has to be shown, not assumed, that they are better at it, and that this cannot be changed at reasonable cost in reasonable time. 35 One might think that this criterion treads dangerously closely to the democratic social egalitarian pluralism rejected earlier (7.3.1). Tat impression is mistaken, because there is no case at all for the more egalitarian system if the non-​egalitarian system uniquely fulfls the epistemic criterion, and thus no case for a permanent trade-​of between democratic equality and substantive social justice. Tat possibility of a permanent trade-​of is the relevant feature of pluralism.

220 Implications

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

injustice one should accept in the name of fuller political equality. Tat depends both on its gravity—​e.g., in terms of basic liberties—​and its duration. Nevertheless, the normative criterion indicates the kind of reasoning which should inform reform of the political system. Tere is, of course, always a further normative consideration: any move away from political equality must not itself violate other demands of relational egalitarian social justice, such as non-​domination. Tis will be important particularly for the analysis of possible restrictions of ground-​level equality of political opportunity between ordinary citizens (7.4.3). Taken together, these criteria thus determine which kind of priority political equality enjoys over other requirements of social justice, and demonstrate that it is a form of priority, albeit not an indefeasible one ranging across the board, as it is on ‘non-​domination as democracy’. It is, instead, an exacting presumption. If the argument so far is sound, this is as it should be. Te case for political equality and the criteria for its restriction developed so far do not uniquely determine any concrete conception of democratic decision-​ making. Singling out such a conception, or a set of alternatives, is a task that is beyond the reach of this chapter (and book); it is one for democratic theory. However, the criteria do give good reasons to hold that, generally, the introduction of, or strengthening of existing, special institutions instituting ofces vested with superior power has much better chances of fulflling the presumption than the institution of any ground-​level inequality of political opportunity between non–​ofce holders. Let us discuss these two cases in turn.

7.4.2. Special Institutions Because the case for special institutions is relatively straightforward, we can be brief. First, the performance of special institutions, such as of a constitutional court engaging in substantive review, can be assessed both historically—​how did a given state, or other institutional framework, fare, in terms of social justice, before and afer its introduction?—​and comparatively—​how do different states with and without such institutions fare? Te more clearly defned the mandate of such a special institution is, the easier is such assessment. So, to that extent, there will be a wealth of information which can be run against the epistemic criterion.36 36 Tis is not to say that the assessment is easy, as the criterion requires some counterfactual assessment; but it is easier than in the case of ground-​level inequality.

Political Equality  221 Second, if a special institution reliably does better in terms of identifying and implementing some substantive requirements of justice, and if its mandate is strictly restricted to this area in a way that makes it accountable to the public, so that it can be removed if it ceases to do well, then its members do not dominate ordinary citizens, on a substantive conception of domination. So it will not run up against other requirements of relational egalitarian social justice.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

7.4.3.  Ground-​Level Political Inequality Surviving the presumption for political equality is, against that, much more difcult for measures introducing ground-​level inequality of political opportunity among ordinary citizens. Measures such as plural voting schemes, which ofcially discriminate between citizens based on their supposed unequal moral competence, are out of bounds, because they violate basic respect and thus directly undermine equal standing (4.3.2 in ­chapter 4).37 We could think of a plural voting scheme which is not based on such judgements of competence. It might be based on the possession of expertise and competence in specifc matters, which is, in principle, available to all, just in fact not possessed by all, so that a failure of basic respect is not at stake. But then the question has to be why the rights at issue are not strictly restricted to the area of expertise in question. Why should they be general? Others not so privileged will be entitled to claim, on grounds of non-​domination, that any such special competence be treated as a rationale akin to that needed for special institutions, or ofces: any increased rights ought to be tightly matched to it, and not extend further, and there have to be procedural safeguards making sure that holders of increased rights are indeed constrained to use them in a way that furthers the cause of justice.38 Teir performance has to be

37 More precisely, schemes which discriminate between all those who cross a threshold of basic moral competence are out of bounds. Not enfranchising children, for example, is permissible. 38 For a proposal of ‘voter tests’ checking basic political knowledge, see Moyo 2018. While it is certainly true that better political knowledge enables better judgement, other things being equal, and should be a focus in education, it is likely untrue, and defnitely unproven, that more political knowledge makes all those who have it vote for more just policies across the board. It seems possible that a poor, uneducated person voting for ‘whatever makes people like me better of ’ gets it right at least as ofen as a highly educated person, and it is incumbent on the proponent of plural voting to demonstrate that this is not the case. Furthermore, generally diferentiated voting rights will very likely create additional problems of inequality of esteem-​based social status, see fn. 14 in this chapter.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

222 Implications individually checkable, and remedies, such as loss of these rights, have to be in place if they fail.39 Needed to overcome the presumption in the case of ground-​level political inequality are thus arguments that do not rely on supposed unequal personal competence, basic or not, but nevertheless point to benefcial consequences of allowing some form of political inequality for substantive outcomes. David Estlund has presented an ingenious such argument.40 Estlund’s formulation of it is sophisticated, but the basic thought is fairly simple. It relies on the assumption that increasing input by citizens into political institutions is valuable because it improves the substantive quality of the decisions taken by these in terms of justice: the more input, the better, other things being equal. Leveling down everybody’s input to achieve political equality might not be a good idea, as it might make for input that is too low to achieve qualitatively good decisions. Political equality is also valuable, but we face a trade-​ of between political quality and political equality; sometimes we should encourage unequal political input. Estlund illustrates his case by proposing a voucher scheme for permissible contributions to political campaigns that supposedly improves on a proposal made by Bruce Ackerman, which gives every citizen vouchers of exactly identical value, thus instituting perfect equality:41 under Estlund’s scheme, it is possible to buy more vouchers, but only at a sharply rising marginal price which redistributes the lion’s share of the input gained to others by topping up their vouchers.42 In this way, everybody attains more input. Te argument consciously parallels Rawls’s argument for the diference principle, allowing inequalities of income and wealth as long as they beneft the worst of maximally.43 Tis argument seems to be of the right form to survive the presumption of political equality. However, in substance, it is nevertheless unsuccessful, for two reasons. Te frst is that the argument fails to explore egalitarian avenues for increasing input up to a sufcient volume. Te problem is not

39 Tis is the formal analogue to the informal mechanism for keeping political advantage based on reputation and experience in check (7.3.2), where a decline in performance has to be matched by the informal sanction of losing the ear of other citizens. 40 Estlund 2000. 41 Bruce Ackerman, ‘Crediting the Voters’, http://​prospect.org/​article/​crediting-​voters-​new-​ beginning-​campaign-​fnance, last accessed 17 November 2020. 42 Estlund 2000, pp. 153–​157 (partly reprinted in Estlund 2008, pp. 196f). 43 Estlund 2000, p. 150. Tis scheme might seem not to violate fair equality of political opportunity at all—​if background economic inequalities are small. But the argument, if sound, could also justify some degree of unequal opportunity; and, while equal opportunity permits unequal input, this scheme actively encourages and prizes it.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Political Equality  223 that unequal input will itself lead to epistemic distortions, and thus inferior outcomes in terms of justice, as those with more input can use it merely to advance their own interests. Estlund is sensitive to this problem. He points out that the dimensions of maximizing input, and of its distribution, have to be traded of with each other already within the dimension of epistemic quality itself. While this substantially decreases the degree of inequality that the argument would permit, if successful, Estlund insists that some inequality of input might still lead to better decisions.44 Let us grant this. However, what is more fundamentally problematic about the argument is the underlying assumption that increasing input improves political decisions, at epistemically tolerable levels of inequality of input, in a more or less linear fashion. While it is certainly true that there may be too little input for good decisions—​think of total political apathy among citizens—​from this it does not follow that input can be gainfully increased indefnitely. Indeed, this cannot be the case. Political processes have to lead to binding decisions in real time. Tis implies that there must be a cut-​of point for input, afer which additional input is at best useless, and at worst distracting noise delaying necessary decisions, for example by fooding institutions with data which they cannot process in adequate time. If this is true, it is open to the political egalitarian to contend that, for every inegalitarian proposal to reach a suffcient level of input, there will be a competing egalitarian strategy, such as improving education and motivation to participate in politics among the disadvantaged and less engaged. Te presumption then mandates selecting this policy, because it does better at expressing equal respect for people’s sense of justice. Estlund’s counterargument sufers from the fact that it does not take democracy’s character as a real-​world decision process seriously enough. It remains incumbent on the political inegalitarian to show that her proposal will indeed, in real-​world circumstances, do better in terms of substantive justice than all credible egalitarian proposals. As long as no such concrete case is successfully made, political egalitarians can remain unmoved even if their case for equality is based on democracy’s epistemic qualities for furthering justice in the way the argument of this chapter is. Te second reason is that Estlund’s scheme allows buyers of more political input to dominate those with less input, even if these have more in absolute terms than they would otherwise have had. Or, this is the case at least if there 44 ‘In the context of political participation, inequality tends to diminish epistemic value even as the total quantity of participation, suitably distributed, promotes it’; Estlund 2000, p. 138.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

224 Implications are no checks on how buyers use their extra input. Even if the institution of Estlund’s scheme did, as a matter of fact, lead to more just decisions, perhaps because the buyers of more input indeed happen to be public-​spirited, those with less input are still entitled to demand, on grounds of non-​domination, that it be made sure that the former use their extra input well.45 Otherwise they are asymmetrically dependent on the goodwill of the more infuential. Te ingenious design of Estlund’s scheme makes sure that such domination will not be of great extent; but it does weigh in the balance. Tis discussion illustrates the difculties that any proposal to institute ground-​level political inequality of opportunity in the name of justice will encounter, from the point of view of liberal relational egalitarianism. It does not, and cannot, rule out that no good scheme can be found. Tere are many real and possible schemes of decision rights about which we do not have robust empirical knowledge; and perhaps there are others where such knowledge just is not attainable—​since running controlled experiments with entire societies seems neither feasible nor permissible. However, this thought does not weaken the case for fair equality of political opportunity and its qualifed priority over substantive social justice in the real world. To the contrary, it strengthens it, because in the absence of reliable evidence for the superiority of inequality demanded by the epistemic criterion, we have to err on the side of trusting ordinary citizens. Not only do political rights thus, within liberal relational egalitarianism, enjoy a qualifed priority over other requirements of social justice. We also have good reasons to think that, insofar as substantive equality of political opportunity among ordinary citizens is concerned, this priority is secure in those circumstances we have most reason to care about: the ones characterizing our own societies, or those not too far away from them. And this conclusion has been reached without reliance on any supposed special value of democracy, apart from justice.

45 Te case would be somewhat diferent if buyers bought their input directly from others (which cannot be the case in the voucher scheme). Ten, buyers would arguably not dominate sellers. But they would still dominate third parties; and sellers would probably fail to fulfl their duties of justice to these. Plausibly, these duties prohibit surrendering one’s power to another person without credible assurance that this will lead to more justice. Selling is diferent from asking those whom one has good reasons to trust how to use one’s input.

Political Equality  225

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

7.5.  Conclusion: Dworkin and Democracy To conclude the argument of the chapter, it is worthwhile to show how and where purely recipient-​oriented views of justice of the kind criticized earlier (­chapter 2) fail to account for the value of egalitarian political rights; and how this is the case even for those views in the broader distributivist camp whose stated ambition is to account for dimensions of justice beyond the purely recipient-​oriented one—​namely, Dworkin’s. Dworkin’s conception is, in this respect, the strongest contender within that camp, so it is particularly instructive to see where exactly, on political equality, it falls short.46 Tis strengthens the critique delivered in earlier chapters, as well as the case made so far in this chapter—​that the way in which liberal relational egalitarianism recognizes both the intrinsic, expressive value of political equality and its orientation towards social justice overall strikes the right balance, and best accounts for their interplay. On purely recipient-​oriented views (­chapter 2, 2.2), democratic relations are conceived of as exclusively instrumental to distributive justice;47 such views entirely deny their non-​instrumental value. Pluralist egalitarians, such as G. A. Cohen, seek to accommodate such non-​instrumental value by arguing that egalitarian distributive justice is just one of a plurality of values, which has to be balanced against competing others in order to fnd out what we ought to do, all things considered. On such an account, it seems natural to regard the ideal of democracy as another such value extrinsic to, and, where necessary, to be balanced with, distributive justice. However, our ambition (motivated in 2.4 in ­chapter 2) was to develop more than pro tanto principles, and to show how diferent dimensions of justice are connected, and shed light on each other. Against that, such pluralist views are in danger of overlooking the connection between the instrumental and expressive value of democracy: expressing respect for people’s sense of justice by granting them equal political rights is required only if it indeed leads to adequate results in terms of substantive justice, but, as seen, this does not imply that such expressive

46 See c­ hapter 2, fn. 12 and 69, for clarifcation that the argument of that chapter applies straightforwardly only to Dworkin 2000, chs. 1 and 2. 47 See Arneson 2004, especially p. 46: ‘If one has [ . . . ] a moral right [to exercise power over the lives of others through one’s say in democratic decisions], this will be so only because one’s having the right is more conducive to the fourishing of all afected parties than any feasible alternative’; see also Arneson 2008b.

226 Implications value is not a genuine rationale for such equal rights—​that it is fully reducible to its instrumental value.48 Dworkin’s view is diferent from pluralist views. His theory regards justice as equality as the primary value, prior to others, which ‘government’, or the ‘political community’ has to respect, through showing ‘equal concern for the fate of all those citizens over whom it claims dominion [ . . . ]’,49 and such ‘equal concern’ is not exclusively distributive. Accordingly, Dworkin accepts, and frequently reiterates, that a unifed theory of justice must aim at weaving together the best account of equality by paying attention to all relevant dimensions of equality.50 His theory gives no thought to equal non-​domination, so it is not surprising that relational egalitarians and neo-​ republicans will fnd it unsatisfying in this regard; it is therefore not interesting to pursue this point in any detail. However, Dworkin does emphasize the non-​instrumental importance of political equality:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

A genuine society of equals must aim at equal stake as well as equal voice and equal status for its citizens. We must build conceptions of these different dimensions of equality that ft with and draw upon one another, not suppose that either economic or political or social equality is more fundamental than the others.51

Accordingly, on the level of theoretical ambition, there is, to this extent, no disagreement between the approach pursued in this book and Dworkin’s consolidated theory. Terefore, whatever their other disagreements, it is worthwhile to analyse how Dworkin’s view and liberal relational egalitarianism come apart on the question of what it takes to express proper respect for people’s political agency. Te remainder of this section will demonstrate that Dworkin’s conception of ‘political equality’ does not live up to its ambition: despite noting the non-​instrumental value of participation rights, he endorses an arbitrarily undemanding account of such rights. He thus ends

48 To be sure, pluralists can maintain that democracy also has further intrinsic value unrelated to results; see fn. 23 in this chapter. 49 Dworkin 2000, p. 1. 50 Dworkin 2000, pp.  11f., 65, 184. See Lippert-​Rasmussen 2018b, pp.  227f., on the relational aspects of Dworkin’s overall ideal of equality, and Schefer, 2003a, pp. 35–​37, and 2003b, for the objection that it still amounts to a merely ‘administrative’ conception—​without, however, much analysis of Dworkin’s specifc conception of political equality. 51 Dworkin 2003, p. 190.

Political Equality  227 up seriously underplaying their importance, and unduly privileging purely recipient-​oriented considerations, afer all. In his essay on ‘Political Equality’, Dworkin contrasts ‘dependent’ and ‘detached’ conceptions of democracy.52 A detached conception

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

insists that we judge the fairness or democratic character of a political process by looking to features of that process alone, asking only whether it distributes political power in an equal way, not what results it promises to produce.53

Dependent conceptions, on the other hand, hold that democratic decision-​ making must be generally geared towards substantively correct—​ i.e., just—​decisions. Tis defnition of a dependent conception, however, is instrumentalist in a wide sense, since Dworkin acknowledges that the democratic process itself serves to realize participatory goals that are of non-​instrumental importance for citizens’ lives, such as having one’s agency valued, and that dependent conceptions have to account for this.54 Given this inclusion of participatory goals on the side of the dependent, not the detached, conception, the conception of democracy put forward in this chapter falls on the dependent side of Dworkin’s distinction: it recognizes prior standards of social justice that democratic decision-​making has to respect (even though it obviously disagrees with Dworkin about their content). Dworkin’s discussion comes down strongly against detached conceptions of democracy, arguing that distributing infuence equally would make it impossible to institute any kind of safeguards for the quality of democratic decisions regarding justice-​relevant concerns, such as the extent and shape of people’s rights to basic liberties, and the distribution of resources.55 Tis conclusion seems unassailable, if infuence is understood simply as the probability to get one’s way in political processes, without any qualifcations as to the reasoning requirements that such processes have to fulfl.56 Indeed, it seems so unassailable that it is doubtful whether his adversary, a pure ‘detached conception of democracy’, is more than a strawman.57 52 Dworkin 2000, ch. 4.  For a more recent of the argument along essentially similar lines, see Dworkin 2011, pp. 388f. 53 Dworkin 2000, p. 186. 54 Ibid., pp. 186f. 55 Ibid., pp. 194f. 56 Ibid., p. 192. 57 He does not cite any advocates of this view.

228 Implications Dependent conceptions, on the other hand, have to strike the proper balance between assigning democratic procedures the importance they deserve, according to justice, and achieving substantively just decisions: [A]‌benevolent tyranny, in which none of our assumptions about democracy held, might nevertheless produce a just property scheme, and might otherwise respect the distribution goals of the right conception of equality; indeed, it might produce a more egalitarian distribution than a democracy could. But no tyranny could advance the participatory goals any egalitarian community would also aim to secure.58

Engaging in a democratic process constitutes a signifcant opportunity for every individual to ‘treat politics as an extension of his moral life’,59 and, specifcally to express one’s concern for justice:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Just as someone denied opportunity to worship according to his or her own lights is denied a foundational part of religious life, so someone denied opportunity to bear witness to his concern for justice, as he understands what the concern requires, fnds his political agency stultifed, not merely bounded.60

Furthermore, Dworkin also acknowledges that political processes ideally ought to bring to light ‘the best arguments, the most convincing reasons for taking up or working for one political cause rather than another’.61 Regarding some fundamental characteristics of democratic processes, Dworkin’s view thus also seems to be in agreement with the view developed earlier. In terms of the implications of this view, however, Dworkin limits himself to arguing that substantive equality of opportunity for infuence in the unqualifed sense noted earlier is not required by the agency goals of politics. He points out that, in order to equalize opportunity for infuence so understood, we would have to endorse measures that level down the political capacity of especially gifed or motivated individuals, e.g., by limiting freedom of political speech and association, and notes that ‘we certainly do not want to eliminate extra infuence that comes from a powerful mind or infectious

58 Ibid., p. 187. 59 Ibid., p. 203. 60 Ibid., p. 202. 61 Ibid., p. 198.

Political Equality  229 idealism’.62 Tese considerations motivate him to embrace a sufciency conception of such opportunity instead:

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te emphasis is [ . . . ] on the opportunity for some infuence—​enough to make political efort something other than pointless—​rather than on the opportunity to have the same infuence as anyone else has. [ . . . ] Tat is a threshold notion, and nothing in it takes equality of infuence to be an ideal toward which we should strive.63

Equality thus ends up playing no role in Dworkin’s conception of political opportunity. As long as people can assure themselves of a minimal chance to infuence others’ political views and choices, and generally have equal voting rights, the requirements of democracy are fulflled. However, on his way to this conclusion, Dworkin simply overlooks more plausible, but nevertheless demandingly egalitarian, conceptions of rights to political participation, beyond equalizing unqualifed infuence. Even though he recognizes that political agency ought to express concern for justice, and that political processes be reason-​respecting, he fails to notice the possibility of changing the metric of political equality to opportunity to contribute to deliberative processes in accordance with one’s sense of justice (7.3.2).64 Accordingly, his conception leaves no space for fair equality of opportunity to hold public ofce, and for other measures to equalize opportunities to develop political skills mentioned earlier. Tese are chucked overboard along with the obviously perverse levelling-​down measures he mentions. To be fair, part of Dworkin’s motivation for endorsing political sufciency rather than equality is his desire to salvage the possibility of assuring more 62 Dworkin 2003, p. 196. 63 Dworkin 2000, p. 203. 64 In the same passage in which Dworkin argues that people ought to look for ‘the best arguments’ in politics, he also argues that we should not try to educate people not to attempt to infuence others [ . . . ] except in ways that do not rely on special advantages they might have, in experience, commitment, or reputation, and also to attempt to resist being infuenced by other people whose arguments might have special force traceable to such advantages. People could not succeed in following that advice, because in political argument it is impossible to separate dancer from dance; ibid., p. 197. Tis passage illustrates well how he overlooks more plausible, nuanced positions. In one sense, we should certainly try to educate ourselves to separate between speaker and argument in order to avoid being persuaded by a very bad argument just because the speaker has a fantastic reputation. In another sense, as argued in section 7.3.2, certain inferences from dancer to dance, based on commitment, experience, and reputation, are fully reasonable, and not a threat to equal political opportunity of a desirable kind. Te relevant question of justice is that of the extent to which inegalitarian social and economic background processes have contributed to turning some into better dancers than others.

230 Implications

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

political infuence to those who sufer grave substantive injustice, where this is a good way of addressing such injustice. His example is that of inhabitants of very poor US inner-​city districts.65 But, frst, a deliberative process aiming at justice is already by defnition more able to give this fact its due than Dworkin’s adversary ‘equality of infuence’. And, second, the view developed earlier demands not only, as noted, that special measures be taken to equalize the political opportunity of the individuals concerned in terms of education and training, which, given their overall disadvantage, will likely be signifcantly impaired. In the case of grave present injustice, or past injustice with still sizeable efects, it may also require special representation for the victims of injustice, until injustice and its traces are eradicated.66 It thus appears possible to address the problem by temporarily ofsetting one inequality by another; it does not necessitate a move to permanent sufciency of political opportunity.67 At frst glance, Dworkin seems more sensitive to the connection between economic inequality and political inequality. He argues that we should eliminate ‘to the degree we can, extra infuence that comes from money’;68 and stresses that excessive economic inequality leads to the ‘particularly deplorable consequence’ of allowing the economically advantaged to ‘perpetuate and multiply their unfair advantage’.69 But the fact that that advantage is already unfair here once again makes clear that Dworkin does not regard political inequality as an independent injustice.70 Te ‘extra infuence’ comes from money that people ought not to have, anyway. All he is saying is thus that, if ‘equality of resources’ is absent, resulting inequalities in the political system 65 Ibid., p. 188. 66 Many of Dworkin’s very poor inner-​city inhabitants are blacks descended from slaves, and historically are discriminated against also in many other ways. 67 In later works, Dworkin treats this case as a legitimate deviation from equality of the worth of one’s vote; Dworkin 2011, p. 391; but he continues to emphasize—​due to his misplaced worries about political levelling down—​that no measures need to be taken to equalize political opportunity in a wider sense; ibid., p. 389. 68 Dworkin 2003, p. 196. 69 Dworkin 2000, p. 195. 70 See ibid., p. 196: ‘When we object to the distribution of power in our society, is power really the root of our complaint? Or are we really objecting to other, independently unjust features of our economic or political or social organization in a forceful way, by calling attention to a very unfortunate consequence of these features?’ (my emphasis). In later works, Dworkin notes that ‘[w]‌e regret some people’s special infuence because it is grounded in wealth, which we think should make no diference in politics’; Dworkin 2011, p. 389. However, it remains unclear to which extent he thinks this yields genuinely political equality-​based reasons to worry about unequal wealth. Te only example of political inequalities connected to wealth which he explicitly mentions and objects to are formal measures that tend to make voting harder for the worse of—​such as measures requiring ID documents for voter registration, which many poor people typically lack; ibid., p. 393.

Political Equality  231

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

will be used to perpetuate that absence.71 Tis is plausible, but no answer to the question of why citizens who want to put into action their sense of justice should rest content with minimal equality of political opportunity even if ‘equality of resources’ should be present. Dworkin’s position hence stands in outright contradiction to his own admonition that ‘[we] must [ . . . ] not suppose that either economic or political or social equality is more fundamental than the others’.72 Whatever limits his overall theory imposes on permissible political inequality appear to be side efects of limits on economic inequality required for independent reasons. Contrary to Dworkin’s avowed aims, considerations of political equality end up playing no role in justifying, or modifying, his theory of justice, beyond ruling out a ‘benevolent tyranny’. His conception of participation rights seems little more than an underdeveloped aferthought to a theory which still, in substance, represents a purely recipient-​oriented conception of equality. While his account of democracy pays lip service to the idea that institutions not only have to distribute justly, but have to treat people as equals while doing so, the theory does not take seriously Dworkin’s own acceptance of the claim that political rights express respect for the moral agency of citizens, and, in particular, for their concern for justice, because it fails to address the question of why that respect should not be substantively equal. Against that, the argument of this chapter has demonstrated how respect for political agency and concern for substantive justice can be reconciled within a demanding conception of political equality.

71 It is also worth pointing out that it is not clear how much economic equality Dworkin’s view requires; it demands that ‘people be made equal, so far as this is possible, in their opportunity to insure or provide against bad luck’; Dworkin 2003, p. 191. But we do not know how high insurance cover would be. Te assurance that the distribution so generated will sufce to sustain political conditions guaranteeing its continuity rings somewhat hollow if it is not clear how egalitarian that distribution actually is. 72 Ibid., p. 190.

8 Distributive Justice

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

8.1.  Introduction As noted in c­ hapter  1, relational egalitarianism frst gained traction as a source of objections to distributive conceptions of equality, such as luck egalitarianism. It is not clear what relational equality, understood as a demand of social justice, itself requires in distributive terms. A common worry in the literature, raised especially against Elizabeth Anderson’s view, is that it vacates much of the terrain of distributive justice in favour of a minimalist, sufciency view. So we need to investigate the distributive implications of the core requirements of liberal relational egalitarianism with regard to power and status identifed, and justifed, in Part I, and of the resulting conception of political equality argued for in the preceding chapter, and demonstrate that, in its case, such a worry is unfounded. Tis chapter argues that a justice-​ based conception of relational equality yields powerful expressive and instrumental reasons to care about distributive inequality in socially produced goods—​despite its according centre stage to just social relations, and not to the distribution of goods per se. It does so by way of an internal argument against Anderson’s view, as this is the most developed rival account of the distributive implications of relational equality.1 Samuel Schefer, on the other hand, has not developed the distributive requirements of his view on relational equality in much detail, so far.2 Te argument develops over fve sections. Section 8.2 introduces the problem of connecting relational equality to distributive justice, explains Anderson’s approach to the problem, and objections to it. Section 8.3 argues that, given a commitment to an ideal of society as a cooperative scheme among equals, relational egalitarians ought to hold that there are intrinsic reasons of an expressive kind in favour of limiting distributive inequality in socially produced goods, even where risks of 1 Anderson 1999, 2004a, and 2008. 2 Schefer 2003a, 2005, and 2015.

Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0008

Distributive Justice  233 relational equality between individuals are not at stake.3 Sections 8.4 and 8.5 deliver additional instrumental reasons for relational egalitarians to limit inequalities of income and wealth: such inequalities engender risks of domination and political inequality, and encourage the formation of inegalitarian social status norms that are damaging to the self-​respect of the worse of, and deprive them of important social opportunities. Section 8.6 concludes by applying the arguments made in these sections to the case of equality of opportunity to attain desirable social positions, and shows that justice-​based relational egalitarianism is committed to a demanding principle of equality of opportunity so understood.

