The Oxford Handbook of Distributive Justice 9780199645121, 0199645124

Citation preview

T h e ox f o r d h a n d b o o k o f

DI ST R I BU T I V E J U ST IC E

The Oxford Handbook of

DISTRIBUTIVE JUSTICE Edited by

SERENA OLSARETTI

1

3 Great Clarendon Street, Oxford, ox2 6dp, United Kingdom 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 in certain other countries © in this volume the several contributors 2018 The moral rights of the authors‌have been asserted First Edition published in 2018 Impression: 1 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 licence or under terms agreed with the appropriate reprographics rights organization. Enquiries 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 Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2017959074 ISBN 978–​0–​19–​964512–​1 Printed and bound by CPI Group (UK) Ltd, Croydon, cr0 4yy Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Contents

Contributors  Introduction: The Idea of Distributive Justice  Serena Olsaretti

ix 1

PA RT I   C ON T E M P OR A RY A P P ROAC H E S 1. Rawls on Distributive Justice and the Difference Principle  Samuel Freeman

13

2. Dworkin and Luck Egalitarianism: A Comparison  Richard Arneson

41

3. Equality versus Priority  Michael Otsuka and Alex Voorhoeve

65

4. Sufficiency and Needs-​Based Approaches  Gillian Brock

86

5. The Capability Approach  Ingrid Robeyns

109

6. Libertarianism, Left and Right  Joseph Mazor and Peter Vallentyne

129

7. Desert-​Based Justice  Jeffrey Moriarty

152

PA RT I I   DI ST R I BU T I V E J U S T IC E A N D OT H E R V I RT U E S 8. Retributive Justice  Larry Alexander

177

vi   Contents

9. The Good Society  Steven Wall

195

10. The Ethics of Care  Virginia Held

213

11. The Theory and Politics of Recognition  Colin Bird

235

PA RT I I I   F O U N DAT IONA L DI SP U T E S 12. Distributive Justice and Human Nature  Paula Casal

259

13. Political and Distributive Justice  Chad van Schoelandt and Gerald Gaus

283

14. Consequentialism, Deontology, Contractualism, and Equality  Jonathan Quong

306

15. Ideal Theory  David Schmidtz

327

16. Constructivism, Intuitionism, and Ecumenism  Aaron James

346

17. Conceptual Analysis and Distributive Justice  Matthew H. Kramer

367

PA RT I V   T H E A P P L IC AT ION OF DI ST R I BU T I V E J U S T IC E 18. Gender  Anca Gheaus

389

19. The Family  Colin Macleod

415

20. Education  Matthew Clayton

438

21. Health  Shlomi Segall

460

Contents   vii

22. Discrimination  Kasper Lippert-​Rasmussen

479

23. Race  Bernard R. Boxill

498

24. Work  Paul Gomberg

513

25. Exploitation  Benjamin Ferguson and Hillel Steiner

533

26. Public Goods  David Miller and Isaac Taylor

556

27. Cultural and Religious Minorities  Peter Jones

576

28. Language  Alan Patten

597

29. Justice across Borders  Michael Blake

619

30. Migration  Sarah Fine

640

31. Climate Change  Simon Caney

664

32. Future Generations  Rahul Kumar

689

Index 

711

Contributors

Larry Alexander is the Warren Distinguished Professor at the University of San Diego School of Law. He is the author or editor of ten books and 235 published articles, primarily dealing with legal and moral theory. He is a co-​editor of the journal Legal Theory and is on the editorial boards of Ethics, Law & Philosophy, and Criminal Law & Philosophy. He is also an Executive Director of the Institute for Law & Philosophy at his university. Richard Arneson works mainly in moral and political philosophy. His recent research includes essays on the justification of democracy, egalitarian theories of social justice, and act consequentialism and its critics. He is Distinguished Professor (Professor above Scale) in the Department of Philosophy at the University of California, San Diego, where he has taught since 1973. He holds the Valtz Family Chair in Philosophy at UC San Diago. Colin Bird is Associate Professor of Politics and Director, Program in Political Philosophy, Policy and Law, University of Virginia. He is the author of The Myth of Liberal Individualism (CUP 1999), An Introduction to Political Philosophy (CUP 2006), and articles on a wide variety of topics, including state neutrality, the scope of ‘public reason’, propaganda, democratic theory, toleration, the role of religion in public life, respect, and self-​respect. He is currently completing a book on the role of arguments about human dignity in political theory, tentatively entitled After Respect: the Use and Abuse of Dignitarian Humanism in Political Argument. Michael Blake is Professor of Philosophy and Public Affairs, and former Director of the Program on Values in Society, at the University of Washington. He writes on international distributive justice, the ethical foundations of foreign policy, and on the ethics of migration. He is the author of Justice and Foreign Policy (OUP 2013) and, with Gillian Brock, of Debating Brain Drain: May Governments Restrict Emigration? (OUP 2015). He is currently finishing a book on the relationship between justice, mercy, and migration. Bernard R. Boxill is Professor of Philosophy Emeritus at the University of North Carolina at Chapel Hill. His essays on self-​respect, protest, race, justice, reparations, and affirmative action have appeared in leading journals and collections. In 2017 his book Blacks and Social Justice (Rowman & Littlefield 1992) was awarded the Lippincott prize. Gillian Brock is Professor of Philosophy at the University of Auckland in New Zealand and currently also a Fellow at the Safra Center for Ethics, Harvard University. Her most recent work in Philosophy has been on global justice and related fields. Her

x   Contributors books include Debating Brain Drain (OUP 2015 with Michael Blake), Cosmopolitanism versus Non-​Cosmopolitanism (OUP 2013), Global Heath and Global Health Ethics (CUP 2011), Global Justice: A Cosmopolitan Account (OUP 2009), and Necessary Goods: Our Responsibilities to Meet Others’ Needs (Rowman & Littlefield 1998). She also has many interdisciplinary interests, some of which lie at the intersection of philosophy and public policy. For instance, during 2013–​2015 she took up a fellowship from the Edmond J. Safra Center for Ethics at Harvard University to research institutional corruption. Simon Caney is Professor of Political Theory at the University of Warwick. He works on issues in contemporary political philosophy, and focuses in particular on issues of environmental, global, and intergenerational justice. He is completing two books—​Global Justice and Climate Change (with Derek Bell) and On Cosmopolitanism—​both of which are under contract with Oxford University Press. He is the author of Justice Beyond Borders (OUP 2005). Paula Casal is ICREA Professor at Pompeu Fabra University, Barcelona, associate edi­ tor of Politics, Philosophy and Economics, co-​editor of Law, Ethics and Philosophy, and President of ASAP-​Spain and of The Great Ape Project, Spain. She has published in journals such as Ethics, Economics and Philosophy, Hypatia, Journal of Medical Ethics, Journal of Moral Philosophy, Journal of Political Philosophy, Political Studies, and Utilitas, on social and global justice, multiculturalism, procreation, animals, and enhancement.  Matthew Clayton is Professor of Political Theory at the University of Warwick. He is the author of Justice and Legitimacy in Upbringing (OUP 2006), and has co-​edited The Ideal of Equality (Palgrave Macmillan 2000) and Social Justice (Blackwell 2004). Benjamin Ferguson is Assistant Professor of Ethics at Vrije Universiteit, Amsterdam, and was formerly a lecturer in Political Philosophy at Universität Bayreuth. His research focuses on ethical issues raised by fraud, exploitation, and colonialism. Sarah Fine is Senior Lecturer in Philosophy at King’s College London. She specializes in issues relating to migration and citizenship. Her forthcoming book, Immigration and the Right to Exclude (OUP), sets out to challenge the idea that the state has a moral right to exclude would-​be immigrants. She has co-​edited a new collection of essays, Migration in Political Theory: The Ethics of Movement and Membership (OUP 2016) with Lea Ypi. Her publications include ‘Freedom of Association is not the Answer’ in Ethics. Samuel Freeman is Avalon Professor of the Humanities and Professor of Philosophy and of Law at the University of Pennsylvania. He is the author of Liberalism and Distributive Justice  (OUP 2018),  Justice and the Social Contract (OUP 2006)  and of Rawls (Routledge 2007). He edited John Rawls’s Collected Papers (Harvard University Press 1999) and his Lectures in the History of Political Philosophy (2008). Freeman also edited the Cambridge Companion to Rawls (CUP 2003) and was co-​editor of Reasons and Recognition: Essays on the Philosophy of T. M. Scanlon (OUP 2011). Gerald Gaus is the James E.  Rogers Professor of Philosophy at the University of Arizona, where he directs the program in Philosophy, Politics, Economics and Law.

Contributors   xi He is the author of a number of books, including The Order of Public Reason (CUP 2011), Justificatory Liberalism (OUP 1996) and Value and Justification (CUP 1990). His most recent book is The Tyranny of the Ideal, published by Princeton University Press 2016). Anca Gheaus is Ramon y Cajal researcer at teh Universitat Pompeu Fabra in Barcelona. She is interested in the relevance of caring relationships for theories of distributive justice, and has published work on gender justice, parental rights and duties, the value of the family, and methodological issues in political philosophy. She edited a special issue of the Journal of Applied Philosophy on the nature and value of childhood, and is a coeditor of The Routledge Handbook of the Philosophy of Childhood and Children. Paul Gomberg taught philosophy at the University of Missouri–​St Louis 1971–​1978 and Chicago State University 1985–​2014. Final drafts of his contributed chapter were completed while a Visiting Scholar in the Department of Philosophy at the University of Illinois at Chicago. He is currently Research Associate in the Department of Philosophy at the University of California at Davis working on a book entitled American Racial Injustice: How It Arose, Why It Persists, How It may End. Virginia Held is Professor of Philosophy emerita at the City University of New York Graduate Center. Among her books are How Terrorism is Wrong:  Morality and Political Violence (OUP 2008), The Ethics of Care: Personal, Political, and Global (OUP 2006),  Feminist Morality:  Transforming Culture, Society, and Politics (Chicago 1993), Rights and Goods:  Justifying Social Action (Free Press 1984), and The Public Interest and Individual Interests (Basic Books 1970). Her edited collections include Justice and Care: Essential Readings in Feminist Ethics (Westview Press 1995), and Property, Profits, and Economic Justice (Wadsworth 1980). In 2001–​2002 she was President of the Eastern Division of the American Philosophical Association. Aaron James is Professor of Philosophy at the University of California, Irvine. He is author of Fairness in Practice: A Social Contract for a Global Economy (OUP 2012) and numerous articles on meta-​ethics, moral theory, and political philosophy. He has been an ACLS Burkhardt Fellow, a fellow at the Center for Advanced Study in the Behavioral Sciences, Stanford University, and Visiting Professor of Philosophy at  New York University. Peter Jones is Emeritus Professor of Political Philosophy at Newcastle University, UK. Much of his recent work has focused on issues associated with differences of belief, culture, and value, including those of toleration, accommodation, compromise, recognition, freedom of expression, and discrimination law. He has also written on various aspects of rights, including human rights, group rights, and welfare rights, and on democracy, self-​determination, and international justice. Matthew H. Kramer is Professor of Legal and Political Philosophy at the University of Cambridge and a Fellow of Churchill College, Cambridge. He is a Fellow of the British Academy, and is the Director of the Cambridge Forum for Legal and Political

xii   Contributors Philosophy. He is the author of sixteen books and the co-​editor of four further books. His most recently published book is H. L. A. Hart: The Nature of Law (Polity Press 2018). Rahul Kumar is Associate Professor of Philosophy at Queen’s University, Ontario. He is the author of several articles on contractualist moral theory. His current research concerns non-​consequentialism and intergenerational obligations. Kasper Lippert-​Rasmussen is Professor of Political Theory at University of Aarhus, Denmark, and Professor II in Philosophy at University of Tromsø, Norway. He works primarily in the fields of political and moral philosophy and has published papers in journals including Journal of Political Philosophy; Ethics, Philosophy & Public Affairs; Philosophical Studies; and Economics and Philosophy. He is the author of Luck Egalitarianism (Bloomsbury 2015) and Born Free and Equal? (OUP 2013). Presently, he is working on a book on affirmative action. He is associate editor of Ethics. Colin Macleod is Professor of Philosophy and Law at the University of Victoria. His research focuses on issues in contemporary moral, political, and legal theory, with a special focus on distributive justice and equality; children, families, and justice; and democratic ethics. He is the author of Liberalism, Justice, and Markets (OUP 1998); co-​author with Ben Justice of Have a Little Faith: Religion, Democracy, and the American Public School (University of Chicago Press 2016), and co-editor with David Archard of The Moral and Political Status of Children (OUP 2002). Joseph Mazor is a visiting academic at the Centre for Philosophy of Natural and Social Science at the London School of Economics. He works on issues of distributive justice, environmental ethics, philosophy of welfare economics, and democratic theory. David Miller is Professor of Political Theory at the University of Oxford and a Senior Research Fellow at Nuffield College. His research interests include social justice, nationality, and global justice. His most recent books are Justice for Earthlings (CUP 2013) and Strangers in our Midst: The Political Philosophy of Immigration  (Harvard University Press 2016). Jeffrey Moriarty is Associate Professor and Chair of Philosophy at Bentley University. His research interests lie in political philosophy and business ethics, and at the intersection of these fields. He is especially interested in questions of just distribution in state and organizational contexts. Publications to feature his work include Business Ethics Quarterly, Journal of Business Ethics, Noûs, Philosophical Studies, and Social Theory and Practice. Serena Olsaretti is ICREA Research Professor at the Universitat Pompeu Fabra, Barcelona. Her current research is mostly focused on family justice. She is the author of Liberty, Desert and the Market (CUP 2004), and the editor of Preferences and Well-​Being (CUP 2006) and Desert and Justice (OUP 2003). She has published articles in various journals, including Philosophy and Public Affairs, the Journal of Political Philosophy, Analysis, Economics & Philosophy and Utilitas.

Contributors   xiii Michael Otsuka is a Professor in the Department of Philosophy, Logic and Scientific Method at the London School of Economics. In addition to prioritarianism and egalitarianism, his current research interests encompass the morality of imposing risks, harming, and saving from harm; the benefits of risk-​pooling and other forms of cooperation; and the virtues of left-​libertarianism versus social democracy. His articles have appeared in Philosophy and Public Affairs, Ethics, and the Journal of Political Philosophy, among other places. Alan Patten teaches political theory at Princeton University. He is the author of Equal Recognition:  The Moral Foundations of Minority Rights  (Princeton University Press 2014) and of Hegel’s Idea of Freedom (OUP 1999). From 2010 to 2017 he served as editor of the journal Philosophy & Public Affairs. Jonathan Quong is Associate Professor of Philosophy at the University of Southern California. He taught previously at the University of Manchester, and has held visiting positions at the Australian National University, Princeton University, and Tulane University. His areas of research are political and moral philosophy. He is the author of Liberalism without Perfection (OUP 2011). Ingrid Robeyns holds the Chair in Ethics of Institutions at Utrecht University. She has written extensively on the capability approach and various problems of social and distributive justice. David Schmidtz is Kendrick Professor of Philosophy, Eller Chair of Service-Dominant Logic, and Department Head of Political Economy and Moral Science at the University of Arizona. He is also Editor-in-Chief of Social Philosophy & Policy. His book with Harry Brighouse on Markets in Education is forthcoming from Oxford University Press. Chad van Schoelandt is an Assistant Professor at Tulane University. He works primarily on social and political philosophy, particularly related to social norms and the public reason tradition. His recent works appear in Philosophical Studies, Philosophical Quarterly, and Law and Philosophy. Shlomi Segall is a Professor of Political Philosophy and the Chair of the Program in Politics, Philosopy, and Economics (PPE) at the Hebrew University of Jerusalem. He is the author of Health, Luck, and Justice (Princeton University Press 2010), Equality and Opportunity (OUP 2013), and Why Inequality Matters (CUP 2016). Hillel Steiner is Emeritus Professor of Political Philosophy at the University of Manchester and a Fellow of the British Academy. He is the author of An Essay on Rights (Blackwell 1994) and co-author (with Matthew Kramer and Nigel Simmonds) of A Debate over Rights: Philosophical Enquiries (OUP 2000). His current research concerns the concept of ‘the just price’, and the application of libertarian principles to global and genetic inequalities. Isaac Taylor is a Scholar in Residence at the Center for Western Civilization, Thought and Policy, University of Colorado Boulder. His research interests focus on questions

xiv   Contributors surrounding global justice (especially relating to the provision of global public goods) and security policy. He is currently completing a book manuscript on the ethics of counterterrorism. Peter Vallentyne is Florence G.  Kline Professor of Philosophy at the University of Missouri. He writes on issues of liberty and equality in the theory of justice (and left libertarianism in particular) and, more recently on enforcement rights (rights to protect primary rights). He is an associate editor of the Journal of the American Philosophical Association and of Social Choice and Welfare. Alex Voorhoeve is a Professor in the Department of Philosophy, Logic and Scientific Method at the London School of Economics. He works on distributive justice, healthcare justice, and rational choice theory. His articles have appeared in Philosophy and Public Affairs, Ethics, and Economics and Philosophy, among other places. Steven Wall is Professor of Philosophy at the University of Arizona, where he is a member of both the Center for the Philosophy of Freedom and the Philosophy, Politics, Economics and Law Program. He is the author of Liberalism, Perfectionism and Restraint (CUP 1998), and the editor of The Cambridge Companion to Liberalism. He is a co-​editor of Oxford Studies in Political Philosophy.

I n t rodu ction The Idea of Distributive Justice Serena Olsaretti

As any student of contemporary political philosophy can attest, theorizing about distributive justice has played a considerably large role in the discipline over the past half-century. Distributive justice has concerned political philosophers of other historical periods,1 but nobody can deny—​indeed, this is by now a well-​worn refrain—​that since the publication of John Rawls’s A Theory of Justice in 1971, debates about how we should arrange our social and economic institutions so as to distribute fairly the benefits and burdens of social cooperation have proliferated. Nor can anyone deny that these debates address some of the deepest and most pressing questions in political philosophy. Together with the question of the legitimacy of the state or of political authority, that of distributive justice lies at the heart of our attempt to identify criteria for evaluating and justifying to each other our shared political practices and institutions.2 This volume is a reflection of the wealth of issues that contemporary debates about distributive justice have been treating, and continue to treat. The chapters it comprises provide an overview of the state of those debates and identify the trajectory in which they are, or—​according to the philosophers who have written these chapters—​ought to be, moving. Before providing an outline of what the volume includes, this introduction offers some remarks on the idea of distributive justice: how do theorists of justice, including the ones who contribute to this volume, conceive of distributive justice, as opposed to other types of justice, and as opposed to other, non-​justice-​based, demands? Like the idea of justice simpliciter, with which it is often used interchangeably, the idea of distributive justice has been taken to refer to different things: theorists of justice have adopted different views, mostly without any explicit acknowledgement or defence of 1  For two recent treatments of the history of the ideas of justice and of distributive justice, see Fleischaker 2005 and Raphael 2003. See also Jackson 2005. 2  For a discussion of the difference and relation between the questions of legitimacy and of justice, see Christiano 2008 and Pettit 2012.