8.2.  Relational Egalitarianism and Distributive Justice:  Te Problem

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Relational equality requires protecting individuals against domination by others at varying levels of intensity, and combating inegalitarian norms of social status which facilitate domination, wound people’s self-​respect (also in other ways), and deny them important social opportunities through collectively enacted, self-​reinforcing devaluation of their skills or other traits. Terefore, the most important and stringent distributive requirements that follow from this ideal of relational equality must be precisely those that enable individuals to avoid such relations of domination and status inequality. As Anderson puts it, [n]‌egatively, people are entitled to whatever capabilities are necessary to enable them to avoid or escape entanglement in oppressive relationships. Positively, they are entitled to the capabilities necessary for functioning as an equal citizen in a democratic state.4

Tis maps on well enough to the aforementioned requirements, as well as to the requirements of political equality spelled out in ­chapter 6. However, the 3 As these reasons are expressive, and arise from processes of social cooperation, one might object that they are therefore not intrinsic, as intrinsic reasons would fasten on the objectionable nature of distributive inequality as such, no matter how it arises. ‘Intrinsic’, for the purposes of this chapter, means ‘non-​instrumental’. 4 Anderson 1999, p. 316. Anderson spells out the distributive implications of her view of relational equality in terms of capabilities, and, in discussing her view, I will stick to this term. But, for the purposes of the argument of this chapter, nothing hinges on endorsing capabilities, resources, or social primary goods as the currency of justice. Sections 8.4–​8.6 will discuss the particular goods of income, wealth, and opportunity in more detail.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

234 Implications question is whether this is all there is to relational egalitarian requirements of distributive justice, and how egalitarian, all things considered, these requirements are likely to be. Neither Anderson nor Schefer has so far developed a full-​fedged account of the distributive requirements of justice-​based relational egalitarianism. Anderson’s sketch of such an account spells out three dimensions in which individuals have to have adequate means in order to be able to stand in the required relations of equality to each other: (a) as human beings—​this covers, for example, aspects of physical and mental well-​being, freedom from (treatable) illness, and freedom of movement; (b) as a worker in a system of cooperation—​covering, inter alia, the education necessary to carry out a role in the division of labour, and freedom of occupational choice; and (c) as a participant in democratic politics—​covering equal rights to political participation and the means necessary to exercise such rights in a meaningful manner.5 As already noted, the main worry that this account faces in the current debates about the value of equality, and its connection to social justice, is that it displays too little sensitivity to the comparative distribution of material goods in society among its members—​that, by seeking to refocus egalitarian theorizing on matters of social and political relations rather than distributions, it makes the concomitant claim that the distributive sphere is of little importance. Tis is because Anderson’s account of distributive justice seems to commit her to two claims. First, as several critics have noted, to a sufciency conception of distributive justice:6 such a conception merely demands that individuals be given enough to cross the relevant threshold set by the respective theory of justice. In the case of more orthodox sufciency views,7 this is ofen referred to as ‘a decent life’; in the case of relational egalitarianism, it is standing as an equal to others in the relations mentioned. Distributive inequalities between individuals who are above the threshold are then of no intrinsic concern to social justice.8 5 Anderson 1999, pp. 317f. However, given her focus on social cooperation, Anderson can only be committed to the idea that people are owed equal capabilities to function as human beings as a matter of social justice in virtue of their capacity to participate in a system of cooperative production, not simply in virtue of their humanity. Tis does not rule out duties to non-​cooperators; it only rules out that these duties are signifcantly egalitarian; see fn. 21 in this chapter. 6 Casal 2007; Arneson 2006. 7 Raz 1986, ch. 9; Frankfurt 1987, 1997, and 2000; Crisp 2003 and 2004. Te argument of this chapter does not apply directly to these views, which are sufcientarian ‘all the way down’; it is that relational egalitarians cannot be distributive sufcientarians. 8 Sufcientarianism would not necessarily be an independent rival position to egalitarianism if it merely claimed that the requirement to raise everybody over the sufciency threshold enjoyed

Distributive Justice  235 Second, Anderson seems to be committed to the claim that the distributive requirements of relational egalitarianism are sufcientarian also in an all-​ things-​ considered sense, including possible instrumental rationales for equalizing distributions. Even if distributive inequality is not by itself regarded as a concern for justice, it could be that people stand to each other in the required egalitarian social relations only under conditions of strict distributive equality; or that this can only be achieved if stringent limits are set to material inequality (as will indeed be argued in sections 8.4–​8.6). But Anderson seems to hold—​in places, at least—​that a reasonably low cut-​of point can be found for distributive concerns instrumental to achieving such relations. For example, she argues that

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

[m]‌ost able-​bodied citizens [ . . . ] will get access to the divisible resources they need to function [as an equal in the three dimensions mentioned earlier] by earning a wage or some equivalent compensation due to them on account of their flling some role in the division of labor9

and that ‘one mechanism for achieving a decent minimum would be a minimum wage’.10 Such claims are decidedly unambitious—​the possibly vast disparities between diferent wages and remunerations for diferent positions within the societal division of labour are not mentioned. Tis seems to vindicate the suspicion that relational egalitarianism vacates a large part of the terrain of distributive justice. Among contemporary political theorists, only hard-​nosed right-​libertarians deny that members of society are owed at least a ‘decent minimum’. In other places, Anderson’s view seems somewhat more demanding:  she also holds that concern for egalitarian relations requires constraining top incomes, and argues that egalitarians should prefer that ‘individuals be crowded in the middle of the distribution’—​that is, a large middle class.11 Schefer, on the other hand, suggests that a relational understanding of equality can plausibly underpin egalitarian conceptions of distributive justice such as Rawls’s ‘justice as fairness’.12 He also argues that in a relationship priority over other justice concerns. Te additional claim that inequalities above that threshold are of no relevance to justice is needed to turn it into such a distinct position; see Casal 2007, pp. 299f.

9 Anderson 1999, p. 321; see also Anderson 2004a, p. 251.

10 Anderson 1999, p. 328 (emphasis added). 11 Anderson 2008, p. 267.

12 Schefer 2003a, pp. 22f., 31.

236 Implications between equals, partners are committed to regarding each other’s interests as equally important, and that this, in many cases at least, implies that distributing the goods and bads arising within such a relationship equally should be the partners’ preferred solution—​even if there is no algorithm leading from relational equality and equal importance of interests to egalitarian distributions of specifc goods and bads in every case.13 Te task of this chapter is to spell out the arguments for, and implications of, such a view, when applied to the overall relation of social cooperation among free and equal individuals, and to show that relational egalitarianism, properly understood, has more demanding distributive implications than Anderson seems to think. Such a position does indeed claim that a plausible conception of social justice should regard the avoidance of the inegalitarian social relations specifed in Part I as the dominant egalitarian aim of society, and should not regard distributive equality per se as a value of similar stringency. However, as this chapter will go on to show, a sufciency distribution cannot satisfy the requirements of relational equality; on the best understanding of such a view, distributive inequality raises concerns of justice for both expressive (section 8.3) and instrumental reasons (sections 8.4–​8.6).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

8.3.  Justifying Distributive Inequality Te argument of this section is straightforward:  Given that relational egalitarians such as Anderson and Schefer follow Rawls in specifying participation as equals in reciprocal cooperation as the foundational relation of egalitarian social justice,14 they have an intrinsic reason to limit inequality in the goods produced by such cooperation. Relational egalitarians should thus endorse a—​defeasible—​presumption of distributive equality, because such a presumption expresses equal respect for participants in cooperation who jointly produce basic social goods. Such a position regulates distributive inequality for intrinsic reasons, but difers from purely distributive views such as luck egalitarianism, which regards distributive inequalities for which individuals are not responsible through their own choices as unfair, and hence pro tanto unjust, no matter



13 Schefer 2015, pp. 24f.

14 Anderson 1999; Schefer 2003a and 2005, passim.

Distributive Justice  237 how these inequalities otherwise came about.15 What it judges to be just or unjust, in the frst instance, are not distributions of a given amount of goods among individuals. Tis Rawls calls the problem of ‘allocative justice’, and distinguishes it from the problem of distributive justice, understood as regulating ‘a fair system of social cooperation over time’.16 Distributive justice so understood applies to the social relations that govern the production and distribution of goods: these relations are regulated by the ‘basic structure of society’, that is, a society’s main social, political, and economic institutions.17 Tese institutions operate so as to guarantee

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

what we may call pure background procedural background justice. Te basic structure is arranged so that when everyone follows the publicly recognized rules of cooperation, and honours the rules the claims specify, the particular distributions of goods that result are acceptable as just [ . . . ] whatever these distributions turn out to be.18

A system of production can follow pure procedural justice in this sense, and still set overall limits to the distributive inequalities which socio-​economic institutions may generate, as long as, in doing so, it follows general rules which specify acceptable overall distributions in advance, and does not seek to determine which individual gets what exactly. Anderson follows Rawls in understanding the problem of distributive justice as (mainly) one of the set-​up of the basic structure, and agrees that rules implementing ‘range constraints’ on inequality—​for example, through a system of social insurance—​are compatible with pure procedural justice so understood.19 Now, the argument of this section is that individuals’ status as equals in cooperation yields an intrinsic reason to introduce range constraints, because 15 Radical luck egalitarians do not ascribe any foundational importance to participation in cooperation, but regard their conception of distributive equality as a direct implication of moral equality, according to which all persons, qua persons, enjoy a basic equality of moral standing (2.2). Arguably, Cohen 1989 and Arneson 1989 are examples of such a view. Other egalitarians, such as Dworkin, single out a diferent foundational relationship, that of being subject to the coercive laws of the state; Dworkin 2000, p. 1. On relational equality and global justice, see the Conclusion. 16 Rawls 2001, p. 50. 17 See Rawls 1999a, p. 6. 18 Rawls 2001, p. 50. 19 Anderson 1999, p. 309, and 2008, p. 241. As shown in ­chapter 4, section 4.5, liberal relational egalitarianism can remain agnostic about whether we have good reason to adopt a general basic structure restriction for requirements of social justice. What is relevant here is merely that the implementation of range constraints on acceptable inequality of wealth and income is certainly a task that only a basic structure can carry out efectively, whether or not individuals also have duties going beyond fully supporting its eforts.

238 Implications

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

they express this equal standing:  if the institutions governing production have to display egalitarian concern for participants in the enterprise which they regulate, then this means that they have to aim at distributing advantages and disadvantages that are socially produced equally, unless there are suffcient reasons for an unequal distribution. Unequal distributions of such goods (and bads) constitute unequal treatment on the part of the basic structure, and have to be justifed by justice-​relevant reasons.20 It also means that unequal impact of these institutions on other goods that are ofen considered natural goods, for example, on health (­chapter 9) and genetic endowments (e.g., through socially caused pollution), requires such justifcation.21 Tis Rawlsian view that equal treatment in cooperation yields a presumption of equality in social goods (and bads) has, of course, been contested by non-​ egalitarians, and has been subject to intensive debate over the last fve decades. Tis section does not seek to defend it against such rival views. Possibly the most notable of these is the view that equal treatment demands rewarding people according to their specifc individual contribution to cooperation, that is, according to their distributive desert. What it argues instead is that justice-​based relational egalitarians should subscribe to the Rawlsian view. Regarding desert, both Anderson and Schefer are indeed sceptical about the existence of independent, pre-​institutional criteria that could determine what people individually deserve in distributive terms: they argue that the highly complex division of labour that characterizes advanced economies makes it impossible to determine the value of individual contributions to cooperative production at a fundamental level. Social goods thus have to be regarded as jointly produced.22 If such scepticism were unfounded, that would 20 For a similar position on socially produced goods, but in the context of an argument focusing not on relational egalitarianism as understood here but on a conception of equality that can serve as the basis of an ‘overlapping consensus’ along the lines laid out by Rawls in his later work, see Heath 2008. For a detailed exploration of the presumption of equality, but without reference to underlying relations of social cooperation, see Gosepath 2004, pp. 200–​211. 21 Restricting requirements of limiting inequality to socially produced goods, and to social impact on other goods, does not imply that there are no further duties of assistance to those who are naturally disadvantaged—​especially those who are permanently excluded from cooperation due to, for example, severe mental disability, see fn. 5 in this chapter, and section 9.4 in ­chapter 9. For an argument that an entitlement to a social minimum for non-​cooperators can be made to ft within a Rawlsian framework, see Stark 2007; for an application of relational egalitarianism to disability more broadly, see Brown 2019. Tat, on a relational egalitarian view, natural inequality must not be compounded by social means also gives reasons to seek to sever, or at least weaken, the links between achievement, social status, and self-​respect, insofar as achievement is partly determined by natural abilities—​see section 8.6. 22 Anderson 1999; Schefer 2001, pp.  190f. For a detailed argument that relying on considerations of distributive desert is incompatible with relational egalitarian justice, see Anderson 2008, pp. 247–​254. While Schefer argues that scepticism about distributive desert need not carry over to

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  239 create a signifcant problem for relational egalitarians, for several reasons. If distribution could and should be undertaken according to diferential desert, then people’s just deserts might confict with relational equality where, for example, the superior rewards of the economically deserving put them in a position to dominate the less deserving. It seems that a person whose just economic deserts were traded of against such other considerations would have a complaint of justice against this which is not eliminated by that trade-​of being desirable in other respects. Or, even if that complaint were eliminated, because desert might only be a pro tanto consideration which can be traded of against others without remainder, the fact that a trade-​of needs to be struck would still speak against the particular stringency of demands of non-​ domination argued for so far. Finally, on the conception of non-​domination put forward here, which ranges over (prima facie and fnal) claims of justice (see 3.3.2 in ­chapter 3), the economically more deserving might even themselves come out as dominated, where they face the danger of having their superior rewards curtailed by powerful agents, such as state ofcials. Similar considerations would apply to status inequality and inequality of desert: there would be no mandate of justice to curtail economic inequality where it leads to status inequality (section 8.5) if that economic inequality tracks unequal economic desert.23 However, if the reasons for rejecting foundational, pre-​ institutional conceptions of diferential economic desert advanced by relational egalitarians are sound, then a relational egalitarian position which seeks to exclude distributive inequality in socially produced goods as such from the scope of social justice is a non-​starter. Given people’s standing as equal producers of social goods, caring only about the instrumental consideration that distributive inequality may cause individuals to encounter each other in particular unjust relations, such as domination, would be arbitrarily narrow. What then needs to be specifed is which kinds of reasons exactly sufce to defeat the presumption of equality. Rawls endorses a stringent interpretation of the presumption. His ‘general conception of justice’ mandates that retributive desert (ibid., p. 192), Scanlon goes further and argues that desert itself can only ever be a ground for praise and condemnation, and that any further policy consequences, whether of diferential economic reward or of criminal punishment, can never be justifed by it, but only by the further benefts which these may yield; Scanlon 2018, pp. 123f. 23 A pluralist social egalitarian arguing for pre-​institutional criteria of diferential economic desert is David Miller (1999, chs. 2 and 7); this is one of the reasons why he argues that social equality has to be traded of against justice (5.3.1).

240 Implications

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

[a]‌ll social values [ . . . ] are to be distributed equally unless an unequal distribution of any, or all, of these values is to everyone’s advantage.24

As is well known, inequality-​justifying reasons according to Rawls’s conception of justice fall into two broad categories: frst, rights to basic liberties constraining the possibility of maintaining strict equality (Rawls’s frst principle of justice); and second, reasons of efciency, demanding a regime of distributive inequality if it leaves the worse of better of than under equality—​ the rationale of the diference principle, which requires that inequalities in income and wealth be to the greatest possible advantage of the worst of. As mentioned before, Schefer’s approach is congenial to Rawls’s. Anderson’s position is, against that, once more ambiguous: on the one hand, she stresses the ideal of society as a cooperative enterprise among equals throughout, and argues for ‘a conception of reciprocity that would squeeze the gap between the highest-​and lowest-​paid workers’.25 On the other hand, she explicitly argues for range constraints on top and middle incomes beyond safeguarding sufciency for those at the bottom of the distribution exclusively on the basis of their instrumental contribution to avoiding unjust power relations and social esteem hierarchies.26 She thus seems to overlook the intrinsic reason for limiting distributive inequality: that it expresses respect for people’s equal status in the overall relation of social cooperation. What drives her view is an emphasis on the capacity of market arrangements to deliver efcient outcomes, because, for many goods, market prices signal more successfully than any other economic arrangement how they may be efciently allocated.27 Such efciency may be a reason to accept some distributive inequality, and to do so not only where this benefts the worst of maximally. However, Anderson herself, when acknowledging the need for range constraints on income distributions on instrumental grounds, argues that distributions can be signifcantly compressed without endangering the signal function of prices.28 Her argument against the diference principle—​and against the general conception of justice, since it relies 24 Rawls 1999a, p. 54 (my emphasis). Te full quotation includes Rawls’s list of primary goods as the ‘social values’ in question. It is omitted here because relational egalitarians need not accept it; see fn. 4. 25 Anderson 1999, p. 326. 26 Anderson 2008, pp. 263f. 27 Ibid., p. 242. 28 Ibid., p. 262.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  241 on the same reasoning—​is a diferent one: it is the well-​known worry that ‘[i]‌n giving absolute priority to the worst of, the diference principle might require considerable sacrifces in the lower middle ranks for trifing gains at the lowest level’.29 Tis worry has some plausibility.30 Furthermore, it needs to be acknowledged that Rawls’s case for the diference principle is based not just on its expressing equal moral status in cooperation, but on a variety of diferent arguments whose connection to relational egalitarianism is less clear. Tese are the argument that the diference principle would be chosen in the original position, based on the assumption that risk aversion is the most reasonable attitude towards choice under conditions of radical uncertainty,31 and his objection to letting distributive outcomes be determined by ‘morally arbitrary factors’ such as natural abilities, which inspired luck egalitarianism.32 However, even if these points are conceded, the expressive rationale for limiting distributive inequality in social goods still rules out a sufciency conception of distributive justice. One possibility would seem to be to drop the absolute priority of the worst of in favour of only relative priority, and hence to endorse a prioritarian position.33 How much distributive equality exactly relational egalitarianism requires on expressive grounds may thus remain an open question. Tere is ample space for reasonable disagreement about it; such disagreements should be resolved in egalitarian democratic procedures. Tis section has established that relational egalitarians ought to accept a presumption of equality in social goods, because it expresses respect for people’s standing as equals in social cooperation, and has shown that Anderson overlooks this intrinsic rationale for limiting distributive inequality. However, this does not sufce to conclude that only an interpretation of the presumption that is as stringent as Rawls’s is capable of expressing due respect for equality in social cooperation, without drawing on additional reasons for limiting distributive inequality. Fortunately, relational egalitarianism can draw on such additional criteria: on such a view, there are also strong instrumental reasons for limiting 29 Anderson 1999, p. 326. 30 Hard-​nosed Rawlsians might, however, want to deny that such cases are indeed practically possible, and ask why, in any given case, better placed groups could not in turn compensate the second worse of. Te general conception of justice rules out unjustifed inequality between all groups, not just between the worst of and the second worse of; see Rawls 1999a, pp. 71f. 31 Rawls 2001, pp. 106f. 32 Rawls 1999a, pp. 64f. 33 See Parft 1991.

242 Implications inequality of income, wealth, and opportunity. In particular, a commitment to an ideal of society as a cooperative enterprise among equals requires paying particular attention to two areas of concern: frst, what distributive inequality enables members of society to do to each other. Distributive inequality might, in instances, be judged unobjectionable, when taken in isolation, but might end up being judged impermissible because of the risk of domination and political inequality it gives rise to (section 8.4). Second, distributive inequality might pose a threat to individuals’ social status in society, according to prevailing norms, which, sometimes at least, also threaten their self-​respect (section 8.5). Finally, both these rationales come together in requiring a demanding conception of equality of opportunity (section 8.6).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

8.4.  Distributive Inequality, Domination, and Political Inequality Tis section deals with the frst instrumental rationale for limiting permissible distributive inequality:  distributive inequalities tend to bring about inequalities of political opportunity, and such political inequality is unjust where it institutes domination, or is disrespectful for other reasons (see section 7.3). Tis delivers a yardstick with which the presumption of distributive equality discussed earlier can be flled out with more defnite content, and the remainder of the section delivers a frst set of reasons why this content will likely be demandingly egalitarian, in material terms. As argued before (section 4.5.1 in ­chapter 4), three kinds of policies are generally open to a society committed to the ideal of relational equality to ward of domination. First, procedural protection, which ofers institutional safeguards against the abuse of holdings and power by those in superior social positions. Second, socio-​psychological protection: the ‘engineering’ of social attitudes, for example, in school and through the media, which instil a practical sense of basic equality, and of the limited moral importance of de facto inequalities in talents, capacities, and social functions and professions, into the members of a cooperative scheme, so that those occupying superior positions become less disposed to make use of any opportunities of abuse that their positions might ofer them, and those subject to their power more disposed to resist such abuse. And, third, distributive policies, which aim at preventing the abuse of power by limiting material inequalities, so that the advantaged have comparatively less means to buy power. Tis section is

Distributive Justice  243 devoted to the third strategy. Its aim is to show that there are good instrumental reasons to think seriously about setting stringent limits to distributive inequality on grounds of non-​domination, and equality of political opportunity. Tis might seem perplexing; afer all, as seen, many relational egalitarians present their conception as an alternative to distributive egalitarianism, and some of their statements seem to imply that they regard distributive equality as of lesser importance. Anderson claims that

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

the degree of acceptable income inequality would depend in part on how easy it was to convert income into status inequality—​diferences in the social bases of self-​respect, infuence over elections, and the like. Te stronger the barriers against commodifying social status, political infuence, and the like, the more acceptable are signifcant income inequalities. Te moral status of free market allocations is strengthened the more carefully defned is the domain in which these allocations have free rein.34

Tis quotation can be read as not only pointing out that distributive inequality is especially troubling if and when it endangers relational equality, but also as making an additional, stronger claim: it seems to express a policy preference for the ‘careful defnition of domains’. It thus seems to argue that relational egalitarians should assign priority to strategies and policies which erect barriers against the conversion of wealth and income into status and power; that is, that they should assign priority to socio-​psychological and procedural strategies over distributive ones. Even though Anderson endorses, as seen, some range constraints on distributions, she holds that these are merely second-​best, and only come into play where conversion barriers fail.35 Now, socio-​psychological and procedural strategies indeed have one crucial advantage over distributive ones: if they are in place and function well, then they implement non-​domination directly. Strategies of more distributive equality, on the other hand, implement non-​domination indirectly, that is, only if additional empirical assumptions about how people will use their distributive shares are true. Take access to the media for political purposes as an example: strategies which bar the rich from using their wealth to buy media time implement, if successful, political equality directly. Against that,

34 Anderson 1999, p. 326.

35 See Anderson 2008, p. 266.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

244 Implications a strategy of distributive equality is based on the assumption that people will at least also use their more equal shares to gain access to the media, rather than to merely buy additional consumer goods for themselves. But this point does not sufce to establish that relational egalitarians should prefer procedural policies over distributive ones. It only makes clear that distributive policies, if not working by themselves, have to be supplemented by additional strategies; such as, to stay in the example, a socio-​psychological strategy emphasizing the importance of widespread adequate media access for political equality, and the unimportance of being able to acquire additional consumer goods, once a certain level of afuence is reached—​for example, through ‘citizenship education’ in school. On any given issue, the choice of the right policy depends to a large extent on context and empirical circumstances. Sociological and socio-​psychological research are needed for a comprehensive assessment of the opportunities for relational inequality that given social scenarios ofer to the advantaged, and to give informed guidance as to how these can best be countered. But political theory can make clear that there are several good reasons for denying that procedural and socio-​psychological strategies ought to enjoy priority over distributive strategies as a matter of principle, and can thus give guidance for more detailed empirical investigation about which policy to pursue. Tis can be shown by focusing on Anderson’s claim about the necessity of blocking exchanges between wealth and other goods, such as status and power. Clearly, some such barriers need to exist.36 To some extent, diferent goods pose diferent conditions on the adequacy of their distribution and transfer. We want people of superior capacities of political reasoning and rhetoric to be able to enjoy better access to such positions than others, because we think that, other things being equal, the political system works better if it is directed by such people, rather than by people who have simply bought their positions of infuence (section 7.3.2 in c­ hapter  7). Complete convertibility of wealth into political infuence would not be acceptable even if all had equal substantive opportunity to buy such infuence. But the relevant question for this section is to what extent convertibility must be blocked in order to make sure that those with greater distributive means do not obtain opportunities to dominate others—​and to which extent this aim can and should be achieved through other strategies, such as that of equalizing 36 Anderson’s claim is reminiscent of Walzer’s theory of ‘complex equality’, according to which the distribution of social goods ought to track the social meaning of these goods in the society in question; Walzer 1983.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  245 distributions. Two arguments speak in favour of a policy of more distributive equality for relational egalitarian reasons, and against the general preferability of a policy of blocked exchanges. First, distributive policies might be less intrusive. Strategies of limiting distributive inequality can be preferable to procedural and socio-​psychological strategies, where the latter involve heavy interference with liberty. Te enforcement of procedural protection might involve potentially problematic invasions of privacy, insofar as it might require close monitoring of the spending patterns of the advantaged. Whether that is especially troubling seems to depend largely on the individual case—​there is surely no problem with rigidly enforcing complete transparency regarding fnancial contributions to political candidates and campaigns. But these are only some of the more obvious ways in which wealth may be used to infuence politics. What about, for example, throwing parties where important political connections may be made? Te problem thus cuts deeper than that: blocking exchanges between different goods cannot be wholly achieved by procedural protection and law enforcement. Tese strategies require expansive control of individuals and groups by state agents; control that, for a large part, will have to come afer the fact occurred, in the form of sanctions. Other things being equal, a better way to erect conversion barriers is socio-​psychological, via self-​and peer-​ control, via the entrenchment of social norms regarding the unacceptability of exchange between diferent goods. But this constitutes a problem of potential illiberality: relational egalitarians such as Anderson and Schefer are committed to a Rawlsian principle of liberal neutrality,37 according to which justice merely sets constraints to the pursuit of individual and collective conceptions of the good; avoidance of domination is one of these constraints. An attempt to fx the social meaning of goods so as to block exchanges restricts the capacity of individuals and groups to fgure out for themselves what diferent goods mean to them. From a liberal point of view, it should be up to people, to the greatest extent possible, to decide what money can and cannot buy.38 Of course, it should not directly buy political ofce, due to the considerations of functionality and competence mentioned earlier; but, if 37 Anderson 1999, p. 330. For Schefer, this is implied by the closeness with which his account tracks Rawls’s; see Schefer 2003a, p.  25, and Schefer 2005, pp.  18f. However, he also briefy mentions the possibility of a relational egalitarian ideal that draws on a more comprehensive account of what is good for people; ibid., p. 19. 38 Waldron 1995, pp. 167f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

246 Implications somebody whose conception of the good ascribes paramount importance to engagement in politics consequently decides to devote more resources to the development of her own political skills and talents than others do, is it unjust if, other things being equal, she gains more infuence?39 It seems that this is not the case, as long as plutocracy is avoided, and the richer are not also the more powerful, across the board, who use the political system merely to pursue their own interests. As far as the danger of domination is concerned, the problem here really seems to be constituted by too much distributive inequality and its political consequences, not by the possibility of conversion itself. One might then think about fostering social norms which condemn specifcally only the conversion of greater wealth into political power. But the question then becomes: how much wealth exactly may be converted? It is hard to see how such norms could be both strong enough to be reasonably efective, and precise enough to apply only to a circumscribed range of cases, and hence avoid illiberality.40 Tis leads over to the second point: limiting distributive inequality might be a more efective way to implement relational equality than procedural and socio-​psychological strategies. If procedural and socio-​psychological strategies work, distributive inequality will be relationally harmless. But the efectiveness of these two strategies will always be, to some extent, precarious, since superior distributive means can and will be used to infuence the political process itself, in open attempts to change procedural and socio-​ psychological norms, or to prevent them from being enacted, for example, through media pressure, or covert attempts to make their subversion easier, by lobbying for loopholes and deliberately vague legislation. Keeping distributive inequality in place thus means leaving the weapons in the possession of the advantaged, and merely prohibiting their use—​as opposed to real disarmament. Furthermore, unequal distributive scenarios, such as concentration of the ownership of the means of production, introduce greater risks of misrepresentation into the political system. If the means of production are concentrated among a section of the population, and this section manages to act as a political unit, this makes it easier for its members to present their sectarian interests as the interests of ‘the economy’ in general, and thus press for legislation which introduces opportunities for them to attain dominance

39 Dworkin 2000, p. 195. Tis, in turn, must be distinguished from a scheme institutionalizing direct avenues for buying more political input, such as Estlund’s (7.4.3). 40 Recall the similar argument about social norms policing tax morality made in 4.5.2.

Distributive Justice  247 over others.41 Superior distributive means thus give the advantaged the instruments to hollow out the system of relational protection in various ways. More distributive equality, if coupled with enough political alertness of citizens in general, can help prevent such scenarios from arising in the frst place.42 Neither of these two considerations sufces to fully determine policy. Tey need to be validated further, and supplemented by, detailed empirical research. However, they do make clear that relational egalitarianism might have much more demandingly egalitarian consequences in terms of distributions than is presently thought in the literature, both by relational egalitarians, such as Anderson, and their adversaries.43 At the very least, the ideal of relational equality delivers a set of interesting questions to ask, for empirical researchers as well as for decision-​makers concerned about distributive policies. In cases where unacceptable power diferentials can best be tackled by reducing distributive inequality, relational egalitarian social justice mandates this reduction.44

8.5.  Social Status and Distributive Inequality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Te second group of instrumental arguments against inequality of income and wealth centres on cases where such inequalities give rise to inegalitarian 41 Such problems of ‘concentrated interests’ harming the public good exist not only in the case of the rich: for concentrated interests, it is always easier to convert money into infuence than for difuse interests. But the problem is worse if the concentrated interests in question can additionally draw on much more money than the difuse interests. 42 For reasons such as these (and others), the later Rawls regards redistributive welfare capitalism, which leaves the ownership of the means of production in the hands of relatively few, as an insufcient implementation of ‘justice as fairness’, and endorses egalitarian ‘property-​owning democracy’, characterized by widely spread ownership of such means; Rawls 2001, pp. 131, 139f. For criticism that not even a ‘property-​owning democracy’ can fully do the job, see Schemmel 2015b, pp. 403f. 43 In Anderson’s defence, it needs to be pointed out that part of the ambition of her writings on relational equality is to make policy proposals for real-​world scenarios, particularly for the United States. Some of her claims thus incorporate judgment about which more modest reforms seem feasible in the US, given rooted ideological traditions and potential unwillingness of sections of the population to go along with more demanding reforms; see Anderson 2004a, pp. 252, 255. It is hard to disagree that large-​scale redistribution of the means of production is not possible in the US at the moment. But then, whether the successful blocking of conversion of wealth into power is currently any more feasible also seems an open question. 44 Among sufcientarians, Frankfurt harbours most sympathy for relational egalitarian considerations: ‘[I]‌t may be maintained that inequalities in the distribution of economic benefts are to be avoided because they lead invariably to undesirable discrepancies of other kinds—​for example, in social status, in political infuence, or in the abilities of people to make efective use of their various opportunities and entitlements’; Frankfurt 1987, p. 24.