2   Serena Olsaretti them, about what characterizes and delimits the demands of justice as opposed to other moral demands (for example, the demands of legitimacy, community, efficiency, or stability, to mention a few central ones). They have also adopted different views of what characterizes distributive justice as opposed to other types of justice. Some, for example, have assumed or claimed that justice, as opposed to humanitarian concerns or charity, concerns our perfect duties, that is, duties owed to specific individuals that leave no room for discretion on the part of duty-​holders in deciding how to discharge them (see Buchanan 1987). Some have held that justice regards the negative duties we have to not harm others, as opposed to any duties to assist or aid others (Campbell 1974). Alternatively, or additionally, some have assumed that what characterizes duties of justice is that they are enforceable, that is, they are duties which a legitimate authority may use coercion to ensure are fulfilled (Nozick 1974). As far as distributive justice is concerned, some have assumed that what distinguishes it from other types of justice is that it is justice in the distribution of material or economic advantages only, or that it only concerns the allocation, as opposed to the production, of given goods; others have instead equated the idea of distributive justice with that of social justice, and used it to refer to all the principles regulating the balancing of individuals’ claims to all of the possible benefits of social cooperation (Rawls 1999; Bedau 1978). These different usages of the ideas of justice and of distributive justice reflect different views of what characterizes these social values and distinguishes them from other moral demands, and in what follows I identify some key dimensions along which such views vary. As a preliminary to that discussion, it is helpful to clarify how the variation I am drawing attention to here relates to the more familiar variation among different interpretations of the demands of justice, or between competing principles of justice. Theorists of justice widely endorse shared, abstract concepts of justice and of distributive justice: they agree that justice consists in giving each person his or her due, or treating like cases alike; and that distributive justice is justice in the distribution of benefits and burdens to individuals, or consists in the balancing of the competing claims persons make on the benefits that are up for distribution.3 But, as is often observed, theorists of justice disagree about how to interpret these abstract ideas and, accordingly, formulate different conceptions of justice and of distributive justice.4 Crucially, these conceptions reflect different understandings of what considerations are relevant for treating like cases alike and different cases unalike, or for determining a balance of claims. For example, is people’s deservingness, or their neediness, relevant for giving individuals their due? Is treating people equally necessary for settling fairly their competing claims? These questions are the staple of many debates among theorists of justice. 3  I take these definitions of justice, and of distributive justice, respectively, to be roughly equivalent. These familiar definitions of justice and distributive justice are widely endorsed. See, for example, Hart 1961, Rawls 1999, Cohen 2008. 4  See Hart 1961; Rawls 1999; Dworkin 1986.

Introduction: The Idea of Distributive Justice    3 Besides disagreeing on what justice demands, theorists of justice also disagree about which other features, if any, of justice and distributive justice, apart from those that characterize the abstract ideas captured by the shared concepts, are essential to understanding these ideas5 and to demarcating them from other moral demands.6 This variation is what interests us here: what do theorists of justice mean by saying, and what follows from their saying, that a particular principle they defend as the most defensible interpretation of the concept of justice (for example, desert, need, or equality) is a principle of distributive justice, rather than, say, one of corrective justice or a humanitarian principle?7 In identifying the different views theorists adopt of what characterizes distributive justice, it is helpful to note that there are four main and inter-​related dimensions along which they tend to vary, which concern, respectively, (i) the preconditions; (ii) the subject; (iii) the object; and (iv) the normative significance of distributive justice.8 (i) The preconditions of distributive justice are the conditions that must obtain for considerations of distributive justice to be pertinent at all. David Hume’s ‘circumstances of justice’ are a case in point:  most contemporary theorists agree with Hume that questions of distributive justice arise only when there is relative material scarcity (neither great abundance nor extreme scarcity in the resources people need and want). Under these conditions, there is both an identity and a conflict of interests among individuals that make the quest for principles needed to resolve conflicting claims equitably both necessary and possible. Understanding distributive justice as involving a balancing of competing claims over what is distributable, as was suggested earlier, reflects acceptance of the view that claims of distributive justice only arise if the circumstances of justice obtain. This point is widely shared among theorists of distributive justice, but alternatively, or additionally, some think that the existence of social cooperation is necessary for the demands of distributive justice to arise, in that it is only in the context of relations of reciprocity that individuals can assert claims to sharing fairly the goods that social

5  It is possible to think that some feature typically accompanies justice but that it does not contribute to making justice what it is (see Cohen 2008: 148). 6  As Ronald Dworkin remarks, at the level of abstraction at which we identify a concept ‘agreement collects around discrete ideas that are uncontroversially employed in all interpretations’, while at the level of conceptions, ‘the controversy latent in this abstraction is identified and taken up’ (Dworkin 1986: 71). 7   It is possible to characterize the idea of distributive justice by reference to a specific interpretation of its demands, and this is sometimes done when it is suggested that distributive justice is justice in accordance with desert, or a matter of negative rights (see Campbell 1974). This way of characterizing distributive justice is unhelpful, however, as it makes it impossible to even raise the question we are focusing on here—​that is, what does it mean to say, and what follows from saying, that desert, or respect of negative rights, are principles of justice, specifically, as opposed to being a non-​justice-​based moral demand? For another objection to identifying justice with the substantive principle of desert, see Vallentyne 2015. 8  Some theorists present some of the claims I cite below as claims about justice, rather than distributive justice in particular, but I take it that they think this is what characterizes distributive justice, specifically—​that is, distributive justice as opposed to other types of justice (such as criminal justice), as well as to other non-​justice-based moral demands (such as the demands of beneficence).

4   Serena Olsaretti cooperation makes available (Rawls 1971). A different view holds that considerations of distributive justice are only pertinent where there are shared institutions through which we exercise coercion over each other, or which speak in our name (Nagel 2005), as only these practices trigger a demand for justification which can only be met by making those practices just. We could furthermore believe that, within the context of shared institutions, only disadvantage that is intentionally and avoidably caused by those institutions, rather than the result of natural causes, is unjust (Nagel 1997). (ii) Discussions of distributive justice also conceive of what characterizes it differently in line with what they take the primary subject of distributive justice to be (see Bedau 1978): is it individuals’ acts that are primarily just and unjust, all social practices, or only certain institutions? Famously, Rawls’s view of justice is institutionalist, in the sense that for Rawls the principles of justice are principles that regulate primarily the basic structure of society. Drawing on Rawls, various theorists now assume that what characterizes the demands of justice is precisely the fact that they are demands which (certain) social institutions, specifically, must satisfy (see, for example, Scanlon 1998; Tan 2004). Demands of justice, on this view, identify a subset of the moral considerations that concern what we owe to one another, where what helps demarcate them is the fact that they are to regulate a particular domain.9 Alternatively, we could think of the demands of just­ice as applying primarily to distributions of whatever burdens and benefits are thought relevant; on this view, legal institutions, social norms, and individual acts can all be assessed as just or unjust, depending on whether they help to bring about, or disrupt, just distributions (Cohen 2008). (iii) Thirdly, different theorists of justice take different views of the object of distributive justice. On a doubly narrow interpretation of the object of distributive justice, to focus on distributive justice is to focus on the justice of the mechanisms and procedures that only allocate a given amount of goods, and only a subclass of distributable goods, namely, distributable economic goods like income and wealth. A wider interpretation of the idea of distributive justice adopts a more generous view of the goods whose distribution raises concerns of justice, and/​or considers productive mechanisms, as well as allocative ones, as subject to the demands of justice. For example, we could think that a theory of distributive justice bears on how distributable goods other than economic ones are distributed; or, more broadly still, that it bears on how individuals fare with regard to any aspects of advantage that we think morally relevant (for example, how happy individuals are, or whether they enjoy recognition). These types of advantage may not be themselves distributable, but it is true both that individuals can enjoy them, or have access to them, to different degrees, and that we can affect the degree to which persons can access or enjoy them, and these two facts make it intelligible and sensible to apply justice considerations to the distribution of these types 9  The view on which this is by itself sufficient to demarcate distributive justice, so that justice considerations are any and all considerations that bear on how social institutions should be arranged, is implausible, as Vallentyne 2015 points out. Like Vallentyne, I assume that these usages of the idea of justice (and a fortiori, of distributive justice) are unhelpfully loose.

Introduction: The Idea of Distributive Justice    5 of advantage. A wide interpretation of distributive justice can also take as its concern the productive mechanisms that affect which and what amount of distributable goods there are in the first place, rather than focusing merely on the mechanisms for the allocation of pre-​given goods. The idea of distributive justice in this wider sense, which Rawls explicitly endorses (Rawls 1971: 88), is often associated with that of social justice. Although most theorists of justice are silent on whether they conceive of the object of distributive justice as narrow or wide in this sense, their principles often have implications for what productive processes, as well as narrowly distributive mechanisms, should be in place. (A simple example is a principle enjoining maximal equal opportunities for welfare as a demand of justice: different productive arrangements, as well as allocative schemes, affect how great the range of welfare opportunities people enjoy is, and realizing the demands of justice thus understood therefore requires setting up some, rather than other, productive schemes.) (iv) Finally, and crucially, theorists of distributive justice conceive of it differently depending on what view, sometimes explicitly but mostly implicitly, they take of the normative significance of distributive justice claims. On most views, distributive justice considerations offer us very weighty reasons for action. Even more strongly, injustice is on most views a decisive reason for altering arrangements: as Rawls famously stated, ‘laws and institutions no matter how efficient and well-​arranged must be reformed or abolished if they are unjust’ (Rawls 1971: 3). On some views, as was mentioned earlier, justice considerations offer us reasons for action that leave no room for discretion in deciding what exactly we must do for others, and on most views, moreover, they offer us reasons for action that are enforceable, that is, that can justifiably be backed up by force by a designated legitimate authority. On other views, however, justice-​based reasons are not essentially action-​guiding, and identifying an injustice is thought to be primarily an evaluative task, one that is carried out independently of whether there are reasons to do something about the injustice and indeed, of whether it is possible for anyone to remedy it (Cohen 2008). Reasons of justice, here, track primarily what we have reason to regret, or find disvaluable. Detecting variations in the usage of the concept of justice along the lines just sketched is helpful for two main reasons. First, because theorists of justice have used importantly different concepts of distributive justice, and have generally done so implicitly, without clearly stating what they mean when they affirm or deny that something is a demand of distributive just­ ice, they—​and their critics—​have sometimes argued at cross-​purposes. Some defenders of the politics of identity, for example, who reject the ‘distributive paradigm’ (Young 1990), rely on an understanding of distributive justice according to which it is concerned only with the distribution of material resources. This is a narrower understanding than one that is held by many theorists of distributive justice. Similarly, it has been argued, the anti-​constructivist critique of Rawls developed by G. A. Cohen partly relies on Cohen’s using a different concept of justice from Rawls’s (Willams 2008). For Rawls, principles of justice are action-​guiding, and more specifically, they are principles for facilitating citizens’ cooperative interaction with one another, so

6   Serena Olsaretti they must be ones which citizens can understand, and which citizens can verify are being followed by others. For Cohen, by contrast, justice considerations need not play this particular social role. (For another diagnosis of Cohen’s critique of Rawls as premised on the use of different concepts of justice, see Anderson 2012.) Noticing that philosophers have used different concepts of distributive justice reveals that some disagreements are more apparent than real. Bringing the diversity of uses of the concept of justice and of the substantive commitments that underlie that diversity to the fore is also important for another reason, this one directly relevant from the point of view of introducing this volume. Once we notice that the idea of distributive justice can be and has been used in many ways, we get a clearer picture of the vast range of questions that can be tackled by debates on distributive justice. It becomes apparent, for example, that a concern with distributive justice can inform our stance on what productive arrangements a just society should host, as much as what stance we take on the allocation of whatever a just society produces; or that theorists of justice can be as concerned with individuals’ unequal enjoyment of recognition as they are with their unequal access to resources. As Michael Walzer points out: ‘[t]‌he idea of distributive justice has as much to do with being and doing as with having, as much to do with production as with consumption, as much to do with identity and status as with land, capital, or personal possessions’ (Walzer 1983: 3). In line with Walzer´s remarks, the choice of topics for this volume reflects a generous understanding of the purview of distributive justice. The volume opens, in Part I, with discussions of the main competing interpretations of the demands of distributive justice as advanced in contemporary debates—​what I earlier referred to as ‘the staple questions’ for theorists of justice. While all contemporary theories of justice are premised on the assumptions that all persons have equal moral status and ought to be treated as equals, thereby sharing an ‘egalitarian plateau’ (Kymlicka 1990: 5), they diverge substantially over exactly what treating people as equals requires. They diverge, centrally, in line with what pattern in the distribution of advantage the demands of justice are supposed to help create, and in line with what the currency of justice is, that is, what aspect of people’s situations should command our attention when assessing whether or not there are distributive injustices among them. As far as the pattern of distributive justice is concerned, some theorists favour redistributive policies with a view to mitigating or eliminating the gap between the better off and the less well off (whom we might call distributive egalitarians, or egalitarians simpliciter, discussed in Chapters 2 and 3), while others hold that these must only ensure that the badly off have enough, or have their basic needs met (Chapter 4), and yet others that improvements in the situation of the worse off are given priority (Chapter 3). Yet other theorists view redistributive policies as required by justice insofar as they help ensure that people are as well or as badly off as they deserve to be (Chapter 7), while some reject any redistributive policies as unjust because the only rights people have are rights to use, control, and exchange at full tax immunities justly acquired private property rights (these are right libertarian views, discussed in Chapter 6).

Introduction: The Idea of Distributive Justice    7 Theories of justice also take a stand on what the currency of distributive justice is: according to some, what resources people have access to is what is relevant for justice (see Chapters 1 and 2), while on other views what matters is the opportunities for welfare people have (Chapter 2), or the effective freedom to achieve valuable states of being and doing (or ‘capabilities’ to function, such as, for example, the capability of being well nourished, or escaping morbidity; Chapter 5 discusses the capability approach in general, in its egalitarian and non-​egalitarian variants). These two orthogonally related sets of questions, about the pattern and the currency of distributive justice, have structured many debates among defenders of the main competing contemporary conceptions of justice, and broadly guide the division of topics among the chapters in Part I. Since John Rawls and Ronald Dworkin have offered the two best worked-​out contemporary egalitarian theories (both of which take resources to be relevant for justice), the two opening chapters focus on discussing those and related approaches (those of ‘luck egalitarians’) that have taken their inspiration from one or the other of these theories. Other important related questions the chapters of Part I address include the role of personal responsibility for justice, the relevance of individuals’ subjective assessment of their situation as compared to others’ for determining whether they are unjustly advantaged or disadvantaged, and the possibility of reconciling egalitarian commitments with the endorsement of robust rights of private ownership over one’s body and mind (and hence, a strong presumption against paternalism), which animates the left libertarian project. Parts II and III treat issues, some substantive and some methodological, that are less often treated in the context of debating distributive justice. As was mentioned earlier, all philosophers acknowledge that distributive justice, however important, is not the only social virtue that we have reasons to endorse, so questions arise about how its demands relate to those of other central values which a society should promote or protect. The chapters of Part II address these questions. They ask how the requirements of justice in punishment, which are often thought to require conferring punishment in accordance with desert, relate to those of justice in the goods of social cooperation (Chapter 8); whether and how the promotion of impersonal values such as excellence in the arts or sciences or environmental goods, while not itself demanded by justice, ought to be pursued in a good society, and whether this would be in tension with justice (Chapter 9); whether a concern with the value of care and caring relationships ought to be endorsed alongside justice (Chapter 10); and whether a just society is also one which heeds individuals’ and groups’ multifaceted demands for recognition, including recognition of people’s distinguishing identities (and hence their differences), social acceptance of their authentic selves, and appreciation of their worth (Chapter 11). As emerges from some of the discussions, and in line with what was said in the opening pages of this chapter, it is possible to conceive of the idea of distributive justice more or less encompassingly, and depending on how wide or narrow an interpretation of justice we embrace, our view of its place relative to other social virtues will differ. (People’s demands for recognition, for example, can themselves be thought to be something which people have a justice-​based claim to having met; by contrast, on a

8   Serena Olsaretti narrower understanding of the idea of justice, the latter is claimed to be both different from, and potentially in tension with, the values championed by defenders of the ethics of care.) The chapters of Part III turn to some central questions concerning the nature of our theorizing about distributive justice; these are questions that any theory must take a stand on, implicitly or explicitly. Do our theories of distributive justice presuppose a particular view of human nature? What would follow, both for the viability and the defensibility of a theory, from acknowledging that certain human dispositions and desires have an evolutionary explanation (Chapter 12)? Do demands of justice arise only in a context in which political institutions exist, and how does the demand that our political institutions be legitimate—​which, as I mentioned earlier, is thought to be a fundamental demand that we should make on our shared institutions—​relate to the concern with just­ ice (Chapter 13)? What assumptions about the nature of moral demands do competing theories of distributive justice make, and what do we learn about such theories when we keep in view crucial distinctions in normative ethics, between consequentialists and nonconsequentialists, deontological and teleological theories, agent-​neutral and agent-​ centred requirements, and forward-​looking and backward-​looking views (Chapter 14)? The last three chapters of Part III address overtly methodological questions which theorists of justice have been paying increasing attention to. What facts should our theorizing about justice abstract from, and is the attempt to formulate principles that would guide institutions of an ideally just society—​one in which we assume that everyone would comply with the principles we formulate, and in which conditions for realizing justice are favourable—​a useful and necessary part of a theory of justice, or an unnecessary and potentially misleading idealization (Chapter 15)? What is the defensible method for justifying principles of distributive justice? How do the ostensibly different methods of justification that are deployed by different theorists of justice—​ constructivism, intuitionism, and conceptual analysis—​relate to each other (Chapters 15 and 16)? The final part of the volume, Part IV, turns to discussions of the demands of distributive justice in various areas of social, economic, and political life. The titles of these chapters are self-​explanatory. Many of these pieces deal with different but importantly related topics and could be helpfully read together—​this is true, for example, of the chapters on gender, on the family, and on education, or of the chapters on cultural and religious minorities, language, justice beyond borders, and migration. With regard to some of the social and political issues discussed here, there is already widespread agreement that the challenges they raise can fruitfully be analysed through the lens of a distributive justice approach. This is the case, for example, with respect to the distribution of educational opportunities, employment, access to health, and the claims to resources by individuals beyond borders. (Saying this, as was already noted above, does not amount to saying that distributive justice concerns are the only concerns we have reasons to acknowledge with respect to these issues.) With other challenges we face, such as those presented by the persistence of social divisions along racial lines and the need to heed the claims of future generations, but also the wrongs of exploitation and

Introduction: The Idea of Distributive Justice    9 of discrimination, whether the questions we face are helpfully viewed as questions of distributive justice is disputed. The chapters on these topics discuss this important issue among others. While the topics covered by the chapters of Part IV are not supposed to exhaust the areas of our social lives that give rise to problems of distributive justice, they constitute a sizeable selection of the central cases. The debates they survey, and the discussions they contribute, are a good reflection of how rich and wide ranging distributive justice is as an area of political philosophy.

References Anderson, E. (2012). ‘The Fundamental Disagreement between Luck Egalitarians and Relational Egalitarians’. Canadian Journal of Philosophy, Supplementary Vol. 36: 1–​23. Bedau, H. (1978). ‘Social Justice and Social Institutions’. Midwest Studies in Philosophy, III: 159–​75. Buchanan, A. (1987). ‘Justice and Charity’. Ethics 97(3): 558–​75. Campbell, T. D. (1974). ‘Humanity before Justice’. British Journal of Political Science 4: 1–​16. Christiano, T. (2008). The Constitution of Equality:  Democratic Authority and Its Limits. Oxford: Oxford University Press. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Dworkin. R. (1986). Law´s Empire. Cambridge, MA:  The Belknap Press of Harvard University Press. Fleischaker, S. (2005). A Short History of Distributive Justice. Cambridge, MA:  Harvard University Press. Hart, H. L. A. (1961). The Concept of Law. Oxford: Oxford University Press. Jackson, B. (2005). ‘The Conceptual History of Social Justice’. Political Studies Review 3: 356–​73. Kymlycka, W. (1990). Contemporary Political Philosophy:  An Introduction. Oxford University Press. Nagel, T. (1997). ‘Justice and Nature’. Oxford Journal of Legal Studies 17: 303–​21. Nagel, T. (2005). ‘The Problem of Global Justice’. Philosophy and Public Affairs 33: 113–​47. Nozick, R. (1974). Anarchy, State, and Utopia. Oxford, UK and Cambridge, MA: Blackwell. Pettit, P. (2012). On the People’s Terms:  A Republican Theory and Model of Democracy. Cambridge: Cambridge University Press. Raphael, D. D. (2003). Concepts of Justice. Oxford: Oxford University Press. Rawls, J. (1999 [1971]). A Theory of Justice, revised edn. Oxford: Oxford University Press. Scanlon, T. (1998). What We Owe to Each Other. Cambridge, MA: Harvard University Press. Tan, K. (2004). Justice without Borders:  Cosmopolitanism, Nationalism and Patriotism. Cambridge: Cambridge University Press. Vallentyne, P. (2015). ‘Justice, Interpersonal Morality, and Luck Egalitarianism’, in A. Kaufman (ed.) Distributive Justice and Access to Advantage. Cambridge: Cambridge University Press, pp. 40–​9. Walzer, M. (1983). Spheres of Justice:  A Defence of Pluralism and Equality. Oxford, UK and Cambridge, MA: Blackwell. Williams, A. (2008). ‘Justice, Incentives and Constructivism’. Ratio XXI(4): 476–​93. Young, I. M. (1990). Justice and the Politics of Difference. Princeton, NJ:  Princeton University Press.

Pa rt  I

C ON T E M P OR A RY A P P ROAC H E S

Chapter 1

R awls on Dist ri bu t i v e Justice and t h e Difference Pri nc i pl e Samuel Freeman

1.1 Introduction Justice is associated with equality:  equal treatment, equal distribution, equal social status, or the formal requirement to treat similar cases similarly. But equality is not the only value of justice. A democracy that denies all citizens personal freedoms of conscience, association, and expression is oppressive, hence unjust. An egalitarian society is also unjust if it is able but unwilling to provide educational, cultural, and diverse career opportunities, enabling citizens to develop their capacities and skills and choose from a wide range of activities and life plans. John Rawls said early on that his principles of justice express a complex of three ideas:  liberty, equality, and rewarding contributions that promote the common good.1 He also says that justice as fairness incorporates and reconciles two different conceptions of equality: equality of distributions and equality of respect for persons (Rawls 1999a: 447). Equality regarding distributions, Rawls says, is incorporated into the difference principle’s reciprocity requirement—​that departures from equality must benefit everyone, starting with the least advantaged. Equality of respect for persons is, however, more fundamental; it is owed to humans as moral persons and is grounded in their possessing the moral powers of rationality and justice. Equality of respect for moral persons, Rawls says, is exhibited by the equal basic liberties and their priority, fair equality of opportunities, and such natural duties as mutual respect. Rawls also appeals

1 

“Justice as Fairness” (Rawls 1999b: 48). On the difference principle and the common good, see also “Distributive Justice” (Rawls 1999b: 153).

14   Samuel Freeman to the ideal of free and equal moral persons cooperating on grounds of reciprocity and mutual respect to explain why inequalities of economic distributions are justified to guarantee the worth of the basic liberties and fair opportunities for all citizens. G.A. Cohen, Rawls’s most trenchant egalitarian critic, contends that Rawls’s account of distributive justice is a confused amalgam of egalitarian and welfarist/​prioritarian considerations that sacrifices justice-​as-​equality for the sake of efficiency (Cohen 2008). But if we take seriously, Rawls’s position that justice is based in equality of respect for free and equal moral persons and accordingly must guarantee equal liberties and their worth for all, fair equality of diverse opportunities, and also promote the common good, then Cohen’s interpretation of Rawls seems shortsighted. The several values of justice incorporated into Rawls’s principles put into perspective the complexities of Rawls’s account of distributive justice. Rawls’s account of distributive justice has several key components. I begin by clarifying the role of the basic structure (section 1.2), and then discuss the distributive role of the basic liberties and fair equality of opportunity (section 1.3). I next clarify the broad and narrow requirements of the difference principle (section 1.4) and discuss whether inequalities are permissible or mandatory under it (section 1.5). In section 1.6 I discuss the scope of the maximizing provision, and why the difference principle is not a consequentialist prioritarian principle, and section 1.7 sets forth principles for applying the difference principle to ideal and non-​ideal conditions. I  conclude with a discussion of why the difference principle justifies property-​owning democracy rather than welfare-​state capitalism (section 1.8).

1.2  The Basic Structure as the Primary Subject of Justice Social cooperation and reciprocity are for Rawls fundamental to distributive justice. It is the role of principles of social justice to specify the standards for the fair distribution of primary social goods that attend and result from social cooperation in any functioning society. The primary goods include rights and liberties; opportunities, powers, and positions of office and responsibility; income and wealth; and the social bases of self-​respect. Rawls assumes that certain basic social institutions are necessary to sustain social life in any modern society and to guarantee the creation, distribution, and secure possession and enjoyment of these primary social goods. Basic social institutions include the political constitution, whose role is to make and enforce laws and adjudicate disputes; the legal institution of property, broadly conceived as rights and powers with respect to things; the economic system of production, transfer, and distribution of goods and services; and the family, which is the primary institution for reproducing society from one generation to the next. How these institutions are designed into a social system constitutes a society’s “basic structure.” The “primary subject” to which principles of

Rawls On Distributive Justice    15 social justice apply is the design of basic institutions into the basic structure of society (Rawls 1999a: 6–​7). One reason Rawls gives for focusing on the basic structure is its profound and pervasive effects on individuals’ life prospects, including its effects in shaping their primary aims and aspirations, their characters, and their self-​conceptions.2 The effect of the political constitution and the rights it guarantees on our prospects and aspirations is readily apparent by comparing liberal political systems that protect basic liberties with non-​ liberal ones that deny them. Also, differences in natural endowments, the social class one is born into, and accidental misfortune during one’s lifetime (illness, accident, unemployment, economic crises) profoundly affect inequalities of life prospects. How the institutions of the basic structure should respond to these three arbitrary contingencies is a fundamental question of social and economic justice. The second reason for the primacy of the basic structure is that a correctly designed basic structure is necessary to maintain “background justice” in a liberal economic system that relies on a “social process” of “pure procedural justice” to determine fair distributions. Rawls contrasts background justice in his social process view with Locke’s and Nozick’s “historical process view,” where just distributions are regarded as the accumulated outcome of free and purportedly fair agreements and consensual transfers among private owners (Rawls 2001: §15). Historical process views do not correct for the three arbitrary contingencies that cause inequalities in life prospects. As a result, “the invisible hand guides things in the wrong direction and favors an oligopolistic configuration of accumulations” (Rawls 2005: 267). The normal tendency of unregulated market transactions is increasing inequality and concentration of wealth in fewer hands (Piketty 2013). The background conditions necessary for fair exchange, fair equal opportunities, and the fair value of political liberties are then increasingly undermined. In order to maintain a fair background for the economic system and a just distribution of income and wealth, the accumulation of property and market power must be regulated by taxation, and laws governing inheritance, gift, and other means for acquiring market power, income, and wealth.

1.3  Basic Liberties and Fair Equality of Opportunities: Distributive Effects Rawls’s two principles of justice say: (a) Each person is to have the same indefeasible claim to a fully adequate scheme of equal basic liberties, which scheme is compatible with the same scheme of liberties for all; and, 2 

See Rawls (1999a: 6–​7); see also, Rawls (1999b: 138) and Rawls (2001: 55–​7).

16   Samuel Freeman (b) Social and economic inequalities are to satisfy two conditions: first they are to be attached to offices and positions open to all under conditions of fair equality of opportunity; and second, they are to be to the greatest benefit of the least advantaged members of society (the difference principle). (Rawls 2001: 42) Rawls uses the term “distributive justice” “in the narrow sense” (Rawls 2001: 61) to refer to the difference principle and the economic distribution of income, wealth, and economic powers. Still he emphasizes that the other principles have significant distributive effects. Indeed, Rawls says the difference principle cannot be taken seriously apart from the first principle and fair equality of opportunity (Rawls 2001: 46n). He says this partly in response to the frequent challenge that the difference principle puts no restrictions on overall inequalities. For example, G. A. Cohen argues that the difference principle permits the enormous inequalities typical of capitalism so long as inequalities marginally increase the share going to the least advantaged class (Cohen 2008: 138). But there are restrictions on inequalities imposed by both the first principle and fair equality of opportunities (FEO). Inequalities in income and wealth are unjust if they dilute the “fair value” of equal rights of political participation, or if they undermine the adequacy of educational, professional, and cultural opportunities and citizens’ ability to take advantage of formally equal opportunities and compete on fair terms with others.3 Inequalities in wealth, then, cannot be so great that they seriously dilute the “full and equally effective voice” and political influence of the less advantaged, or distort the political process and its agenda to favor the interests of the more advantaged (Rawls 2005: 361). The less advantaged should be in a position to participate in politics and influence public life on a par with the more advantaged. For them to do so, Rawls says, “property and wealth must be kept widely distributed” (Rawls 1999a: 198). He provides no specific formula or threshold to determine the limits on economic inequalities needed to preserve political equality.4 There may be no specific answer to the question of limits on economic inequality needed to preserve political equality; the question may depend on a society’s political culture. There are considerable inequalities in Germany, Sweden, and other European countries, but they do not have the decisive influence on politics that wealth has in the United States. Because of the “curse of money,” Rawls suggests, our politics is dominated by corporate and other organized interests that distort, if not preclude, public discussion and deliberation (Rawls 1999b: 580). 3  Rawls says: “Background institutions must work to keep property and wealth evenly enough shared over time to preserve the fair value of the political liberties and fair equality of opportunity over generations” (Rawls 2001: 51). 4  Instead, Rawls discusses public funding of forums for free public discussion and of political parties and campaigns, and strict limits on private contributions to neutralize the corrupting effects of money on politics (Rawls 1999a: 198–​9; 2001: 149–50; 2005: 328, 357).

Rawls On Distributive Justice    17 Regarding fair equality of opportunity: Rawls again says that economic inequalities are to be restricted when they reach a point that subverts the fair distribution of (formally) equal opportunities to compete for open educational and career positions and take advantage of the benefits of culture: Fair equality of opportunity means a certain set of institutions that assures similar chances of education and culture for persons similarly motivated and keeps positions and offices open to all . . . It is these institutions that are put in jeopardy when inequalities of wealth exceed a certain limit. . . .The taxes and enactments of the distribution branch are to prevent this limit from being exceeded. (Rawls 1999a: 245–​6)

Again, Rawls provides no formula to decide limits on inequalities needed to maintain fair equal opportunities. He discusses progressive income, estate, inheritance, and gift taxes as means to prevent excessive concentrations of wealth and private power (Rawls 1999a:  245–​6; 2001:  51, 53, 64, 161). Perhaps Rawls is less specific on these matters since he assumes that the economic system justified by the principles of justice is not capitalism but a property-​owning democracy, “a democratic regime in which land and capital are widely though not presumably equally held” (Rawls 1999a: 247). An essential feature of property-​owning democracy is that, unlike capitalism, the preponderance of productive resources is not controlled by a small sector of society. The absence of extreme inequalities in ownership of capital in property-​owning democracy eliminates the primary impediment to fair equality of opportunity that exists in a capitalist society. Rawls says that equality of opportunity is not the authorization to leave the less fortunate behind (Rawls 1999a: 91). This is a problem with formal equality of opportunity and “careers open to talents” where access to social positions is governed by the principle of efficiency. Fair equality of opportunity addresses this problem to some degree since it requires that the economically less advantaged be afforded generous educational opportunities, enabling them to compete fairly for positions with those equally talented regardless of social class. But on its “liberal interpretation,” when combined with the principle of efficiency, fair equality of opportunity also leads to a meritocracy that leaves the less talented behind, for then resources for educational development are allotted mainly according to talents and the return on productive abilities and skills (Rawls 1999a: 73, 91). Rawls says that on its democratic interpretation fair equality of opportunity does not lead to a meritocratic society, for then it is combined with the difference principle. FEO then requires that the less talented and less favored have ongoing educational, career, and cultural opportunities from early on and throughout their lifetimes, so that they can develop their capacities and take advantage of the benefits of culture (Rawls 1999a: 92, 265). This is especially important in view of the “essential primary good of self-​respect” (Rawls 1999a: 91).

18   Samuel Freeman Resources for education are not to be allotted solely or necessarily mainly according to their return as estimated in productive trained abilities, but also according to their worth in enriching the personal and social life of citizens, including here the less favored. As a society progresses the latter consideration becomes increasingly more important. (Rawls 1999a: 92)

The redistributive demands of fair equality of opportunity on its democratic interpretation potentially can be quite extensive. For example, consider evidence that children in upper-​middle-​class families acquire, well prior to their formal schooling, educational and cultural advantages due to constant attention lavished upon them by their parents (Lareau 2011). T. M. Scanlon has suggested that to achieve fair equality of opportunity, considerable efforts must be made to bestow similar advantages upon poorer children whose parents are not in a position to provide such educational and cultural benefits.5 This would support, for example, publicly funded day care for all children soon after infancy, designed to stimulate their capacities and develop their mental abilities and social skills. Also family allowances may be required by FEO as well as the difference principle, so that families can afford to expose children to social and cultural experiences otherwise reserved for parents who can afford such advantages (cf. Rawls 1999a: 243). There are many other options that might be applied to lessen inequalities in fair opportunities that stem from family circumstances. The point is that, if the aim of fair equality of opportunity is to both extend the benefits of culture to all and neutralize the effects of social class as much as is reasonably possible consistent with equal basic liberties and preserving the benefits of the institution of the family, it requires much more than simply supplying talented children born to poorer parents with educational advantages so that they can compete for desirable positions in a meritocratic system. As Rawls says, fair equality of opportunities applies to all citizens, whatever their degree of talent, to enable them to develop their capacities, take advantage of cultural opportunities, and enrich their personal and social lives. The distributive demands of FEO, so conceived, can be quite extensive, independent of considerations of the reciprocity claims of the less advantaged under the difference principle.6

5  Scanlon (2018), Ch. 5, “Substantive Opportunity.” He sees FEO as imposing a strong egalitarian requirement. 6  Phillipe van Parijs also discusses the significant demands of FEO, and suggests a paradox: the resources needed to fully satisfy basic liberties and FEO, may leave no resources for the difference principle (Van Parijs 2003: 225–​6). However, as the first principle does not require maximizing equal basic liberties but instead a scheme “fully adequate” to exercise the moral powers, FEO also is not a maximizing principle requiring “perfect equality of opportunity,” or perfectly equal chances in life (Rawls 1999a: 265, 448). Rather, it requires all should have a “fair chance” to attain offices and social positions (Rawls 2001, 43), with educational and training opportunities adequate to enable the less advantaged to fully develop and exercise their capacities so they can fairly compete for open positions and take full advantage of cultural and social life.

Rawls On Distributive Justice    19 Finally, regarding inequality of opportunity for women due to childbearing and their assuming the predominant burden of childrearing: Rawls says that childrearing within the family is “socially necessary labor,” and that women’s (or men’s) share of childrearing either should be equalized, or they should be compensated for it (Rawls 1999b: 600). This would be personal compensation in addition to family allowances intended to benefit children and required by FEO and the difference principle (Rawls 1999a: 243).

1.4  The Difference Principle: Its Broad and Narrow Requirements Though Rawls says it is a principle of “distributive justice,” the difference principle is more generally a principle of economic justice that regulates the design of economic systems, for the difference principle applies to institutions other than those directly affecting the distribution of income and wealth. It is the fundamental standard to assess and reform a wide range of economic institutions necessary for economic production, commerce, and consumption. Many of these institutions concern the fair and efficient functioning of the economic system and only indirectly bear on economic distributions of income and wealth: the law of contracts, securities, bank and finance regulations, business corporations and partnerships, commercial law, labor law, liability and remedies in economic torts, and so on. For example, the difference principle applies to restrict predatory “pay-​day” loans, and to regulate interest rates, late payment penalties, and other terms of consumer credit loans; or to regulate the terms of mortgages and restrict banks from investing customers’ deposits in risky derivatives, mortgage securities, and the other speculative ventures that were behind the 2007 financial crisis.7 Distributive justice is often depicted as the problem of dividing pre-existing accumulations of goods or of income and wealth, allocating them to individuals who need not stand in cooperative relationships. Rawls calls such approaches “allocative” accounts of distributive justice (Rawls 1999a: 56, 77; Rawls 2001: 50). Familiar examples are:  To each equally, or according to need, or merit, or to maximize welfare (aggregate, average, or weighted in favor of the less advantaged). The difference principle is not an allocative principle. It applies “at the front end,” to assess and reform economic institutions (property, contracts, and others mentioned above) that make possible the production and transfer as well as the distribution of income and wealth. Distributions of income and wealth are just when they result from individuals’ actual compliance with the institutions of an “ideal social process” structured by the difference principle (Rawls 2001: 54). Individuals’ shares are then decided according to “pure background procedural justice,” by the outcomes of the workings of just background institutions (Rawls 2001: 57). 7 

See Blinder (2013) on such practices, and measures needed to avoid them.