248 Implications norms of status which, apart from contributing to domination, can also harm self-​respect, or deprive the worse of of important social opportunities (see sections 6.3 and 6.4 in ­chapter 6). Tis section thus focuses on the general norms of social acceptability which tie status to the possession of economic resources (whether or not any particular group in society can be held especially responsible for the existence of these norms). It asks whether and how limiting distributive inequality can address the problem. Chapter 6 has already disentangled this problem from the connected, but diferent question of when people are disposed to feel envy at others’ superior holdings, and why, if at all, that should matter for justice. Unlike Rawls’s attempt to deal with the problem of unequal social status, neither the argument of that chapter nor its application to income and wealth undertaken in this section hinges on the psychological feeling of envy. Rawls argues that

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

[s]‌ociety may permit such large disparities in [ . . . ] goods that under existing social conditions these diferences cannot help but cause a loss of self-​ esteem. For those sufering this hurt, envious feelings are not irrational; the satisfaction of their rancor would make them better of. When envy is a reaction to the loss of self-​respect in circumstances where it would be unreasonable to expect someone to feel diferently, [ . . . ] it is excusable.45

However, as seen, liberals have strong reasons to maintain that it can never be justifed to experience a loss of self-​respect, and a consequential feeling of envy, merely because of the fact that others have more than oneself. Tis conviction is, among others, what drives the view that, if, in material terms, all have enough to satisfy their personal needs and individual interests, the appropriate reaction to material inequality is indiference.46 Focusing on envy instead of social status norms gets the problem back to front. What is largely missing in Rawls’s account is an explanation of how exactly ‘existing social conditions’ can make it reasonable and justifed to feel envy as a reaction to ‘large disparities in goods’47—​and why that is a problem of justice even if the overall distribution of goods fulfls the criteria set out in the preceding sections. 45 Rawls 1999a, p. 468. 46 See Frankfurt 1987, pp. 22–​23. 47 Rawls’s treatment of the relationship between material inequality, status, and self-​respect is ambiguous: as seen, he is sensitive to the problem, but in places also claims that, ideally, equality of basic liberties as demanded by his frst principle of justice should sufce to take care of it; see Rawls 1999, p. 478. See also ­chapter 6, fn. 25.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  249 Te answer to this is that even inequalities that would otherwise be acceptable may give rise to the formation and maintenance of inegalitarian status norms. Distributions infuence collectively enacted, self-​sustaining norms of social acceptability, for example setting standards of appearance and dress in public, which in turn govern the appropriateness of emotions like shame that accompany such norm-​based judgments of acceptability.48 Markers of social acceptability can be so narrow and exclusive of others, and thus restrict people’s opportunity to acquire and maintain robust self-​respect to such an extent, that they have to be classifed as threats to it, not just social contingencies which people must be asked to deal with themselves, if circumstances are otherwise just (see c­ hapter 6, 6.3.2). Feelings such as envy and shame (or apathy and resignation) among the worse of are reasonable and justifed if they can be traced to such norms, and are not if they cannot. Furthermore, such norms can be unjust also, and more simply, where those falling foul of them are thereby deprived of important social opportunities (6.4). Te tendency to the formation of dominant norms of social status based on diferential possession of income and wealth is, of course, not a conceptual truth, but based on an empirical assumption about human societies, especially about those characterized by extensive market economies. Status norms and their relation to distributions vary across societies; detailed empirical research is needed in order to answer the questions ‘When exactly does distributive inequality give rise to unjustly inegalitarian norms of social status?’ and ‘When does more distributive equality constitute the appropriate remedy?’ Relational egalitarians have to fnd the right balance of socio-​psychological and distributive strategies; only strategies of procedural protection can clearly be judged less appropriate, since law and political institutions do not constitute very efective tools for changing norms governing social status, but can only rein in further opportunities for domination of the status disadvantaged by the advantaged. Tere is no reason to suppose that strategies of distributive equalization should be ruled out, or resorted to only once others have failed—​even if remedies to the impact of unequal distribution on status inequality can cut both ways. Take an example involving inequalities in income and wealth: if income and wealth are sufciently reined in through a strategy of ‘blocked exchanges’, and more money can serve predominantly only as a means for the acquisition of more consumer goods, and—​as must always be the case—​other

48 See Smith 1976 [1776], Vol. 2, p. 399; Anderson 1999, p. 320, and 2008, p. 266.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

250 Implications justice-​relevant reasons for permitting distributive inequalities exist (see 8.3), then a socio-​psychological strategy can be appropriate, challenging such norms; for example, by pointing out in school that there is no good reason to link social acceptability to the wearing of brand name clothes. We might hold that the fundamental problem here is not inequality in income and wealth, but a—​historically and socially contingent—​spirit of ‘possessive individualism’,49 which connects status too closely to possession of material goods. Parallel to challenging such norms, attempts should be made to encourage the formation of a plurality of diferent social groups with divergent standards for assigning status, thus discouraging the formation of society-​ wide and uniform status norms that facilitate the emergence of objectionable status hierarchies.50 Tis then suggests that the limits set on economic inequality by concern about status norms are not easily pinned down, and might turn out to be quite permissive. On the other hand, paralleling the argument of the preceding section, a strategy of more distributive equality will make it less likely that problematically inegalitarian social norms emerge, in the frst place, since, unless the conversion of wealth into status is constantly and efectively blocked, the better of will continue to enjoy greater opportunity to form society-​wide social norms that exclude and marginalize others who cannot live up to the standards of living set by these norms. If the better of were a clearly circumscribed, not too large group, with group standards of their own, but these did not impact on society-​wide social norms any more than those of any other group, the problem would indeed be smaller. But it is hard to see how a society could guarantee not only blocked exchanges, but also the right sizes of the relevant groups in question that efectively prevent the formation of problematic status norms. To be more precise, it is hard to see how this could be guaranteed without becoming intolerably intrusive, without violating the limits of state coercion that a liberal conception of social justice must respect. Just think about the scenario where the better of are the majority of the population, and a societal underclass, even though enjoying what might be regarded as sufcient material means in absolute terms, still falls so far short of the wealth level of the majority that they cannot meet society-​ wide standards of social acceptability as equals.51 It seems wrongheaded to 49 Tis label is borrowed from Macpherson 1962. 50 See Rawls 1999a, p. 470 on ‘non-​comparing groups’; Scanlon 2018, pp. 36f; Anderson 2008, p. 264. 51 Tis scenario is arguably what drives one important methodology of poverty assessment in affuent Western societies: for example, the OECD statistics on poverty in developed Western countries

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  251 maintain that the right response here is to seek to get the majority not to create such status norms, rather than to lif the wealth level of the disadvantaged group, even if the consequence is not that the norm vanishes, but only that all satisfy it. Tus, relational egalitarian distributions will have to vary with the impact of the overall level of income and wealth of a given society, and their distribution, on status norms. It might be true that a society of individualists without any tendency to form status norms would be better, from many points of view. But as said, achieving such a society has to fall within the limits of what liberal institutions are permitted to do to those who are subject to their power. In conclusion, it is important to double-​check to what extent this is a problem of ‘ideal theory’, and to what extent it is one of ‘non-​ideal’ theory—​ and distributive solutions to status problems are thus only second best. It is not a problem of non-​ideal theory because people acting on status norms are, in some way, individually engaged in intentional, unjust, discrimination against the worse of, with the aim of wounding their self-​respect or depriving them of opportunities. As seen in c­ hapter 6, neither status norms, nor those acting on them, need to have such discrimination as their purpose. Te norms may simply express what people in the society in question regard as desirable aims for their lives. It is, however, true that, ideally, people ought to be aware of the existence of pernicious status norms, or the risk of their arising, and seek to stop acting on them, if they do have consequences for the self-​respect or opportunities of the worse of. So requiring them to do so would indeed be one available, purely ideal-​theoretical, solution. But not only is it sometimes hard to know for individuals that these are the efects of such norms (sometimes it might not be). Te fundamental problem is that these norms are self-​sustaining, because they are collectively—​albeit informally—​enforced (see 6.2.3 in ­chapter 6), so that would-​be non-​conformers face pressure to conform. Of that pressure, distributive strategies attacking the root conditions for such norm formation or maintenance can relieve all individuals. And it can do so more painlessly than an alternative strategy, which would consist of ruling out as altogether unjust all those aims whose widespread adoption might lead defne relative poverty as having less than half of the median income in the society in question; see Foerster and Mira d’Ercole 2005, p. 3. Another reason for a relative poverty measurement is that, under income inequality, markets might stop producing the goods that the worse of need; see Ryan 2002, p. 238. In such a case, money income itself becomes, to that extent, a positional good. For a discussion of positional goods, see Brighouse and Swif 2006.

252 Implications to the formation of inegalitarian status norms, and of policing this prohibition. Tis would no longer be a liberal strategy for achieving social equality. So the distributive strategy may well be second best, but only if it is judged from the vantage point of an excessively ideal ‘ideal theory’, which is insufciently attentive to pressures which dominant social norms generate on all, not just the disadvantaged.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

8.6.  Relational Equality and Equality of Opportunity How does the instrumental framework laid out in the preceding two sections for the case of income and wealth apply to the good of opportunity to attain desirable social positions? First, there is a strong link between inequality of such opportunity and scenarios of domination, stretching beyond the link between political inequality of opportunity and political domination analysed in the preceding chapter. Second, there is good reason to regard the connection between such opportunities and social status as less contingent than the one between status and income and wealth. Tese opportunities thus turn out to be the easier case for relational egalitarianism. It stringently requires a demanding conception of equality of opportunity to attain desirable social positions. Much of the recent debate on equality of opportunity has focused on equality of opportunity in education.52 As noted, this section has a more limited focus: it focuses on equality of opportunity to obtain social positions and ofces to which social and economic advantages are attached,53 and applies to equality of opportunity in education only insofar as it is instrumental to the former. Te previous sections have made clear that relational egalitarianism does not support a case for complete equality of opportunity for overall attainment of desirable goods in life, of which complete equality of educational opportunity would be an instance; it relies instead on a defeasible presumption that social goods, including education, ought to be distributed equally.54

52 Swif 2003 and 2004; Anderson 2004b and 2007; Satz 2007; and Brighouse and Swif 2009. For an empirically well-​informed internal argument that ‘adequacy’ in education as argued for by Anderson actually requires equality, see Macfarlane 2018. 53 Rawls 1999a, p. 72. 54 Defenders of an ideal of complete educational equality of opportunity based on luck egalitarian fairness also concede that their ideal has to be balanced against other values; see Brighouse and Swif 2009, p. 118.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  253 It does, however, require strictly that every member of society is given the educational resources necessary to be able to avoid relational injustice.55 Similarly, justice does not require substantive equality of opportunity between individuals of equal talent and commitment across all positions in the socio-​economic division of labour, as long as the social, economic, and political spheres are generally organized in a manner that is benefcial to everybody (see section 8.3). Tis includes stringent measures against nepotism and corruption, and the efective eradication of discrimination on grounds of pre-​identifable group characteristics, such as sex, ethnic origin, sexual orientation, and disability (‘formal’ equality of opportunity). But it is no more a social injustice if equally talented and committed people do not make it into equal social positions because they are diferentially lucky at, for example, meeting the right people at the right time, than it is if two people of equal attractiveness end up with unequally attractive partners for reasons beyond their control. However, this is merely to repeat that relational egalitarianism difers from luck egalitarianism. Relational egalitarianism yields a stringent rationale for a demanding conception of equality of opportunity, in one very important sense which is prominent in political discourse—​regarding access to desirable social positions and ofces. First of all, there is a special case for equality of opportunity based on the requirement to avoid domination. Tis concerns access to positions which confer decision-​making power, and applies in particular to top positions—​the positions of decision-​makers in politics and the bureaucracy, in infuential media, and in the economic sphere, such as corporate executives. Generally, relational egalitarianism mandates that decision hierarchies are as fat as possible, and ofer only minimal opportunities for power abuse (see section 3.2 in ­chapter 3); but, for cooperation to be successful, some positions of power have to exist, and insofar as they do, they will likely, at least as far as formal regulation is concerned, always give some people some opportunity to abuse this power, no matter how well controlled and circumscribed they are (­chapter 4, 4.5.1). For reasons of avoidance of such domination, it is then imperative that this elite (if it may be so called in a relational egalitarian society) be highly permeable, and not constituted by persons predominantly drawn from a narrowly confned social class. Otherwise, it might degenerate into a closed sub-​society of family dynasties and personal acquaintance, with its own

55 Anderson 1999, pp. 318f, 328.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

254 Implications norms of behaviour and social intercourse, which serve to exclude people from other social layers. In such a scenario, a body of privileged people in positions of power emerges, or maintains itself, who are neither sincerely willing nor able to properly take into account the interests of the people affected by their decisions.56 Tis is a reason to give special attention to the development of talents from families and social backgrounds that have not hitherto had access to top positions; the stringency of this requirement is not based on considerations of individual fairness, but on the wider social and political importance of these positions. At the same time, special attention to talents from such backgrounds clarifes the permeability of society not only to the talented people in question, but to all members of their social layers, and hence serves to discourage an attitude of servility and deference,57 which in turn makes it less likely that powerful individuals can get away with using the opportunities for abuse that their positions will ofer. Second, paralleling the argument of the preceding section, there is an instrumental case for a demanding conception of equality of opportunity based on considerations of social status. Te link between status and desirable social positions is arguably stronger, and less contingent, than in the case of inequality of income and wealth. Even if it should be possible, through a strategy of de-​emphasizing the social importance of money and encouraging the formation of non-​comparing groups, to largely sever the link between possession of income and wealth and social status, a parallel strategy for severing the link between positions of infuence and social status seems much less promising, for reasons laid out in ­chapter 6. To be sure, many positions are assigned social status that they should not have, due to unjustifable, but historically entrenched norms governing the perceived worth of the profession in question; relational egalitarianism objects to such status norms, as every liberal theory does. Te proper response here is to eliminate the norm; ensuring fair opportunity to attain such a position can always only be second best. However, insofar as the positions in question rightly confer power, it is almost impossible to think that they should not also confer status, or at least a privileged opportunity to attain it (6.3.1 in c­ hapter 6): for social cooperation to function, the power attached to decision-​making positions must not be exercised as ‘naked’ power, but as power based on perceived authority, which unavoidably constitutes a privileged opportunity for its bearers to

56 Anderson 2007, pp. 598–​606, 616. 57 Satz 2007, p. 625.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Distributive Justice  255 acquire higher social status. It is thus less convincing to claim that people should not assign status to such positions than it is in the case of money. Tis phenomenon need not be restricted to top positions, as long as reasonable diferentiations between diferent groups of social positions can be made: for example, jobs involving neither complex skills which are difcult to acquire nor powers over others will likely always fail to command much social status. Tere is thus a status problem that cannot reasonably be regarded as fully eliminable, even if all positions in question are properly designed and circumscribed, all irrational status norms, discrimination, and economically detrimental practices (such as nepotism and corruption) are abolished, and the link between wealth and status is broken. Te problem exists even if institutions have done everything they can to make sure that the people who occupy them really are the most qualifed. It appears in two guises: frst, it might be the case that the people who end up being most qualifed for the positions in question predominantly come from certain social groups and layers—​say, the middle classes. People from disadvantaged backgrounds are efectively denied access, because they (or their parents) lack the resources to develop their talents properly. Tis, in turn, has consequences for the formation of status norms, and self-​respect and self-​esteem, even where undue deference and servility might not be a problem. Knowing that one just does not belong to the class of people who can make it into such positions, or at least that one will constantly have to make eforts of a kind which are not required of the more privileged to have some chance, is a threat to self-​respect whatever one’s own particular talents and abilities: it constitutes a kind of marginalization, and is apt to engender a sense of disenfranchisement.58 Tis is another reason for relational egalitarians to devote substantial resources to ofsetting the efects of disadvantaged background on the development of talents.59 58 It need not always produce this sense in the psyche of individuals. It is possible that, for example, particularly combative and ambitious individuals manage to escape or repress it. Te fact that some people overcome this burden is not a justifcation for its existence. 59 How far exactly this requirement approaches Rawlsian ‘fair equality of opportunity’, which demands the removal of all diferential efects of social class on the development of natural talent (Rawls 1999a, p. 63), is a question for further discussion. On the one hand, the right to a meaningful family life speaks against the removal of all family-​transmitted advantage, at least insofar as it is non-​ fnancial. Tis is at least implicitly acknowledged by Rawls; ibid., pp. 64, 448. On the other hand, as Scanlon points out, what should count as relevant talent, in the frst place, does not depend on any ‘natural’ criteria, but purely on the nature of the diferential social positions that our best conceptions of social cooperation require us to institute in the interests of all; Scanlon 2018, pp. 48, 70. So the category of ‘natural talent’ lacks much specifc content, and a relational egalitarian society has considerable leeway regarding the kinds of developed talents it may require and whose development it may encourage.

256 Implications But, second, what about those who, even under highly permeable social arrangements, would sufer from low self-​respect, because they simply lack the talents needed to occupy any desirable position,60 no matter how good a training they might be given? As said, it would be unreasonable to hope that social status diferences can be entirely abolished, and, in the case of status attached to positions of power, fewer counter-​strategies seem available than in the case of income and wealth. But precisely because status diferences cannot be fully eradicated, a relational egalitarian society should consciously and constantly seek to keep alive a sense of fundamental equality independent of people’s particular characteristics and achievements, and, in this way, seek to shore up robust self-​respect and self-​esteem. Any ideal of social and political ideal equality must emphasize ‘the irrelevance of individual diferences for fundamental social and political purposes’.61 Tis can happen symbolically, for example, in political rhetoric, and in norms of everyday intercourse that infuence the assignment of social status. Ensuring that material rewards attached to desirable social positions are moderate has an important symbolical dimension as well.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

8.7.  Conclusion Taken together, the arguments of this chapter demonstrate that the distributive implications of relational egalitarianism are signifcantly more demanding than many of both its proponents and opponents think. It requires setting stringent limits to distributive inequality: it recognizes both intrinsic reasons of justice, of an expressive kind, in favour of distributive equality in social goods between participants of equal standing in social cooperation, and a set of additional instrumental reasons for limiting distributive inequality, in order to avoid risks of domination, and the emergence of status hierarchy. At least two important issues need further investigation. First, the argument for the expressive perspective made in Part I implies that a complete account of just relational egalitarian distributions must be sensitive to the particular ways in which social and political institutions generate and engender desirable distributive outcomes. As the expressive content of their 60 Scanlon 2018, pp. 35. 61 Schefer 2003a, p. 22. Bernard Williams discusses a similar tension between equality of opportunity and equality as a humanitarian ideal; Williams 1973, p. 247.

Distributive Justice  257

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

actions cannot always be fully reduced to the latter, there can be no guarantee that two sets of institutions doing equally well, in outcome terms, in fulflling the distributive requirements developed in this chapter will do equally well in expressive terms, overall. Terefore, any more concrete institutional proposals for the implementation of relational egalitarian distributive justice will, beyond being guided by these requirements, have to stay alert to this possibility.62 Te second issue concerns the extension of relational egalitarian distributive requirements to goods that are not purely socially produced. Probably the most important such good is health. Tis will be the subject of the fnal chapter.

62 One issue for enquiry is, for example, whether universal egalitarian welfare policies (see Schemmel 2015b) have better expressive qualities than policies specifcally targeting the worst of, which cannot be fully explained by the worst of tending to attain higher levels of benefts under the former. Tese systems might (also) do particularly well in terms of making clear to all that all are dependent on societal cooperation for securing basic goods.

9 Health Inequality

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.1.  Introduction If the argument of this book so far has been successful, it not only has established that there are stringent demands of social justice to eradicate inegalitarian social and political relations of domination and status, and has shown how exactly the stringency of these demands depend on the relation at stake. Te preceding chapter has also demonstrated how the social cooperation-​ based framework underlying liberal relational egalitarianism yields a plurality of rationales for limiting distributive inequality: such inequality is not only unjust where and because it engenders unjust relations between particular individuals, or groups; respect for people’s equal standing as participants in social cooperation requires that all inequality in socially produced goods and bads be justifed to them. Is there a good case, then, that this framework not only can vindicate the place of requirements of relational equality as particularly demanding and stringent egalitarian requirements within an overall conception of social justice, but also can serve as the basis for a reasonably complete such conception? Tat would be a conception that can account for the justice or injustice of a full range of social processes, including their distributive outcomes, without having to draw on any fundamentally diferent, and possibly competing, fundamental principles—​such as purely recipient-​oriented ones (­chapter  2), which determine the (in)justice of distributive states of afairs without any reference to how exactly these were brought about. It may seem doubtful that it can. Te preceding chapter focused mainly on the social goods of income, wealth, and opportunity. Even if its argument that inequality in such goods is problematic not only where it leads to particular inegalitarian social relations, but always stands in need of justifcation by justice-​relevant reasons, is sound, one might be tempted think that the position still sufers from a kind of ‘social relation fetishism’. Tat is, one may think that while it is certainly appropriate to include the overall relation of co-​production of social goods shared by participants in a cooperative scheme in its focus, this is insufcient Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0009

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  259 to render it a plausible basis for a complete conception of social justice. Tis is because it would seem to fail to give plausible, and reasonably complete, guidelines for the distribution of goods which are of elevated importance for individuals’ lives, but which are not, or not exclusively, co-​produced in social cooperation, but have a very signifcant natural dimension. Health is the most important such good. So, if liberal relational egalitarianism failed to give such guidance on the justice or injustice of health inequalities, this would be a failure which would, at the very least, call for reliance on diferent principles of social justice alongside relational egalitarian ones, leading to a kind of pluralist egalitarian position (2.4). Such principles may, for example, be luck egalitarian ones, according to which all individually unchosen inequality in health is unfair. Accordingly, the task of this chapter is to show that liberal relational egalitarianism in fact yields plausible, signifcantly egalitarian, and reasonably complete guidelines for the distribution of health, including principles for the distribution of healthcare. It would be too much to ask of this book to develop a complete conception of social justice; but if liberal relational egalitarianism does well for health, this is a good reason to think that there is little, if anything, to such a charge of ‘relation fetishism’.1 Te chapter demonstrates that it indeed does well, and that one important reason for this is a good ft between its focus on hierarchical relations and research in public health on the ‘social determinants of health’ (SDH). It thereby completes the argument of Part II of this book, which reinforces the argument of Part I by showing that liberal relational egalitarianism has determinate, demandingly egalitarian implications for political justice and distributive justice, overall. Te chapter is structured as follows. Section 9.2 defnes health, for the purposes of our cooperation-​based framework. Section 9.3 delivers the main argument: it shows that liberal relational egalitarianism can draw on three diferent approaches for incorporating demands of health and healthcare. It does not have to choose between them, but instead assigns the diferent resulting requirements a plausible order of priority. Tese approaches are, in this order, (a) an indirect, or derivative, approach, according to which health inequality caused by unjust social relations is itself unjust; (b) a direct approach, according to which all health inequality caused by social processes and practices which cannot be justifed by justice-​relevant reasons is unjust; 1 On matters of global and intergenerational justice, where a related charge might resurface, see the Conclusion.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

260 Implications (c) a natural vulnerability–​based approach to justifying universal access to healthcare, independently of how health defciencies are caused. Te section goes on to demonstrate how these three approaches apply. First, for some SDH, there is a strong justifcatory ft between relational egalitarianism and the respective social causes of health inequality: the reasons for which we should object to the hierarchical relations singled out in earlier chapters latch on well to the causal mechanisms through which these produce health inequality. Second, other SDH can be incorporated into liberal relational egalitarianism under the direct approach, up to the point where possible remedies to socially caused health inequality run up against prior requirements of liberal social justice. Tird, and fnally, it would indeed be fetishistic, and unjustifable, to care exclusively about socially caused health defcits, or about other health defcits only where and because they render individuals vulnerable to domination or status inequality. No cooperation-​ based framework can be plausibly restricted in this way. Liberal relational egalitarianism is no exception, and recognizes a separate, vulnerability-​based rationale for universal healthcare. Te section concludes by comparing the relational egalitarian account with Norman Daniels’s seminal account of how to incorporate concern about health and healthcare into a Rawlsian conception of social justice, and demonstrates how the clear ordering of health demands following from the former avoids an important incoherence besetting the latter. Section 9.4 considers potential objections to this relational egalitarian approach to health and healthcare. Two are particularly salient. Te frst points out that the approach is, despite its inclusion of equal access to universal healthcare, and the attention it pays to how social arrangements seize on, and can exacerbate the consequences of, diferential natural endowments, less generous to the victims of purely natural bad health than some rival views, such as luck egalitarianism. Tis is true, but turns out to be an inconclusive objection, because this feature of such views, in turn, generates corresponding, and very signifcant, objections to these, too. A more pressing, because more internal, objection argues that the view does not actually yield universal access to healthcare which is equally open to all, because it must be committed to treating patients sufering from identical health conditions unequally within the healthcare system, depending on how their condition was caused. It thus seems committed to a form of discrimination between patients which is problematically expressively inegalitarian, and therefore at odds with a central commitment of relational egalitarianism

Health Inequality  261 itself (­chapter 2). Te response to this objection points out that the priority ordering of the requirements issuing from the three diferent approaches to health inequality established in section 9.3 must indeed inform a society’s public health eforts as well as the level of coverage of its healthcare system, but need not translate into any form of discrimination between patients with identical health needs at the point of delivery of treatment. It is only such discrimination which is problematic—​if the overall argument of the chapter is sound.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.2.  Health and Social Cooperation Before developing the implications of liberal relational egalitarianism for health inequality, we should briefy delineate the concept of health most appropriate for this enterprise. Tis is necessary because the concept of health is hotly contested, and there is a wide variety of competing defnitions. It would be unrealistic to expect that this contest can be settled here, but we need to identify a working defnition that both fts the theoretical framework adopted in this book and is independently plausible. Tat defnition cannot be a very wide one, such as the one adopted by the World Health Organization (WHO), according to which ‘[h]‌ealth is a state of complete physical, mental, and social well-​being, and not merely the absence of disease or infrmity’.2 On this kind of view, the concept of health is dissolved into a general notion of well-​being, and accordingly, it would seem that there cannot really be any special requirements for health. Tere could only be applications of whatever general principles for the distribution of well-​being a conception of justice puts forward to issues that are more conventionally understood as health problems. It is thus better to adopt a narrower concept of health, more akin to the one from which the WHO explicitly distances itself. According to Norman Daniels, health is the ‘absence of pathology’. Pathologies include, but are not restricted to, diseases; they are any biological impediments to ‘normal functioning for our species’.3 Tis functioning is determined in the following, ‘biostatistical’ way:

2 https://​www.who.int/​about/​who-​we-​are/​constitution, last accessed 18 November 2020. 3 Daniels 2007, p. 37, with reference to Christopher Boorse’s work, especially Boorse 1997.