20   Samuel Freeman Since the difference principle applies to institutions “at the front end” of this process, it also determines the justice of distributions of ownership and control of means of production. This includes ownership of productive capital as well as rights to exercise “powers and prerogatives of offices and positions of responsibility” (Rawls 2005: 181). This is essential to Rawls’s argument for property-​owning democracy (discussed in §8). Since the difference principle has multiple roles, it is not simply an allocative principle of distributive justice. It is the fundamental standard of economic justice for a democratic society. Rawls regards the difference principle as a “principle for institutions” that directly applies only to the institutions of the basic structure. It “is not meant to apply to small-​scale situations [but] is a macro, not a micro principle” (Rawls 1999b: 226). The difference principle is not, then, an individual rule of conduct that we observe in making economic decisions; there is no natural duty to choose to maximize the circumstances of the least advantaged. Instead, it applies to the economic framework and the policies, laws, and other norms that regulate economic transactions. Rawls has been widely criticized for restricting the application of the difference principle to the basic structure (Cohen 2008; Murphy 1998). His reasons are complex:8 they connect with Rawls’s liberalism, which guarantees individuals’ special commitments and freedom to pursue their conception of the good, and with the difference principle’s being a principle of reciprocity rather than consequentialist or “end-​state” (discussed in section 1.7).9 By “greatest benefit to the least advantaged” Rawls means that “inevitable inequalities”—​in income, wealth, and powers and positions—​are to be arranged to make the least advantaged class better off than they would be under any alternative social and economic system (Rawls 2001: 59, 63). A fundamental role of the difference principle, then, is the comparison and assessment of the justice of different economic systems. Rawls says justice as fairness “is a conception for ranking social forms viewed as closed systems” (Rawls 1999a:  259). Existing economic systems are to be critically assessed according to how closely they approximate to the ideal of a “well-​ordered society.” The second principle of justice is, then, to be used as the ultimate standard to make and reform laws, regulations, and other economic norms in order to eventually realize the economic system the second principle prescribes. This raises the complex question of how the difference principle is to apply in ideal and non-​ideal circumstances (see section 1.7 ). Rawls says, “In appraising institutions we may view them in a wider or a narrower context,” either assessing the justice of institutions separately, one-​by-​one, or assessing the justice of the social system of institutions as a whole (Rawls 1999a: 50). This suggests two separate but complementary requirements in the difference principle. First, there is a narrower local requirement: for any alternative economic measure (laws, conventions, regulations, or general economic policies), its justice depends upon the degree to which 8 

For discussion, see Freeman (2014: 88–​111). Robert Nozick says the difference principle is an “end-​state” “patterned” principle (Nozick 1974: 202). Rawls said in conversation that he did not understand how Nozick could say it is end-​state. Rawls also denies it is patterned (see Rawls 1999b: 229). 9 

Rawls On Distributive Justice    21 it maximally advances the position of the least advantaged members of society, given the existing background of laws and institutions of which it is (or is to be made) a part. As a practical matter this requires that in deliberating, for example, upon the rate of taxation, a legislative body should choose the tax program that makes the least advantaged better off in the foreseeable future than other alternatives. Similar critical assessments apply with respect to regulations of securities and financial institutions, the specification of property rights, corporate law and labor relations, and all other significant measures bearing on economic production, transfer, exchange, and consumption. Most discussion and criticism of the difference principle has focused on this local requirement and the narrow application of the difference principle. This makes sense practically, from our perspective now, since legislative and other institutional changes normally proceed in piecemeal fashion, one step at a time. The narrow application of the difference principle and gradual reform of the status quo in the direction of greater economic justice is perhaps as much as we realistically can hope for under current circumstances. But this can obscure the second and even more fundamental requirement of the difference principle:  this is the broad systemic requirement presupposed by its narrower local application. Strictly applied, the difference principle says that a society has the duty to put into place the economic system of basic institutions that makes the least advantaged members of society as well off as they can be, consistent with preserving the equal basic liberties and fair equal opportunities. This means that the difference principle requires, in the first instance, that societies take steps to reform their institutions and enact measures that put into place that organized combination of institutions (laws, conventions, regulations) that maximally benefit the least advantaged members of their society over their lifetimes. Imagine “a rough continuum of basic structures” (Rawls 2001: 70), each of which is an efficient arrangement that specifies a particular division of social advantages. “The problem is to choose between them, to find a conception of justice that singles out one of these efficient distributions as just” (Rawls 1999a: 61). Simplifying, there is a range of feasible economic systems (Table 1.1), each discussed by Rawls at some point (cf. Rawls 2001: 136), except the social democratic welfare state: Table 1.1   Command economy state socialism

Market socialism

Propertyowning democracy

Social Welfare democratic state capitalism welfare state

“Liberal equality” fair equal opportunity + safety net

Classical liberal laissez-​ faire “system of natural liberty”

Libertarian laissez-f​aire

On the far right is the ideal libertarianism of Robert Nozick and others, which provides for absolute property rights and unfettered freedom of contract. Income, wealth, and all property rights are then distributed solely by consensual transfers through market exchanges, gifts, bequests, and gambling. Next is the laissez-​faire

22   Samuel Freeman “System of Natural Liberty” (Rawls 1999a: 57, 62) of the classical economists. Unlike libertarianism it allows for some regulation of negative externalities (price collusion, etc.) and taxation for essential public goods; but otherwise income, wealth, and rights in things are distributed as libertarians profess. Next is the system of “Liberal Equality” (Rawls 1999a: 57, 63) which, by incorporating a degree of fair equality of opportunity, funds education, a social “safety net,” and other measures intended to neutralize the effects of social class on starting positions and life chances. This is endorsed by Friedrich Hayek and moderate classical liberals, also to a lesser degree by Milton Friedman (whose proposed negative income tax supplies a social minimum at the upper end of the poverty level).10 Then comes the capitalist welfare state, which puts in place a more robust social minimum, universal healthcare, and perhaps (as in Ronald Dworkin’s “equality of resources”) a compensatory social insurance system that addresses individual misfortune and disadvantage, but otherwise allows income and wealth to be distributed by market and other consensual transfers (Dworkin 2002). Rawls contrasts the capitalist welfare state not with a social democratic welfare state, but with property-​owning democracy (POD). POD largely dissolves the capitalist distinction between the owning and laboring classes by widely distributing economic wealth among all society’s members. Also it allows a wider distribution of economic powers among society’s members, and generally requires that economic distributions satisfy some more egalitarian principle than the capitalist welfare state. Next is liberal socialism, which provides for market allocations of publicly owned means of production, and normally enforces an egalitarian principle of distributive justice.11 Finally, there are command economy arrangements as Marx envisioned, which dispense with private economic ownership and with markets in allocating factors of production, including labor. The difference principle’s wide systemic requirement says a society is to put into place the economic system that makes its least advantaged class better off than other alternatives (Rawls 2001:  59–​60). Then, society is to fulfill the difference principle’s narrow requirement and continually make local adjustments and economic reforms needed to improve and maintain the position of the least advantaged. The crucial point is that the difference principle is not satisfied in non-​ideal conditions such as our own simply by making local adjustments to particular institutions that marginally improve the position of the least advantaged—​especially not if inequality is increased. It requires instead widespread across-​the-​board revision of any economic system that is not designed to maximally benefit the least advantaged—​taking the necessary measures of reform to guarantee that “the least advantaged are better off than they are under any other scheme” (Rawls 2001: 60). 10 

See Friedman (1962: 191) on the negative income tax. Hayek similarly endorses a “minimum income” as “a necessary part of the Great Society” (Hayek 1979: 55, 187 n13). 11  See Roemer 1994, who advocates equality of opportunity for welfare, and “coupon socialism” wherein all individuals own rights to receive dividends from publicly held shares that they can trade on markets but cannot bequeath or sell to others.

Rawls On Distributive Justice    23 This addresses G.  A. Cohen’s criticism that the difference principle is compatible with the enormous inequalities typical of capitalism and requires exaggerating these inequalities if they benefit the least advantaged.12 Cohen assumes the status quo of the capitalist welfare or safety-​net state in the US and argues that many of capitalism’s gross inequalities are justified by the difference principle, for they make the least advantaged better off than they otherwise would be in the absence of those inequalities. This overlooks Rawls’s contention that the inequalities resulting from our capitalist system are unjust, and that the difference principle requires instead a property-​owning democracy or a liberal socialist system.13 The difference principle necessitates systemic reforms to capitalism, and these reforms are not advanced by the Pareto measures Cohen assumes that only reinforce and increase gross inequalities. For example, assume (as conservatives argue) that the least advantaged in the US would enjoy less (employment, earned income tax credits, public assistance, healthcare, etc.) were it not for the lower tax rates imposed under the Reagan and Bush Jr presidencies, with ensuing extraordinary rewards going to corporate executives, “high-flying” financiers, and other extremely wealthy persons. Even if true, this in no way means that Rawls’s difference principle sanctions these tax measures or the activities and gross inequalities that result. For the position of the least advantaged surely could have been improved far more by some degree of increased taxation of capital gains and extreme salaries and bonuses, using increased tax income for public goods and transfers to those much less advantaged. But more important, according to the difference principle’s systemic requirement, the current economic system in which people are allowed to engage in such “high-​flying” practices and reap extreme benefits from pure ownership is itself unjust and should be reformed in the direction of a property-​owning democracy, or at least an equitable welfare state. These inequalities are unjust, not simply because the rate of taxation to the wealthy and benefits to the least advantaged are set far too low, but because an economic system which encourages this sort of economic activity and the attitudes or “ethos” that sustain it, and which allows such extremely unequal distributions of income, wealth, and economic powers, is unjust according to the difference principle. Our capitalist economic system currently is one where few conscientious efforts are being made to even advance, much less maximize, the position of the least advantaged. The “trickle-​down” economics institutionalized since the Reagan–​Thatcher era is directly at odds with the difference principle.

12 

Cohen contends the difference principle is consistent with “a maximizing ethos [that] will produce severe inequalities and a meager level of provision for the worst off ” (Cohen 2008: 138). He argues that the difference principle can even sanction distributions like the United States where the least advantaged are worse off than they are in a society where people have a more egalitarian ethos, like Germany (p. 143). This conflicts with Rawls’s claim that the least advantaged are to be better off than in all other systems. If a more egalitarian ethos would make the least advantaged better off, then the difference principle would require institutions, consistent with freedom of conscience, that encourage people to develop such an ethos (see Freeman 2007: 121; see also Cohen 2001). 13   Rawls (1999a: xiv–​xvi); Rawls (2001: 8n, 135–​40). The local application of the difference principle assumes in the ideal case the background institutions of property-​owning democracy (see Rawls 1999a: 67–​68, 242).

24   Samuel Freeman It is not true, then, that local improvements to the least advantaged that allow increasing inequalities of the kind Cohen envisions are sanctioned by the difference principle. Not just any improvement of the position of the least advantaged is justified; strictly speaking only very few are. The difference principle is not to be confused with a Pareto-​like principle that says that any measure that improves the position of the least advantaged is permissible, regardless of how much it increases inequality.14 Strictly applied, the difference principle requires taking those measures that, from among existing alternatives, maximally benefit the least advantaged, “other things being equal” (Rawls 2001: 63). “Other things” are not equal under our non-​ideal conditions, when policies involve narrow measures that may benefit the least advantaged in the short run, but nonetheless increase and reinforce enormous inequalities; for narrow measures are not designed to modify the economic system in the direction of a more just basic structure, whether it be property-​owning democracy, liberal socialism, or an equitable welfare state.

1.5  Is Maximin Optional? Consider now in more detail how the difference principle applies in different circumstances of justice. Rawls says a system is “perfectly just” when the expectations of the least advantaged group (LAG) are maximized; this is the “best arrangement.” A  system is “just throughout” when the expectations of the least advantaged would diminish if those of the most advantaged groups (MAG) were lessened, but would improve still further if the expectations of the MAG also were to improve. And a system is unjust when worsening the expectations of the most advantaged could improve those of the least advantaged (Rawls 1999a: 68; Rawls 1999b: 138). The US (unsurprisingly) is unjust according to Rawls’s definition. For there are multiple ways to improve the position of the least advantaged group that would also diminish inequality. Instead of adopting these measures, the usual approach involves, at best, measures that further increase inequality, in hopes that modest benefits will “trickle down” to the less advantaged. Which if any of these (alleged) Pareto improvements are just? Suppose Congress proposed further tax cuts for the wealthy 14  G. A. Cohen says the “lexical difference principle” is the “canonical version” (Cohen 2008: 17, 156f., 161, 164f., 320). Rawls, citing Sen’s suggestion, briefly mentions the lexical principle if close-​knitness fails, but in the revised edition of A Theory of Justice he dismisses it as irrelevant: “when the greater potential benefits to the more advantaged are significant, there will surely be some way to improve the situation of the less advantaged as well. The general laws governing the institutions of the basic structure insure that cases requiring the lexical principle will not arise” (Rawls 1999a: 72). Accordingly, I forgo discussion of the lexical principle. Cohen’s criticisms heavily rely on it. He argues Rawls’s difference principle is a “strong Pareto Principle” that justifies benefiting the more advantaged but not the less advantaged (Cohen 2008: 29, 158). This conflicts with Rawls’s statement, “the difference principle is a strongly egalitarian conception in the sense that unless there is a distribution that makes both persons better off . . . an equal distribution is to be preferred” (Rawls 1999a: 65–​6).

Rawls On Distributive Justice    25 that would benefit everyone. Such Pareto measures cannot be said to be required by justice under unjust conditions since there are many other measures that would benefit the less advantaged more (e.g., income supplements and family allowances) and would also reduce inequality and the excessive economic expectations of the most advantaged. Are Pareto improvements then at least permissible? If under unjust conditions Pareto improvements further increase substantial inequalities, they would be permissible only if a necessary short-​term feature of a program that substantially improves the LAG’s position and reduces inequality in the longer run. Generally, under unjust circumstances where there are ways to improve the LAG while reducing inequalities, Pareto improvements that increase inequality are not justifiable under the difference principle. Only measures that decrease inequality while raising the level of the least advantaged are authorized by the difference principle. Pareto measures that do not increase inequality may be permissible so long as they are not part of a policy designed to avoid reducing inequalities. But in unjust conditions of substantial inequality, Pareto measures are unjust when they further increase permanent inequalities. Justice always is prior to (Pareto) efficiency and is consistent with it only in a “perfectly just” economic system where the least advantaged position is maximized (Rawls 1999a: 69). A more difficult question: Assume no further improvements can be made for the LAG by reducing inequality through transfers or other measures. Are local maximin measures, those that given existing alternatives maximally improve the LAG under the circumstances, then optional or required by the difference principle? What if maximin measures substantially increase inequality for the indefinite future? Under these circumstances normally there should be several alternative courses of institutional action that over time would result in less inequality and still make the least advantaged group better off than they are under the existing economic system. Does the difference principle nonetheless require taking only those measures that maximally benefit the least advantaged, even when other alternatives result in less inequality and leave the less advantaged only marginally worse off than the maximin alternative? Could a society simply forgo any further measures that increase inequality, even if they are needed to improve the least advantaged position? Finally, are maximin measures even permitted under circumstances when they permanently increase substantial inequalities? Answers to these questions depend on whether the basic structure is just or unjust according to justice as fairness. Suppose there were a fairly adequate social minimum in the United States (e.g., a $35,000 guaranteed income supplement plus child allowances for the LAG), and that further redistribution would worsen the LAG’s position. Only by reducing the capital gains tax on wealth from 20% to 15% (its level during the Bush, Jr era) could we further increase the social minimum; but this would create substantially greater discrepancies between the most and least advantaged. In an unjust basic structure where there is a fairly adequate social minimum, surely the difference principle does not require the tax reduction on capital gains when it further increases substantial inequalities indefinitely. One reason is the potentially bad effects of increased inequality on the self-​respect of the least advantaged, which outweigh the benefits of marginal increases in their income.

26   Samuel Freeman But let’s assume self-​respect does not decide the issue (the increased inequality would not further decrease LAG’s self-​respect). Rawls suggests that in making decisions that increase the social minimum, a society may choose sub-​optimal measures involving less inequality. For example, referring to his diagram of the difference principle curve (Rawls 2001: 62), (Fig. 1.1) he says that when the difference principle is satisfied, “Society would always be on the upward-​ rising part or at the top of the OP curve” (Rawls 2001: 64). This might explain his ceteris paribus qualification: “Other things being equal, the difference principle directs society to aim at the highest point [D]‌of the OP curve of the most effectively designed system of cooperation” (Rawls 2001: 63). Other things are not equal when a society is an unjust system and local measures that optimize the position of the least advantaged within that system require substantial increases to inequality. The requirement to strive to be at the highest point, D, on the curve strictly applies only within “the most effectively designed system”—​presumably a property-​owning democracy. In unjust circumstances typified by gross inequalities it is at least optional, and sometimes may be required, for a society to forgo local measures that optimally improve the least advantage position when they also increase existing inequalities. Thus, in the example above, where the LAG have

y = LAG

D

J

J

N B

F P

45° 0

x = MAG

LAG = Less advantaged group OP = Production curve

MAG = More advantaged group

D = Efficient point nearest equality

B = Bentham point, where sum of utilities is maximized N = Nash point, where product of utilities is maximized F = Feudal point, where OP curve becomes vertical 45° line = Equality

fig. 1.1

JJ = Highest equal justice line

Rawls On Distributive Justice    27 a fairly adequate social minimum ($35,000) and the maximin measure (reducing the capital gains tax) would substantially increase existing inequality, society should take the alternative measures necessary to reform the system in the direction of a property-​ owning democracy, even if this comes at the expense of delaying immediate increases to the social minimum. Moreover, local short-​term maximin measures would not even be permitted if they not only increased but cemented extreme inequalities, thereby impeding the transition to a more just basic structure. In unjust conditions, even local maximin measures can sometimes be unjust, when conditions of extreme inequality obtain. I return to these difficult questions regarding the application of the difference principle in non-​ideal or unjust conditions in section 1.7, but first, consider more ideal circumstances and a system that is not as unjust and beset with such gross inequalities as is the US. Consider an “effectively designed” economic system, a property-​owning democracy that is “just throughout” in Rawls’s sense: the prospects of the least advantaged would decline if those of the most advantaged group declined, but would improve if the expectations of the most advantaged also improved. I’ll call “Optional Inequality” the interpretation of the difference principle which says that (further) increases in inequality that (maximally) benefit the least advantaged are not required by the difference principle, but are optional and to be left to democratic decisions. At the limit, Optional Inequality suggests that a society does not have to institute any inequalities; democratic citizens might instead choose strict equality. Strict equality might seem an option, given the difference principle’s wording. Rather than mandating inequalities, it addresses “permissible inequalities in the basic structure” (Rawls 1999a: 56), and specifies how “social and economic inequalities are to be arranged” (Rawls 1999a: 72). Still the strict equality option is not supported by Rawls’s suggestion (above) that society is to be on the upwardly rising incline of the OP curve.15 It also conflicts with Rawls’s assertion that the “basic structure of society is perfectly just when the prospects of the least fortunate are as great as they can be” (Rawls 1999b: 138; see also Rawls 1999a: 68). This parallels Rawls’s statement: “Taking the two principles together, the basic structure is to be arranged to maximize the worth to the least advantaged of the complete scheme of equal liberty shared by all. This defines the end of social justice” (Rawls 1999a: 179). I’ll call this the “Mandatory Maximizing” interpretation; it says inequalities are not optional but required whenever they benefit the least advantaged, and that from among feasible alternatives society must enact those measures that maximally improve the position of the least advantaged.16

15 

Rawls says that strict equality is irrational (2001: 151). Further support for Mandatory Maximizing lies in the “equal justice” lines JJ (Rawls 2001: 62, Fig. 1; Rawls 1999a: 66, Fig. 6) which suggest that any distribution higher on the OP curve provides the least advantaged with greater benefits and is more just than one lower that provides them with less, since it is on or closer to a higher equal justice line. 16 

28   Samuel Freeman Rawls says various things that can seem to support both the Optional Inequality and Mandatory Maximizing readings of the second principle. The truth appears to lie somewhere in between. Taken by itself and independent of prior principles, the difference principle alone may indeed imply that economic inequalities are optional. But Rawls says the difference principle’s “meaning is not given by taking it in isolation” (Rawls 2001: 158n): it must be considered in conjunction with the demands of prior principles and the need to guarantee the worth of the basic liberties and fair equal opportunity.17 For example, suppose a well-​ordered democratic society legally mandated and sustained an egalitarian agrarian/​artisan society, similar to the society the Amish voluntarily reside in, assiduously avoiding the modern infrastructure needed for higher education, cultural achievements, and the benefits of modern technology. This would have adverse consequences for the diversity of career and cultural opportunities available to society’s members, depriving them of the conditions needed to fully educate themselves and take advantage of careers and pursuits enabling them to develop their talents and capacities and pursue a wide range of experiences and conceptions of the good. The worth of citizens’ equal basic liberties and fair equal opportunities would be significantly diminished in this stubbornly agrarian, egalitarian society, to the point of injustice. This is implied by Rawls’s statement that the “end of social justice” is to maximize the worth of the basic liberties to the least advantaged.18 The crucial point is that justice, including distributive justice, cannot simply be a matter of instituting equality of social goods—​of liberties, opportunities, income, wealth, and economic powers—​or equality of welfare. Justice also requires that a society provide fully adequate resources and opportunities to enable citizens to develop their capacities, effectively exercise their basic liberties, and pursue a wide and diverse range of conceptions of the good. This requires substantial public investment in public goods, including infrastructure, education at all levels, public health and universal healthcare, and so on. This is a crucial difference between Rawls’s liberal egalitarian position and strictly egalitarian accounts of distribution. Justice for Rawls requires not equality of (access to) resources, opportunities, or welfare, regardless of material and social conditions, but fair equal access to a wide range of career and cultural opportunities, and sufficient economic resources to give fully adequate, if not maximum, worth to all citizens’ equal basic liberties. This distribution and the social circumstances and institutions needed to effect it are for Rawls the “end of social justice,” not strict equality or equalizing the consequences of good and bad luck. 17 

“[T]‌he parts of the two principles are designed to work in tandem and apply as a unit” (Rawls 2001: 46n). Moreover, “the meaning of the difference principle is determined in part by its ranking as subordinate to the first principle of justice. That meaning is not given by taking it in isolation” (Rawls 2001: 158n). 18  Rawls also says that “the priority of fair opportunity” over the difference principle means “that we must appeal to the chances given to those with lesser opportunity [and] hold that a wider range of opportunity is open to them than would otherwise be the case” (Rawls 1999a: 265).

Rawls On Distributive Justice    29 The Optional Inequality interpretation of the difference principle must then be significantly qualified. It can only apply under conditions where the least advantaged citizens have sufficiently adequate career and cultural options and economic resources to enable them to take full advantage of their basic liberties and fair opportunities and therewith develop their capacities and pursue a wide range of life plans. Otherwise strict equality of income, wealth, and economic powers would not be just.

1.6  Mandatory Maximizing and the Scope of the Difference Principle What about the mandatory maximizing interpretation, suggested by Rawls’s remark that “the end of social justice” is to maximize the worth of basic liberties to the least advantaged” (Rawls 1999a:  179)? The extreme version of this interpretation would impose a strict requirement that all social policies not regulated by prior principles be assessed by the difference principle and designed to maximally benefit the LAG. Maximizing the position of the LAG then would be a kind of constrained dominant end guiding all social policies (except those covered by prior principles) much like a “mixed conception” with a similarly constrained principle of utility (Rawls 1999a: §49). Call this the ‘Strict Mandatory Maximizing’ interpretation: that all social policies not covered by prior principles are to maximally benefit the least advantaged. There are many social policies for which the difference principle does not appear to be the appropriate standard of assessment: foreign policy and non-​trade treaties; many public goods decisions, such as highways and other infrastructure; public investments in scientific research and technological innovation (e.g., the National Aeronautics and Space Administration (NASA), or computing and the internet); support for public museums and parks, or fine arts and athletic programs in schools; most family, child custody, and marriage law; ordinary negligence and determinations of fault and remedies in non-​economic torts and other areas of law not integral to economic cooperation; redressing disabilities and care for the severely handicapped; preserving the environment for aesthetic reasons; and many other policies. In designing these and other policies, legislators should avoid creating adverse consequences for the least advantaged; but this does not mean these policies should be tailored to maximally benefit them, for the difference principle does not respond to the nature of the issues raised by these policies. Of course, when these and other non-​economic measures do result in significant inequalities, such as divorce laws that result in female poverty, there is a problem of unjust inequality. But the difference principle is no more appropriate to questions of the fair division of assets in divorce proceedings than it is to ascertaining fault and equitable remedies in auto negligence cases. Regarding divorce Rawls endorses giving spouses an equal share in income and the increased value of family assets acquired during marriage

30   Samuel Freeman (Rawls 1999b: 600; 2005, 473); but this measure does not rely on the difference principle but on other considerations of fairness. Like determinations of fault and remedies in negligence cases, it would be unfair to require that assets between divorced spouses be divided so as to maximally benefit the least advantaged class, since they are normally entirely unrelated individuals. The difference principle responds to a different set of issues than family and marriage law, and most negligence and other tort law. What are these issues? In any society there is the problem of designing the social institutions that make economic production and commerce possible and fairly dividing the social product among those who contribute to productive activity. These institutions include property and legal control of means of production, markets and other conventions of transfer, and economic contracts and other commercial transactions. Utilitarianism in large part originated in Hume’s argument that these “conventions of justice” should be designed to promote public utility (Hume 1975: 183–​204). Rawls is addressing a similar problem but offers a different solution: the basic institutions necessary to productive economic cooperation should be designed so as to achieve economic reciprocity, or fairness in the division of the social product among “normal and fully cooperating members of society” (Rawls 2001: 8). By “normal and fully cooperating citizens,” Rawls means those who are actively engaged in economic cooperation and who do their fair share in contributing to social and economic life. Or as Rawls says, “This assumption implies that all are willing to work and do their part in sharing the burdens of social life” (Rawls 2001: 179, emphases added). This explains why Rawls comes to define the “least advantaged” as the lowest paid, least skilled workers, and not as the severely disabled or even the poorest citizens.19 The difference principle is designed to address the question of the fair and efficient distribution of economic powers and responsibilities in production and the equitable distribution of ensuing income and wealth among members of society engaged in productive social and economic activity. It is not conceived as a principle of redress or assistance to meet the basic or special needs of citizens. Their circumstances require principles that are specifically tailored to their conditions. Thus, Rawls says that the two principles of justice presuppose a principle of basic needs, ensuring that citizens’ basic needs are met so that they are able to “understand and to be able fruitfully to exercise [basic] rights and liberties” (Rawls 2005: 7). Meeting all citizens’ basic needs is a “constitutional essential” in any liberal society (Rawls 2005: 166, 228ff.) and in “decent societies” as well (Rawls 1999c: 38). Rawls does not enumerate such principles of basic needs or assistance for the special needs of those severely handicapped over their lifetimes; these are to be decided at the constitutional and legislative stages according

19  On unskilled workers as the least advantaged class, see Rawls (1999a: 67–​8, 83–​4). On the presumption that the least advantaged work, see Rawls (1999b: 455 n7 and 2005, 182 n9): “Those who are unwilling to work . . . must find a way to support themselves.” On society’s duties to the severely handicapped, see Rawls 2001: 176n.

Rawls On Distributive Justice    31 to principles yet to be specified (Rawls 2001: 176n). The important point is that the difference principle has a different role than such principles of redress: to maintain democratic reciprocity, or “reciprocity at the deepest level” among citizens who are “normally and fully cooperating” and thus engage in socially productive activity “over a complete life” (Rawls 2001: 49). This distinct role of the difference principle provides further reasons to reject the strict mandatory maximizing reading of the difference principle. For what could be the justification for requiring that all laws and social policies maximally promote the position of least paid workers when this would come at the expense of other equally if not more important social needs and interests, such as meeting the basic needs of all citizens and the special needs of the disabled? Assuming then that the difference principle does not provide the proper standard of assessment for all laws and social policies, the question remains whether, with respect to the basic institutions and economic measures to which it does apply, it mandates maximizing the least advantaged position. The difference principle is often regarded as an aggregative consequentialist principle that requires maximizing the sum of advantages for the least advantaged class. “Prioritarianism” is such a position sometimes attributed to Rawls’s difference principle.20 On the prioritarian interpretation the difference principle would require maximizing the absolute level of the least advantaged regardless of their position relative to those better off or the resulting degree of inequality. The Qualified Optional interpretation implies that the difference principle is not prioritarian, for it says that a society may forgo increases in inequality if the least advantaged already have an adequate share of relevant primary social goods. This corresponds to Rawls’s assertion that the difference principle does not require that society strive for ever-​increasing growth in order to maximize the expectations of the least advantage indefinitely. “That would not be a reasonable conception of justice” (Rawls 2001: 63–​4, 159). Also against the prioritarian reading is Rawls’s assertion, that “even if it uses the idea of maximizing the expectations of the least advantaged, the difference principle is essentially a principle of reciprocity” (Rawls 2001: 64). This suggests that the difference principle has a different purpose than maximizing the absolute sum of benefits for the LAG —​namely economic reciprocity, or fairness in the division of the social product. Unlike consequentialist principles, as a principle of reciprocity the difference principle determines distributive shares relationally, not in absolute terms:  the fairness of each person’s entitlements is ascertained relative to those of others with the justice of distributions to the least advantaged decided by how well off they are compared to the most advantaged. If the absolute position of the least advantaged were all that mattered, then any distribution where the least advantaged were better off would

20 

See Parfit 1991 and 1997. “On the Priority View, we are concerned only with people’s absolute levels” (Parfit 1997: 214).

32   Samuel Freeman be more just than distributions where they had less. But this is clearly not the case under the difference principle. A distribution affording less to the least advantaged quite often can be more just than alternatives affording more. For example, assume a moderately unequal society is “just throughout” (on the upwardly rising slope of the OP curve), with a $40,000 social minimum. The same society could have adopted policies (or might yet still) resulting in a $42,000 social minimum, but with substantial inequalities that are unjust since, were they diminished, the social minimum could be improved further (to $44,000). Society when it is “just throughout” with a $40,000 minimum is more just than if it had aimed for a $42,000 minimum with substantial unjust inequalities. Moreover, under ideal conditions of a just society with a fully adequate social minimum, a society with a lesser minimum might even be equally just as it would be had it adopted policies with a greater minimum requiring greater inequalities. Under ideal conditions, when society is just throughout, the difference principle requires that, from among alternative measures that increase inequality, they must choose those that make the least advantaged better off at that point than other alternatives. But this does not mean a society continually has to adopt measures that increase inequality in order to maximize the absolute sum going to the LAG. Referring again to the OP curve (Rawls 2001: 62), a society is not required to aim for that distribution (Point D) that maximally benefits the least advantaged; instead, it can decide on a point on the upwardly rising incline of the OP curve that improves their position but involves less inequality. In this regard too, Rawls’s difference principle is not prioritarian. Thus, under ideal conditions of a just, well-​ordered society, where the worth of the basic liberties and fair opportunities is fully adequate for all, a society may democratically decide not to increase inequality any further and decline to maximize the absolute sum going to the least advantaged. Return to Rawls’s claim that the difference principle does not require ever-​increasing growth to maximize the expectations of the least advantage indefinitely (Rawls 2001: 63–​4, 159). He says this to not rule out “Mill’s idea of a just stationary state where real capital accumulation may cease.”21 Suppose the ideal case, a property-​owning democracy where all citizens own a fair share of productive capital and the least advantaged have incomes of $150,000 per year. Assume these resources are fully adequate and enable them to effectively exercise their basic liberties and take full advantage of a wide range of professional and cultural opportunities. Then surely justice would not require that this society take measures to increase the social minimum still further, especially if this would increase existing inequalities. This may even apply to measures that do not increase inequalities. In either case, important social values such as limiting adverse effects on the environment can justify limits on economic growth and the social minimum.

21  Rawls (2001: 159). Mill argued that society would finally be able to direct its energies towards social ends more worthwhile than incessant pursuit of material wealth.

Rawls On Distributive Justice    33

1.7  Ideal vs Non-​Ideal Theory and the Difference Principle Rawls’s argument for the difference principle does not ultimately rely on the maximin rule of choice observed by rational parties in the original position (Rawls 2001: 94–​5). Instead, his argument is contractualist in that it appeals to the perspective of reasonable persons in a well-​ordered society: the difference principle is justifiable and generally acceptable to free and equal moral persons morally motivated by a sense of justice and self-​respect. Rawls’s arguments for the difference principle are based mainly in considerations of democratic reciprocity, the strains of commitment, and the conditions needed to maintain the self-​respect of free and equal citizens. The ideal of a well-​ordered society grounds ideal theory for Rawls, including his assumption of full compliance: it is a society where all reasonable persons willingly endorse and normally comply with demands of justice. Critics claim that Rawls’s ideal theory means that the principles of justice only apply to a “perfectly just” society, but not to non-​ideal circumstances of injustice.22 But justice as fairness would have no point if the principles of justice did not apply to assess the injustices of non-​ideal circumstances. Clearly the equal basic liberties and fair equality of opportunity apply to our circumstances, for Rawls discusses their application at length (Rawls 2005: Ch. 8). Moreover he is explicit that ideal theory guides the application of the principles of just­ ice to non-​ideal situations (Rawls 1999a: 267): “Until the ideal is identified . . . nonideal theory lacks an objective, an aim, by reference to which its queries can be answered” (Rawls 1999c: 90; see also, Rawls 2001: 13). The relevant question then is, how should the difference principle be applied to assess non-​ideal circumstances and guide reform under unjust conditions where the economic systems meeting its demands are not in place? Questions of reforms needed to bring about a just basic structure are largely empirical and strategic, especially given entrenched interests that benefit from injustice, but certain rules of application still apply. Here in broad outline is a proposed summary of rules for applying Rawls’s difference principle in ideal and non-​ideal circumstances. These suggestions need further elaboration and refinement. I. Application of the difference principle under ideal conditions of a well-​ordered society. Assume the difference principle’s broad requirement is satisfied: “The most effectively designed system” is in place, and the least advantaged are better off than in any other system. The alternatives Rawls proposes, property-​owning democracy and liberal socialism, do not by nature involve the substantial inequalities typical of capitalism. There

22  See Sen (2009: 16, 100), arguing that Rawls’s “transcendental principles” apply only within a “perfectly just society” and are irrelevant to our non-​ideal circumstances.

34   Samuel Freeman is no privileged class of persons who own and control the preponderance of productive wealth. Economic markets are regulated to prevent taking unfair advantage of citizens and undermining economic efficiency. Moderate inequalities exist, but since there is fair equality of opportunities and an adequate social minimum with income, wealth, and economic powers being widely distributed, these inequalities are both necessary and fair: inequalities realize “reciprocity at the deepest level” since society is on the upwardly rising slope of the OP curve. Consequently the self-​respect of the less advantaged is not adversely affected by these moderate inequalities. Under these conditions, first, a well-​ordered society in its economic policies normally is to adopt local measures regarding basic social institutions that maximally benefit the least advantaged. The exception would be the rare case where maximin measures involve a substantial increase in inequality; if so, then a society may choose alternative measures that improve the position of the LAG but which involve less inequality. Second, a well-​ordered society is to continue to increase the position of the least advantaged in its economic policies up to a point where further increases in the social minimum would not substantially add to the effective exercise of the basic liberties of the least advantaged (and presumably other income groups) or their ability to take full advantage of a wide range of educational, career, and cultural opportunities. At this point a society may decide the social minimum is fully adequate and devote further increases in social wealth entirely to improving infrastructure and other public goods, increasing benefits and services for the handicapped and disabled, providing assistance to poorer societies beyond what is already required by justice (Rawls 1999c: 106), setting apart additional savings for future generations, and promoting other democratically legitimated public values compatible with public reason (exploring space, preserving nature and improving the environment, promoting cultural values, etc.). Alternatively, assuming all its duties of social and political justice are satisfied (to the disabled, burdened societies, and future generations), a society may choose to maintain the economic status quo and its fully adequate social minimum, and forego increasing national wealth any further. This is Mill’s stationary state, where further increments to real capital accumulation have ceased (Rawls 2001: 64). Under these conditions, a society should replenish existing capital, and maintain its infrastructure and public goods for future generations. II. Application of the difference principle under non-​ideal conditions, such as the capitalist welfare or “safety-net” state. There are many complexities in non-​ideal conditions: In addition to strategic problems involved in applying the principle of justice because many citizens reject it, a society has to weigh off narrow measures that (maximally) improve the position of the least advantaged now, with broader measures needed to bring about “the most effectively designed economic system” (a property-​owning democracy). How strong is society’s duty to promote a just basic structure compared with its duty to maximally benefit the least advantaged in the short run? Clearly the difference principle does not allow society to sacrifice the basic liberties or well-​being of the least advantaged for the sake of maximally promoting the greater well-​being of increased numbers of the least advantaged in future generations (Rawls 1999a: 263–​4). But can it forgo local

Rawls On Distributive Justice    35 maximin measures and adopt alternatives that benefit existing generations somewhat less in order to enact broader economic reforms? This does not seem unreasonable if it is the only way to realize needed reforms of an unjust basic structure. And yet we saw in section 1.5 that non-​maximin Pareto-​improving measures that benefit the least advantaged are not authorized by the difference principle if this increases inequality, since there are alternative measures that decrease inequality and benefit the least advantaged still more. But if these inequality-increasing Pareto measures are the only way to achieve necessary reforms of an unjust system, then things are different. Finally, we saw (section 1.5) that not just Pareto measures but even measures that maximally benefit the least advantaged in the short run can be unjust when they also substantially increase and solidify gross inequality and/​or delay indefinitely the achievement of reforms that would move society in the direction of less inequality and a just basic structure. The following principles address these and other complexities in broad outline, though they may raise as many questions as they resolve. The first two can result in conflicting demands, but many of these can be resolved by an added qualification. First, as under ideal conditions, a society normally is to adopt those local economic policies that maximally benefit the least advantaged, if not in the short run, then intermediately and during a substantial portion of their lifetime. The exception, just noted, is maximin measures that aggravate and solidify substantial inequalities, and impede reform in direction of a more just economic system. Second, a society also has a duty to reform the economy by adopting broad measures that are most effective or (the qualification) at least moderately effective in bringing about a just basic structure, so long as these measures eventually improve the position of the least advantaged of current generations. Of course, not all local measures that maximally benefit the LAG now or in the near future are also broad measures that most effectively bring about the “most effectively designed system” of justice, and vice versa. Hence, the qualification allows for local maximin measures that are less than maximally effective in promoting a just society if they significantly improve the position of the least advantaged now (e.g., immediate income tax credits for the LAG), but still it excludes local maximin measures that considerably delay or make a just basic structure much less likely in the future. Thus, ruled out are further tax cuts for the wealthy that maximally benefit the LAG in the short run, but solidify substantial inequalities and impede a transition to POD. Third, if measures that both maximally benefit the least advantaged in the short run and also promote a future POD would substantially increase existing inequalities, a society may choose alternative measures causing less inequality that improve the position of the less advantaged (so long as society remains on the upwardly rising part of the OP curve, Rawls 2001: 64), and still are consistent with advancing just institutions in the future. This might be justifiable on grounds of maintaining the sense of self-​respect of the less advantaged—​substantial increases in economic inequality must always be carefully assessed for these reasons. But even when measures have no detrimental effects on self-​ respect, still a society is not required by justice to adopt local maximizing measures that substantially increase inequalities when there are alternatives causing less inequality

36   Samuel Freeman that also benefit the LAG. The latter measures causing less inequality may even be required by justice if more likely to advance a just basic structure. Fourth, the most difficult case: What about broad reforms that are necessary to eventually institute a just basic structure (POD) but that worsen temporarily the condition of the least advantaged? For example, the difference principle requires at some point a (gradual) restructuring of the property system with redistribution of economic wealth so all (working) citizens own a fair share (e.g., of shares in firms). This likely will result in economic downturn, increased unemployment, and considerable loss of market value of shares (e.g., similar to the 50% decline of the US stock market from October 2007 to March 2009 (Blinder: 2013)). Such measures are permissible, if not required, when their short-​term effects on the less advantaged are not too severe and longlasting; otherwise more gradual measures should be devised. Moreover, such measures must eventually benefit existing members of less advantaged classes, and they must reduce existing inequality, hence they involve sacrifices to the most advantaged as well. This “one step backwards—​two steps forward” approach seems inevitable if serious economic reform of an unjust economic system is to be possible. It is a complex matter requiring far more discussion than can be given here.