262 Implications [A]‌biological function can be defned as a causal contribution to a species-​ typical goal, such as survival or reproduction, and it is the task of the biomedical sciences, broadly conceived, to characterize these functions of organisms and their parts. A  departure from normal functioning is then simply a statistical deviation from the causal contribution of the relevant part.4

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

However, stronger normative reasoning about what counts as ‘normal species functioning’ must come in, at least at the following point: while certain kinds of biological shortcomings will be threats to ‘normal species functioning’ for all kinds of possible societies, clearly not all are. For example, being able to run at a certain top speed might have been essential for survival in the Stone Age, where being chased by a Smilodon might have been a relatively regular occurrence. Now, it is not essential. Conversely, a disposition to severe dyslexia is not a departure from proper functioning in a pre-​literate society, but is one in our current conditions, where literacy and numeracy are, for the most part, essential requirements for being able to participate fully in social cooperation. It is not implausible to maintain that, absent other problems connected to the biological causes of low top speed, such as, e.g., severe pain caused by bone deformities, a very slow individual can be fully healthy for us, but could not be in the Stone Age.5 Tis thought might seem to invite the conclusion that health is, in an important sense, ‘conventional’6: basically, societies just have to decide which impediments count as health impediments. Tese will simply be those for 4 Daniels 2007, p. 38. 5 Daniels is aware of such complications, and concedes that ‘what counts as a disease [or, more broadly, pathology, C.S.] category’ can be ‘relative to some features of social roles in a given society, and thus to some normative judgments, provided that the basic notion of normal functioning is lef intact’, ibid., p. 42. Tat would seem to take care of the dyslexia case, but, depending on what that ‘basic notion’ exactly consists of, it might still be that the very slow runner falls short of it. Daniels might then concede that he indeed has a health problem, but not one that should concern us in any way. Tat seems the right result, in terms of practical implications, but is still conceptually implausible. Tis person is functioning fully normally, for all our purposes; it should be precisely the task of a concept of health to indicate when that is not the case, due to biomedical causes. It is doubtful that we can get to enough of a ‘basic notion of normal functioning’ without already giving some consideration to social roles. What is crucial is a broad enough, and non-​arbitrary, account of such roles; see what follows in the main text. As opposed to that, it would not be a good objection to the ‘normal functioning’ account to argue that it is not informative because normality depends on social and technological context, in the sense of what can actually be achieved, or regularly is: for example, for most of human history, a ‘very high infant mortality rate’ was absolutely normal everywhere; Weinstock 2011, p. 426. Tis objection confuses what is statistically normal for diferent societies given technological capacities with biologically normal functioning of individual organisms. 6 Weinstock 2011, p. 428.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  263 which they wish to ofer remedies, such as individual treatment, or more collectively aimed public health interventions.7 However, that would be to give up too quickly. Te preceding observations suggested that health must be indexed to social cooperation, and refer to biologically normal functioning within it. To be sure, if we simply took a certain cooperative framework as given, that would indeed lead to a concept of health that is, in an important sense, ‘conventional’—​health problems are, then, those biomedical impediments that keep you from functioning as the currently dominant modes of cooperation demand. However, that is not a problem encountered by a conception of social justice which, as the liberal view does, anchors the notion of social cooperation in the fundamental idea of society as a fair cooperative enterprise between free and equal participants, aimed at enabling the full development and exercise of the two fundamental moral powers for all. Tese fundamental ideas have to inform the choice of the ‘dominant cooperative framework’8; in particular, cooperation must be so structured as to be as fully inclusive as possible of all those endowed with the moral powers. It must take a wide view of what counts as full cooperation, and thus also of the ‘benefts and burdens’ such cooperation produces. For example, those who, given currently typical economic organization, could not work ‘proftably’, in the sense of generating a sizeable surplus of societal income and wealth (which might then be open to egalitarian distribution) might nevertheless count as fully cooperating. Tis is the case, at least, if they participate in a broad range of social activities that generate social benefts, on such an appropriately wide understanding. We have seen in previous chapters that such benefts include the ‘social bases of self-​respect’, generalized and particular trust, opportunities for friendship (­chapter 5, 5.4.3), contributions to child-​rearing, and the constitution and maintenance of a wide range of societal subgroups which make distinctive cultural contributions to overall social life. It would be unworkable, for a liberal society, and unduly exclusionary and restrictive of freedom, to draw up a ‘balance sheet’ of goods production on which those who are to count as full cooperators have to ‘tick all boxes’, in the sense of generating net positive contributions. 7 Weinstock argues that liberal structures on the proper scope of health policy, which is otherwise in danger of inappropriately colonizing other policy domains (over which it tends to claim priority), should lead us to adopt a suitably narrow concept of health; ibid. Otherwise, such a ‘conventional’ approach could, in principle, also lead to a very wide concept, such as the one used by the WHO. 8 Buchanan et al. 2000, p. 259.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

264 Implications Tis makes clear that what counts as a health defcit must be specifed with such a wide notion of cooperation in mind. Health defcits are those biomedical impediments which keep one from participating fully in social cooperation in that, wide, sense. Tey are thus always problematic, and the question of justice in health and healthcare is the question of which kind of health defcits ground which claim to counteraction and remedy. In the remainder of this chapter, and especially where reference to empirical studies concerning the SDH are made, examples of health defcits will largely be more classical ones, which can be accounted for by the biostatistical model, without the complications introduced by the reference to a justifed dominant cooperative framework. For the purposes of the argument of this chapter, these ft well enough.9 However, it is essential to have this more complex concept of health fully in view from the outset. Tis is because, as shown, it makes clear that the kinds of capacities which must be demanded for full participation in social cooperation themselves need to be justifed at the bar of justice, and need to respect individuals’ fundamental claim to full inclusion, where it can at all be had. To overlook this is to risk inadvertently licencing severe forms of unjust relational inequality. Of course, in real social life, dominant cooperative frameworks are not chosen, as they are in hypothetical Rawlsian original positions. Tey evolve over time, are subject to signifcant path-​dependency, and are thus always likely to display constellations of required capacities of which only some are hard and fast necessities, while others are mostly due to the contingencies of historical productive development (but nevertheless hard, because very costly, to change radically). For example, that societies institute an intricate division of labour (3.2 in c­ hapter 3), and that such complexity requires a signifcant level of literacy and numeracy among all its participants for them to be able to navigate such societies and fnd their place in them, is, on a liberal view, such a necessity. It is hard to see how, otherwise, societal cooperation could generate both the freedoms and the resources needed to institute the rich possibilities for diverse conceptions of the good life that liberals prize. But how exactly diferent liberal societies organize this division of labour has to be open—​for example, regarding the physical 9 As does measuring overall health inequalities by using common health-​adjusted life year (HALY) metrics, and referring to common partial indicators such as mortality, life expectancy, and morbidity. Determining how exactly these might have to be adjusted, at times, to account for the considerations about cooperation just adduced is not a task for this chapter, but for a fuller theory of justice in health and healthcare.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  265 capacities deemed ‘normal’, and thus used for designing standard tools and machines for work, or the organization of the working day.10 Contemporary advanced liberal societies, at least, have reached a level of material abundance and technological sophistication which can allow for a high degree of organizational fexibility, and these considerations require putting it into practice. Te need to justify which kinds of cooperative capacities are required thus makes clear that those who have difculty cooperating fully within a given framework, but could very well do so in a diferent, equally justifable one, have a claim to inclusion and assistance in adapting which is, itself, a claim of social justice, not one of benefcence—​and one that can also require transforming the dominant cooperative framework itself. It is not the case that such individuals sufer from a kind of natural disadvantage, and it is, in this sense, open whether it is the individuals that need transforming, or the framework. Nor need they sufer from any health defcit; as seen, health defcits are those that impede cooperation under viable and justifable frameworks. It is easy to see how overlooking the interplay of cooperative requirements and risks of exclusion in favour of insisting on a kind of ‘normality’ which simply replicates what is currently needed to participate in dominant existing practices risks severe forms of domination. Somebody, whether it be an ofcial, a medical practitioner, or a political decision-​maker, who has the power to deny individuals full participation in social cooperation based on procedures of assessment of their capacities which are not constrained so as to give the fundamental claim to inclusion proper weight, has a particularly dangerous, and potentially pernicious, kind of arbitrary power. A misguided exercise of that power deprives the subjected individuals of much more than merely the end products of cooperation (so that even compensating the excluded with generous other kinds of benefts would do little, if anything, to address this problem). None of this is to say that which capacities to require, and which remedies and assistance to ofer, will not ofen be subject to reasonable disagreement. 10 For example, a ‘night owl’ with a late circadian rhythm might be at a much higher risk of chronic headache, and fatigue, when all her available work options are based on a ‘standard working day’ which starts too early for her. Her primary claim is not to recognition of her ‘pathology’, and help in adjusting her rhythm, nor to remedial medication for chronic headache and fatigue (which are, of course, pathologies), but to more accommodating work options. It is only if, on balance, these cannot be ofered, that the remedial claim to medication, or other ways of adaptation, should become our main focus (see further 9.3.3); see Aas and Wassermann 2016, pp. 600f., for a similar argument regarding the accommodation of disabilities.

266 Implications But this is not the same as to say that decisions about these matters, in the end, have to be conventional. Te considerations about inclusive cooperation just adduced do give guidance for proper deliberation about the subject, and make clear that those whose prima facie claims of justice are at stake (3.3.2 in c­ hapter  3) are subject to potentially severe domination already when they are not included in that deliberation on a footing of full equality (7.3 in ­chapter 7).11 With the concept of health as absence of biomedical impediments to full cooperation in hand, we can now turn to the implications of liberal relational egalitarianism for diferent kinds of health inequalities.

9.3.  Unjust Health Inequalities 9.3.1.  Tree Approaches and Justifcatory Fit

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Tere are three diferent ways in which relational egalitarianism can incorporate concern for health and healthcare, and account for the injustice of different kinds of health inequalities.12

11 Nor is it to say that a concept of health that is appropriately indexed to inclusive cooperation avoids all problem cases. Here is one: very severe pain at childbirth is, unfortunately, part of ‘normal species functioning’, and not as such pathological; Segall 2010, pp. 354f. And it occurs even in clinical settings where otherwise all risks of bad health outcomes to both women giving birth and their babies are fully taken care of, and the pain itself, and possible reactions to it, thus do not pose any further danger. However, holding that this may not be a health problem is a bullet which the view can bite. First, some women seem to agree that such pain is not unhealthy and, on balance, not a bad, and opt for ‘natural’ birth procedures that allow it—​they object precisely to what they see as the pathologization of childbirth. Second, reasonable cooperators can, and should, still agree that we should all insure each other against severe pain, at least when it is unwanted, even if it is not a health problem, and make appropriate remedies available for each other, socializing the cost. A somewhat related, but contrasting case, that of severe menstrual pain, is certainly included under health, because it impedes one’s participation in cooperation and, unlike childbirth, does not occur in a setting where the sufferer already has a claim to be safeguarded against all other connected health risks. A diferent kind of objection observes that any cooperation-​based framework will have difculty grounding claims to remedies for those with severe health defcits which render them permanently unable to cooperate, under any reasonable cooperative framework—​such as is the case for very severe, irremediable cognitive disabilities. Tis is not an objection for the present section, whose task was to single out a suitable concept of health; that concept correctly diagnoses a severe health defcit in these cases. It will be taken up in section 9.4. 12 For an excellent overview and analysis of these three diferent approaches, see Voigt and Wester 2015.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  267 Te frst is to take an ‘indirect’13, or ‘derivative’14, approach. On this approach, unjust health inequalities are those which are caused by unjust inegalitarian relations, that is, by those kinds of relations, such as domination, which are already condemned by our conception of relational equality, on independent grounds. It is intuitive both that any bad efects of such relations on a further justice-​relevant good such as health are, thereby, also unjust, and that these bad efects, in turn, increase the overall injustice of these relations—​at least if there is a reasonable match between the grounds advanced for condemning these relations, and the causal mechanisms through which they produce such efects (on which more in a moment). Terefore, this approach must play an important role within our overall approach. However, it would seem to be merely a piece of good luck if it, in fact, covered all health inequalities we have good reasons to object to; and it is unclear why one should be able to object health inequality only if it is already attached to other injustice. Tus, there must be scope for diferent approaches leading to further requirements. Te second approach is more direct, and relies on the principle, developed and defended in the previous chapter, that all socially produced inequalities in salient goods and bads require justifcation.15 Health is a very salient such good. Terefore all socially produced health inequalities require such justifcation, and do so even if there should otherwise be nothing unjust in the process at hand. Tis, too, is thus an approach which will have to play an important role. As it is potentially very wide—​a vast array of diferent social processes are able to produce some health inequality—​it will be important to indicate the considerations that limit it, and serve to pin down which place health concerns should occupy in comparison to other goods and bads. Tird, we can take recourse to the expressive perspective with regard to health inequalities also in a more general fashion, beyond already unjust inegalitarian relations, and health inequality which is socially produced in other ways.16 We can ask what a society expresses about the standing of its members when it fails to respond to their health needs, and, for example, leaves a remediable health defect untreated, in whichever way it was caused. 13 Ibid., p. 220. 14 Gopal Sreenivasan, ‘Justice, Inequality, and Health’, Stanford Encyclopedia of Philosophy, https://​plato.stanford.edu/​entries/​justice-​inequality-​health/​, last accessed 18 November 2020. Te most prominent example of such an approach is Normal Daniels’s—​or, to be more precise, one of the two approaches Daniels puts forward (see 9.3.5). 15 Voigt and Wester 2015, pp. 214f. 16 Ibid., pp. 211f, 225f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

268 Implications Tat may well be an unjustifable form of neglect. However, we have seen that the expressive perspective is not self-​interpreting. One of the main lessons of c­ hapters 2 and 3 was precisely that it needs to be flled out with a principled framework for deriving requirements of social justice, drawing on background ideas of the person and society: wider requirements of health and healthcare need to be justifed by reference to these, drawing on how natural vulnerability to health risks endangers individuals’ capacity to participate in social cooperation and pursue their plans and projects. Te main contention of this chapter is that liberal relational egalitarianism does not have to choose between these three approaches, but can draw on all of them. Te task is thus to indicate what each covers, as well as how their respective demands stand to each other. Tey stand in an order of priority from the frst to the third: health inequalities caused by unjust relations are more unjust than other socially caused ones, which are foreseeable, and avoidable at reasonable cost, which are, in turn, more unjust than those not so caused, other things—​that is, the magnitude of the health inequality in question—​being  equal. Before working out their reach and place in more detail, however, it is useful to spell out a few methodological considerations about justifcation. As noted earlier, the aim is not merely to apply liberal relational egalitarianism to health and healthcare, but to investigate whether what it has to say about these goods in fact strengthens the case for it as a conception of social justice, overall. Tis will be so when empirical research about health inequality, its causes, and its possible remedies, coheres well with its requirements. As Norman Daniels notes, with regard to the ‘indirect’ or ‘derivative’ approach, this is the case if the same inequalities that the conception so far has singled out as unjust, for independent reasons, also turn out to be harmful to health.17 It is the case, for the other two approaches, if, given what we know about (other) causes of health inequality and possible remedies, the conception promises to account for enough of these, at reasonable cost, without any need to resort to extraneous principles. A theory of social justice issuing policy recommendations that lead to satisfying results in more

17 Daniels 2007, pp. 100f. For an elaboration of the underlying method of justifcation, see Daniels 1996. Fit with empirical evidence is standardly classifed as one way of ‘widening’ refective equilibrium (a ‘narrow’ equilibrium is one between a set of considered moral judgements and a set of moral principles accounting for them, not yet taking into account competing moral principles, or social theory and empirical research which might bear on the acceptability of the principles).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  269 policy areas than a rival one is, other things being equal, the better theory: it has greater reach. However, we must also note that, for a theory to draw a strong form of support from such convergence, the convergence must not be merely accidental, in the sense that the theory recommends a degree of socio-​economic equality which would also likely lead to a suitably egalitarian distribution of health, and good health overall, but for reasons that have little to do with the causal mechanisms underlying good health. Take the example of an ‘individualist’ distributive conception of equality of resources, which requires limits to economic inequality on the grounds that it is fair if all can buy equal amounts of goods for themselves. However, if it turns out that more equal opportunity to buy things (including healthcare) is not what makes people healthier in more egalitarian societies (as many public health researchers indeed argue),18 then the support generated by the fnding that the degree of equality demanded by this conception also makes for (roughly) equal health would be weak.19 We have to require a higher degree of what may be called justifcatory ft: as noted, the reasons for which the theory in question objects to inequality must latch on suitably to the causal mechanisms which produce an inegalitarian distribution of health. If that is so, the greater reach of the theory is a theoretical virtue because the diferent reasons that underlie its policy recommendations shed light on, and mutually reinforce, each other (2.4 in ­chapter 2). Tis ft then increases our justifed confdence in the theory overall. It is, for example, a reason to turn to it also when approaching further domains of justice, where we do not have confdence yet in our judgements and intuitions.20 So we need to check whether liberal relational egalitarianism applied to health exhibits such a ft, examining how it draws on the three diferent approaches outlined earlier.

18 See the references on research into the SDH in the next subsection. 19 Tis ‘individualist’ conception of resource equality is merely an illustrative example; it is not meant to refer to Dworkinian ‘equality of resources’. Te latter is best interpreted not as demanding that everybody has equal opportunity to buy healthcare if they wish, but as requiring compulsory collective health insurance issuing in universal healthcare, and as understanding collective public health measures as another form of such insurance; see Dworkin 2000, ch. 8. 20 Tis could, for example, be the domain of genetic modifcation beyond issues of health. Of course, this reason is defeasible; the theory might turn out to do badly for it. What it does is give rise to a justifed ‘head start’ over other theories which have not yet demonstrated similar reach and ft.

270 Implications

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.3.2.  Sickening Hierarchies: Derivatively Unjust Health Inequality It is not only the case that those health inequalities which are caused by unjust relations are, therefore, unjust, and that this increases overall injustice. Te argument for these specifc health inequalities also being themselves more unjust than others is straightforward: in a slogan, making you sick in unjust ways is the worst possible way of making you sick. Somewhat more elaborately put, a cooperative scheme causing health defcits, or exposing individuals to signifcantly increased risks thereof, by subjecting them to the arbitrary power of others, or to inegalitarian status norms instituting a socially coordinated and self-​sustaining process of devaluation of their characteristics, talents, or ways of life, thereby also shows a particularly pernicious kind of disregard for their interest in good health. Tis way of disregarding that interest is worse than other ways, even if resulting health defcits are of equal extent. So much for the principled argument; the salient question now is whether there is indeed a good ft between the central requirements of liberal relational egalitarianism and important and well-​documented social causes of health defcits. Te most straightforward way to argue for such a ft draws on the ‘psychosocial model’ of the SDH, developed and put forward especially by two of the most prominent pioneers of research into the connection between inequality and public health, Richard Wilkinson and Michael Marmot.21 According to this model, social inequalities of position, such as those instantiated by hierarchies of power and status,22 put those at their wrong end at risk of more or less constantly experiencing increased anxiety, stress, and fear. Constant exposure to stress factors reduces the capacity to return to unproblematic levels of stress hormones. Tis is bad for health because, for example, it increases the risk of coronary heart disease.23 Crucially,

21 Marmot and Wilkinson 2005. 22 Wilkinson cautions that it is hard to distinguish between the efects of hierarchies of social positions and those of economic inequality in resources and wealth: ‘Material standards are so closely associated with social position because, among animals and humans alike, the allocation of scarce resource in ranking systems is based on diferences in power. Tis makes it almost impossible to design studies capable of distinguishing, unambiguously, between the efects of social position and material wealth in human populations’ (Wilkinson 2005, p. 342). We will return to such difculties later, when noting objections to the psychosocial model. 23 For a detailed explanation of the biological pathways that connect individuals’ reactions to being at the bottom of hierarchies—​such as fear, and resulting impulses to fght-​or-​fight—​to long-​term health damage, see Brunner and Marmot 2005.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  271 such risks are not confned to the very bottom of social hierarchies, or, in the case of economic inequality, to those who clearly qualify as poor. Instead, overall, health displays a social gradient across all levels of socio-​economic status, as measured by standard indicators such as income and education;24 with the gradient being steeper at the lower ranges of the distribution than at the upper levels.25 Health inequalities also seem to be present within hierarchies among the relatively well-​to-​do and well-​educated, as evidenced by the well-​known Whitehall studies, which found a gradient in health according to rank among British ministerial civil servants. All of these enjoyed both adequate material living conditions and equal access to the healthcare system.26 However, it is important to note that the psychosocial model, while very well established in public health research, is not uncontested. Tis is not surprising, as its attempt to explain health inequalities by way of a direct and relatively uniform causal link leading from macro-​level inequalities in social position through negative reactions such as fear and stress to worse health outcomes is very ambitious. Critics point out that it does not account suffciently for how the impact of such inequality in resources and social position can be mediated by diferent kinds of material conditions experienced by the better and the worse of, and is thus also in danger of overplaying the impact of perceptions of inequality on health.27 Similarly, others argue that diferent kinds of social factors, such as income and education diferentials, which, as just noted, are ofen grouped together under an umbrella measure of ‘socioeconomic status’ in public health studies, may actually cause health inequalities through diferent pathways not matching any reasonably unitary explanatory model.28 Fortunately, for relational egalitarianism to draw sufciently strong support from research into the SDH, the psychosocial explanation does not have

24 Note that this is a diferent concept of status from the one employed in ­chapter 6, which is based on social esteem, though resulting status classifcations may ofen coincide. 25 Daniels 2007, p. 85. From this it follows that, other things being equal, achieving a greater degree of socio-​economic equality would not only make the worse of healthier, but ‘would improve aggregate health and would have little efect, if any, on the best-​of groups’, ibid. 26 Marmot et al. 1984; Marmot et al. 1991. 27 Lynch et al. 2000. See p. 1202: ‘Under a neo-​material interpretation [their alternative explanatory model, C.S.], the efect of income inequality on health refects a combination of negative exposures and lack of resources held by individuals, along with systematic underinvestment across a wide range of human, physical, health, and social infrastructure’. 28 For a critical overview, see Cutler et al. 2011. Such pathways can also be quite indirect, for example, where diferential education leads to diferential health-​related behaviours, difering ability to process relevant information, and difering capacity to appropriate new health technologies; ibid., pp. 141f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

272 Implications to be true across the board. On the one hand, at least for some important environments which are already highly relevant for relational egalitarians for independent reasons, the link from power and status inequality to stress is well-​documented. Tis is so for experiences of racial discrimination,29 of hierarchical workplace relations, which accord those at the bottom little autonomy to design their own contributions, and control about what it is they have to do when, leaving them at the mercy of superiors,30 and of the loss of social position and status linked to becoming unemployed, extending to precarious employment and the consequent constant fear of becoming unemployed.31 On the other hand, other salient inequalities, such as especially macro-​ level inequality of political infuence (­chapter 7), can produce worse health for the less infuential through mechanisms not centrally drawing on stress and fear, without such cases being any less central for relational egalitarians. Consider the following cases: the rich may simply use their superior political power to deny those in worse positions equally good background conditions for health, such as equally good education, equally good infrastructure, and equally healthy and pollution-​free residential areas.32 Tat power is dominatory, whether or not its existence or exercise stresses the worse of, and whether or not that stress itself (also) causes worse health. Similarly, racial discrimination results in, or reinforces, residential segregation, with those discriminated against tending to be confned to worse neighbourhoods, in terms of housing quality and environmental health risks;33 racial bias among medical practitioners results in worse access to needed treatments for those discriminated against,34 and practitioners not constrained to avoid and eradicate such biases dominate their (would-​be) patients;35 migrants without proper knowledge of their entitlements, or of the local language, or without papers, are ofen assigned the most dangerous and unhealthy jobs, with insufcient protection.36 29 Nazroo and Williams 2005; see p.  257 on how fear of becoming a victim of racism already worsens health. 30 Marmot et al. 2005. For some doubts that rank-​induced stress is signifcantly at work also in relatively privileged work environments such as Whitehall (see fn. 26), see Deaton 2013, p. 271. 31 Bartley et  al. 2005; see p.  87 for fndings of health damage due to job insecurity:  health deteriorates before people actually lose their job, due to anxiety. 32 Deaton 2013, p. 277. 33 For an overview of the evidence, see Braveman et al. 2011, pp. 387f. 34 Cutler et al. 2011, p. 150. 35 Tis, of course, assumes that such healthcare provision is a requirement of justice (9.3.4): unconstrained racial bias constitutes domination as soon as any such provision is required. 36 For a vivid and disturbing account of the experience of many Turkish ‘guest workers’ in Germany in the 1980s, see Wallraf 1988.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  273 Finally, it is important to note that the derivative approach to health inequalities is, in principle, also capable of covering aspects of such inequalities which are more proximately caused by diferential individual behaviour, such as smoking or not. Tere is a socio-​economic gradient in smoking; if unjust inequality makes it harder for the disadvantaged to quit smoking, or not to start, the derivative approach applies. Factors at work may include, again, stress induced due to multiple constraints, such as time and budget constraints, disproportionately faced by poorer people, which can also increase the desire for at least some instant gratifcation (‘I really need a smoke now’), even at the price of future harm,37 and can reduce volitional strength to undertake lifestyle changes; and low self-​esteem, which is a well-​known risk factor for drug addictions.38 Of course, in line with the caveats outlined earlier, such psychosocial models of the causes of individual behaviour have to identify the exact pathway from social disadvantage through individual psychological dispositions such as low self-​esteem to consequent harmful behavior. Furthermore, for the derivative approach to apply, we also need to inquire whether the conceptions of self-​esteem drawn on in the respective studies latch on well enough to the specifc conception of self-​respect as a capacity for robust orientation in diferent, including challenging, circumstances that, as seen, liberal relational egalitarians have special reasons to care about (6.3.2 in c­ hapter 6). Only deprivations of the social bases of this kind of self-​respect are already unjust, as demanded by that approach. However, even if that should not be so, such ways of causing health defcits may fall under the second, direct approach to health inequality, to which it is now time to turn. Derivatively unjust health inequalities are not the only unjust ones; they are the ones that give rise to the most stringent claims for removal and counteraction.

9.3.3.  Other Social Determinants of Health:  Te Direct Approach Tis second, direct and more general, approach rests, as seen, on the requirement that all socially produced inequalities in salient goods and bads requires a justifcation by justice-​relevant reasons. Such reasons are those

37 Marmot 2004, pp. 45f.

38 Jarvis and Wardle 2005, pp. 224, 231.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

274 Implications grounded in the background conception of society as a fair enterprise of cooperation among free and equal cooperators. As health is a very important personal good, for participation in cooperation as much as for individual plans and projects (whose enablement cooperation serves), production of health inequality must be capable of being, on its own, a reason to judge a social process unjust—​and to judge so conclusively, if no countervailing considerations can be found. To rely exclusively on the derivative approach is to underplay the importance of health, no matter how good and far-​reaching the ft between our independent requirements of relational egalitarian justice and the SDH may be. We not only care about threats to our health due to unjust causes. Why should health be treated so diferently from other goods, such as liberties and appropriately equal opportunities? Instead, social processes which cause health inequalities but are not derivatively unjust should be assigned a lower injustice ranking than health inequalities of equal magnitude condemned by the derivative approach. So much, again, for the principle; as before, the task is to give some indication as to when and how requirements based on this direct approach apply. Te thrust of the investigation must, in an important sense, be opposite to the one regarding the derivative approach and the SDH, where the task was to show that their ft is good, and wide enough. It must show how such requirements can be limited. Tis is because they could be, in principle, enormously wide—​depending on the weight put on health, they could condemn any social process or practice which produces some health inequality, no matter how well it abides by other egalitarian constraints, and how benefcial it is otherwise, overall. Tis would amount to treating health as a fully dominant good; that cannot be compatible with a liberal conception of social justice. So we need to develop some guidance for carrying out trade-​ofs between health and other justice-​relevant goods, including guidance on when they must be ruled out. To illustrate, public health researchers ofen stress the importance of ‘social cohesion’ and positive, community-​based relations for people’s health—​ across society at large, but also within particular neighbourhoods.39 If it is true that such cohesive environments produce better health, then there is, prima facie, a liberal justice-​based case for creating them, and facilitating their maintenance. Tis is one important way of incorporating concern for a more positive, communitarian, ideal of social equality, beyond

39 See, for example, Stansfeld 2005 and Shaw et al. 2005.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  275 non-​domination and the prohibition of pernicious status hierarchy, into a liberal conception. What matters is not the intrinsic—​perhaps also impersonal (5.3.3 in c­ hapter 5)—​value of such positive relations but, instrumentally, their impact on health. However, this can be taken too far, as social cohesion can be brought about and maintained in ways incompatible with basic liberties of the liberal and democratic kind, or other key liberal requirements. For example, Wilkinson celebrates the spirit of ‘comradeship’ which he thinks responsible, among other things, for the improvements in aggregate level and equal distribution of health in communist Eastern European states between the end of the Second World War and the early 1970s (afer which public health apparently failed to improve further despite continued economic growth).40 Tese states maintained such cohesion at the expense of basic liberties. Similarly, but somewhat less dramatically, Marmot, in an analysis of Japan’s excellent health results, in terms of both aggregate level and egalitarian distribution, celebrates the place of ‘tradition’ in Japanese economic life,41 as well as a general reluctance, in the resolution of social conficts, to rely on the claims of individuals, rather than on communal considerations. His example is one of neighbours complaining about loud music not because it is a nuisance to them, but because it supposedly brings shame on the neighbourhood.42 Even where the state does not withhold basic liberties, dominant social norms of an individuality-​efacing kind can present serious obstacles to people’s capacity to develop and realize their own conception of the good. While basic liberties must serve as a hard and fast constraint on eforts to limit health inequality,43 illiberal aspects of more difuse social norms may be able to be tackled in diferent ways, while seeking to keep their salubrious efects in place as much as possible. How exactly to strike the trade-​of between the loss of some, non-​basic, liberties brought about by such policies and norms and their good efects on aggregate and equal health is not something that can be fully determined at the level of principle, to then deliver it to democratic decision-​making as a blueprint to be implemented. Te considerations introduced by this second approach should function as inputs to democratic deliberation which can help to structure it, and select between 40 Wilkinson 1996, pp. 123f. 41 Marmot 2004, p. 175. 42 Ibid., pp. 176f. 43 To be more precise, the most stringent requirement here is that of keeping ofcials from dominating over the basic liberties in the name of health (4.3.1 in ­chapter 4).

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

276 Implications diferent possible trade-​ofs. Cases of social processes and practices which produce some health inequality but are benefcial overall, and do abide by other egalitarian constraints, also call for intermediate solutions, which keep such practices in place, but introduce remedies for their bad health efects—​ e.g., through more stringent and far-​reaching health and safety measures, or increased remedial healthcare.44 One may be tempted to pursue a more ambitious line of defense against restrictions of liberties on grounds of health; one that is more internal to health as such, and denies that there is a signifcant trade-​of, in the frst place. Ardent liberals might push the argument that what is ultimately valuable about health is not that it enables us to live as long as possible, but that it enables us to carry out life plans and projects of our choosing—​and that is the very same ability which is curtailed by such restrictions. Tis line of argument is successful against the limiting case of a health regime which denies individuals exercises of all liberties which might carry some health risk. Such a restrictive regime would indeed have lost sight of why health matters, on a liberal view. However, it is unsuccessful against any regime which safeguards a core area of basic liberties, and admits some trade-​ofs between health and liberty beyond that; denying that such trade-​ofs might have to be struck amounts to a kind of liberal perfectionism. Liberal neutralists have to accept that some individuals may really only value health for tranquillity and longevity, and care less about extended options. Furthermore, it is also a position with very unpalatable implications in cases where more liberty for all is connected to bigger health risks for some (e.g., due to possible interaction efects with residual social and economic inequalities which are not independently unjust). In this case, our extended options may put you at risk of losing some years of life, but not me. Denying the relevance of the trade-​of then seems breezy, to say the least. Tis second approach thus has considerable scope to condemn, and require remedies for, health inequalities caused by social processes and 44 Furthermore, as causal pathways from overall social and economic inequality to health inequality are, as seen in the previous subsection, complex and ofen contested, we may, in many cases, have higher causal confdence in such remedial measures, to start with. Tis is not to doubt that we should sometimes intervene at both ends at the same time. For example, it is a consistent fnding of public health research that parental disadvantage is strongly associated with increased health risks for children. Bad childhood health then ofen reverberates through later life, resulting in a cycle of reproduction of signifcant disadvantage; Deaton 2013, pp. 266f. Whether or not such parental disadvantage is already unjust for other reasons, this grounds a case for both improving targeted health services and measures for children, and limiting overall inequality.