1.8  Property-​O wning Democracy vs the Welfare State According to the difference principle, the least advantaged is the class whose distributive shares, or “index” of primary social goods, is the least. In addition to having less income and wealth, the least advantaged have the fewest social and economic powers and responsibilities. Rawls says that, for the sake of simplicity in making interpersonal comparisons we should rely on representative persons’ share of income and wealth, since one’s share of powers, responsibilities, and social bases of self-respect normally track one’s share of income and wealth. He assumes the LAG is the “class of unskilled laborers” (Rawls 1999a: 67). A question Rawls does not adequately address is, how is the index that determines representative persons’ share of primary goods to be composed. How much (what percentage) is assigned to social and economic powers and responsibilities and self-​respect compared with income and wealth? (Rawls suggestively says that self-​respect is perhaps the most important primary social good [Rawls 1999a: 386]). Normally this issue is of little consequence, assuming income and wealth are a reliable indicator of one’s relative position with respect to other primary social goods (Rawls 2001: 65). But it may be important in assessing Rawls’s arguments for property-​owning democracy versus welfare-​ state capitalism. He argues that one of the decisive reasons in favor of POD is that it affirms the self-​respect of citizens who are least advantaged, since they own a share of capital or total productive wealth in society (Rawls 2001: §42). I assume too that part of

Rawls On Distributive Justice    37 the argument for property-​owning democracy relies on all citizens also having at least some share of economic rights, powers, and responsibilities in their workplace (such as co-​determination rights, rights to alter daily tasks to reduce monotony, move around freely in the workplace, take periodic breaks, etc.) (Freeman 2007: 133–​6). One place these questions are especially relevant is the hypothetical case of welfarestate capitalism (WSC) where the least advantaged have no economic wealth, powers, or responsibilities, but the social minimum of income and wealth to meet basic needs may be generous so that it potentially exceeds the income and wealth of the least advantaged in a property-​owning democracy where the least advantaged own capital (e.g., shares of firms) and have some degree of economic powers. Though the index of primary goods is greater in POD, those in WSC have, let’s assume, greater income. What is the argument for POD in this case? Why shouldn’t there be a permissible trade-​off of economic wealth and powers for a greater income for the LAG? The best way to address this problem is by examining the reasons Rawls relies upon to argue for POD and against WSC. He primarily relies upon economic reciprocity and the primary good of self-​respect, the same reasons relied upon to argue for the difference principle. Rawls argues that one of the major reasons that POD satisfies the difference principle is that, unlike WSC, it maintains reciprocity “at the deepest level.” Such democratic reciprocity exists when increasing benefits are shared by all members of society and maximally benefit the least advantaged (hence society is on the rising slope of the OP curve, Fig. 1.1). This affirms the self-​respect of the least advantaged, since they know that their interests are not being sacrificed for the sake of greater gains to the more advantaged or the general welfare, as in WSC. A problem with WSC is that, once basic needs are met, there are no guaranteed further gains for the least advantaged even though they continue to contribute to the social product. The LAG are thus put in a position where they justifiably believe they are left behind by society and are not fully members of it. This is partly because in WSC the social minimum is a conventional assessment of basic needs, a vague idea inevitably subject to disagreement, giving rise to political conflict between the more advantaged and least advantaged members of society. The least advantaged are a minority, and in the absence of a public understanding and agreement that democratic reciprocity requires that their position be maximized, arguments that the social minimum should be increased or sustained are likely to be met with the reply that the least advantaged are a drag on society’s productivity, do not deserve the share they now have, and that if anything it should be reduced. (Such disputes are familiar in the US, where one party now argues that Food Stamps and other benefits for the poor, 40% of whom are children, should be decreased if not eliminated, allegedly in order to maintain the self-​respect of the poor.) But what if the least advantaged in a POD are willing to sacrifice the economic powers and modest share of wealth they enjoy for the sake of a marginally greater social minimum of income? Suppose their share of total income in wages and dividends in a POD is $40,000, while were they to agree to surrender their shares of stock and economic powers to their employers and other entrepreneurs, their wages plus income supplements eventually would rise 10% to $44,000. Why shouldn’t the least advantaged be allowed to make this decision?

38   Samuel Freeman One reason not to sacrifice the wealth and economic powers of the least advantaged is that it reinstitutes the separation and conflict between economic classes of capitalists vs labor. There may be income classes in a POD too, but unlike capitalism, there is not a division of social classes between owners who control most real and liquid capital, exercise predominant economic power, and whose interests conflict with the class of productive laborers who have neither wealth nor economic control with its attendant powers. The hypothetical assumes that in such a class-​divided society of capitalists versus labor, less advantaged workers can maintain their economic agency and sense of self-​respect, in spite of the fact that they have been deprived of any economic powers or any share of capital wealth. Why wouldn’t giving up their share of wealth and economic powers eventually have similar adverse effects on the sense of self-​respect of the least advantaged as would property qualifications on the franchise, where the least advantaged are deprived of political rights? Many of the poor today are so politically despondent that they would sell their right to vote, but this does not justify such alienability. Similarly, it would not be justified if the least advantaged were given greater income supplements and in return alienated their rights to fair equal opportunities to develop their capacities. Rawls assumes that social equality and self-​respect of citizens depends upon a fair and widespread distribution of all primary goods to all social classes. This includes, not just equal basic liberties, fair opportunities, and income, but also ownership and some degree of control of economic wealth, and exercise of economic powers and responsibilities in the workplace (Freeman 2013). WSC effectively bars the least advantaged from access to a share of economic wealth, powers, and responsibilities, and this undermines social equality and their sense of self-​respect. The argument for property-​owning democracy over WSC is in this regard an extension of the fundamental significance assigned to social equality, democratic reciprocity, and the social bases of self-​respect that informs Rawls’s arguments for equality of political and other basic liberties, fair equal opportunities, and the difference principle.

1.9 Conclusion The difference principle is but one part, even if the most significant, of Rawls’s account of distributive justice. In addition to distributive measures that guarantee the fair value of the political liberties and fair equality of opportunities, there is the right to healthcare guaranteed by FEO, the Just Savings Principle applying to future generations, the duty of assistance owed to burdened peoples in the Law of Peoples, the duty of assistance owed to those with significant disabilities who are unable to make economic contributions, and finally, presupposed by both principles of justice, the fundamental requirement of basic justice that society meet the basic needs of all citizens so they can effectively exercise their basic rights and liberties. All of these requirements have priority over the difference principle, and their claims must be satisfied before the social minimum

Rawls On Distributive Justice    39 required by the difference principle can be fully settled. A fuller account of Rawls’s complex conception of distributive justice would include discussion of them.

Acknowledgments I am grateful for their helpful comments to T. M. Scanlon, Joshua Cohen, Samuel Scheffler, Philippe van Parijs, Thomas Christiano, Christine Korsgaard, Arthur Kuflick, Kok Chor Tan, Justin Bernstein, Pierce Randall, and also to members of the Philosophy Department and the Law School at the University of Arizona, the Philosophy Department at University of Vermont, the Global Justice Program at Yale University, and to participants at the conference on the Heritage of Rawls at the Ignatium Academy in Krakow, Poland, June 2014. I am especially grateful to Serena Olsaretti for her extensive comments and helpful suggestions.

References Blinder, A. (2013). After the Music Stopped: The Financial Crisis, the Response, and the Work Ahead. New York: Penguin Press. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Cohen, J. (2001). “Taking People as They Are.” Philosophy and Public Affairs 30: 363–​86. Dworkin, R. (2002). Sovereign Virtue:  The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Freeman, S. (2007). Rawls. New York: Routledge. Freeman, S. (2013). “Property Owning Democracy and the Difference Principle.” Analyse & Kritik 9(1): 9–​36. Freeman, S. (2014). “The Basic Structure of Society as the Primary Subject of Justice,” in J. Mandel and D. Reidy (eds) A Companion to Rawls. Oxford: Blackwell, pp. 88–​111. Friedman, M. (1962). Capitalism and Freedom. Chicago, IL: University of Chicago Press. Hayek, F. (1979). Law, Liberty, and Legislation, Vol. 3:  The Political Order of A  Free People. Chicago, IL: University of Chicago Press. Hume, D. (1975). Enquiries concerning Human Understanding and concerning the Principles of Morals, third edn. Oxford: Oxford University Press (originally published in 1777). Lareau, A. (2011). Unequal Childhoods, revised edn. Berkeley, CA: UC Press. Murphy, L. (1998). “Institutions and the Demands of Justice.” Philosophy & Public Affairs 27(4): 251–​91. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. Parfit, D. (1991). Equality or Priority? The Lindley Lecture. University of Kansas. Parfit, D. (1997). “Equality and Priority.” Ratio 10(3): 202–​21 (shorter version of Parfit 1991). Piketty, T. (2013). Capital in the Twenty-​ First Century. Cambridge, MA:  Harvard University Press. Rawls, J. (1999a). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press (originally published in 1971). Rawls, J. (1999b). Collected Papers. Edited by Samuel Freeman. Cambridge, MA:  Harvard University Press. Rawls, J. (1999c). The Law of Peoples. Cambridge, MA: Harvard University Press.

40   Samuel Freeman Rawls, J. (2001). Justice as Fairness:  A Restatement. Edited by Erin Kelly. Cambridge, MA: Harvard University Press. Rawls, J. (2005). Political Liberalism, expanded edn. New York: Columbia University Press. Roemer, J. (1994). A Future for Socialism. Cambridge, MA: Harvard University Press. Scanlon, T. M. (2018). Why does Inequality Matter? Oxford, UK: Oxford University Press. Sen, A. (2009). The Idea of Justice. Cambridge, MA: Harvard University Press. Van Parijs, P. (2003). “Difference Principles,” in Samuel Freeman (ed.) The Cambridge Companion to Rawls. Cambridge: Cambridge University Press, pp. 200–​40.

Chapter 2

Dworkin and Lu c k Egalitaria ni sm A Comparison Richard Arneson

In 1981 Ronald Dworkin published two magisterial essays on “What is Equality?” that initiated a trend in political philosophy that eventually came to be called “luck egalitarianism.” The name of the trend refers to a distinction that Dworkin drew between option luck and brute luck. The latter is good or bad fortune that simply falls on a person in ways that are entirely beyond that person’s power to control or—​in another formulation of the idea—​in ways that are not alterable by any reasonable action the person might have chosen and pursued. Option luck is good or bad fortune that lies within the individual’s power to control or is alterable by some reasonable course of action the person might have chosen and pursued. The distinction between the two kinds of luck is one that varies by degree. Roughly speaking, luck egalitarianism holds that justice requires that people be made equal in the benefits and burdens that accrue to them via brute luck but not in the benefits and burdens that accrue to them via option luck. Several political theorists including G. A. Cohen (1989), Thomas Nagel (1991), John Roemer (1993 and 1998), and Larry Temkin (1993) have at one stage or another of their careers embraced some version of luck egalitarianism. Dworkin was credited with articulating a powerful defense of broadly egalitarian views of social justice in the face of a challenge that had been issued by Robert Nozick, an adherent of right-​wing Lockean libertarianism (Nozick 1974). Nozick presented a thought experiment. Suppose that your favorite theory of distributive justice is fully implemented; for simplicity, let’s suppose this is flat equality—everyone should have the same. Starting from that ex hypothesi fair distribution, people will make choices and take actions as they go about their lives that will, barring a miracle, result in a later unequal distribution. But if people are acting fairly, violating no one’s rights, imposing no wrongful harm on others, voluntarily choosing to pursue their own ends, surely the later unequal distribution must be fair. But if distributive justice, as we initially assumed,

42   Richard Arneson requires equality, then the later distribution is unfair. Hence there is a contradiction. The distribution can’t be both fair and unfair. Nozick urges that to avoid the contradiction we should give up our initial assumption that distributive justice requires an equal distribution, and since nothing in the argument turns on the particular initial choice of fair distribution pattern, the conclusion is that no pattern of distribution—​neither equality nor any other pattern—​is a distributive justice requirement. (A patterned distributive justice requirement demands that the distribution of goods across persons must conform to a specified shape. Everyone having the same is one. Everyone having a share that is proportional to her moral deservingness would be another.) More broadly, a common and plausible right-​wing objection to left-​wing theories of justice is that requiring that everyone have the same or anything remotely close to that fails to pay adequate heed to the importance of personal responsibility in any sensible account of what we owe one another. People can become badly off, fall into misfortune, in any of a wide variety of ways, and how misfortune comes to attach to you plays a big role in determining the extent to which other people might be morally bound to come to your rescue. You might have fallen into misfortune by reckless, or imprudent, or self-​ abnegating, or vicious, or willfully self-​destructive conduct, so that your current plight is reasonably held to be your responsibility, not the responsibility of others to fix at cost to themselves. The same goes for good fortune: If you worked hard and long and made prudent and savvy choices and scrimped and saved, or took a gamble when the odds were favorable, your personal responsibility for your good fortune stands in the way of the egalitarian demand that we should take from you in order to help the needy. Dworkin’s answer to the “What is Equality?” question he poses can be read as a reply to Nozick. Dworkin suggests that insofar as we care about equality in the distribution of privately owned goods, we should favor an initial distribution of material resources that is the equilibrium outcome of an auction in which all such resources are up for bidding and all have equal resources for bidding, with the auction supplemented by two hypothetical insurance markets, one for marketable talent and one for handicap. In the former, individuals who do not know the market value of the personal talents they possess have the opportunity to purchase insurance against having low marketable talent. In the market for handicap insurance, individuals who know the incidence of various disabilities that might befall people and the remedies available for such disabilities (e.g., eyeglasses for nearsightedness) but do not know their individual probability of having any disabilities, have the opportunity to purchase insurance against finding oneself with a disability. The insurance market is conducted so that the books balance; what people would purchase yields an amount of funds that is the total payout to which individuals making these purchasing decisions would be entitled. After an initial distribution of resources, individuals then interact as they choose in a framework of market exchange, with criminal law, tort law, and contract law set so that people do not harm others in ways these laws forbid. Individuals pursue their projects in this fair framework, and whatever pattern of distribution emerges over time through people’s lifetime from this fair starting point is morally acceptable according to this “equality of resources” proposal. So Nozick’s concern that preserving any fair

Dworkin and Luck Egalitarianism    43 distributional pattern of resource holdings must squelch individual liberty is met. Apart from the once-​a-​generation distribution that sets the fair starting point, no pattern is imposed. Nozick’s concern that egalitarian justice doctrines are insensitive to personal responsibility is also addressed. Under Dworkin’s equality of resources scheme, people who start on a fair footing, choose well or badly, and have good or bad luck in the gambles that inevitably accompany decisions to interact with others when outcomes are not known in advance with certainty, have no complaint of injustice that calls out for compensation or further redistribution when outcomes are not to their liking or as they had hoped. (A complication is that the scheme cannot literally be put in place, because the hypothetical insurance markets cannot actually be administered, and generations do not succeed each other at discrete points of time, but come into being by continuous births and deaths, so Dworkin proposes that we ought to institute arrangements including a tax and transfer scheme that mimics as closely as possible the result of the hypothetical equality of resources procedure for the population over time.) The views on social justice that Dworkin appeared to champion attracted criticisms. Meanwhile, Dworkin refined his views. In 2000 he published Sovereign Virtue:  The Theory and Practice of Equality, which offered a broad interpretation of social justice requirements. Its starting point is that governments (unlike private individuals) have a duty to treat each of the citizens under its jurisdiction with equal concern and respect. This requires action from the complex conviction that it is important that the life of each member of the political community should go well, and equally important that each member’s life goes well, and yet that each individual person has a nondelegable responsibility to shape her life in her own way. On the topic of distributive justice, the book reprints the two 1981 essays and adds chapters that explore what the theory requires by way of practical political measures to make advances in fair treatment of individuals in a society that has not implemented equality of resources. The criticisms that the luck egalitarian trend of thought attracted were various (see especially Anderson 1999; Scheffler 2003a and 2005). Some object to the “luckism” or personal responsibility component of the view (Fleurbaey 1995 and 2008). Others object to its egalitarianism (Miller 2015; Sher 2014). Against the residual egalitarianism, Lockean libertarians and other conservatives can protest that there should be no moral presumption in favor of equality of condition, not even one qualified by personal responsibility. People have some moral rights, which should be honored, but nothing like a claim to the same property or same condition as others enjoy. The rejection of the idea that equality as morally desirable in itself is also shared by other moral perspectives: sufficientarians hold that justice demands that everyone have a good enough condition, not an equal condition, and prioritarians propose that it is morally more important to provide benefits for people, the worse off in absolute terms they otherwise would be. Another line of criticism objects that insistence on equality of condition can require policies that shrink the pie and make us all worse off over the long run, and such policies cannot be sensible or just. Another view is that the right interpretation of the personal responsibility idea renders any version of egalitarianism otiose: distributive justice requires that the good fortune that people obtain in

44   Richard Arneson life should correspond to their deservingness: the virtuous saints should do better than the nonvirtuous sinners, and the greater one’s virtue, the more one should outstrip the nonvirtuous (Kagan 1998). Against the luckism or personal responsibility component, some hold that luck egalitarian doctrines support wrongfully moralistic, ungenerous policies toward those whose own choices and behavior have landed them in trouble. People in misery deserve help and surely often have a moral right to be helped, even if they could have avoided their present plight by appropriately prudent conduct on their part. The thought that luck egalitarianism preaches wrongful stinginess toward those who have come to be in peril through their own fault or choice sometimes combines with endorsement of sufficientarianism:  each person should be enabled to have enough, come what may, throughout life, whatever the source might be of threats to the person’s maintenance of a good-​enough condition. Still others hold that it tends to be counterproductive to uphold a view of justice that requires classifying people into sheep and goats, sorting the deserving from the undeserving poor and the deserving from the undeserving rich for purposes of deciding what treatment people are owed. It would be better to avoid these classification exercises unless they happen to be cost-​effective instruments for advancing entirely distinct and separate moral goals. The treatment of personal responsibility in Dworkin’s doctrine and in those of some luck egalitarians following his lead involves crucially the idea that the standards of distributive justice should be calibrated in terms of resource shares, not the life outcome people reach by their uses of these resource shares. Once a fair distribution of resources, liberties, and opportunities is in place, and then sustained in appropriate ways, individuals themselves bear responsibility for how they use their fair shares and what happens to them as a result. On this view, justice might uphold rights to the pursuit of happiness but not to any degree of success in this pursuit. Whether resources or welfare is the proper measure of people’s condition is partly an intramural dispute among luck egalitarians, but one that reflects large disputes about what we owe one another in the fundamental ways that should register in social justice requirements. So far I have treated Dworkin as a founder of luck egalitarianism and one of its leading figures. But Dworkin’s fascinating response to critics who regard his work as a canonical member of the luck egalitarian family of views is that his views do not belong within this family of doctrines as the critics characterize it. Responding to a critic, he denies he adheres to the core luck egalitarian idea that “inequalities deriving from unchosen features of people’s circumstances are unjust.” Instead, he holds that his core idea is that distributive justice requires making people’s circumstances “equal under some appropriate version of the envy test.” He adds that this test properly understood requires “that people be made equal, so far as this is possible, in their opportunity to insure or provide against bad luck before it has occurred, or, if that is not possible, that people be awarded the compensation it is likely they would have insured to have if they had had that opportunity” (Dworkin 2002 and 2003; Scheffler 2003a and 2003b). To some of his followers (including the author of this chapter), Dworkin’s protest that he is not a luck egalitarian sounded as startling as if the Roman Catholic Pope had

Dworkin and Luck Egalitarianism    45 responded to objections against central Church doctrines by announcing that he is not actually a Roman Catholic. However, we should not be too concerned about what labels to attach to a candidate conception of social justice. This chapter eventually addresses these questions: Are the criticisms of what is called “luck egalitarianism” sound? Do the main criticisms of luck egalitarianism, whether good or bad, apply to Dworkin’s social justice views fairly interpreted? If Dworkin’s doctrines differ from the main run of luck egalitarianisms, are Dworkin’s distinctive doctrines morally acceptable or unacceptable?

2.1  Resources or Welfare as the Distributive Justice Measure? Dworkin’s conception of distributive equality has two main building blocks. One is that the metric for equality should be resources, not welfare (for further articulation of resource-​oriented distributive justice views that eschew interpersonal comparisons of welfare, see Fleurbaey 2008). What matters from the standpoint of justice is whether people have the same resources, not whether they enjoy the same welfare. Among resource-​oriented theorists Dworkin’s position is notable for its expansive view of what kinds of things count as resources for individuals. The other main building block is that people should be made equal, so far as this is possible, in their prospects, not in their outcomes. We should be seeking ex ante not ex post equality. This section explores the rationales that Dworkin offers for equality of resources over equality of welfare. Section 2.2 of this chapter considers the issue of ex ante versus ex post equality. The idea of equality of welfare is that justice demands that everyone ends up with the same welfare so far as this is feasible. “Welfare” is a placeholder for whatever in itself makes a person’s life go better for that very person. Dworkin suggests that when we seek to pin down the appropriate conception of welfare to serve as distributive justice metric, we find that none is appealing. A related argument is that people disagree about what makes people’s lives go better, what is worthwhile and choiceworthy in human life, and any attempt by government to improve individuals’ lives according to one or another inevitably controversial specification of welfare will be aiming to promote what some affected people will see as irrelevant to their interests or even as hindering their pursuit of their interests. The government, even though claiming to promote their advantage, would be treating them disrespectfully. Dworkin considers three broad types of notions of welfare. One is preference satisfaction or life aim satisfaction. A second is hedonism broadly construed. According to this latter notion, the good for a person consists in a desirable mental state or quality of conscious experience. A third type is objective. The idea is that there is a fact of the matter as to whether something you get or achieve makes you per se better off. Whether something you get or achieve increases your welfare level is independent of your subjective attitudes or beliefs regarding your getting or achieving that thing. So if achievement,

46   Richard Arneson knowledge, and enjoyment are objectively good, gaining some of any of these items improves your life. (The reader might respond that the distinctions are crosscutting: one might hold that getting pleasure is in itself objectively good for one, independently of one’s subjective attitudes toward this state of affairs. I set this problem to the side.) Dworkin considers and rejects unrestricted preference or desire satisfaction views. People’s desires can range widely, and take impersonal objects, so that their satisfaction does not intuitively make the desirer’s life go any better. So consider the restricted form of this proposal, which identifies welfare with satisfaction of preferences regarding one’s own life and circumstances. Dworkin further distinguishes relative and overall success. One can achieve success in fulfilling one’s preferences but regard this fulfillment as more or less important. Relative success is fulfilling a greater rather than a smaller proportion of one’s preferences, and overall success has to do with a person’s overall judgment of how successful his life has been given his relative success. With this distinction in mind, we can see that overall personal success is the better candidate for the idea of welfare in which we should all be made equal. But whether persons have overall success or not depends on their judgments of the value of their life aims, and these can vary for idiosyncratic or arbitrary reasons. Jack and Jill can be leading lives that look to an outside observer to be identical in any respect that might be relevant to their welfare, but Jack has optimistic attitudes and believes his life is rich and valuable, while Jill has a pessimistic philosophy of life and believes her life to be pretty much worthless. Trying to make people equal in their overall judgments of how well their lives are going, in light of this example, looks to be a fool’s errand. Comment:  the Jack-and-Jill example as developed by Dworkin undermines the overall success version of a preference satisfaction conception of welfare in a way that supports the idea that there is an objective fact of the matter as to how well off or badly off a person is in given circumstances. That is, there is an objectively correct account of individual welfare, which determines how well off a person is independently of her subjective opinions on this matter or her attitudes toward her circumstances. If Jack and Jill have lives that are the same in work, achievement, friendship, love, enjoyment, and other components of welfare, they are equally well off, regardless of whether either one subscribes to odd beliefs or philosophies of life and happens to have some inflated or depressive opinion about how well his or her life is going. So it will be important to see what Dworkin has to say about objective conceptions. (Not much, it turns out.) (It should be noted that there is another possible ground available to Dworkin to support the claim that distributive justice demands no compensation from Jill to Jack or vice versa. In the example neither would prefer to have the resource bundle possessed by the other. The no-​envy test is satisfied. But in the argument under review the issue is this: is there a plausible conception of welfare that could be the standard for a viable principle of distributive equality? The concern being voiced is that Dworkin’s arguments do not rule out objective list accounts as a plausible conception of welfare for this purpose.) Dworkin then turns to broadly hedonistic or quality-​of-​experience conceptions of welfare. Against them he makes the reasonable objection that feeling good or, more broadly, having desirable experiences is very plausibly a component of living well but

Dworkin and Luck Egalitarianism    47 not plausibly regarded as all of it. Individuals reasonably seek, for example, not merely to have the experience as though having a friend but actually to have a friend, and not merely the experience of writing a good novel or surpassing some athletic achievement threshold but actual success in these endeavors. It might seem that once one acknowledges that there are different and opposed conceptions of human good (human welfare) and that competent adult individuals disagree as to which one, if any, is correct, for a political community to embrace any conception of good and seek to bring it about that its members are equally well off according to that inevitably controversial conception is to treat its members with disrespect (Rawls 1982; see also Ripstein 2007; also Fleurbaey 2008). But this issue is delicate. There is a danger here of just begging the question against the view that at the end of the day questions of value admit of right and wrong answers. Some things are genuinely valuable, some are not. If that is true, then it is not at all obvious that a political community treats its members wrongly when it seeks to bring it about that people have the opportunity to flourish according to that conception, even when that differs from some individuals’ own views. For a comparison, it is not at all obvious that a political community does wrong in trying to bring it about that people treat each other fairly according to the correct or most reasonable account of fairness, even if some of their members conscientiously adhere to some mistaken conception of fairness. Dworkin thinks he has a master argument that defeats all types of equality of welfare views at one stroke. When do I plausibly have a complaint against society on the ground that my opportunity for a good life is meager? I cannot complain just on the ground that my fulfillment is not as great as I could imagine or wish. To have a complaint I must be able to voice reasonable regret about my opportunities. Dworkin suggests that the only reasonable baseline for this necessary reasonable regret is that I have had less than an equal or fair share of the resources that others enjoy and for this reason my opportunities for welfare are not what they should be. But this move gives the game away. If any attempt to specify a welfarist notion of a person’s equal share of resources already must presuppose an independent notion of fair or equal shares of resources, then the project of identifying a plausible conception of equality of welfare is doomed. Far from being decisive, this objection is question-​begging. An equality of welfare view must include a measure that tells us when some are worse off than others and how large the shortfall is between better off and worse off. No notion of “reasonable regret” tied to resource shares is needed. We need a measure of welfare, not one of “reasonable regret.” The same goes for a doctrine of equal opportunity for welfare. According to views in the category, people have (roughly, we need make no false pretenses of fine degrees of measurement) equal opportunity for welfare when their circumstances are such that if they conduct their lives with the same rational prudence their welfare prospects (their expected lifetime welfare) are the same. If we take on board the idea that someone’s bad brute luck may include a poor genetic endowment and poor early socialization that make it more difficult for him than for others to figure out and follow a rationally prudent life course, we might adjust the formulation of welfarist equal opportunity: individuals have equal opportunity for welfare when it is the case that if each

48   Richard Arneson conducted her life as prudently as it would be reasonable to expect given her prudential capacities, each would have the same lifetime welfare prospects. Dworkin presses yet another objection against equality of welfare that has more substance, and cuts through to deeper issues. If we accepted the ideal of equality of welfare, we would accept that a person with expensive tastes, who needs the best oysters to get the pleasure others readily obtain from canned tuna fish, would be entitled to redistributive transfers so that he stays equal in welfare to the level of other people. Balking at this implication of the ideal, we should see that we do not, and should not, accept the ideal. Here is a parallel story about equality of resources. Suppose that a person squanders whatever resources are bestowed on her. She starts with equal shares of resources along with all members of her community, but the resources slip through her fingers like so much water. To stay at a level of resources that is continuously equal to what others have, she needs continuous infusions of more and more resources. Rejecting this implication of the equality of resources ideal, we should see that we do not and should not, accept the ideal. But, of course, Dworkin never embraces that flat equality of resources ideal that this story criticizes. He embraces something closer to equal opportunity for resources. Having had a fair initial share of resources, the person described in the previous paragraph bears responsibility for how he use them, and does not merit compensation for future resource deficits he suffers that lay within his power to control. The advocate of “equality of welfare” can take a similar tack (Arneson 1989 and 2004; Cohen 1989 and 2004). We should distinguish expensive tastes that simply fall on a person and those that are beyond her power to control. Perhaps her taste buds are substandard, so ordinary food and water are unpleasant for her to eat, though pleasant for everyone else. If this is the source of her shortfall in welfare stemming from expensive tastes, it merits equalizing compensation. In contrast, if a person negligently or recklessly fall into a lifestyle that emulates that of the idle rich and results in her being afflicted with expensive tastes, her faulty conduct that predictably brings about her expensive tastes undercuts her demand that justice requires compensation for them. The same goes if a person deliberately cultivates expensive tastes for no good reason. If a person enjoys equal opportunity for welfare, and then squanders her opportunities in ways for which she is reasonably held responsible, equality of opportunity for welfare does not identify her as entitled to equalizing compensation. Dworkin raises interesting doubts about this response in defense of a welfarist distributive justice metric. One is that a sensible person who develops expensive tastes supposes that doing so will make her life go better even if she becomes pinched for resources. If society by its welfare metric judges she has become worse off, this judgment flies in the face of the person’s own appraisal of her situation after her deliberate cultivation of expensive tastes. It would be inappropriate for a person to demand compensation for a condition she embraces and views as enhancing her life, so it would be inappropriate for society to offer or impose compensation. Compensation for expensive tastes violates a justice constraint: there should be no compensation for an individual’s condition if the person regards that condition as an enhancement, not an affliction.

Dworkin and Luck Egalitarianism    49 Dworkin adds that even when an expensive taste is voluntarily cultivated, this action will proceed from underlying judgments and other features of the individual’s personality that are not voluntary at all. So the division between voluntarily cultivated and not voluntarily cultivated expensive tastes is rather superficial, and cannot bear the weight the advocate of equal opportunity seeks to place on it. Dworkin proposes a different way of distinguishing tastes, preferences, and ambitions for which the individual should bear responsibility and the rest: if an individual is glad not sad to have a preference, that preference falls on the side of her choices and ambitions for which she bears responsibility, rather than on the side of her unchosen circumstances. The advocate of a welfarist distributive justice metric need not abandon her ground in the face of these worries. The glad-​to-​have test is flawed. That I choose certain values and form certain preferences and am glad to have them may simply reflect my poor native endowment of choice-​making and value-​forming ability. Even if I cannot coherently ask to be compensated for what I take to be a benefit, other people, charged with determining whether I am badly off in a way that triggers requirements that people offer me aid, can coherently judge that some of my preferences and values and ambitions are warped through no fault of my own, in such a way that I am entitled to compensation for the worsening of my life they have induced (for an opposed view, see Williams 2002). The claim that the line between deliberate or careless cultivation of expensive tastes and other such cultivation is superficial raises issues about free will and determinism and responsibility that are entirely orthogonal to the dispute between the resourcist and welfarist versions of equal opportunity. If hard determinism is the truth, then responsibility will play at most an instrumental role in distributive justice theory. If any version of free will libertarianism or soft determinism is the truth, then there is room for responsibility to play more than a merely instrumental role in distributive justice theory (so that the theory might judge, for example, that it is intrinsically morally better when saints rather than sinners get more of scarce benefits). The thought that the problem of determinism poses some special difficulty for welfarist as opposed to resourcist views rests on confusion (but see Scheffler 2005 for an opposed view; also Hurley 2003). Finally, in trying to figure out whether welfare or resources should be the metric for assessing people’s condition for purposes of deciding who owes what to whom as a matter of distributive justice, we should separate the issue of welfare as the metric and the issue of the appropriateness of equality as the distributive principle. If some people, through no fault or choice of their own, suffer from grievous afflictions that make them very poor transformers of resources into welfare, then equality of welfare and equal opportunity for welfare as well will keep recommending resource transfers from other people to these grievously afflicted individuals, even if the resource transfers greatly diminish the welfare of those who give up resources, and just barely improve the welfare of the grievously afflicted. These people then become a basin of attraction for resources, and the level of transfer supposedly required by justice looks to common sense to be excessive. But this intuition should perhaps persuade us that equality is not the proper and fair distributive principle, and leave intact our conviction that the proper measure of how badly off a person is for purposes of determining distributive justice obligations

50   Richard Arneson is how badly off the person really is according to the correct objective account of what makes someone’s life go better or worse for her. (Knight 2009 and Segall 2013 defend equal opportunity views.) Dworkin holds that people should be made equal in resources, not welfare. So the negative claim is that, in assessing a distribution of resources, we should not look beyond it to the welfare outcomes that individuals gain using those resources. Each individual has a nondelegable responsibility for how his or her own life goes, for better or worse, against a backdrop of a fair initial distribution of resources and a fair framework in which individuals can freely decide how to cooperate together and how to transact with others. Dworkin’s view about equality of resources also includes a positive view about how to conceive of resources. Resources for an individual are means to achieving that person’s goals. The value of a resource that is assigned to one individual is what others would pay for it—​the opportunity cost of not assigning it to another individual when the scheme for registering willingness to pay is fair. So, if a bundle of resources is distributed to a group of individuals by giving each equal bidding power (equal money to use in the auction), and conducting an auction in which no bids are final until no one wants to change any of her bids given the bids others are making, then if the auction concludes, and resources are given to the highest bidder, the distribution satisfies equality of resources as conceived by Dworkin. (In this way an ideal market procedure defines the equality of resources ideal.)

2.2  Ex Ante versus Ex Post Distributive Principles Imagine that a group of people is going to establish a fair initial distribution of resources. Following this distribution, people will interact in circumstances such that the outcomes of their choices will not generally be known at the time of choice. Dworkin holds that equality in the relevant sense obtains when no one envies (prefers to have) the circumstances of any other individual. If a pile of divisible resources is distributed according to an ideal equal auction, then no one envies the bundle of resources anyone else possesses, so equality prevails. Now suppose people are free to choose a course of action so long as they do not harm others, the outcome of anyone’s actions being dependent on the actions of others via market processes. Each knows the choices of others but the outcome of one’s own choices is not known in advance; one knows only that one of several outcomes might obtain, and perhaps the probability that this, that, or the other outcome will ensue given one chooses a certain course of action. An ex ante principle of distributive justice assesses people’s situations prior to the resolution of the uncertainty in the outcomes of their choices. An ex post principle assesses people’s situations after uncertainty is resolved and the outcomes for individuals are known.

Dworkin and Luck Egalitarianism    51 The conviction that shapes Dworkin’s view is that provided the initial set-​up is fair, the relevant perspective of assessment for social justice is ex ante rather than ex post. For example, people may have different proclivities for risk-​taking. Given a fair distribution of resources, some may follow play-​it-​safe strategy, and some may accept the chance to win big accompanied by a chance of losing big. Whatever their preferences, in the face of an uncertain future, each person has a responsibility to choose a course of action and accept the consequences that ensue. Choosing a course of action can include purchasing insurance against undesired future contingencies or taking actions that provide the equivalent of insurance. Having made a choice from a fair array of options, a competent agent must take responsibility for the advantages or disadvantages that accrue to her from that choice, in the sense that she is not entitled to demand that others are obligated to make good her losses if the outcome is bad for her. It is hard to adjudicate the dispute between advocates of ex post and ex ante distributive principles (Fleurbaey 2008 versus Dworkin 2002 and 2011). From the ex post perspective, the choices that people make under uncertainty are just choices they make when ignorant of information that is relevant for sound choice. If the test of fairness is some version of the idea that an arrangement is fair when no one prefers the situation of another, then ex ante fairness makes people’s preferences regarding states of the world that will not occur relevant to the determination that they have been fairly treated, and the advocate of taking an ex post perspective will deny that this is appropriate. One might think that the advocate of ex post fairness must be embracing principles that give people perverse incentives. If it is known in advance that justice requires compensating people who gamble and lose for their big losses, I may have the incentive to take a very bad gamble with a small chance of a spectacularly good outcome and a large chance of a very bad outcome, since I  will not have to bear the adverse consequences if I lose. But this thought is a mistake. Ex post principles might reward or punish people depending on the quality of their choices regardless of the outcomes that ensue, so that a prudent individual whose good bet turns sour, an altruist who sacrifices her prospects to get great gains for others, and an imprudent person who gambles recklessly, will be treated differently if they all end up with the same poor outcome. If the ex post advocate happens to favor equal outcomes for the equally virtuous (see Eyal 2007; more broadly, Kagan 1998 and 2012), then she will favor ex post equalization across persons who choose equally prudently and end up better and worse off, and across persons who equally opt for virtuous self-​sacrifice and end up better and worse off. Recall that the initial luck egalitarian view is that it is morally bad if some are worse off than others through no fault or choice of their own. “Fault or choice” suggests two different norms. “Choice” asserts an ex ante perspective: people should be free to act as they choose by their own lights and absorb the consequences that ensue, whether the outcomes are certain or uncertain (Sher 2014). “Fault” points toward the idea that we ought to bring it about that people get what they deserve, or more weakly that deservingness partly determines rightful distributive shares. This idea consorts naturally with an ex post perspective.