Health Inequality  277 practices which are not, in themselves, unjust in any other way. But it can also be appropriately reined in by liberal constraints.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.3.4.  Natural Health Inequality and Healthcare Taken together, the indirect (or derivative) and the direct approach to socially created health inequalities will go a long way towards accounting for many of the most signifcant such inequalities besetting our societies. However, this does not sufce to lay to rest the worry about ‘social relation fetishism’ raised in the Introduction. Caring exclusively about socially created health inequality would be unmotivated. Similarly, it would be unjustifable to care about diferently caused health inequality only where and to the extent that it engenders relational inequalities among individuals.45 Tat should be a particularly stringent reason to remedy it, but not the only one. On any plausible understanding of what societal cooperation is about, the protection and maintenance of health must be counted as among its main objectives, in a general fashion, whatever the causes of bad health. Tis is so quite simply because all humans are naturally vulnerable to injuries and diseases, which threaten their capacity to participate in social cooperation, and may thwart their personal plans and projects at any time. Health protection and maintenance are thus, from the start, among the core goods at whose provision any reasonable conception of societal cooperation must aim. Accordingly, liberal relational egalitarianism must include a general requirement of health protection, through traditional measures of public health, as well as the delivery of healthcare. Failing to respond to health needs which can be met, even if they are not caused by social institutions and processes tackled by the frst two approaches, is thus rightly interpreted as expressing objectionable neglect. A relational egalitarian society has to institute a universal healthcare system. Furthermore, such a system must not rely on merely residual general health 45 Berkey 2018. For an insightful exploration of how socially caused health inequality can in turn lead to stigmatization and loss of social status, for example for those unable to fnd unemployment due to bad health, see Haverkamp et al. 2018. Not all the resulting injustices identifed by Haverkamp et al. qualify as salient inter-​individual relational inequalities according to the criteria developed in preceding chapters, however: a signifcant, socially patterned ‘risk to enjoy fewer pension years in good health’ (ibid., p. 316) than others is, in itself, more plausibly regarded as a simpler distributive unfairness which a cooperative scheme should avoid, according to the general presumption of equality defended in 8.3 in c­ hapter 8.

278 Implications insurance: making access to, and level of, healthcare contingent on variable social factors such as employment, as happens in societies organizing healthcare as part of job packages (such as the US), entrenches status inequalities,46 renders individuals vulnerable to domination by ofcials and bosses, and obstructs their capacity to plan their lives and projects, as it introduces uncertainty about the kind of provision they will be entitled to at any point in their lives. What the argument of this chapter so far has demonstrated is that bringing about health defcits by way of relational injustice, or other social processes and practices, amounts to expressing something worse than mere neglect, so the requirements resulting from the indirect approach, from the direct social approach, and from the natural vulnerability-​based approach have to stand in an order of decreasing priority. Tat priority is not lexical, but, as noted, holds for health defcits of equal magnitude. It would be crazy to think that a socio-​economically induced, avoidable propensity to get a cold is more unjust than letting somebody die of a Spanish fu which could be easily treated, but is randomly contracted, or vulnerability to which was enhanced by some particular genetic disposition.47 With this third, most general requirement in place, liberal relational egalitarianism can claim to deliver reasonably complete and comprehensive guidance for public health and healthcare. To confrm that this guidance really is plausible, we need to tackle potential objections to it.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.3.5. Daniels’s Conception of Just Health Before delving into these, however, it is worthwhile to compare the position arrived at to a very infuential rival conception of just health, the one developed by Norman Daniels. Tis is particularly instructive as Daniels’s conception is similarly cooperation-​based, and broadly Rawlsian, and the account just developed is evidently indebted to it. Te upshot of this comparison is that liberal relational egalitarianism gives a clearer and more coherent 46 Voigt 2018, p. 455. 47 One might seek to object that this priority order thus ends up giving too little, or too imprecise, guidance for health policy; too much will still be lef to case-​by-​case balancing in democratic deliberation and health practice. However, stricter priority rules would not only lead to unacceptable results; nothing within the argument speaks for them. Te proposed order is clearer and more coherent than the one issuing from the most prominent competing cooperation-​based approach to health justice, as we will see in the next subsection.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  279 account of the injustice of health inequalities, and accordingly yields clearer recommendations. It thus represents a signifcant improvement over Daniels’s account. In an important part of his account as presented in Just Health,48 Daniels takes an indirect (derivative) approach, and argues that the degree of socio-​ economic inequality permitted by Rawls’s conception of ‘justice as fairness’ would lead to few, if any, health inequalities which ought to be considered unjust—​and that this strengthens the case for it.49 Good aggregate and equal health would thus be a—​particularly welcome—​‘byproduct’50 of Rawlsian justice. It would do well in addressing the SDH, despite not having been developed with this issue in mind. Tis is correct; a very signifcant part of the argument made in subsection 9.3.2 about the justifcatory ft between relational egalitarianism and evidence on the SDH applies to ‘justice as fairness’, too. It stresses the importance and priority of political equality, for example, dedicates much attention to the ‘social bases of self-​respect’, and limits distributive inequality through the diference principle, which would also help to reduce or eliminate domination and status inequality (see ­chapter 8)—​even if, unlike relational egalitarianism, it does not rest on any sustained treatment of domination and unequal social status, which, as seen, can be fruitfully connected to much of such evidence. However, the objection to be pressed here is not that the justifcatory ft between relational egalitarianism and the SDH is better than the one achieved by ‘justice as fairness’. We have seen that the psychosocial model of the SDH, on which that argument would largely have to rest, is not uncontested, so a better ft cannot be demonstrated conclusively. Daniels’s position faces a different objection. Tis objection is, simply, that the ordering of the demands of just health issued by his position is unclear in principle. His initial position, and original infuential contribution to the political philosophy of health, was to incorporate concern with health under an extended Rawlsian principle of ‘fair equality of opportunity’.51 Te analysis of the SDH and ‘justice and fairness’ just mentioned is primarily meant to double-​check that this derivative account converges with the original opportunity account on the kind and magnitude of health inequalities they rule out. Given that it does, 48 Daniels 2007. 49 Ibid., pp. 95f. 50 Daniels et al. 2000a. 51 Daniels 2007, c­ hapter 2; see Daniels 1985 for the original extension of fair equality of opportunity to healthcare.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

280 Implications the justifcation for the overall, two-​pronged, Rawlsian approach is supposedly strengthened further.52 However, critics have noted that it is not evident that these two accounts could not come apart, and have urged Daniels to clarify their relationship.53 Te objection broadly follows this line of criticism, and points out that it is not clear how much work the incorporation of a requirement of meeting health needs into a principle of fair equality of opportunity actually does in Daniels’s consolidated account, as presented in Just Health. As is well known, in Rawls’s original framework, ensuring fair equality of opportunity is lexically prior to limiting general socio-​economic inequalities by way of the diference principle. Accordingly, if health was already incorporated into that prior principle, then, for example, double-​checking, as Daniels does, that the diference principle would not engender any objectionable health inequalities would be superfuous, since these would already be taken care of. Daniels shies away from this conclusion, and accepts that the equality of opportunity-​account does not actually require equal health, all things considered.54 Tere is good reason for this, since such a stringent interpretation of the extended fair equality of opportunity principle would indeed be implausible. Fair equality of opportunity is lexically prior to the diference principle in the original Rawlsian framework because it is doubly restricted: it requires only the removal of socially created obstacles to equal opportunity (as far as this is consistent with the existence of the family), not that of other obstacles, such as sheer lack of talent;55 and it only demands equality of opportunity in the competition for desirable ofces and positions. Daniels argues that health must be regarded as important for equal opportunity not primarily understood in this competitive sense, but in a wider sense, simply because it is of great importance for being able to participate in cooperation, and to choose and carry out one’s life plan.56 Tis is plausible (9.2). However, in the case of ‘justice as fairness’, it is all of the principles taken together that instantiate such a wide conception of (roughly) equal opportunity to choose and pursue one’s life plan. Broadly speaking, ensuring such opportunity is one of the primary ambitions of all liberal egalitarian 52 Daniels et al. 2000b. 53 Anand and Peter 2000. 54 Daniels 2007, p. 99. 55 Recall, however, that the idea of natural talent does not have much specifc content; see fn. 59 in ­chapter 8. 56 Daniels 2007, pp. 59f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Health Inequality  281 conceptions of social justice, which distinguishes them from others.57 In Rawls’s case, the diference principle is explicitly designed to provide individuals with the material means to do so, and to ensure that inequality of such means is limited. It seems that it should then also be the entire conception that is extended to the case of health. Te natural way of doing so would be to explicitly add the latter to the list of ‘primary goods’ it ranges over. Daniels wants to refrain from adding health to the index of goods covered by the theory in this manner, because a longer index complicates the task of weighing the goods fguring on it against each other, which is necessary to be able to make interpersonal comparisons.58 However, his position would avoid this problem only if it incorporated health into equality of opportunity as it stands, with its lexical priority over the diference principle—​and we have just seen that he does not do this, since it not only would make his discussion of ‘justice and fairness’ and the SDH superfuous, but also would simply be implausible. Instead, one is lef wondering just which health inequalities the extended and watered-​down version of fair equality of opportunity as applied to health now covers, and to which extent. Te merit of incorporating health frst and foremost into fair equality of opportunity is then illusory.59 At the same time, this principle cannot simply be dropped from Daniels’s account in favour of an exclusively derivative approach to health inequality, either. Not only does that run counter to Daniels’s explicit aims; it would be unjustifable because, as demonstrated, exclusive reliance on the derivative approach necessarily underplays the importance of health. Compared to Daniels’s position, the relational egalitarian approach is more coherent, issues a clearer verdict on diferent kinds of health inequality, and therefore gives more unambiguous guidance on which health inequalities we have most reason to avoid.

57 See c­ hapter  1 (1.2.), and, for contrast, neo-​republican conceptions of social justice such as Pettit’s, which jettison this ambition and hold social justice to be more exclusively directed at non-​domination  alone. 58 Daniels 2007, p. 56. 59 For a similar line of argument, see Kelleher 2013, who argues that Daniels downgrades fair equality of opportunity as extended to health to a pro tanto principle.

282 Implications

9.4. Objections: Undercoverage, Exclusion, and Discrimination? Even if the view developed so far possesses the virtues of clarity and coherence, it might seem that it nevertheless has unattractive implications which might endanger the central ambition of the argument of this chapter, to demonstrate that liberal relational egalitarianism as applied to health and healthcare yields requirements which are both reasonably complete, and independently plausible. Potential objections center on the general, vulnerability-​based rationale for universal healthcare, and hold that its subordinated place within the hierarchy of health demands outlined earlier implies that pressing health needs will go entirely unmet, or that there will be problematic discrimination between equally needy patients—​or that the rationale is itself too limited, leading to the same results.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.4.1.  Crowding Out Universal Healthcare? According to a frst objection, the priority of the demands issued by the frst two approaches, which focus on socially caused health inequalities, implies that eforts to eradicate these inequalities will, in fact, exhaust all available resources which a relational egalitarian society can devote to health—​even if that priority ofcially holds, as outlined, only for health inequalities of equal magnitude. However, regarding the content of these demands in principle, it is unclear why this should be the case. Liberal relational egalitarianism demands the eradication of domination and of hierarchical norms of social status, and of the health inequalities associated with these, and that health inequalities brought about by other kinds of social processes and practices are justifed. Tere is no reason to suppose that meeting these requirements would, in itself, have to be so costly that no resources are lef for other pursuits and requirements—​in particular, as they are, as argued throughout, compatible with an intricate division of labour-​enhancing productivity, which is needed to generate, among other things, the means to deliver healthcare. Tere is more plausibility to the suggestion that combating such hierarchies, and seeking to meet such justifcatory demands, might in practice—​ here and now—​be so difcult, and consume so much of our eforts at bringing about justice, that there is a danger of not enough being lef for other concerns. Tat difculty would then not be due to the inherent

Health Inequality  283 demandingness of these requirements. It would be mainly due to powerful social actors seeking to block these eforts. However, if the arguments advanced so far are sound, they would do so without good reasons. So this is a problem of ‘non-​ideal theory’. Removing the social causes of bad and unequal health will in fact be politically very difcult, since it requires a significant egalitarian restructuring of societies, which will be fercely resisted by political actors representing the interests of the currently privileged. Now, given this fact of expectable non-​compliance by powerful actors, it would indeed be unreasonable to insist that all eforts should nevertheless go into such an attempt. We should keep on reserving resources for tackling other health deprivations, in the meantime, just as we would under just circumstances. However, while doing so, we should continue to insist that it is merely self-​ interested resistance without good moral reasons that stands in the way of meeting the pressing demands of those whose health is sacrifced due to political and social inequality.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.4.2.  Exclusion of Non-​Cooperators A stronger objection points out that the general requirement of equal access to healthcare is itself limited by the framework of liberal relational egalitarianism, according to which egalitarian demands are justifed by reference to the background ideas of a fair structuring of social cooperation among free and equal individuals, and aim at full inclusion within such cooperation on such fair terms. We have seen in section 9.2 that this view does not rely on any conventional idea of ‘normal’ cooperative abilities, but, to the contrary, yields an imperative to structure such cooperation in the most inclusive fashion possible, thus leading to a wide view of what must count as ability to participate in cooperation. It directs our attention precisely to the question of how social practices seize on diferential abilities, and transform them into social advantages and disadvantages. Nevertheless—​the objection presses on—​this does not change the fact that it is the capacity to cooperate which grounds the egalitarian entitlement to healthcare, aiming at its restoration. Tere will be cases of individuals who lack the ability to cooperate under any reasonable and justifed cooperative framework, no matter how inclusive—​for example, due to very severe, irremediable cognitive disabilities. If that is so, these individuals have no equal right to healthcare.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

284 Implications It is indeed an implication of the cooperation-​based framework underlying liberal relational egalitarianism that it does not cover such cases. However, there are two reasons not to regard this as a decisive objection. Te frst is that there is nothing in this framework which rules out other duties in these cases; in particular, duties of benefcence.60 What it rules out is that such duties are duties of relational equality. Te framework is compatible with these duties aiming at meeting absolute needs of non-​cooperators. It is also compatible with such duties being mandatory in the sense of being amenable to collective regulation and enforcement, where they do not confict with requirements of social justice (on which more in a moment).61 It is not the ambition of this book to account for all enforceable duties, but, more modestly, to make a case for duties of relational equality as the most stringently and demandingly egalitarian requirements of a liberal conception of social justice. Furthermore, it is important to note that this does not imply that duties of relational equality are only pro tanto duties, afer all, and need to be balanced with the requirements of other values (2.4 in ­chapter 2), such as, in this case, benefcence. A  relational egalitarian conception of justice can, and should, leave space for individuals to fulfl such other duties. It need not claim the resources needed for this fulflment; its demands would clash with those of such other values throughout only if it did. But it is not clear why, in principle, it should not be possible to fulfl duties of benefcence aiming at meeting absolute needs consistently with all cooperators retaining the means to develop and carry out their conception of the good, and implementing all relational egalitarian constraints on social cooperation. If it is, then duties of benefcence are complementary, not conficting. 62 In this, its requirements difer from, for example, an unrestricted luck egalitarian principle, whose demands are indeed so far-​reaching that they clash with such other requirements across the board (2.5 in c­ hapter 2). Diferential brute luck is everywhere, and stems from all kinds of diferent sources. Liberal relational egalitarianism can thus admit the demands of other values,

60 See c­ hapter 8, fns. 5 and 21. 61 For an instructive discussion of how duties of benefcence aiming at meeting basic needs can be allocated and enforced, see Buchanan 1984, pp. 68f. 62 To be sure, such general consistency judgments have to rely on background assumptions about the size of the cooperative surplus which can be achieved, especially by economically advanced societies (see 1.4 in ­chapter 1; for subsistence societies, the demands of cooperative justice and benefcence may well confict throughout). Furthermore, even if they are sound, hard cases where demands do confict certainly remain possible.

Health Inequality  285 and can do so without taking recourse to other purported egalitarian principles of justice, such as the luck egalitarian one. Tis leads to the second reason why this objection is not decisive. As seen, any rival view insisting on egalitarian demands for non-​cooperators conficts with the core idea of the liberal view that the terms of social cooperation have to be fair for its participants, as they are the ones who produce the benefts and resources that are to be distributed. So it will either have to identify a fair cut-​of point for these other demands in a diferent way, or accept that there is none.63 And as far as health is concerned, this is a practically very important worry, due to the potentially very high costs of healthcare. None of this is to say that the objection that liberal relational egalitarianism does not ground rights to healthcare for non-​cooperators should be taken lightly; as seen, it calls for other duties, and it would be a signifcant problem if, contrary to what was argued earlier, it turned out that these duties could not be made consistent with the approach. It is merely to point out that it is matched by corresponding objections for rival views.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

9.4.3.  Discrimination between Patients with Identical Health Needs? Tere is, however, an objection which is, in an important sense, more threatening, because it is more internal. It is somewhat similar to the frst objection tackled earlier in that it also focuses on the priority given to combating socially caused health inequalities: it holds that this will give rise to objectionable discrimination between patients within the healthcare system. It seems that it is an implication of this priority ordering that those who contracted illnesses, or sufered injuries, due to social inequalities condemned by the derivative (9.3.2) or direct social (9.3.3) approach have a stronger claim to remedial treatment by the healthcare system than those who sufer from the same kind of illnesses or injuries due to other reasons. So it seems that a relational egalitarian health system might be forced to operate a kind of triage, according to which the health needs of the latter might not be met afer all, or meeting them will at least have to be postponed.

63 Recall that G. A. Cohen, for example, conceives of luck egalitarianism both as only giving rise to pro tanto demands, and as tempered by an ‘agent-​centred prerogative’—​whose limits, however, are not specifed; 2008, p. 220. See earlier discussion in c­ hapter 2, fn. 37.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

286 Implications Tis is a salient objection; it would be merely speculative to reply that this would not happen because there will be enough resources to avoid any such triage at any point. Tere might be, or there might not be; it depends on the overall amount of resources which a society has at its disposition, and decides to dedicate to its healthcare system.64 What is more, even the possibility of such discrimination seems odious. Tis can be well explained with the help of the expressive perspective: it would be somewhat heroic to hold that a society which turned away, or which would be prepared to turn away, individuals with pressing health needs because it did not cause these needs, while it did cause the needs of others, would not express an attitude of neglect towards the former, and would not in fact communicate to them that they are worthy only of lesser concern. A very sick patient needing urgent treatment would be right to be unimpressed by a healthcare ofcial pointing out to her that there is no problem with her claim being subordinated to that of a diferent patient sufering from the same condition because research into the SDH establishes that the latter was at greater risk of contracting it due to social inequality. However, it is important to note that the intuitive strength of this objection crucially depends on the fact that the individuals in question are being in turned away despite being in pressing need, while others in the same situation are not. It does not apply, or at least not with similar strength, to a situation where such needs have not yet arisen, and institutions are assessing the risks that diferent groups of individuals face, and deciding how to respond. If the overall argument of this chapter is sound, they must give priority to combating health risks due to independent injustice, or which are socially caused in other, unjustifable ways.65 Furthermore, this consideration must also be able to infuence the decision of which kinds of treatments are generally to be ofered by the healthcare system, and which are not. Knowledge about socially induced risks may

64 To put it somewhat diferently: while we can be confdent that these prior requirements will not eat up all resources for healthcare (9.4.1), we cannot be confdent that there will not be cases of intuitively objectionable discrimination because of them. 65 Take, as an example, screening procedures: it is not objectionably discriminatory for a society to devote more resources per capita to screening a group of people facing socially patterned risks of contracting a certain disease for early onset indicators than to screening the remainder of the population (it could, for example, be a typical occupational hazard for the former). Tis does not rule out that those facing a heightened genetic risk of contracting a certain disease also have a claim to early screening. However, if the argument of this chapter is sound, screening obligations in the former type of case are more stringent, and thus must also inform the allocation of funding for research developing viable screening procedures, to start with.

Health Inequality  287

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

lead to the inclusion of treatments for certain conditions within a society’s universal healthcare system, while treatment for other conditions remains outside. Tat is a justifable way of accounting for the increased stringency of obligations to avoid and remedy conditions that are typically caused by factors such as social and economic inequality. Nothing objectionable is expressed by such an arrangement; to the contrary, a failure to institute it would constitute a failure to acknowledge the special nature of the responsibility borne by social and political institutions towards victims of the latter. A would-​be patient for whose rare condition the healthcare system does not ofer costly treatment, while it does for a diferent one which is typically socially caused, even if it is equally rare, equally damaging, and equally costly to treat, thus does not have a complaint of unjust discrimination. Once, however, a justifable decision about general coverage has been made, discriminating between individual patients within the healthcare system must be avoided, for expressive reasons internal to relational egalitarianism. Strength of individual need is, at that point, the only justifable criterion for diferential treatment. Te general argument of the last section about the priority relations between diferent health demands thus permits responding to individual cases with the appropriate nuance; there is no need to revise it. Liberal relational egalitarianism yields a plausible, ordered set of diferent egalitarian requirements for health and healthcare, on which a fuller conception of just health and healthcare can be built.

Conclusion Te aim of this book has been to develop a conception of relational equality as a matter of liberal social justice, and to show that its demands are plural, far-​reaching, and stringent. It has not been to refute competing egalitarian theories and ideals. Tree issues ought to be addressed by way of conclusion. Te frst is how this conception could be challenged from the point of view of such competing theories and ideals, by briefy recapitulating how we arrived at its main demands. Te second is to give an indication of the distinctive advantages it promises to have when it comes to applying these demands to concrete cases of relational egalitarian policies and institutional set-​ups, beyond the general indications for political and distributive justice developed in Part II of the book. Te third is which possibilities it ofers for extending it, beyond domestic society, to pressing social egalitarian concerns in international, transnational, and global scenarios.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

C.1.  Challenging Liberal Relational Egalitarianism Te key contention of this book is not only that liberalism gives rise to demanding requirements of relational equality, but that these requirements ought to enjoy special stringency. However, this work has not ofered a foundational justifcation of the liberal background conceptions of the person and society as a fair scheme of cooperation on which these demands rest. Tis liberalism has mostly been assumed, and argued for only indirectly—​by showing how well it coheres with relational egalitarian demands, compared to prominent alternatives. Nor have I attempted any kind of conclusive demonstration that there really are no other egalitarian demands of justice, or that, should there be any, they should not be similarly stringent. Both these tasks would have distracted from the specifc, constructive task of the book. What this work has done instead is to give, at each point, some indications of the problems which rival (distributive or social) egalitarian positions face, as well as positive reasons why, whatever speaks for or against Justice and Egalitarian Relations. Christian Schemmel, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/​oso/​9780190084240.003.0010

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Conclusion  289 such other demands, one should accept the liberal relational egalitarian ones, as well as their stringency. For example, liberals should accept that requirements of non-​domination are particularly stringent—​more stringent than those based on hierarchies of social status alone, or ‘simpler’ demands of distributive fairness based on cooperation—​because being subject to the superior and arbitrary power of other individual agents is a particularly insidious relational inequality. Such other persons are one’s moral equals, and therefore should be one’s social and political equals, too. A determinate hierarchy of arbitrary power between individuals denies that equality of standing in a more conspicuous and forceful manner than, for example, not getting one’s fair share of goods, compared to these others. However, for all that has been said, critics naturally remain free to submit a case for the similar importance of other egalitarian demands, such as those based on luck egalitarian fairness, or on the impersonal value of social equality—​or to challenge the liberalism of the conception, or its connections to relational egalitarian requirements. As regards the former, the challenge for them is to go beyond intuitive claims to justify the existence and importance of such other demands, as well as to rebut the objections which were made, in ­chapters 2–​6, pertaining especially to how rival conceptions struggle to capture what is most important about demands of relational equality, and to account for their full range. As regards the latter, direct objections focus on whether the demands of relational egalitarianism really follow from the liberal background conceptions of the person and society. On this, no more can and ought to be said here. Either the arguments submitted to critical scrutiny over the course of this book can hold their ground, or they cannot. Indirect objections challenge these background conceptions themselves. Perhaps others are more plausible, and can lead to the same, or even more demanding, requirements of relational equality. Perhaps, however, requiring individuals to treat each other as free and equal in social cooperation, in a way that gives rise to the comprehensive and stringent kind of demands on the fair structure of such cooperation argued for throughout, is itself problematic, in a way which speaks for accepting more strategic and more dominantly self-​interested attitudes towards cooperation as the basic ‘material’ which political theory has to work with. If that is so, however, it seems hard to see how we could arrive at any comparable demands of relational equality at all. We have encountered rival, slimmer views of expressive egalitarianism which aimed at equal concern, or equality of standing, such as those put forward by Ronald Dworkin and

290 Conclusion Philip Pettit, and have seen where and how these fall short in various important relational egalitarian dimensions—​non-​domination, political equality, and avoiding hierarchies of social status. Dworkin’s and Pettit’s conceptions fail to account for the full range of relational egalitarian demands without accepting more strategic and self-​interested attitudes toward cooperation and membership in society; if these are accepted, it seems the results can only be even less socially egalitarian overall. Against that, alternative conceptions of what the—​personal or impersonal—​good of living together as social equals consists of certainly have the potential to lead to much more encompassing and far-​reaching ideals of egalitarian relations. But they do not answer the question of why aiming at these is required in practice—​or even permissible, should individuals not already agree with these conceptions. Still, that Rawlsian liberal conceptions of the person and society lend themselves to a demanding ideal of relational egalitarian social justice does not amount to a direct argument for them. Even if they ultimately fail to be justifed, however, the general argument for the anti-​hierarchical thrust of the conjunction of the expressive perspective on justice, and basic moral equality (­chapters 2 and 3) remains in place. It would then urge us to devise and scrutinize alternative conceptions of relational egalitarian justice.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

C.2.  Applying Relational Egalitarianism Relational inequality of the kind which matters for justice typically subordinates members of some group to the members of other groups, in virtue of their group membership: because salient social and political processes seize on groups and their features in bringing about, or entrenching, such inequality. Tis fact, coupled with the fact that, in order to identify unjust relational inequality, we do not need to look at how well or badly individual members of such groups otherwise fare (in some measurable currency), renders relational egalitarian principles of justice relatively easily applicable to policies and institutions, and promises, to that extent, a desirable refocusing of egalitarian theorizing to practical matters. Some relational egalitarians, especially those with an explicit and sustained diagnostic focus on severe real-​world injustice (see 1.4 in ­chapter 1) go as far as defning salient relational inequality as ‘durable group inequality sustained by social structures such as laws, organizations and social norms.’1 Liberal 1 Anderson 2017b, pp. 36f.; see also Anderson 2012, p. 42.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Conclusion  291 relational egalitarianism does not go quite as far; we have seen that liberal non-​domination requires protection also for those who would otherwise be dominated by more powerful others, such as criminal gangs (­chapter 4, 4.2), even without any authority or norms, or even within personal relationships they may have entered voluntarily, and perhaps without inegalitarian norms constraining their choices (4.3.1). It sees no reason to withhold a verdict of unjust relational inequality where such relations occur, or prove too difcult to exit, because of insufcient protection. Social status inequalities, on the other hand, are indeed necessarily group-​based, insofar as the relevant norms must have some level of generality regarding the features of individuals they seize on, to count as norms, in the frst place (­chapter 6, 6.2). However, neither domination-​enhancing nor ‘simple’ status norms need to fasten on to any ‘thick’ ascriptive identity traits, such as of ethnicity, gender, or sexual orientation—​although, evidently, they very ofen do.2 However, in any case, for applying relational egalitarian demands, tackling durable group inequalities in both dimensions will always have the highest priority, whether it is necessary by defnition or because these are simply in fact the most salient inequalities, where the diferent dimensions in question also tend to interact in the most pernicious fashion. Te distinctive advantage of relational egalitarianism is that it can directly connect to empirical social science about group inequalities, in order to read the relevant groups and mechanisms of it, as it were. Tere is thus a wealth of opportunities to test and check how real institutions and norms in diferent domains constitute and entrench hierarchies of power and status, and devise directly targeted remedial policies and institutions. Tis contrasts with distributive theories of egalitarian justice, because these processes can, ofen at least, be identifed and addressed without frst having to enquire which exact distributive impact they have, as measured in some currency of justice, on the individuals engaged or subject to them. Whether or not it is a sensible philosophical enterprise to search for a generally valid such currency, it is, in practical terms, somewhat debilitating to have to accept that, in principle, one would frst have to have a frm answer on how well individuals are doing overall in distributive terms, before being able to focus on the kinds of policies and institutions that tend to cause, or would remedy, the so ascertained unjust inequality.

2 Ibid.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

292 Conclusion Of course, sometimes we can make do with rough-​and-​ready assessments of overall distributive advantage, and justice-​based, liberal egalitarians will also sometimes have to resort to quite intricate measures to ascertain who is subject to most domination (­chapter 4, 4.6), or which kinds of outcomes, in terms of goods generated or systematically infuenced by social cooperation, count as failing to respect people’s standing as equals in cooperation (­chapters 8 and 9).3 Furthermore, to be sure, when it comes to applications, distributive approaches mostly focus on groups, too. Here, however, there is a diference to relational views which is even more salient than the one pertaining to the need for measuring: for distributive theories, these groups can merely be groupings of people in the same distributive situation—​no matter how each of its members actually got there. A well-​known example are the Rawlsian ‘worst of ’; Rawls accepts that his theory of justice, in order to give appropriately publicly accessible and checkable action guidance, has to focus on groups, and directs our focus to that group whose members are worst of in terms of expectations of income, wealth, and the social bases of self-​respect, integrated into an index.4 However, whether or not Rawls happens to propose the correct measure for disadvantage, the mechanisms and social processes responsible for placing individuals in that group could be very diverse indeed. So, even if one’s measure is in fact the best one, it may still tell us very little about how one should remedy the problem. Relational egalitarianism does not encounter this problem, or not to the same extent. Not only do we, in many cases, not have to assess who is worst of according to some overall index to be able to judge with confdence that some hierarchical arrangements express disrespect for people’s equal standing. Te institutions and norms which create and maintain the groups in question, as well as their subjection, are more directly in focus, because it is they who express that disrespect. Teir precise identifcation takes place in collaboration with the empirical social sciences, such as sociology, political science, anthropology, educational research, and empirical political economy. Te investigative focus of studies in these disciplines and felds will 3 It is also important to note that distributive theories have in fact made signifcant advances on the question of how to identify the worst of. See only, as an example of sophistication and progress within the literature on capability theory, Wolf and De Shalit 2007. However, the question of who is worst of naturally leads to a focus on cases of multidimensional, severe, and self-​reinforcing disadvantage (ibid.). While these cases certainly call for most urgent attention, and a broad array of possible remedial policies (this is an issue on which the two approaches need not come apart; see 1.4 in ­chapter 1) it is, to that extent, questionable whether that focus is any longer on inequality, and what, if anything, is distinctively wrong with it. 4 Rawls 1999a, pp. 78f.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Conclusion  293 itself ofen be directed by normative assumptions about where we ought to look for the most problematic hierarchical mechanisms and arrangements. Collaborative, multidisciplinary research into potentially unjust hierarchy is thus a natural focus for relational egalitarians. Tis is well demonstrated by the number of such studies already undertaken, both regarding general social science on causes and processes of hierarchical group formation, diferentiation, and reproduction,5 and studies of important sites for, and mechanisms of, generating, perpetuating, or countering socially and politically salient hierarchies, beyond formal political institutions, such as education and the formation of social networks in and around school,6 micro-​practices and norms governing everyday informal social interaction,7 workplace relations,8 public and community land ownership and banking,9 welfare state arrangements and beneft policies,10 including unconditional grant policies,11 public health and healthcare policies,12 and broader policies of racial integration.13 Te distinctive contribution of liberal relational egalitarianism to such debates of application is to show how, ofen, far-​reaching egalitarian policies, going beyond formal institutions, can be justifed on relatively standard liberal premises alone—​which, at the same time, serve to circumscribe their respective mandates (­chapter 4, 4.5; ­chapter 6, 6.4), and can thus help us to fne-​tune their shape appropriately. Its liberal nature also makes clear that the analytically and diagnostically necessary focus on groups and hierarchies between them does not necessitate in any way advocating remedial policies whose point it is to somehow shore up and strengthen existing, less powerful or esteemed, groups, as groups—​as opposed to strengthening their individual members by aiming at dissolving or integrating (top or bottom) groups into larger groups or broader society, for example, by obstructing the emergence of potentially pernicious norm-​based processes of group diferentiation (­chapter 6, 6.4). Te right policy depends on the specifc nature of the problem (­chapter 8, 8.5).



5 See Anderson 2017b and Wolf 2017 and 2019, pp. 14f., for example. 6 Macfarlane 2018. 7 McTernan 2018.

8 For example, Anderson 2017b. 9 Guinan and O’Neill 2019.

10 For example, Schemmel 2015b; O’Neill et al. 2016. 11 For example, Bidadanure 2019; Tomas 2017.

12 See references in ­chapter 9; especially Voigt and Wester 2015, and Haverkamp et al. 2018. 13 Anderson 2010a.

294 Conclusion

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

C.3.  International Justice and Beyond Te cases and applications just mentioned, however, all seem to be, at least primarily, about domestic policies and institutions. What about international, transnational, and global justice? In the course of this book, we have encountered a great variety of diferent social egalitarian positions, which came apart on the question of why egalitarian relations are valuable, and/​or required by justice. It is thus clear that the question of what social egalitarianism as such implies for these domains is not one that admits of a unitary answer. For justice-​based positions, such as the liberal relational egalitarian one, it is natural to think of requirements to set up signifcant social and political relations on an egalitarian footing as triggered primarily, or even exclusively, by patterns of already existing, non-​trivial, social interactions. Tat is, it is natural to demand relational content of social justice on the basis of a relational grounding of it.14 If persons are already engaged in social relations, it is an obvious question to ask how these should be structured so as to be justifable, and fair, and if persons are each other’s moral equals, then such relations will have to have a recognizably egalitarian quality. Tere is, however, no reason at all to require such existing relations to already take the form of state-​based cooperation, or even democratically organized cooperation, for relational egalitarian justice to apply.15 Whether we should aim at creating egalitarian relations where no relations exist is, against that, an entirely different question (perhaps the unrelated are content in their mutual isolation, and pose no threat to each other). Tis book has taken state-​organized cooperation as its primary focus and thus arrived at determinate requirements of relational equality, not only in the development of implications undertaken in Part II, but in determining the substance of fnal or prima facie claims of justice which non-​domination has to range over (­chapters 3 and 4), and in specifying the injustice of inegalitarian status norms as enabling domination, threatening self-​respect, and impeding access to signifcant opportunities (­chapter 6). More fundamentally still, it has relied on liberal background conceptions of the person, based on possession of the two Rawlsian moral powers, and of society as a

14 See Miklosi 2018, p. 112. For the distinction between relational and non-​relational theories of global justice, understood in this grounding sense, see Sangiovanni 2007, p. 5. 15 For frst attempts to ground and apply social egalitarian requirements beyond the state, see Nath 2011 and 2015; Ip 2016; Cloarec 2017; and Amighetti 2019.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Conclusion  295 fair scheme of cooperation. It is not self-​evident that these can travel, without modifcations, to the international, transnational, and global arenas. Tis is not to say that it would be a theoretical problem for relational egalitarian justice beyond the state that such background conceptions are not in fact accepted by people around the globe. We have seen that the liberal conception is not ‘practice-​dependent’ in a way that would presuppose basic ideas such as that of liberal personhood to be already shared among participants in the social practices of cooperation to which it applies, or that would draw on specifc interpretive material informing existing state-​ based cooperative practices for arriving at its requirements (such as state constitutions, for example).16 All it claims, and needs, are that its basic ideas are shareable (­chapter 3, 3.1), and that these requirements do follow from them. So the same holds for applications beyond the state. Still, this does not guarantee that the liberal background conceptions used are the appropriate ones for these scenarios (though they may well be). However, as seen, what we can afrm with full confdence is that there is no necessity whatsoever for relational egalitarian requirements beyond the state to require a context of interaction and cooperation that is as dense and comprehensive as that of a state.17 It is only that the nature of that context will make some diference. For example, even in those cases where cooperative practices do not themselves stretch beyond the boundaries of states, that there be such boundaries at all requires justifcation. Isolating one’s cooperative practices from others, and restricting the access of outsiders to it, requires making sure, to the extent that one can, that others’ cooperative schemes can function properly, too.18 What exactly the content of that requirement is we must leave open here. However, whether it is, in any recognizable sense, egalitarian, and thus requires that such other schemes can perform at levels of economic productivity and democratic guarantees that are similar to one’s own, or whether it is more restricted, and only requires that others be able to erect and maintain functioning basic institutions,19 it is 16 On practice-​dependence, see James 2005; Sangiovanni 2008; and Valentini 2011. 17 For some thoughts on how diferent broadly relational egalitarian views presupposing some shared institutional arrangements might require demanding global reforms, see Schemmel 2012, pp. 442f. 18 It also commits states, and thus their ofcials, to take pains to express that, even if they have special responsibilities to their own members, all individuals everywhere are moral equals; see ­chapter 2,  2.6.3. 19 As, for example, Rawls’s ‘duty of assistance’ to ‘burdened societies’ has it; Rawls 1999b, pp. 106f. Whatever its content, such a duty must be conceived of as a duty of justice proper (see what follows in the main text). It cannot be a duty of benefcence to non-​cooperators of the kind we encountered in

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

296 Conclusion important to notice that stringent requirements of protection against domination will range over them. A powerful state, or a club of states (analogous to a gang; ­chapter 4, 4.2), which has unconstrained and superior power over the international institutional arrangements that should implement these duties, severely dominates others. Whether or not their content is egalitarian does not determine how egregious that domination itself is, and how stringent protections against it should be. As such duties are conditions of justifability for running one’s own cooperative scheme, they should be very stringent indeed—​possibly more stringent than internal requirements governing that scheme itself. Furthermore, we should note that, in such a scenario of international domination, relational inequality between states ofen cuts through to the individual level:  individual members of more powerful states dominate, other things being equal, individual members of less powerful ones, insofar as the former each have more power—​even if individual shares of power are small—​over international rules. Of course, the ‘other things being equal’ clause is very important. Tings are not equal, for example, when, in realistic scenarios of signifcantly indented state regulatory capacity, there are transnational coalitions of powerful people, such as within a global social class of capital owners, which press for further weakening the protective capacity of states everywhere (though their efect on weaker states can certainly be much worse than on stronger ones), and hollow out the substantive opportunity of ordinary citizens to have an efective say on how their own state is organized. And they are not equal in the case of non-​democratic states, whose individual members can then be doubly dominated, both by more powerful other states, and their members, and by their own strongmen. Accordingly, the precise content of international claims to non-​domination may vary very signifcantly, and cannot simply be carried over, as it were, from the domestic case. Where cooperative practices, in particular of economic production, are themselves already integrated across borders, beyond international ground rules, the nature of these practices will infuence the content of relational egalitarian requirements even more directly, and they could come to approach those holding within a state. However, there is, again, no shortcut to the conclusion that transnational and global relational egalitarian claims c­ hapter 9, 9.4.—​because members of other cooperative schemes can cooperate within ours, too, if we let them.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Conclusion  297 of justice should have the same content as domestic ones throughout—​that the two ideals will indeed end up being very similar in their concrete shape, and resulting recommendations—​because strengthening, (re-​)empowering, and democratizing states might ofen be part of the solution to many existing cases of global and transnational domination, exploitation, and status hierarchy.20 Similar considerations apply to the question of what the appropriate democratic forum is for deliberating and deciding about claims of justice beyond the state on which there is reasonable disagreement. While explicit attempts to apply relational egalitarianism beyond the domestic case are still few,21 there is a rich neo-​republican literature on international, transnational, and global justice, and progress can be made by integrating the two. Finally, it is imperative to note that, while there are global institutions, and a lot of global interaction, actual interaction is not strictly needed to trigger relational egalitarian concern. An agent dominates another already when she has the capacity to intervene in the other’s choices in an arbitrary, or uncontrolled, manner, and it is not necessary that this capacity be actually exercised. Sharing common institutions and practices, or being engaged in other kinds of interaction, may give rise to all kinds of domination, depending on how these are structured, and erecting new institutions can create new such kinds. But it is not necessary for there to be domination: powerful agents—​ individuals or states—​can dominate others simply by having the opportunity to intervene in their afairs, and such opportunity can arise in the absence of institutions; or even precisely because of it.22 Tat possibility is enough. It is all the social context which domination needs, because it need not be based on de facto authority, or public in any other way (­chapter 4, 4.2). Somewhat similarly, while diferential social esteeming of the kind mandated by, and in turn upholding, pervasive status norms, is not possible without actual interaction (­chapter 6, 6.2), it can certainly happen without the involved persons, (dis-​ )esteemer and (dis-​ )esteemed, sharing any 20 Nath rightly points out that the typically superior consumer power of members of richer states within the global economy can institute objectionable relational inequality between them and typically poorer individuals in poorer states, and certainly does so where that divergence is extreme; 2011, p. 602. Again, however, from this diagnosis we cannot read of that the right remedial policy is to seek, in some reasonably direct way, to make these individuals richer—​as opposed to aiming at structurally strengthening the regulatory and investment capacity of their states, so that they can forge more of their own way to economic success. Of course, perhaps we can, and should, simply do both. Te point is merely that we cannot reason just by analogy from the domestic case. Tat is the main reason for this book’s restricted focus. It is methodological and in no way implies a conviction that domestic relational inequality is typically worse. Tat would be a very implausible conviction. 21 See fn. 15. 22 See Ronzoni 2009.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

298 Conclusion institutions; and neither do they have to think that they are already somehow engaged in a common enterprise part of which should be to regulate such norms. Using a fair cooperation framework, we should ask what it would mean for parties to such relations to cooperate fairly, which role exactly requirements of combating status hierarchy should play for that, and which form they should take. If a thicker institutional context were absent, then the preferred liberal strategy for avoiding status hierarchy might well be to militate for ‘negative cooperation’; for dissolving norms by making sure that all can go their own way, unconcerned by what others think, and how they express it, rather than trying to make sure their content becomes more harmless (­chapter 6, 6.4), and that all have the means to meet their demands (­chapter 8, 8.5). Tat picture, of course, does not ft with our world at all, which is shaped by centuries of dense structural, and ofen very violent, interaction, due to colonialism and imperialism—​which, in turn, have shaped global status norms. Much of the historically relevant, and pernicious, international status norms have thus fastened on racist belief systems, and possession of military power. It seems plausible that these markers might have shifed, to some extent at least, to possession of wealth and consumer goods. Tere may thus be important similarities to many status norms reigning in the domestic case (­chapters 6,  8). As regards intergenerational justice, relational egalitarian requirements governing our interaction with successive generations can be based on our sharing cooperative schemes with them. It is possible that we can arrive, in this way, at an indefnite iteration of egalitarian demands, as these schemes continue. Relational inequalities of power and status between the older and the younger are distinct phenomena giving rise to distinct forms of intergenerational injustice.23 Against that, where there is no possibility of interaction between agents of the same moral status, as is the case for our relations to distant future generations, relational egalitarianism might not have much to say.24 Not only can there obviously be no intergenerational status norms applying to that relation. On the relational egalitarian case for non-​domination (­chapter 3, 3.3.1), there are no reasons grounding distinct complaints of domination based on distant future generations being subject to our superior, arbitrary power, either. What matters is that their

23 See Bidadanure 2016.

24 See Lippert-​Rasmussen 2018b, pp. 123f.

Conclusion  299

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

entitlements against us are, in the end, made as secure as they have to be to make sure that they fnd themselves with the right kind of heritage. Should we manage to entrench them to that extent, then the fact that our power over them was arbitrary before that moral accomplishment constitutes no distinct injustice to them. Tat is diferent from the case of ongoing interaction, or the possibility thereof, where being at the mercy of others grounds such a distinct complaint. Furthermore, as the case for relational equality made in this book depends on the possibility of cooperation, it yields no other reasons for why the content of the claims of distant future generations against us which need entrenching should be egalitarian, either—​unless, perhaps, we are connected by an unbroken cooperative chain. On this view, we have no egalitarian duties of justice to those with whom cooperation is strictly, and irredeemably, impossible, and where that impossibility is not our fault. To hold that we need the social and political value of equality to justify stringent duties not to endanger the livelihood of our distant successors, including their ability to stand to each other as equals, would be to overestimate its importance. Tat importance is enormous as it is.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography Aas, Sean, and Wassermann, David. 2016. ‘Natural and Social Inequality: Disability and Fair Equality of Opportunity’. Journal of Moral Philosophy 13, 576–​601. Adler, Matthew D. 2000. ‘Expressive Teories of Law: A Skeptical Overview’. University of Pennsylvania Law Review 148, 1363–​1501. Adorno, Teodor W., Frenkel-​Brunswik, Else, Levinson, Daniel J., and Sanford, R. Nevitt. 1950. Te Authoritarian Personality. New York: Harper & Row. Amighetti, Sara. 2019. Relational Equality, Justice and the Boundary Problem. PhD thesis, University College, London. Anand, Sudhir, and Peter, Fabienne. 2000. ‘Equal Opportunity—​A Response to “Justice Is Good for Our Health” ’. Boston Review. Available at http://​bostonreview.net/​archives/​ BR25.1/​anand.html. Anderson, Elizabeth. 1999. ‘What Is the Point of Equality?’ Ethics 109, 287–​337. Anderson, Elizabeth. 2004a. ‘Welfare, Work Requirements, and Dependant-​Care’. Journal of Applied Philosophy 21, 243–​256. Anderson, Elizabeth. 2004b. ‘Rethinking Equality of Opportunity: Comment on Adam Swif’s “How Not to Be a Hypocrite” ’. Teory and Research in Education 2, 99–​110. Anderson, Elizabeth. 2006. ‘Te Epistemology of Democracy’. Episteme 3, 8–​22. Anderson, Elizabeth. 2007. ‘Fair Opportunity in Education:  A Democratic Equality Perspective’. Ethics 117, 595–​622. Anderson, Elizabeth. 2008. ‘How Should Egalitarians Cope with Market Risks?’ Teoretical Inquiries in Law 9, 239–​270. Anderson, Elizabeth. 2010a. Te Imperative of Integration. Princeton, NJ:  Princeton University Press. Anderson, Elizabeth. 2010b. ‘Te Fundamental Disagreement Between Luck Egalitarians and Relational Egalitarians’. Canadian Journal of Philosophy 40, Supplement 1, 1–​23. Anderson, Elizabeth. 2012. ‘Equality’. In: Te Oxford Handbook of Political Philosophy, ed. David Estlund. Oxford: Oxford University Press, pp. 40–​57. Anderson, Elizabeth. 2017a. Private Government—​How Employers Rule Our Lives (and Why We Don’t Talk about It). Princeton, NJ: Princeton University Press. Anderson, Elizabeth. 2017b. ‘Te Problem of Equality from a Political Economy Perspective’. In:  Oxford Studies in Political Philosophy, Volume 3, eds. David Sobel, Peter Vallentyne, and Steven Wall. Oxford: Oxford University Press, pp. 36–​57. Anderson, Elizabeth. 2018. ‘Freedom and Equality’. In: Te Oxford Handbook of Freedom, eds. David Schmidtz and Carmen Pavel. Oxford: Oxford University Press, pp. 90–​105. Anderson, Elizabeth, and Pildes, Richard. 2000. ‘Expressive Teories of Law: A General Restatement’. University of Pennsylvania Law Review 148, 1503–​1575. Aristotle. 2000. Nicomachean Ethics, ed. Roger Crisp. Cambridge, UK:  Cambridge University Press. Arneson, Richard. 1989. ‘Equality and Equal Opportunity for Welfare’. Philosophical Studies 56, 77–​93.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

302 Bibliography Arneson, Richard. 1990. ‘Liberalism, Distributive Subjectivism, and Equal Opportunity for Welfare’. Philosophy & Public Afairs 19, 158–​194. Arneson, Richard. 2000. ‘Luck Egalitarianism and Prioritarianism’. Ethics 110, 348–​358. Arneson, Richard. 2004. ‘Democracy Is Not Intrinsically Just’. In: Justice and Democracy, eds. Keith Dowding, Robert E. Goodin, and Carole Pateman. Cambridge, UK: Cambridge University Press, pp. 40–​58. Arneson, Richard. 2006. ‘Distributive Justice and Basic Capability Equality—​“Good Enough” Is Not Good Enough’. In: Capabilities Equality—​Basic Issues and Problems, ed. Alexander Kaufman. London: Routledge, pp. 17–​43. Arneson, Richard. 2008a. ‘Justice Is Not Equality’. Ratio 21, 371–​391. Arneson, Richard. 2008b. ‘Te Supposed Right to a Democratic Say’. In: Contemporary Debates in Political Philosophy, eds. Tomas Christiano and John Christman. Oxford: Blackwell, pp. 197–​212. Arneson, Richard. 2010. ‘Democratic Equality and Relating as Equals’. Canadian Journal of Philosophy 40, Supplement 1, 25–​52. Arnold, Samuel. 2018. ‘Putting Liberty in Its Place:  Rawlsian Liberalism without the Liberalism’. European Journal of Philosophy 26, 213–​237. Arnold, Samuel, and Harris, John. 2017. ‘What Is Arbitrary Power?’ Journal of Political Power 10, 55–​70. Baderin, Alice. 2016. ‘Political Teory and Public Opinion: Against Democratic Restraint’. Politics, Philosophy & Economics 15, 209–​233. Bakunin, Mikhail. 1996. Bakunin on Anarchism, ed. Sam Dolgof, 2nd ed. Montréal: Black Rose Books. Barry, Nicholas. 2006. ‘Defending Luck Egalitarianism’. Journal of Applied Philosophy 23, 89–​107. Bartley, Mel, Ferrie, Jane, and Montgomery, Scott. 2005. ‘Health and Labour Market Disadvantage: Unemployment, Non-​employment, and Job Insecurity’. In: Marmot and Wilkinson 2005, pp. 78–​97. Benson, Jonathan. 2018. An Epistemic Conception of Deliberative Democracy. PhD thesis, University of Manchester. Benson, Paul. 1994. ‘Free Agency and Self-​Worth’. Te Journal of Philosophy 91, 650–​668. Berkey, Brian. 2018. ‘Relational Egalitarianism and the Grounds of Entitlements to Healthcare’. Les Ateliers de l’Éthique/​Te Ethics Forum 13, 85–​104. Bidadanure, Juliana. 2016. ‘Making Sense of Age-​Group Justice: A Time for Relational Equality?’ Politics, Philosophy & Economics 15, 234–​260. Bidadanure, Juliana. 2019. ‘Te Political Teory of Universal Basic Income’. Annual Review of Political Science 22, 481–​501. Bidadanure, Juliana, and Axelsen, David. 2019. ‘Unequally Egalitarian? Defending the Credentials of Social Egalitarianism’. Critical Review of International Social and Political Philosophy 22, 335–​351. Bird, Colin. 2010. ‘Self-​Respect and the Respect of Others’. European Journal of Philosophy 18,  17–​40. Bloom, Allan. 1975. ‘Justice:  John Rawls vs. the Tradition of Political Philosophy’. American Political Science Review 69, 648–​662. Boorse, Christopher. 1997. ‘A Rebuttal on Health’. In: What Is Disease?, eds. James M. Butler and Robert F. Almeder. Totowa, NJ: Humana Press, pp. 1–​134. Bourdieu, Pierre. 1984. Distinction: A Social Critique of the Judgement of Taste. Cambridge, MA: Harvard University Press.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography  303 Boxill, Bernard. 1978. ‘Self-​Respect and Protest’. Philosophy & Public Afairs 6, 58–​69. Braveman, Paul, Egerter, Susan, and Williams, David. 2011. ‘Te Social Determinants of Health: Coming of Age’. Annual Review of Public Health 32, 381–​398. Brighouse, Harry. 1996. ‘Egalitarianism and Equal Availability of Political Infuence’. Journal of Political Philosophy 4, 118–​141. Brighouse, Harry. 1997. ‘Political Equality in Justice as Fairness’. Philosophical Studies 86, 155–​184. Brighouse, Harry, and Swif, Adam. 2006. ‘Equality, Priority, and Positional Goods’. Ethics 116, 471–​497. Brighouse, Harry, and Swif, Adam. 2009. ‘Educational Equality versus Educational Adequacy: A Critique of Anderson and Satz’. Journal of Applied Philosophy 26, 117–​128. Brown, Jefrey M. 2019. ‘Relational Equality and Disability Injustice’. Journal of Moral Philosophy 16, 327–​357. Brunner, Eric, and Marmot, Michael. 2005. ‘Social Organization, Stress, and Health’. In: Marmot and Wilkinson 2005, pp. 6–​30. Buchanan, Allen. 1984. ‘Te Right to a Decent Minimum of Health Care’. Philosophy & Public Afairs 13, 55–​78. Buchanan, Allen, Brock, Dan W., Daniels, Norman, and Wikler, Daniel. 2000. From Chance to Choice: Genetics & Justice. Cambridge, UK: Cambridge University Press. Carter, Ian. 1999. A Measure of Freedom. Oxford: Oxford University Press. Carter, Ian. 2011. ‘Respect and the Basis of Equality’. Ethics 121, 538–​571. Casal, Paula. 2007. ‘Why Sufciency Is Not Enough’. Ethics 117, 296–​326. Chambers, Clare. 2017. Against Marriage: An Egalitarian Defense of the Marriage-​Free State. Oxford: Oxford University Press. Christiano, Tomas. 2008. Te Constitution of Equality:  Democratic Authority and Its Limits. Oxford: Oxford University Press. Cloarec, Pierre. 2017. ‘Social Equality and the Global Society’. Journal of Moral Philosophy 14, 535–​561. Cohen, G. A. 1989. ‘On the Currency of Egalitarian Justice’. Ethics 99, 906–​944. Cohen, G. A. 1993. ‘Equality of What? On Welfare, Goods, and Capabilities’. In:  Te Quality of Life, eds. Martha Nussbaum and Amartya Sen. Oxford:  Clarendon Press,  9–​29. Cohen, G. A. 1997. ‘Where the Action Is: On the Site of Distributive Justice’. Philosophy & Public Afairs 26, 3–​30. Cohen, G. A. 2004. ‘Expensive Taste Rides Again’. In: Dworkin and his Critics, ed. Justine Burley. Oxford: Blackwell, 3–​29. Cohen, G. A. 2008. Rescuing Justice and Equality. Cambridge, MA:  Harvard University Press. Cohen, G. A. 2009. Why Not Socialism? Princeton, NJ: Princeton University Press. Cohen, G. A. 2011. ‘Fairness and Legitimacy in Justice, and:  Does Option Luck Ever Preserve Justice?’ In:  On the Currency of Egalitarian Justice, and Other Essays in Political Philosophy, ed. Michael Otsuka. Princeton, NJ: Princeton University Press, pp. 125–​146. Cohen, Joshua. 2002. ‘For a Democratic Society’. In: Te Cambridge Companion to Rawls, ed. Samuel Freeman. Cambridge, UK: Cambridge University Press, pp. 86–​138. Cordelli, Chiara. 2015a. ‘Justice as Fairness and Relational Resources’. Journal of Political Philosophy 23, 86–​110.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

304 Bibliography Cordelli, Chiara. 2015b. ‘Distributive Justice and the Problem of Friendship’. Political Studies 63, 679–​695. Costa, M. Victoria. 2009. ‘Rawls on Liberty and Non-​Domination’. Res Publica 15, 397–​413. Crisp, Roger. 2003. ‘Equality, Priority, and Compassion’. Ethics 113, 745–​763. Crisp, Roger. 2004. ‘Egalitarianism and Compassion’. Ethics 114, 119–​126. Cutler, David, Lleras-​Muney, Adriana, and Vogl, Tom. 2011. ‘Socioeconomic Status and Health: Dimensions and Mechanisms’. In: Te Oxford Handbook of Health Economics, eds. Sherry Glied and Peter Smith. Oxford: Oxford University Press, pp. 124–​163. Daniels, Norman. 1985. Just Health Care. Cambridge: Cambridge University Press. Daniels, Norman. 1996. Justice and Justifcation:  Refective Equilibrium in Teory and Practice. Cambridge, UK: Cambridge University Press. Daniels, Norman. 2002. ‘Democratic Equality: Rawls’ Complex Egalitarianism’. In: Te Cambridge Companion to Rawls, ed. Samuel Freeman. Cambridge, UK:  Cambridge University Press, pp. 241–​276. Daniels, Norman. 2007. Just Health:  Meeting Health Needs Fairly. Cambridge, UK: Cambridge University Press. Daniels, Norman, Kennedy, Bruce, and Kawachi, Ichiro. 2000a. ‘Justice Is Good for Our Health’. Boston Review. Accessible at http://​bostonreview.net/​forum/​ norman-​daniels-​bruce-​kennedy-​ichiro-​kawachi-​justice-​good-​our-​health. Daniels, Norman, Kennedy, Bruce, and Kawachi, Ichiro. 2000b. ‘Final Forum Response—​ Why Justice Is Good for Our Health’. Boston Review. Accessible at http://​bostonreview. net/​forum/​justice-​good-​our-​health/​norman-​daniels-​bruce-​kennedy-​ichiro-​kawachi-​ justice-​good-​our-​health-​%E2%80%94. Darwall, Stephen. 1977. ‘Two Kinds of Respect’. Ethics 88, 36–​49. Deaton, Angus. 2013. ‘What Does the Empirical Evidence Tell Us about the Injustice of Health Inequalities?’ In:  Inequalities in Health:  Concepts, Measures, and Ethics, eds. Nir Eyal, Samia Hurst, Ole Norheim, and Daniel Wikler. New York: Oxford University Press, pp. 262–​281. Dworkin, Ronald. 2000. Sovereign Virtue. Cambridge, MA: Harvard University Press. Dworkin, Ronald. 2002. ‘Sovereign Virtue Revisited’. Ethics 113, 106–​143. Dworkin, Ronald. 2003. ‘Equality, Luck and Hierarchy’. Philosophy and Public Afairs 31, 190–​198. Dworkin, Ronald. 2011. Justice for Hedgehogs. Cambridge, MA: Harvard University Press. Elford, Gideon. 2012. ‘Equality of Status and Distributive Equality’. Journal of Value Inquiry 46, 353–​367. Elford, Gideon. 2017. ‘Relational Equality and Distribution’. Journal of Political Philosophy 25,  80–​99. Enoch, David. 2007. ‘Intending, Foreseeing, and the State’. Legal Teory 13, 69–​99. Esping-​Andersen, Gøsta. 1990. Te Tree World of Welfare Capitalism. Princeton, NJ: Princeton University Press. Estlund, David. 1998. ‘Liberalism, Equality, and Fraternity in Cohen’s Critique of Rawls’. Journal of Political Philosophy 6, 99–​112. Estlund, David. 2000. ‘Political Quality’. Social Philosophy & Policy 17, 127–​160. Estlund, David. 2007. Democratic Authority:  A Philosophical Framework. Princeton, NJ: Princeton University Press. Fleischacker, Samuel. 2004. A Short History of Distributive Justice. Cambridge, MA: Harvard University Press.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography  305 Floris, Giacomo. 2019. Moral Status, Equality, and Distributive Justice. PhD thesis, University of Manchester. Foerster, Michael, and Mira d’Ercole, Marco. 2005. ‘Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s’. OECD Social, Employment and Migration Working Papers 22, https://​www.oecd-​ilibrary.org/​docserver/​882106484586. pdf?expires=1573823389&id=id&accname=guest&checksum=648DC8A0EF712E80 E054A271348F80B1. Forst, Rainer. 2012. Te Right to Justifcation: Elements of a Constructivist Teory of Justice. New York: Columbia University Press. Forst, Rainer. 2013. ‘A Kantian Republican Conception of Justice as Non-​Domination’. In Republican Democracy: Liberty, Law and Politics, eds. Andreas Niederberger and Philipp Schink, pp. 154–​168. Forst, Rainer. 2014. Justice, Democracy, and the Right to Justifcation. London: Bloomsbury. Fourie, Carina. 2012. ‘What Is Social Equality? An Analysis of Status Equality as a Strongly Egalitarian Ideal’. Res Publica 18, 107–​126. Fourie, Carina. 2015. ‘To Praise and to Scorn: Te Problem of Inequalities of Esteem for Social Egalitarians’. In: Fourie et al. 2015, pp. 87–​106. Fourie, Carina, Schuppert, Fabian, and Wallimann-​ Helmer (eds.). 2015. Social Equality: On What It Means to Be Equals. New York: Oxford University Press 2015. Frankfurt, Harry. 1987. ‘Equality as a Moral Ideal’. Ethics 98, 21–​43. Frankfurt, Harry. 1997. ‘Equality and Respect’. Social Research 64, 3–​15. Frankfurt, Harry. 2000. ‘Te Moral Irrelevance of Equality’. Public Afairs Quarterly 14, 87–​103. Fricker, Miranda. 2007. Epistemic Injustice:  Power and the Ethics of Knowing. Oxford: Oxford University Press. Garrau, Marie, and Laborde, Cécile. 2015. ‘Relational Equality, Non-​Domination, and Vulnerability’. In Fourie et al. 2015, pp. 45–​64. Gheaus, Anca. 2013. ‘Te Feasibility Constraint on the Concept of Justice’. Te Philosophical Quarterly 63, 445–​464. Gheaus, Anca. 2018. ‘Hikers in Flip-​Flops: Luck Egalitarianism, Democratic Equality, and the Distribuenda of Justice’. Journal of Applied Philosophy 35, 54–​69. Gilabert, Pablo. 2015. ‘Solidarity, Equality, and Freedom in Pettit’s Republicanism’. Critical Review of International Social and Political Philosophy 18, 644–​651. Gilbert, Margaret. 1989. On Social Facts. London: Routledge. Gonzalez-​Ricoy, Iñigo, and Queralt, Jahel. 2018. ‘Political Liberties and Social Equality’. Law and Philosophy 37, 613–​638. Goodin, Robert E. 2007. ‘Why Social Justice Is Not All Tat Matters: Justice as the First Virtue’. Ethics 117, 413–​432. Goodin, Robert E., and Jackson, Frank. 2007. ‘Freedom from Fear’. Philosophy & Public Afairs 35, 249–​265. Gosepath, Stefan. 2004. Gleiche Gerechtigkeit. Frankfurt: Suhrkamp. Govier, Trudy. 1993. ‘Self-​Trust, Autonomy, and Self-​Esteem’. Hypatia 8, 99–​120. Gowder, Paul. 2016. Te Rule of Law in the Real World. Cambridge:  Cambridge University Press. Guinan, Joe, and O’Neill, Martin. 2019. Te Case for Community Wealth Building. Cambridge, UK: Polity.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

306 Bibliography Gutmann, Amy. 2002. ‘Rawls on the Relationship between Liberalism and Democracy’. In:  Te Cambridge Companion to Rawls, ed. Samuel Freeman. Cambridge, UK: Cambridge University Press, pp. 168–​199. Gutmann, Amy, and Tompson, Dennis. 2004. Why Deliberative Democracy? Princeton, NJ: Princeton University Press. Habermas, Jürgen. 1975. Legitimation Crisis. Boston: Beacon Press. Hausman, Daniel M., and Waldren, Matt Sensat. 2011. ‘Egalitarianism Reconsidered’.Journal of Moral Philosophy 8, 567–​586. Haverkamp, Beatrijs, Verweij, Marcel, and Stronks, Karien. 2018. ‘Why Socio-​Economic Inequalities in Health Treaten Relational Justice. A  Proposal for an Instrumental Evaluation’. Public Health Ethics 11, 311–​324. Heath, Joseph. 2008. ‘Political Egalitarianism’. Social Teory and Practice 34, 486–​516. Hill, Tomas E. 1991. Autonomy and Self-​ Respect. Cambridge, UK:  Cambridge University Press. Honneth, Axel. 1995. Te Struggle for Recognition: Te Moral Grammar of Social Conficts. Cambridge, UK: Polity Press. Hosein, Adam Omar. 2014. ‘Doing, Allowing, and the State’. Law and Philosophy 33, 235–​264. Howard, Jefrey. 2019. ‘Te Labors of Justice: Democracy, Respect, and Judicial Review’. Critical Review of International Social and Political Philosophy 22, 176–​199. Ip, Kevin K. W. 2016. Egalitarianism and Global Justice: From a Relational Perspective. London: Palgrave Macmillan. James, Aaron. 2005. ‘Constructing Justice for Existing Practice: Rawls and the Status Quo’. Philosophy & Public Afairs 33, 281–​316. Jarvis, Martin, and Wardle, Jane. 2005. ‘Social Patterning of Individual Health Behaviours:  Te Case of Cigarette Smoking’. In:  Marmot and Wilkinson 2005, pp. 224–​237. Johnston, David. 2011. A Brief History of Justice. Oxford: Wiley-​Blackwell. Jütten, Timo. 2017. ‘Dignity, Esteem and Social Contribution: A Recognition-​Teoretical View’. Journal of Political Philosophy 25, 259–​280. Kant, Immanuel. 1996 [1797]. Metaphysics of Morals, ed. Mary Gregor. Cambridge, UK: Cambridge University Press. Kelleher, Paul. 2013. ‘Real and Alleged Problems for Daniels’s Account of Health Justice’. Journal of Medicine and Philosophy 38, 388–​399. Kittay, Eva Feder. 1996. ‘Human Dependency and Rawlsian Equality’. In:  Feminists Rethink the Self, ed. Diana Meyers. Boulder, CO: Westview Press, pp. 219–​266. Kolodny, Niko. 2014. ‘Rule over None II:  Social Equality and the Justifcation of Democracy’. Philosophy & Public Afairs 42, 287–​336. Kolodny, Niko. 2019. ‘Being under the Power of Others’. In Republicanism and Democracy, eds. Yifhar Elizar and Geneviève Rousselière. Cambridge, UK: Cambridge University Press, pp. 94–​114. Kramer, Matthew. 2003. Te Quality of Freedom. Oxford: Oxford University Press. Kristjànnson, Kristjàn. 2010. Te Self and Its Emotions. Cambridge, UK:  Cambridge University Press. Kropotkin, Peter. 1906. Te Conquest of Bread. London; New York: G. P. Putnam’s Sons. Kropotkin, Peter. 1976 [1902]. Mutual Aid. Boston: Porter Sargent. Kymlicka, Will. 2002. Contemporary Political Philosophy, 2nd ed. Oxford:  Oxford University Press.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography  307 Laborde, Cécile. 2010. ‘Republicanism and Global Justice: A Sketch’. European Journal of Political Teory 9, 48–​69. Lareau, Annette. 2011. Unequal Childhoods, 2nd ed. Berkeley:  University of California Press. Lessig, Lawrence. 1996. ‘Social Meaning and Social Norms’. University of Pennsylvania Law Review 144, 2181–​2189. Levy, Jacob T. 2016. ‘Tere Is No Such Ting as Ideal Teory’. Social Philosophy and Policy 33, 312–​333. Lippert-​Rasmussen, Kasper. 2015. Luck Egalitarianism. London: Bloomsbury. Lippert-​Rasmussen, Kasper. 2018a. ‘(Luck and Relational) Egalitarians of the World, Unite!’ In:  Oxford Studies in Political Philosophy, Vol. 4, eds. David Sobel, Peter Vallentyne, and Steven Wall. Oxford: Oxford University Press, pp. 81–​109. Lippert-​ Rasmussen, Kasper. 2018b. Relational Egalitarianism:  Living as Equals. Cambridge, UK: Cambridge University Press. List, Christian. 2006. ‘Republican Freedom and the Rule of Law’. Politics, Philosophy & Economics 5, 201–​220. List, Christian, and Pettit, Philip. 2002. ‘Aggregating Sets of Judgements: An Impossibility Result’. Economics and Philosophy, 18, 89–​110. List, Christian, and Pettit, Philip. 2006. ‘Group Agency and Supervenience’. Southern Journal of Philosophy 44, 85–​105. List, Christian, and Pettit, Philip. 2011. Group Agency: Te Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press. List, Christian, and Valentini, Laura. 2016. ‘Freedom as Independence’. Ethics 126, 1043–​1074. Lovett, Frank. 2010. A General Teory of Justice and Domination. Oxford:  Oxford University Press. Lovett, Frank. 2016a. A Republic of Law. Cambridge: Cambridge University Press. Lovett, Frank. 2016b. ‘Civic Republicanism and Social Justice’. Political Teory 44, 687–​696. Lovett, Frank, and Pettit, Philip. 2009. ‘Neorepublicanism: A Normative and Institutional Research Program’. Annual Review of Political Science 12, 11–​29. Lovett, Frank, and Whitfeld, Gregory. 2016. ‘Republicanism, Perfectionism, and Neutrality’. Journal of Political Philosophy 24, 120–​134. Lynch, John, Davey Smith, George, Kaplan, George, and House, James. 2000. ‘Income Inequality and Mortality: Importance to Health of Individual Income, Psychosocial Environment, or Material Conditions’. BMJ 320, 1200–​1206. Macfarlane, Kirsty. 2018. ‘Education, Sufciency, and the Relational Egalitarian Ideal’. Journal of Applied Philosophy 35, 759–​774. MacPherson, Crawford. 1962. Te Political Teory of Possessive Individualism:  From Hobbes to Locke. New York: Oxford University Press. Mansbridge, Jane, Bohman, James, Chambers, Simone, Estlund, David, Føllesdal, Andreas, Fung, Archon, Lafont, Cristina, Manin, Bernard, and Martí, José Luis. 2010. ‘Te Place of Self-​Interest and the Role of Power in Deliberative Democracy’. Journal of Political Philosophy 18, 64–​100. March, James G., and Olsen, Johan P. 1989. Rediscovering Institutions: the Organizational Basis of Politics. New York: Te Free Press. Marmot, Michael. 2004. Status Syndrome: How Your Social Standing Directly Afects Your Health. London: Bloomsbury.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

308 Bibliography Marmot, Michael Shipley, and Rose, Geofrey. 1984. ‘Inequalities in Death:  Specifc Explanations of a General Pattern’. Lancet 323, 1003–​1006. Marmot, Michael, Siegrist, Johannes, and Teorell, Tores. 2005. ‘Health and the Psychosocial Environment at Work’. In: Marmot and Wilkinson 2005, pp. 97–​130. Marmot, Michael, Davey Smith, George, Stansfeld, Stephen, Patel, Chandra, North, Fiona, Head, Jenny, White, Ian, Brunner, Eric, and Feeney, Amanda. 1991. ‘Health Inequalities among British Civil Servants: Te Whitehall II Study’. Lancet 337, 1387–​1393. Marmot, Michael, and Wilkinson, Richard. 2005. Social Determinants of Health, 2nd ed. Oxford: Oxford University Press. Marshall, T. H. 1950. ‘Citizenship and Social Class’. In Citizenship and Social Class and Other Essays. Cambridge, UK: Cambridge University Press, pp. 1–​85. Marx, Karl. 2000 [1875]. ‘Critique of the Gotha Programme’, partly reprinted in Selected Writings, 2nd ed., ed. David McLellan. Oxford: Oxford University Press, pp. 610–​617. Mason, Andrew. 2012. Living Together as Equals:  Te Demands of Citizenship. Oxford: Oxford University Press. Mason, Andrew. 2015. ‘Justice, Respect, and Treating People as Equals’. In Fourie et al. 2015, pp. 129–​145. Maynor, John W. 2003. Republicanism in the Modern World. Cambridge, UK: Polity. McBride, Cillian. 2013. Recognition. Cambridge, UK: Polity. McBride, Cillian. 2015. ‘Freedom as Non-​Domination: Radicalisation or Retreat?’ Critical Review of International Social and Political Philosophy 18, 349–​374. McCammon, Christopher. 2015. ‘Domination—​A Rethinking’. Ethics 125, 1028–​1052. McMahon, Christopher. 2005. ‘Te Indeterminacy of Republican Policy’. Philosophy & Public Afairs 33, 67–​93. McMahon, Christopher. 2007. ‘Nondomination and Normativity’. Pacifc Philosophical Quarterly 88, 319–​327. McTernan, Emily. 2018. ‘Microaggressions, Equality, and Social Practices’. Journal of Political Philosophy 26, 261–​281. Miklosi, Zoltan. 2018. ‘Varieties of Relational Egalitarianism’. In: Oxford Studies in Political Philosophy, Vol. 4, eds. David Sobel, Peter Vallentyne, and Steven Wall. Oxford: Oxford University Press, pp. 110–​136. Mill, John Stuart. 1861. Considerations on Representative Government. London: Parker, Son and Bourn. Miller, David. 1990. ‘Equality’. In: Philosophy and Politics, ed. G. M. K. Hunt. Cambridge, UK: Cambridge University Press, pp. 77–​98. Miller, David. 1995. ‘Complex Equality’. In: Pluralism, Justice, and Equality, eds. David Miller and Michael Walzer. Oxford: Oxford University Press, pp. 197–​225. Miller, David. 1998. ‘Equality and Justice’. In:  Ideals of Equality, ed. Andrew Mason. Oxford: Blackwell, pp. 21–​36. Miller, David. 1999. Principles of Social Justice. Cambridge, MA: Harvard University Press. Mills, Charles. 2017. Black Rights/​White Wrongs:  Te Critique of Racial Liberalism. New York: Oxford University Press. Moles, Andres, and Parr, Tom. 2019. ‘Distributions and Relations: A Hybrid Account’. Political Studies 67, 132–​148. Montesquieu, Charles de Secondat, Baron de. 1989 [1748]. Te Spirit of the Laws, eds. Anna Cohler, Basia Miller, and Harold Stone. Cambridge, UK: Cambridge University Press. More, Tomas. 2003 [1516]. Utopia, ed. Paul Turner. London: Penguin.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography  309 Moyo, Dambisa. 2018. Edge of Chaos:  Why Democracy Is Failing to Deliver Economic Growth—​And How to Fix It. New York: Basic Books. Murphy, Liam. 2000. Moral Demands in Nonideal Teory. Oxford: Oxford University Press. Nagel, Tomas. 1991. Equality and Partiality. Oxford: Oxford University Press. Nagel, Tomas. 1997. ‘Justice and Nature’. Oxford Journal of Legal Studies 17, 303–​321. Nath, Rekha. 2011. ‘Equal Standing in the Global Community’. Te Monist 94, 593–​614. Nath, Rekha. 2015. ‘On the Scope and Grounds of Social Equality’. In Fourie et al. 2015, pp. 186–​208. Nazroo, James, and Williams, David. 2005. ‘Te Social Determination of Ethnic/​Racial Inequalities in Health’. In: Marmot and Wilkinson 2005, pp. 238–​267. Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Basic Books. Nussbaum, Martha. 2000. Women and Human Development. Cambridge, UK: Cambridge University Press. O’Neill, Martin. 2008. ‘What Should Egalitarians Believe?’ Philosophy and Public Afairs 36, 119–​156. O’Neill, Martin, McTernan, Emily, Schemmel, Christian, and Schuppert, Fabian. 2016. ‘If You Care about Social Equality, You Want a Big State: Home, Work, Care and Social Egalitarianism’. Juncture 23, 138–​144. Okin, Susan Moller. 1989. Justice, Gender, and the Family. New York: Basic Books. Olsaretti, Serena. 2009. ‘Responsibility and the Consequences of Choice’. Proceedings of the Aristotelian Society 109, 165–​188. Parft, Derek. 1991. ‘Equality or Priority’. Te 1991 Lindley Lecture at the University of Kansas, reprinted in:  Te Ideal of Equality, eds. Matthew Clayton and Andrew Williams. Basingstoke, UK: Palgrave Macmillan 2002, pp. 81–​125. Patton, Paul. 2015. ‘Political Legitimacy’. Critical Review of International Social and Political Philosophy 18, 661–​668. Perelman, Chaim. 1963. Te Idea of Justice and the Problem of Argument. London: Routledge and Kegan Paul. Pettit, Philip. 1998. ‘Reworking Sandel’s Republicanism’. Te Journal of Philosophy 95,  73–​96. Pettit, Philip. 1999a. Republicanism: A Teory of Freedom and Government, paperback ed. with a postscript. Oxford: Oxford University Press. Pettit, Philip. 1999b. ‘Republican Freedom and Contestatory Democratization’. In:  Democracy’s Value, eds. Ian Shapiro and Casiano Hacker-​Cordon. Cambridge, UK: Cambridge University Press, pp. 163–​190. Pettit, Philip. 2001. A Teory of Freedom: From the Psychology to the Politics of Agency. Oxford: Oxford University Press. Pettit, Philip. 2003. ‘Groups with Minds of their Own’. In Socializing Metaphysics:  Te Nature of Social Reality, ed. Frederick Schmitt. Lanham, MD: Rowman & Littlefeld, pp. 167–​193. Pettit, Philip. 2006. ‘Te Determinacy of Republican Policy:  A Reply to McMahon’. Philosophy & Public Afairs 34, 275–​283. Pettit, Philip. 2007. ‘Responsibility Incorporated’. Ethics 117, 171–​201. Pettit, Philip. 2008. ‘Freedom and Probability:  A Comment on Goodin and Jackson’. Philosophy & Public Afairs 36, 206–​220. Pettit, Philip. 2012. On the People’s Terms: A Republican Teory and Model of Democracy. Cambridge, UK: Cambridge University Press.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

310 Bibliography Pettit, Philip. 2014. Just Freedom:  A Moral Compass for a Complex World. New York: Norton. Pettit, Philip. 2015a. ‘Justice: Social and Political’. In Oxford Studies in Political Philosophy, Vol. 1, eds. David Sobel, Peter Vallentyne, and Steven Wall. Oxford: Oxford University Press, pp. 9–​35. Pettit, Philip. 2015b. ‘On the People’s Terms: A Reply to Five Critiques’. Critical Review of International Social and Political Philosophy 18, 687–​696. Pettit, Philip. 2017. ‘Political Realism Meets Civic Republicanism’. Critical Review of International Social and Political Philosophy 20, 331–​347. Pogge, Tomas. 1995. ‘Tree Problems with Contractarian-​Consequentialist Ways of Assessing Social Institutions’. Social Philosophy and Policy 12, 241–​266. Pogge, Tomas. 2000. ‘On the Site of Distributive Justice:  Refections on Cohen and Murphy’. Philosophy & Public Afairs 29, 137–​169. Pogge, Tomas. 2003. ‘Relational Conceptions of Justice:  Responsibility for Health Outcomes’, in Public Health, Ethics, and Equity, eds. Sudhir Anand, Fabienne Peter, and Amartya Sen. Oxford: Oxford University Press, pp. 135–​161. Pogge, Tomas. 2008. World Poverty and Human Rights, 2nd ed. Cambridge, UK: Polity Press. Proudhon, Pierre-​ Joseph. 2011 [1846]. System of Economic Contradictions, or:  Te Philosophy of Poverty, Calgary: Teophania. Quong, Jonathan. 2010. Liberalism without Perfection. Oxford: Oxford University Press. Quong, Jonathan. 2018. ‘Consequentialism, Deontology, Contractualism, and Equality’. In: Te Oxford Handbook of Distributive Justice, ed. Serena Olsaretti. Oxford: Oxford University Press, pp. 306–​326. Rasmussen, Dennis. 2016. ‘Adam Smith on What Is Wrong with Economic Inequality’. American Political Science Review 110, 342–​352. Rawls, John. 1996. Political Liberalism, with a new introduction and the ‘Reply to Habermas’. New York: Columbia University Press. Rawls, John. 1999a. A Teory of Justice, rev. ed. Oxford: Oxford University Press. Rawls, John. 1999b. Te Law of Peoples, with Te Idea of Public Reason Revisited. Cambridge, MA: Harvard University Press. Rawls, John. 2001. Justice as Fairness: A Restatement, ed. Erin Kelly. Cambridge, MA: Te Belknap Press of Harvard University Press. Rawls, John. 2007. Lectures on the History of Political Philosophy, ed. Samuel Freeman. Cambridge, MA: Te Belknap Press of Harvard University Press. Raz, Joseph. 1986. Te Morality of Freedom. Oxford: Clarendon Press. Richardson, Henry. 2002. Democratic Autonomy:  Public Reasoning about the Ends of Policy. Oxford: Oxford University Press. Richardson, Henry. 2006. ‘Republicanism and Democratic Injustice’. Politics, Philosophy & Economics 5, 175–​200. Ripstein, Arthur. 2009. Force and Freedom:  Kant’s Legal and Political Philosophy. Cambridge, MA: Harvard University Press. Ronzoni, Miriam. 2008. ‘What Makes a Basic Structure Just?’ Res Publica 14, 203–​218. Ronzoni, Miriam. 2009. ‘Te Global Order: A Case of Background Injustice? A Practice-​ Dependent Account’. Philosophy & Public Afairs 37, 229–​256. Rothstein, Bo. 1998. Just Institutions Matter: Te Moral and Political Logic of the Welfare State. Cambridge, UK: Cambridge University Press.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography  311 Rousseau, Jean-​ Jacques. 1968 [1762]. Te Social Contract, ed. Maurice Cranston. London: Penguin. Rousseau, Jean-​Jacques. 1984 [1755]. A Discourse on Inequality, ed. Maurice Cranston. London: Penguin. Ryan, Alan. 2002. ‘Does Inequality Matter—​For Its Own Sake?’ Social Philosophy and Policy 19, 225–​243. Sandberg, Sheryl. 2013. Lean In: Women, Work, and the Will to Lead. New York: Knopf. Sanders, Lynn. 1997. ‘Against Deliberation’. Political Teory 25, 347–​376. Sangiovanni, Andrea. 2007. ‘Global Justice, Reciprocity, and the State’. Philosophy and Public Afairs 35, 3–​39. Sangiovanni, Andrea. 2008. ‘Justice and the Priority of Politics to Morality’. Journal of Political Philosophy 16, 137–​164. Satz, Debra. 2007. ‘Equality, Adequacy, and Education for Citizenship’. Ethics 117, 623–​648. Scanlon, T. M. 2008. Moral Dimensions. Cambridge, MA: Harvard University Press. Scanlon, T. M. 2018. Why Does Inequality Matter? Oxford: Oxford University Press. Schefer, Samuel. 1994. Te Rejection of Consequentialism, rev. ed. Oxford:  Oxford University Press. Schefer, Samuel. 2001. Boundaries and Allegiances: Problems of Justice and Responsibility in Liberal Tought. New York: Oxford University Press. Schefer, Samuel. 2003a. ‘What Is Egalitarianism?’ Philosophy and Public Afairs 31,  5–​39. Schefer, Samuel. 2003b. ‘Equality as the Virtue of Sovereigns’. Philosophy and Public Afairs 31, 199–​206. Schefer, Samuel. 2005. ‘Choice, Circumstance, and the Value of Equality’. Politics, Philosophy & Economics 4, 5–​28. Schefer, Samuel. 2006. ‘Is the Basic Structure Basic?’ In: Te Egalitarian Conscience: Essays in Honour of G. A. Cohen, ed. Christine Sypnowich. Oxford: Oxford University Press, pp. 102–​129. Schefer, Samuel. 2015. ‘Te Practice of Equality’. In: Fourie et al. 2015, pp. 21–​44. Schemmel, Christian. 2011. ‘Why Relational Egalitarians Should Care about Distributions’. Social Teory and Practice 37, 365–​390. Schemmel, Christian. 2012. ‘Luck Egalitarianism as Democratic Reciprocity? A Response to Tan’. Te Journal of Philosophy 109, 433–​448. Schemmel, Christian. 2015a. ‘Social Equality—​or Just Justice?’ In: Fourie et al. 2015, pp. 146–​166. Schemmel, Christian. 2015b. ‘How (Not) to Criticise the Welfare State’. Journal of Applied Philosophy 32, 393–​409. Schemmel, Christian. 2019. ‘Real Self-​Respect and Its Social Bases’. Canadian Journal of Philosophy 49, 628–​651. Schmidt, Andreas T. 2018. ‘Domination without Inequality? Mutual Domination, Republicanism, and Gun Control’. Philosophy & Public Afairs 46, 175–​206. Schuppert, Fabian. 2015. ‘Non-​ Domination, Non-​ Alienation, and Social Equality: Towards a Republican Understanding of Equality’. Critical Review of International Social and Political Philosophy 18, 440–​455. Segall, Shlomi. 2010. ‘Is Health (Really) Special? Health Policy between Rawlsian and Luck Egalitarian Justice’. Journal of Applied Philosophy 27, 344–​358.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

312 Bibliography Sen, Amartya. 1979. ‘Equality of What?’, Te Tanner Lecture on Human Values 1979, available at http://​tannerlectures.utah.edu/​_​documents/​a-​to-​z/​s/​sen80.pdf (last accessed 30 November 2020). Sen, Amartya. 1999. Development as Freedom. Oxford: Oxford University Press. Sen, Amartya. 2009. Te Idea of Justice. Cambridge, MA: Harvard University Press. Shapiro, Ian. 2012. ‘On Non-​Domination’. University of Toronto Law Journal 62, 293–​335. Shapiro, Ian. 2016. Politics against Domination. Cambridge, MA: Harvard University Press. Shaw, Mary, Dorling, Danny, and Davey Smith, George. 2005. ‘Poverty, Social Exclusion, and Minorities’. In: Marmot and Wilkinson 2005, pp. 196–​223. Smith, Adam. 1976 [1776]. An Inquiry into Nature and Causes of the Wealth of Nations, eds. Roy Campbell, Andrew Skinner, Oxford, Clarendon Press. Stanczyk, Lukas. Unpublished manuscript. ‘From Each: A Teory of Productive Justice’. Stansfeld, Stephen. 2005. ‘Social Support and Social Cohesion’. In: Marmot and Wilkinson 2005, pp. 148–​171. Stark, Cynthia A. 2007. ‘How to Include the Severely Disabled in a Contractarian Teory of Justice’. Journal of Political Philosophy 15, 127–​145. Stark, Cynthia A. 2012. ‘Rawlsian Self-​Respect’. In: Oxford Studies in Normative Ethics, Vol. 2, ed. Mark Timmons. Oxford: Oxford University Press, pp. 238–​261. Stemplowska, Zofa. 2015. ‘Rescuing Luck Egalitarianism’. Journal of Social Philosophy 44, 402–​419. Stemplowska, Zofa, and Swif, Adam. 2012. ‘Ideal and Nonideal Teory’. In: Te Oxford Handbook of Political Philosophy, ed. David Estlund. Oxford: Oxford University Press, pp. 373–​390. Strawson, Peter. 1974. ‘Freedom and Resentment’. In: Freedom and Resentment and Other Essays. London: Methuen, pp. 1–​25. Swif, Adam. 2003. How Not to Be a Hypocrite: School Choice for the Morally Perplexed. London: Routledge. Swif, Adam. 2004. ‘Te Morality of School Choice Reconsidered: A Response’. Teory and Research in Education 2, 323–​342. Swif, Adam, and White, Stuart. 2008. ‘Political Teory, Social Science, and Real Politics’. In:  Political Teory:  Methods and Approaches, eds. David Leopold and Marc Stears. Oxford: Oxford University Press, pp. 49–​69. Tawney, Richard Henry. 1931. Equality. London: Allen & Unwin. Tomas, Alan. 2017. Republic of Equals: Predistribution and Property-​Owning Democracy. New York: Oxford University Press. Tilly, Charles. 1998. Durable Inequality. Berkeley: University of California Press. Titelbaum, Michael. 2008. ‘What Would a Rawlsian Ethos of Justice Look Like?’ Philosophy & Public Afairs 36, 289–​322. Tomlin, Patrick. 2014. ‘What Is the Point of Egalitarian Social Relationships?’ In:  Distributive Justice and Access to Advantage:  G. A.  Cohen’s Egalitarianism, ed. Alexander Kaufman. Cambridge, UK: Cambridge University Press, pp. 151–​179. Valentini, Laura. 2011. ‘Global Justice and Practice-​ Dependence:  Conventionalism, Institutionalism, Functionalism’. Journal of Political Philosophy 19, 399–​418. Valentini, Laura. 2012. ‘Ideal Teory vs. Non-​ideal Teory: A Conceptual Map’. Philosophy Compass 7, 654–​664. Valentini, Laura. 2016. ‘What’s Wrong with Being Lonely? Justice, Benefcence, and Meaningful Relationships’. Proceedings of the Aristotelian Society, Supplementary Volume XC, 49–​69.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Bibliography  313 Viehof, Daniel. 2014. ‘Democratic Equality and Political Authority’. Philosophy & Public Afairs 42, 337–​375. Viehof, Daniel. 2019. ‘Power and Equality’. In:  Oxford Studies in Political Philosophy, Vol. 5, eds. David Sobal, Peter Vallentyne, and Steven Wall. Oxford: Oxford University Press, pp. 3–​38. Voigt, Kristin. 2007. ‘Te Harshness Objection: Is Luck Egalitarianism Too Harsh on the Victims of Option Luck?’ Ethical Teory and Moral Practice 10, 389–​407. Voigt, Kristin. 2018. ‘Relational Equality and the Expressive Dimension of State Action’. Social Teory and Practice 44, 437–​467. Voigt, Kristin, and Wester, Gry. 2015. ‘Relational Equality and Health’. Social Philosophy and Policy 31, 204–​229. Von Platz, Jeppe. 2014. ‘Are Economic Liberties Basic Rights?’ Politics, Philosophy & Economics 13, 23–​44. Waldron, Jeremy. 1995. ‘Money and Complex Equality’. In Pluralism, Justice, and Equality, eds. David Miller and Michael Walzer. Oxford: Oxford University Press, pp. 144–​170. Wall, Steven. 2007. ‘Democracy and Equality’. Te Philosophical Quarterly 57, 416–​437. Wallraf, Günter. 1988. Lowest of the Low. London: Mandarin. Walzer, Michael. 1983. Spheres of Justice:  A Defense of Pluralism and Equality. New York: Basic Books. Weinstock, Daniel M. 2011. ‘How Should Political Philosophers Tink of Health?’ Journal of Medicine and Philosophy 36, 424–​435. White, Stuart. 2007. Equality. Cambridge, UK: Polity Press. Wilkinson, Richard. 1996. Unhealthy Societies:  Te Afictions of Inequality. London: Routledge. Wilkinson, Richard. 2005. ‘Ourselves and Others, for Better or Worse: Social Vulnerability and Inequality’. In: Marmot and Wilkinson 2005, pp. 341–​357. Wilkinson, Richard, and Pickett, Kate. 2010. Te Spirit Level: Why Equality Is Better for Everyone, new ed. London: Penguin. Wilkinson, Richard, and Pickett, Kate. 2018. Te Inner Level: How More Equal Societies Reduce Stress, Restore Sanity, and Improve Everyone’s Wellbeing. London: Allen Lane. Williams, Andrew. 1998. ‘Incentives, Inequality, and Publicity’. Philosophy & Public Afairs 27, 225–​247. Williams, Bernard. 1973. ‘Te Idea of Equality’. In:  Problems of the Self. Cambridge, UK: Cambridge University Press, pp. 230–​249. Wilson, James Lindley. 2019. Democratic Equality. Princeton, NJ:  Princeton University Press. Winstanley, Gerrard. 1973. Te Law of Freedom and Other Writings, ed. Christopher Hill. London: Penguin. Wolf, Jonathan. 1998. ‘Fairness, Respect, and the Egalitarian Ethos’. Philosophy & Public Afairs 27, 97–​122. Wolf, Jonathan. 2009. ‘Disability among Equals’. In:  Disability and Disadvantage, eds. Kimberley Brownlee and Adam Cureton. Oxford:  Oxford University Press, pp. 112–​137. Wolf, Jonathan. 2010. ‘Fairness, Respect and the Egalitarian Ethos Revisited’. Journal of Ethics 14, 335–​350. Wolf, Jonathan. 2013. ‘Scanlon on Social and Material Inequality’. Journal of Moral Philosophy 10, 406–​425.

314 Bibliography

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Wolf, Jonathan 2015. ‘Social Equality and Social Inequality’. In Fourie et al. 2015, pp. 209–​225. Wolf, Jonathan. 2017. ‘Forms of Diferential Social Inclusion’. Social Philosophy and Policy 34, 164–​185. Wolf, Jonathan. 2019. ‘Equality and Hierarchy’. Proceedings of the Aristotelian Society 119,  1–​23. Wolf, Jonathan, and De-​ Shalit, Avner. 2007. Disadvantage. Oxford:  Oxford University Press. Wright, Anthony. 1987. R. H. Tawney. Manchester: Manchester University Press. Young, Iris Marion. 1990. Justice and the Politics of Diference. Princeton, NJ: Princeton University Press. Young, Iris Marion. 1996. ‘Communication and the Other:  Beyond Deliberative Democracy’. In Democracy and Diference: Contesting the Boundaries of the Political. Princeton, NJ: Princeton University Press, pp. 120–​135.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Index For the beneft of digital users, indexed terms that span two pages (e.g., 52–​53) may, on occasion, appear on only one of those pages.    accountability, 67, 81, 217 Barry, Nicholas, 51n.68 Ackerman, Bruce, 222 basic liberties, 14, 88–​91, 94–​95, 115n.41, action-​guidance 130, 205, 213, 240, 275–​76 and conceptions of justice 18, 20, 30–​31, and moral powers, 90, 101–​3 110, 114–​15, 154n.49, 158 basic needs, 20n.30, 75–​76, 91–​92, agent-​centred prerogative, 35, 285n.63 103n.13, 182–​83, 284n.61 allocative justice, 236–​37 basic structure of society, the, 42–​44, anarchism,  61–​62 46–​47, 50–​51, 61n.6, 69–​70, Anderson, Elizabeth, 23–​24, 33, 43, 47–​48, 186n.37,  236–​38 164n.2, 165n.5, 184n.34, 232–​35, 236, equal political power over, 83–​87 237, 238–​39, 240–​41, 242–​43, 244–​ restriction of duties of justice to, 119–​ 46, 247, 252n.52 23, 192n.48 Aristotle, 6n.1, 7n.7 benefcence, duties of, 238n.21, 284, Arneson, Richard, 24, 25–​26, 160n.64 295–​96n.19 attitudes. See also expressive concept of bias, unconscious, 121n.51, 170n.13 justice Boorse, Christopher, 261n.3 communication versus expression of, Bourdieu, Pierre, 166n.6 47–​48, 53–​54, 166–​67,  209–​10 Britain, 6 conducive to non-​domination (see non-​ brute luck, 10–​11, 22, 25–​26, 284–​85. See domination: ethos of) also luck egalitarianism expressed by institutions, 22–​23,    28–​29,  38–​53 capabilities, 10–​11, 13, 50–​51, 138n.17, expressed in action vs. outcome 140n.25, 233 produced by action, 46–​47 care expressed without social value of relations of, 140–​41 relations,  51–​52 virtue of, 59–​60 expressed versus concealed, 171–​73 Carter, Ian, 108–​9, 160n.62 and intentions, 45–​46 caste, 25, 36 institutional vs. individual, 40–​44 Christiano, Tomas, 216 and meanings of actions, 13, 45–​46 citizenship, 104n.14, 134–​35n.9, 139–​ needed for egalitarian relations, 40n.23, 208n.16, 244 147–​50,  171–​73 claims of justice, prima facie versus fnal, reactive, 43–​44,  47–​48 74–​78, 104–​5, 106–​7, 124, 202 authority. See domination: and de facto class, social, 20, 36, 46–​47, 87n.69, 165–​ authority; social status: and de facto 66, 167–​68, 177–​78, 187, 188–​89, authority 218n.33, 235, 250–​51, 253–​55, 296 autonomy, 9–​10, 209. See also moral Cohen, G.A., 24, 30–​31, 35n.37, 136–​37, powers; self-​respect: and autonomy    141–​42, 225–​26, 285n.63

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

316 Index Cohen, Joshua, 208 collective bargaining, 47, 48–​49 colonialism, 298 communism, 7–​8,  23–​24 community, value of, 4, 12, 127–​28, 132–​ 33, 136–​37, 141–​42, 149–​50, 151–​52, 157n.56,  274–​75 conception of the good, 8–​9, 18–​19, 34, 59–​60, 61–​62, 179, 275 constitutional review, 16–​17, 199–​200, 206–​7, 213,  218–​21 control informal social, 61–​62 and non-​domination (see non-​ domination: and control) Cordelli, Chiara, 156–​61 currency, of justice, 10–​11, 25–​26, 125–​26, 233n.4, 290, 291. See also metric of justice; purely recipient-​oriented conceptions of justice    Daniels, Norman, 208n.16, 260, 261–​62, 267n.14, 268–​69,  278–​81 De-​Shalit, Avner, 140n.25 democracy. See also political equality and non-​domination as democracy direct, 6–​7, 199, 216n.29, 217, 218–​19 epistemic value of, 210, 218–​20, 222–​23 desert, 238–​39, 255n.59 diagnostic value of theories of justice, 20–​ 21, 124, 290–​91 diference principle, 33, 120n.49, 131n.7, 205, 222, 240–​41, 279, 280–​81 Diggers, 6–​7,  132–​33 disability, 24, 76–​77, 238n.21, 253, 265n.10, 266n.11, 283 discrimination, 29–​30, 43, 47n.55, 139, 188n.43, 251, 253, 255 in healthcare, 260–​61, 272, 282, 285–​87 legal, 31, 37, 38–​39, 49–​50, 75–​76, 79–​ 80, 81, 221–​22 racial, 43, 205–​6, 230n.66, 271–​72 division of labour, 4–​5 and concept of health, 263–​65 deliberative,  216–​17 and liberalism, 18–​19, 56–​57, 61–​62 organized by state institutions, 18–​19,  294–​96

and political equality, 62–​63, 84–​85, 199 and unequal social positions, 61–​62,  84–​85 doctrine of double efect, 45–​46, 47n.55 domination. See also non-​domination as arbitrary power, 14, 63–​65, 71–​73, 74–​76, 79, 80–​81n.55, 83, 110–​11, 122–​23,  124 and common knowledge, 94, 97–​98,  125–​26 complaint of, 63–​64, 65–​66, 71, 77, 78, 84, 95, 99, 202–​3, 298–​99 and de facto authority, 95–​97, 175–​77,  297 and fear (see fear) by groups, 95, 100 intergenerational,  298–​99 international,  295–​97 in intimate relationships, 105–​7, 112–​14 mutual, 68 and probability of interference, 67, 68 as relational inequality, 64–​71 respectful protection against, 107–​9 and social status (see social status: domination) by state, 69–​71, 112–​13 structural, 94–​95, 98–​100, 120 in workplace, 84, 91–​92, 117, 182–​83,  271–​72 dominium versus imperium, 65n.15, 81–​82, 112–​13,  129–​30 duty of assistance, international, 295–​96n.19 Dworkin, Ronald, 9–​10, 16–​17, 24, 25–​26, 44–​45, 51n.69, 184n.34, 201, 203, 219n.34, 225–​31, 237n.15, 289–​90    egalitarian movements, 20, 39–​40 envy. See self-​esteem: and envy equal concern, 25–​26, 44–​45, 226, 289–​90 equality of opportunity. See opportunity equality of power, 59–​63. See also political equality esteem,  164–​65 equality of, 173–​74 expressed in action, 166–​68, 171–​73 individual versus social, 163, 193–​95

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Index  317 mutual, 148 and self-​esteem (see self-​esteem) social (see social status: norms of) and social class, 165–​66 versus respect (see respect versus esteem) Estlund, David, 210, 222–​24, 246n.39 ethnic equality, 20, 39–​40, 75–​76, 139, 143,  193–​94 exploitation, 23–​24, 77n.47, 105–​7, 175,  296–​97 expressive concept of justice, 12–​13, 38–​51 and distributive inequality, 236–​42,  256–​57 and healthcare, 260–​61, 267–​68, 277–​78,  286 and hierarchies, 22, 39, 55–​57 and moral equality, 39–​40, 55–​56, 290 and non-​domination, 54, 56, 88, 125–​ 26, 146, 147 and norms of social status, 162–​63,  189–​90 and political rights, 50–​51, 208–​10, 226–​27,  231 expressive wrongs, 20–​21 eyeball test. See non-​domination: ‘eyeball test’    fairness. See also society, liberal conception of of distributive outcomes versus of terms of social cooperation, 10–​11 luck egalitarian, 10–​11, 188, 289 (see also luck egalitarianism) fear and conformity, 167–​68 and domination, 89–​90, 91–​92, 97–​98 and health, 270–​72 feminism,  23–​24 formal justice, and certainty of expectations, 80 Forst, Rainer, 82–​84 France, 6 Frankfurt, Harry, 247n.44 freedom, 6–​7, 9, 245. See also basic liberties versus health, 276 liberal versus republican, 67–​69, 103–​7, 111,  113–​15 of speech, 90, 101, 228–​29

friendship distributing opportunities for, 156–​61 social (see community) Fuller, Lisa, 154n.47    gender equality, 20, 46–​47 norms, 106–​7, 117, 168–​69, 170, 177–​ 78, 181, 191–​86 genocide, 52–​53n.71 Germany, 6, 272n.36 Gheaus, Anca, 151–​56 global justice, 19, 37–​38n.41, 47, 52. See also international justice Gotha Programme of the German Social Democratic Party, 23–​24 group agency, 40–​42, 100 group inequalities, 46–​47, 48–​49, 290–​91,  293 gun ownership, 90    Habermas, Jürgen, 214n.25 habitus, 166n.6, 187n.41 health, 13, 17–​18, 27–​28, 34 and basic liberties, 275–​76 biostatistical model of, 261–​64 concept of, 261–​66 derivative approach to, 259–​60, 267, 270–​74,  279–​80 direct approach to, 259–​60, 267, 273–​77 and dominant cooperative frameworks,  263–​65 and fear (see fear) versus freedom, 276 inequalities,  266–​81 and racial discrimination, 271–​72 and social cohesion, 274–​75 social determinants of, 17–​18, 259–​60, 268–​69, 270–​77, 279–​80,  286 social gradient in, 270–​71, 273 value of, 276 healthcare, 17–​18, 104, 277–​78, 282–​87. See also expressive concept of justice and healthcare discrimination in (see discrimination: in healthcare) for non-​cooperators,  283–​85 hypocrisy, 98n.7, 164n.3, 171–​73   

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

318 Index ideal theory versus non-​ideal theory, 18–​21, 122–​23, 251–​52,  282–​83 impartial spectator, 140–​41 imperialism, 298 imperium. See dominium impersonal value, of social equality, 12, 138–​42, 148n.37, 153–​54,  289–​90 income and wealth convertibility into political infuence,  242–​47 and domination, 17, 242–​47 and norms of social status, 17, 191–​92, 247–​52,  298 indignation, 43–​44, 49–​50, 182 instrumental value of distributive equality for relational equality, 17, 117–​18, 232–​33, 235, 239, 240–​41,  242–​52 of equality of opportunity for relational equality,  252–​56 of political rights, 210–​12, 225–​27 of social equality, 153–​54, 274–​75 insurance against bad luck, 26n.12, 231n.71 health, 269n.19, 277–​78 intergenerational justice, 19, 298–​99 international justice, 19, 52, 294–​99    Japan, 275    Kant, Immanuel, 6n.1, 67n.25, 179n.24    La Rochefoucauld, François de, 98n.7 Laborde, Cécile, 73n.41, 80n.53 legitimacy, 65, 82n.58, 92n.88 Levellers, 6–​7,  132–​33 liberal egalitarianism, 4, 8–​10 Lippert-​Rasmussen, Kasper, 52–​53n.71, 145–​46, 147–​48, 149–​50, 155n.51 List, Christian, 41–​42 literacy, 262, 264–​65 Lovett, Frank, 58–​59, 71–​72, 79–​80, 81, 91n.86, 97n.3, 113n.36, 129–​30 luck egalitarianism, 10–​11, 17, 22, 23–​24, 25–​26, 33, 131n.7, 231n.71, 236–​37, 241. See also fairness and cash transfers, 24

and educational equality, 252n.54 extended to egalitarian relations, 51n.68,  151–​56 and health and healthcare, 258–​59, 260–​61,  284–​85

   Mafa, 97 manipulation, 97–​98, 107, 159–​60 marginalization, 23–​24, 194n.52, 250–​51,  255 Marmot, Michael, 270–​71, 275 marriage, 105–​7, 112–​13, 199, 206–​7 Marx, Karl, 7–​8, 23–​24 Mason, Andrew, 139–​40, 143, 147, 164n.4,  193–​95 McCammon, Christopher, 64–​65n.13, 71–​72,  91n.86 McMahon, Christopher, 84n.64 means of production, ownership of, 47, 75–​76,  246–​47 merit, 6n.1, 7n.7, See also desert metric, of justice, 15, 25–​26, 37, 49, 126, 127–​28, 129, 137–​38. See also currency of justice and purely recipient-​ oriented conceptions of justice relation-​sensitive, 15, 50–​51,  144–​61 microaggressions, 194n.52, 293 Mill, J.S., 209n.18 Miller, David, 134–​36, 138n.17, 138–​39, 140, 142, 239n.23 Montesquieu, Charles de Secondat, Baron de, 116n.42 moral equality, 25, 55–​56, 206, 237n.15, 290 and moral powers, 108–​9, 157–​61 moral powers, 8–​9, 59–​60, 101–​3. See also conception of the good and sense of justice individual diferences in, 108–​9, 158–​60 and moral equality (see moral equality: and moral powers) More, Tomas, 6–​7    Nagel, Tomas, 34 Nath, Rekha, 297n.20 natural inequalities, 33–​35, 277–​78 natural lottery, 33 negative duties, 32

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Index  319 neo-​republicanism, 3–​4, 11–​12, 65–​66, 78, 127–​28, 129–​30, 132–​33, 281n.57, 296–​97. See also domination and non-​domination neutrality, liberal, 8–​10, 113–​14, 118–​19, 206, 212–​13, 245–​46, 276 non-​compliance, 19. See also ideal theory non-​domination. See also domination baseline for, 73–​74, 78, 81, 83–​84 and control, 64–​65, 67, 74n.42, 81–​82, 86, 89, 124, 204 as currency of justice, 125–​26, 146–​47,  292 as democracy, 58–​59, 81–​87, 101, 131–​ 32, 201–​3, 206–​7, 220 and distributive inequality, 117–​18,  242–​47 as efective constraint, 79–​80 ethos of, 14, 116–​22 ‘eyeball test’, 65n.17, 66n.23, 89–​92, 97, 98, 101, 113n.36, 128, 130–​31, 183n.31, 207n.13 intensity of, 14, 65, 75–​76, 77–​78, 89–​ 90, 92–​93, 94–​95,  101–​15 liberal, substantive conception of, 3–​4, 14, 71–​78, 86–​87, 101, 124, 200–​2,  221 maximization of, 14, 94–​95, 110–​15 as paramount value, 65–​66, 68, 80, 129–​32,  201–​2 priority of, over other requirements of social justice, 110, 154, 191–​92, 208,  288–​89 range of, 63–​64, 71–​78, 103 and social cooperation, 56–​57, 67 non-​ideal theory. See ideal theory Nussbaum, Martha, 50–​51    O’Neill, Martin, 138–​39, 140 opportunity educational, 130, 252–​53, 293 equality of, 17, 252–​56 fair equality of, 33, 156–​57, 158, 218n.33, 255n.59, 279–​81 fair equality of political, 199–​200, 213–​ 14, 218, 222n.43, 224, 229 (see also political liberties, fair value of) hoarding, 187

loss through norms of social status (see status, norms of: and opportunities) Organisation for Economic Co-​operation and Development (OECD), 250–​51n.51 original position, 241, 264–​65 ostracism, 176n.18    party invitation example, 194–​95 path-​dependency, 86–​87,  264–​65 patriarchy, 66, 121n.52 perfectionism, 8–​10, 50–​51, 113–​14, 141–​ 42, 151–​52, 174, 211, 276. See also neutrality Pericles, 176n.18 person, liberal conception of, 9–​10, 11–​12, 14, 56–​60, 74–​75, 86, 92–93, 263, 267–​68, 289–​90, 294–​95. See also moral powers Pettit, Philip, 11–​12, 41–​42, 55–​59, 64–​65, 66n.24, 71–​72, 72–​73n.40, 74–​75, 80–​81n.55, 82, 84n.64, 94–​95, 97, 100n.9, 101–​2, 110–​15, 128, 129–​31, 183n.31, 203–​7, 281n.57, 289–​90 Pickett, Kate, 153n.46 Pildes, Richard, 43, 47–​48 Pluralist social egalitarianism, 8n.11, 15, 128–​29, 132–​43, 163, 185, 193–​95, 211, 239n.23 Pogge, Tomas, 22–​23, 26, 28–​29, 32–​33, 34–​35n.36, 35, 37–​38n.41, 43 political equality and basic structure of society, 83–​87 conditions for restrictions of, 213, 214–​17,  218–​24 and distributive inequality, 230–​31,  242–​47 and non-​domination, 16–​17, 212–​13 (see also non-​domination: as democracy) versus sufciency, 228–​30 presupposed in theories of social justice,  203–​7 priority of, over other requirements of social justice, 16–​17, 199–​200, 201–​7,  224 and representation, 216–​17 and sense of justice, 11–​212, 218–​19,  229

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

320 Index political liberties, fair value of, 16, 179–​ 80n.25, 199–​200, 217, 218 possessive individualism, 249–​50 poverty, relative, 249–​50 primary goods, 114n.38, 146–​47, 233n.4, 240n.24,  280–​81 trust and friendship as, 156–​61 prioritarianism, 26n.13, 241 property, protection of, 6, 7 property-​owning democracy, 247n.42 Proudhon, Pierre-​Joseph, 61n.5 public apology, 47–​48, 49, 52–​53n.71,  53–​54 pure procedural justice, 236–​37 purely recipient-​oriented conceptions of justice, 9–​10, 22–​23, 26, 30, 37, 53, 56, 94–​95, 125, 126, 129, 184–​85n.35, 201, 225–​26, 231. See also currency of justice and metric of justice    racial equality, 20. See also discrimination: racial racist social norms, 121, 298 Rawls, John, 8–​10, 22, 33, 34–​35, 57n.2, 80, 92n.88, 101–​2, 103n.13, 120n.49, 131n.7, 179–​80n.25, 180, 184n.34, 186n.37, 199–​200, 203–​4, 205–​6, 208n.16, 210n.21, 212n.24, 217–​18, 222, 235–​38, 239–​41, 247n.42, 248, 279–​81, 292, 295–​96n.19 reasonable contestability, 76–​77, 84, 87, 202, 212–​13, 241, 265–​66, 296–​97. See also claims of justice; prima facie versus fnal reasonable disagreement. See reasonable contestability reciprocity. See social cooperation and society, liberal conception of recognition, 127–​28, 147–​49, 164, 176–​77. See also respect: recognition and respect: versus esteem refective equilibrium, 18, 268–​69 representation, political. See political equality: and representation republicanism. See neo-​republicanism resentment, 43–​44, 47–​48, 121n.51 resources, equality of, 10–​11, 13, 25–​26, 51n.69, 230–​31, 269 relational, 50–​51,  156–​61

respect basic, 107–​9, 158–​60, 181, 221–​22 versus esteem, 164–​65, 181 opacity, 108–​9,  158–​60 recognition, 139–​40, 147, 152–​54 responsibility for choice, 10–​11, 17, 25–​26, 28n.17, See also luck egalitarianism and domination, 152, 155–​56 Richardson, Henry, 75n.43, 86n.67 Rousseau, Jean-​Jacques, 6–​7,  132–​33    Scanlon, T.M., 47n.55, 238–​39n.22, 255n.59 Schefer, Samuel, 117, 138n.19, 232–​33, 234, 235–​36, 238–​39, 240, 245–​46 self-​esteem, 49, 163, 174–​75, 256 concept of, 179 and envy, 183–​84, 248–​49 and health, 273 self-​respect, 15–​16, 49,  137–​38 and autonomy, 15–​16, 180, 182–​85, 186 concepts of, 179 and equality of opportunity, 255–​56 robustness of, 181–​85 role of (see self-​respect: and autonomy) social bases of, 178–​79, 181–​86 and social status, 15–​16, 178–​79, 185–​ 86, 188n.42 standards (see self-​esteem) standing, 179, 182–​83 Sen, Amartya, 50n.67 sense of justice, 8–​9, 16–​17, 59–​60, 113–​ 14, 208–​12, 218–​19, 229 sexual equality, 20, 39–​40, 72–​73n.40,  177–​78 Shapiro, Ian, 73n.41, 80n.53 shop example, 139–​40, 193 slavery, 181 Smith, Adam, 118n.44, 186–​87 social cooperation distribution of goods produced by, 17, 34–​35, 236–​42, 258–​59,  276–​77 and equality of power, 59–​63 exclusion from, 263–​66, 295–​96n.19 fair (see society, liberal conception of ) intergenerational,  298–​99 permanent inability to participate in,  283–​85 social justice concept of, 6, 66

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Index  321 liberal concept of, 8–​9, 18–​19, 60, 61–​62,  280–​81 social status, norms of, 15–​16, 127–​28, 133, 143 and de facto authority, 96–​97, 175–​77 and distributive inequality, 247–​52 and domination, 15–​16, 175–​78 versus habits, 169–​70 and herd behaviour, 167–​68, 175–​76 international,  297–​98 and opportunities, 15–​16, 186–​95 pluralization of, 185, 191, 249–​50 and sanctions, 168–​70 scope of, 170–​71, 290–​91, 293 and self-​respect (see self-​respect: and social status) socialism, 6, 7–​8, 61n.5 society, liberal conception of, 8–​9, 11–​12, 14, 34, 56–​60, 63, 74–​75, 118–​19, 189–​90, 206, 209, 236, 241–​42, 263–​ 64, 289–​90,  294–​96 solidarity, 7–​8, 144–​45,  151–​53 sports,  124–​25 Stark, Cynthia, 186n.37 Steiner, Hillel, 61n.5 sufciency, requirements of justice, 19–​20, 65n.17, 129–​30, 158–​59, 228–​30, 232, 234–​35, 241, 247n.44,  290–​91 swimming pool example, 71–​72, 77 Switzerland, 90n.82    Tawney, R.H., 7–​8, 53–​54 tax morality, 121–​22 tax return example, 72–​73 telic egalitarianism, 140n.25, 1​ 41n.28 Tucydides, 176n.18 Tomlin, Patrick, 140n.25, 141n.28 transnational coalitions, 296 True Levellers. See Diggers trust in citizens’ sense of justice, 218–​19, 224

distributing opportunities for relationships of, 156–​61 general social, 137–​38, 154n.47 and non-​domination,  112–​13

   United States, 90n.82, 91n.85, 167n.8, 247n.43,  277–​78 utilitarianism, 26n.13 V example (vitamins), 27–​30, 31, 33–​34, 36–​37, 38–​39, 41–​43, 46–​47, 56–​57, 67n.26, 69, 80n.51, 165n.5, 188 Value pluralism, 29–​31, 63n.8, 225–​26, 258–​59, 284–​85. See also pluralist social egalitarianism von Platz, Jeppe, 102 vulnerability to domination, 91–​92, 103–​4, 107,  113–​15 to health problems, 34, 259–​60, 267–​68,  277–​78    Walras, Léon, 10 Walzer, Michael, 134–​35n.9, 244n.36 wealth. See income and wealth Weinstock, Daniel, 263n.7 welfare capitalism, 247n.42 welfare economics, 9–​10 welfare, equality of opportunity for, 13,  25–​26 welfare state, 87n.69, 104n.14, 114–​15, 257n.62, 293 Wilkinson, Richard, 153n.46, 270–​71,  275 Williams, Andrew, 122n.53 Williams, Bernard, 256n.61 Wilson, James Lindley, 211n.23 Winstanley, Gerrard, 6–​7 Wolf, Jonathan, 138n.17, 140n.25 World Health Organization (WHO), 261 Young, Iris Marion, 23–​24

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.

Copyright © 2021. Oxford University Press USA - OSO. All rights reserved.