52   Richard Arneson Deservingness might displace egalitarianism altogether. A  noncomparative ideal might hold that for each degree of virtue one might achieve, there is a particular well-​ being level that is fitting, and each person ought to get exactly that amount of well-​being that corresponds to her virtue. A comparative ideal says the virtuous should be better off than the nonvirtuous. Many combinations and variants are possible. One might regard the deservingness that affects the well-​being one should get as global (one’s overall welfare should match one’s lifetime deservingness score), or as local and contextual (if you are undeserving with respect to a situation, you are less eligible for benefits and more eligible for harms in that situation). Deservingness might also play a role in a broadly egalitarian view (Arneson 2007). Roughly, views in this camp hold that the worse off you are, the greater the moral reason to bring it about that your condition improves, and also that the more deserving you are, the greater the moral reason to bring it about that your condition improves. From this standpoint it is always good to improve a person’s condition, but being more deserving can render an individual more eligible for benefits than others if goods to be distributed are limited. Call this view weak deservingness, in contrast to the strong deservingness views that hold it is intrinsically good that the nonvirtuous suffer evil to a degree corresponding to their negative deservingness. Dworkin should be interpreted as holding that taking personal responsibility seriously in the right sort of way just does require interpreting what we owe one another by way of compensation for bad luck from an ex ante, not an ex post perspective. When one gambles for high stakes in a game of chance, or makes a risky investment decision, or decides whether to accept or reject an employment offer with an uncertain package of benefits, or an offer of marriage or friendship when the future is unknown, there is strong moral presumption that one is responsible for bearing the consequences of one’s choice, be they good or bad, such that others have no moral obligation to make good one’s losses if things turn out badly. The presumption can be overcome in various ways. If the array of options available to one is unfairly limited, there may be grounds for compensation ex post. If one is incompetent at the time of choice among options, or predictably will choose so incompetently so that an acceptable norm of paternalism comes into play, then perhaps one should be denied the opportunity to make the choice without constraints to discourage inept choice, and if these constraints are not put in place, and one’s incompetent choice leads to a misfortune, one may be owed compensation to reduce one’s losses. But if society follows a distributive justice principle that systematically rejigs the outcomes of people’s choices in risky situations so that some morally preferred distributive pattern is maintained no matter how one chooses, then personal responsibility as we should understand it is not being properly respected. So run Dworkin’s convictions. Finally, we should note that one might hold that social justice evaluations should combine the ex post and ex ante perspectives. Suppose we have an indivisible good, relief from a lethal threat, that we can give to just one of two persons, who are identical in any way relevant to moral assessment. From an ex post perspective that values lives saved, any rescue strategy that results in one life saved is equally good. From

Dworkin and Luck Egalitarianism    53 an ex ante perspective, we might be concerned to satisfy some standard of ex ante fairness, such as giving each an equal chance of rescue by flipping a fair two-​sided coin and rescuing the coin-​toss winner. Contemplating the example, some might suppose that justice requires somehow splitting the difference between ex post and ex ante evaluations.

2.3  Fair Insurance Extended Dworkin notes that the chapters of Sovereign Virtue that apply the equality of resources theory to public policy issues also modify it. The modification is that the idea of hypothetical insurance markets as dictating what we owe to one another, present already in his Chapter 2 discussion of equality of resources, looms larger in the later chapters and subsequent formulations of his considered view. The fair insurance account of distributive justice could be extended to the point that it entirely supplants the equality of resources idea. In fact, this actually occurs in Dworkin’s post-​1981 writings reasonably interpreted. Let the amount and kind of benefits that we owe one another, to be supplied through government action, be set by fair hypothetical insurance. The background is that people interact freely in a competitive market economy, with market failures appropriately regulated, and contract and tort and criminal law set to facilitate people’s making mutually agreeable deals and not imposing negative externalities on one another. Against this background, people will face significant risks of bad fortune, for which actual insurance markets will not be available. Individuals may find themselves out of work and unable to find paid employment, but familiar moral hazard problems will prevent actual insurance markets from arising in the face of this uncertainty. Individuals may find themselves lacking marketable talent. Individuals may find themselves with genetic predispositions to disease or accident injury. For any such uninsurable variations in good and bad fortune across individuals, we can work out what insurance the average member of the community—​with average wealth and prudence—​would purchase, if insurance were available that covered these various contingencies. For each type of hypothetical insurance, individuals make their decisions behind an appropriate veil of ignorance tailored to the case at hand. In the case of initial-​bank-​ account-​wealth insurance, one might choose coverage against the possibility that one will have low initial wealth, and the cost of this insurance is that one will have to pay to fund this coverage if one is more fortunately situated. Also, the level of coverage provided might diminish or boost people’s incentives to produce and save, and so the total wealth of society that is passed on from one generation to the next may vary depending on the fair insurance chosen. There is also the consideration that a person who has earned money fairly may want to choose to give it to others or bequeath it. Taking these factors into account, the question becomes what initial-​bank-​account-​wealth insurance the average member of the community would purchase in the appropriate

54   Richard Arneson hypothetical insurance market. The answer fixes the character and level of inheritance and gift taxation and regulation in the just political community. Fair insurance enters Dworkin’s distributive justice proposals in two different contexts of discussion. When he elaborates the ideal of equality of resources, the initial thought is to give people equal bidding power and let all available resources be put at auction, the equilibrium result fixing people’s initial shares of goods. But once we notice that personal traits cannot be just divided up and put at auction in this fashion, yet personal traits count as resources for individuals, Dworkin is led to propose hypothetical insurance markets for marketable talent and for handicap as supplements to the equal auction specification of what equality of resources means. In later chapters, Dworkin considers problems of nonideal theory. Suppose we are living in a society that has not implemented equality of resources and is thus unjust in that regard. Now we are considering particular political issues that bear on distributive justice. What advice does the equality of resources ideal give us for policy choice in circumstances in which it is given that there will not be full compliance with the full Dworkin ideal? In this setting of partial compliance, Dworkin suggests that we can invoke the hypothetical insurance mechanism tailored to the particular polity issue that is being addressed. We ask, what insurance would people on the average have purchased had they had the opportunity to do so with equal wealth and full information about facts relevant to the insurance decision, and under the constraint of an appropriate veil of ignorance. Am I twisting Dworkin’s ideas to suggest that fair insurance might come to rule the roost in his theory of justice? Here are some considerations that press him in this direction. First, note that in the initial description of equality of resources, the thought experiment is to imagine shipwrecked sailors who are wondering how fairly to allocate cargo washed ashore among them as resources to be privately owned. These resources are like manna from heaven; there is no question that anyone might have prior claims. But the claim that these resources should be divided equally does not straightforwardly imply that resources that individuals earn in a fair setting should revert to the state at their death and be entirely available as fair shares for the next generation. Dworkin himself suggests that if you fairly earn resources and are free to spend them as you like, why not be allowed to give them to other persons? There is a concern that inheritance and targeted gifts might reproduce class inequality, and a concern that what you get via gift and inheritance is overwhelmingly brute luck, not option luck. But then we have considerations that press in opposed directions and need to be balanced, not just a tidal wave of reasons favoring prohibition of acquiring significant resources by inheritance or gift. There is also a leveling down problem to be considered. Suppose that the desire to provide well for one’s children, or beyond that the desire to give one’s children a “leg up” in social competition or a buffer against losing in social competition, or beyond that a dynastic ambition to establish wealth that will pass from generation to generation among family members, looms large in people’s motivations to contribute productively to the economy, and especially large for the more productive individuals. If any of these

Dworkin and Luck Egalitarianism    55 possibilities or a mix of them should hold, then insistence on an economic regime that forbids significant transfers by gift or inheritance might be a regime in which in each succeeding generation, individuals on the verge of adulthood acquire an equal share that is smaller in absolute terms than what the preceding generation enjoyed. In this scenario there may be compromises with strict inheritance and gift egalitarianism, under which everyone ends up with greater initial wealth when initial inequality is tolerated. If it is permitted that the children of the rich have greater initial wealth than the children of the poor, the children of the poor will end up with greater initial wealth than they would have had if there had been a prohibition on intergenerational transfers. What is clear is that people purchasing hypothetical insurance against the possibility of starting out in life with very little wealth will care about improving their opportunities overall and not at all about attaining equality of initial shares per se. Dworkin asserts that if someone would not purchase insurance at a certain level of coverage against the possibility of being afflicted with some misfortune if one had had the opportunity to purchase such insurance in a fair setting, that person has no valid demand for compensation to that level of coverage if it turns out that the misfortune falls on them. If one would not (in a fair setting) purchase insurance against blindness at a level of coverage that would render one neither glad nor sad to be blind, given one is compensated for blindness at that level of compensation, one does not have a moral right to that level of compensation. This same claim can be applied to the issue of when one is unfairly disadvantaged by having an initial stock of resources that is less than others have. Dworkin does apply this fair insurance test to settle the ideal theory issue of what distribution of resources is fair, and counts as satisfying the ideal of equality of resources, when the domain of resources is expanded to include personal traits in addition to material resources. In this way, the equality of resources ideal as Dworkin elaborates it includes the fair insurance idea, and there is no principled bar to extending the fair insurance idea, so that distributive justice (or “equality of resources” properly conceived) becomes distribution according to the fair insurance model. In fact, the logic of Dworkin’s position leads to dropping equality of resources altogether. Objection: in the various hypothetical insurance markets Dworkin considers, people are always imagined as having equal purchasing power. So there is a kernel of equality of resources inside the fair insurance model. Reply: The general requirement for hypothetical insurance is that people imagined as purchasing insurance are fairly placed. “Fair” need not mean “equal.” The fair initial resources people should be imagined as having are those they are entitled to under the fair insurance scheme regulating inheritance and gifts. We have seen that this insurance market does not yield the outcome that everyone should have equal money on entering adult life. This starting point insurance scheme, then, determines the fair amount of resources people should be imagined as having in facing further hypothetical insurance markets. So the equality-of-resources kernel in fair insurance becomes vanishingly small. (A concern here is, what determines the fair resource share for an initial run of the Dworkin procedures? Dworkin might follow a suggestion made by Nozick and hold that there should be a once-for-all-time equal initial distribution, following which the

56   Richard Arneson proper distributive norm is justice as fair insurance.) Notice also that people engaged in purchasing gift and inheritance “insurance” are directly choosing a tax-and-transfer scheme under ignorance of what gift and inheritance bundle one would have absent that scheme—​the cost of choosing high taxation is that one gives up wealth if one is a lucky inheritance and gift recipient and one faces the cost that high taxation will lower over time the funds available to be taxed and redistributed (one is ignorant of whether one will face the scheme in an early or later cycle of its operation). In this imaginary exercise the choosing parties are symmetrically placed but need not be envisaged as literally purchasing insurance with equal resources of any sort.

2.4  Interpreting Values and Principles as Seamless: The Case of Liberty and Equality Dworkin’s thinking on social justice issues is marked by a strong penchant for seamless unity. The idea is that values and principles that are thought to be opposed can profitably be reinterpreted as united in such a way that one can pursue the amalgam value–​principle wholeheartedly, without facing the need to accept trade-​offs between achieving more of one value and more of the other. We might think equality and liberty are opposed values, but Dworkin finds a clue in the thought that the theoretical auction that defines equality of resources cannot be imagined as functioning unless people know what they are legally at liberty to do with the various resources among which they are choosing and bidding. You do not sensibly know how much to bid for an acre of pine trees unless you know whether you are at liberty to look at them or chop them down for firewood and so on. Dworkin takes this as a sign that liberty is internal to equality, in the sense that we cannot understand the equality of resources ideal without working out what people should be at liberty to do with their resources.1 Dworkin supposes that there is a right answer, from the standpoint of equality of resources, to the question  of what people should be legally at liberty to do with the resources they will possess under an equality of resources regime. The answer is that people should be free to do whatever they choose with their resources so long as they do not wrongfully harm others in ways the criminal law should forbid, and do not impose negative externalities on others in ways that tort law should restrict. For the same

1  This claim should puzzle readers. First, to know what to bid for resources at auction, it would seem that one needs to know, not just what one will be legally at liberty to do with them, but rather what one will actually be able to do with those resources, given actual future circumstances. Second, anyway, perfect foreknowledge is not needed. To know what one should bid for resources at auction, one can proceed provided one can form probabilistic expectations of what one will eventually be able to do with those resources and with what results.

Dworkin and Luck Egalitarianism    57 reason, in the imagined equal auction for resources, people should be free to bid for the smallest bit of any resource they care to have. The ideas just stated are asserted by Dworkin in the form of what he calls the principle of abstraction and the principle of correction. The obvious objection at this juncture is that the correction principle allows what we should regard as nosy preferences and unwarranted disgust and repulsion and moralistic revulsion to affect what counts as imposing external costs on others in a way that triggers claims for compensation and restriction of the liberty of those who are objects of busybody feelings, disgust, repulsion, and revulsion. Not all negative externalities that fall on others as a consequence of what you do constitute any sort of reason to restrict your legal liberty to do that sort of thing. Dworkin accepts this. He endorses a principle of independence that disallows any standing to preferences that involve having contempt or dislike for other persons or groups of persons. So, in a society ruled by Dworkinian principles I could not claim that I am harmed by my neighbors selling my home to people whose skin color I dislike or whose way of life I find contemptible. Nor will the Dworkin hypothetical markets that dictate what we owe to one another register such preferences as affecting the content of these derived social justice norms. This sounds fine as far as it goes, but seems to leave the boundaries of the liberty that ought to be protected largely unresolved. Suppose I have no dislike or contempt for my neighbors but am concerned that they are spoiling the community by (a) homosexual conduct; or (b) recreational use of alcohol or marijuana; or (c) recreational use of heroin or methamphetamine; or (d)  living in accordance with caste and gender hierarchy norms that press women to stay in the home and be subordinate to males. Cases (a)­–(d) seem to differ significantly (to put it mildly) and to merit different legal response, but why and how? I do not see that the norms of treating people as equals and equalizing their comprehensive means to whatever aims they might choose to pursue provide adequate resources for settling the boundaries of acceptable restriction of individual liberty. The welfarist who has identified the best available account of what makes people’s lives go better has a principled basis for settling these boundaries, albeit one that is highly controversial. Here, as elsewhere, the principles Dworkin articulates seem to be more a nice-​sounding vocabulary in which to express his generally sensible and humane policy convictions, rather than genuine principles that dictate those policy convictions.

2.5  Luck Egalitarianism The distributive justice view known as luck egalitarianism takes as central some version of the distinction between what Dworkin has called “option luck” and “brute luck.” In Dworkin’s words, “Option luck is a matter of how deliberate and calculated gambles turn out—​whether someone gains or loses through accepting an isolated risk he or she

58   Richard Arneson should have anticipated and might have declined. Brute luck is how risks fall out that are not in that sense deliberate gambles” (Dworkin 2000: Ch. 2). Luck egalitarianism in a nutshell is the doctrine that social arrangements should be set so that people’s condition stays equal except insofar as inequality between people arises via option luck. Brute luck inequality should be equalized; option luck inequality should not be or need not be equalized. The basic moral appeal is that if people are born into grinding poverty or sheer unchosen affliction and in this way are worse off than others, their being worse off than others is morally bad, unfair. In contrast, if people are born with good prospects, and become worse off than others due to their neglectful or careless, or willfully imprudent, or risk-​seeking, or freely self-​sacrificing behavior, their being worse off than others is not morally bad and unfair (or at least less morally bad and unfair than their being worse off than others as a result of bad brute luck). A few simple distinctions delineate varieties of luck egalitarianism. One distinction is between the doctrine that luck egalitarianism does not require eliminating option luck inequality, and the doctrine that luck egalitarianism requires not eliminating option luck inequality. Hard luck egalitarianism holds that brute luck inequality should be equalized and option luck equality should be left standing, and soft luck egalitarianism holds that brute luck inequality should be equalized and option luck equality may or may not be equalized. A second distinction is between luck egalitarianism understood as one elem­ ent in a pluralistic distributive justice position that includes other elements, and luck egalitarianism understood as the sole determinant of distributive justice requirements. Call the former weak and the latter strong luck egalitarianism. The two distinctions are independent so they mark off four possible views. Another set of distinctions refines the rough distinction between option luck and brute luck. As formulated by Dworkin, the distinction between option luck and brute luck evidently admits of degree. Also, there is the distinction between a risk that one prefers to face and that one prefers to avoid. One prefers to face a risk when one prefers to face a lottery with good and bad possible outcomes than to get the expected value of the lottery. There is also the distinction between a risk (a lottery) that simply falls on one and that one cannot avoid by any course of action one might take and a lottery that one can avoid by one’s choice of action. A further distinction is between a lottery that can be avoided by some reasonable action one could take and a lottery that can be avoided, if at all, only by some unreasonable course of action. These distinctions all vary by degree: one might more or less greatly prefer the lottery to its expected value, one might be able to avoid facing a particular lottery by more or fewer courses of action, one might be able to avoid a lottery by a more or less reasonable course of action and by more or less numerous courses of action of given reasonableness. For further discussion, see Vallentyne 2002; Lippert-​Rasmussen 2001. In Dworkin’s formulation, an option luck risk is one that one either notices and might have declined or that one does not notice but should have (and might have declined if one had noticed it in time). So choice and fault are combined in the idea of option luck. A risk one does not choose can be option luck in character if one is at fault for not anticipating it and might have declined if one had seen it looming. A  further

Dworkin and Luck Egalitarianism    59 complication arises if one holds that a risk is brute luck unless, having noticed it, one could avoid facing it by some reasonable course of action. (If I could avoid attack by an irritated squirrel only by plunging into the raging current, my having the drowning plunge option does not make the attack risk one of option luck. But if my only way to avoid the squirrel attack risk is an unreasonable course of action, then the risk does not cease to be brute luck even if one negligently fails to notice it). For purposes of distinguishing brute and option luck, we should say that one lacks a reasonable alternative to accepting a risk only if there is no acceptable course of action (with reasonable prospects) one can take to avoid facing the risk. So if one is offered a terrifically attractive lottery, that would be most reasonable to accept, the lottery (if one accepts it) still qualifies as option luck provided there is some alternative course of action one might have chosen instead that would have yielded an acceptable (good enough) outcome. On this view, taking a gamble involving a risk can qualify as option luck whether or not accepting this risk is reasonable or unreasonable. On an alternative way of drawing the line between conduct issuing in inequality that triggers a distributive justice demand for equalizing compensation and one that does not, what counts is not whether one had a choice whether or not to accept the risk but rather the degree to which accepting the risk is meritorious or nonmeritorious. Consider supererogatory admirable risk-​taking for the benefit of others. A person who rushes in to save the children trapped in the burning building may face a high risk of suffering grievous injury, but this choice to rescue is morally admirable even though not morally required, and so (on this view) the rescuer who suffers grievous injury and is thus rendered worse off than others should be restored to being as well off as others and is owed this restoration as a matter of distributive justice. All versions of luck egalitarianism as characterized here hold that to some degree, or given that certain conditions obtain, it is in itself morally desirable that everyone’s condition or prospects be the same, according to some favored standard for assessing people’s condition or prospects. The most sweeping rejection of luck egalitarianism in all its forms and varieties insists that equality of condition or prospects simply is not per se morally desirable at all, even conditionally or to a certain degree. A moral position that sometimes mimics egalitarianism in its implications for policy but flatly denies that equality is ever per se morally desirable is prioritarianism. The prioritarian holds that (a) benefits matter more morally, the worse off in absolute terms is the person who gets them; and (b) at least to some degree, we are morally obligated to bring about outcomes of greater rather than lesser moral value. The strictest version of prioritarianism is an act consequentialist doctrine that says (a) we are always bound to do what brings about the best attainable outcome; (b) outcomes are better and worse as they contain more and less aggregate human welfare; and in particular (c) outcomes are better and worse as they contain more and less priority-​weighted welfare, with benefits counting for more, the worse off the persons who get them. But prioritarianism rejects egalitarianism and a fortiori is not a member of the family of luck egalitarian views. This remains the case even if the prioritarian accepts a further priority pertaining to personal

60   Richard Arneson responsibility—​for example, benefits matter more, the more deserving is the person who gets them. Dworkin’s fair insurance view of justice might be regarded as rejecting egalitarianism if that doctrine is construed as affirming that equality of condition of some sort is morally valuable under some conditions and at least to some degree. Dworkin holds that governments, though not private individuals, are required to treat all members of the political community with equal concern and respect. So perhaps Dworkin is committed only to equal concern, not to the maintenance of equality of condition. True, Dworkin’s apparently canonical formulations of his view tend to assert that people should be made equal in their ex ante opportunity to deal with the uncertainty of suffering misfortune. But the logic of the fair insurance approach belies these assertions. This says that people should be equally positioned in the hypothetical insurance markets that fix what resources people are entitled to as a matter of justice. These hypothetical markets determine what is fair in the face of the risk of starting life with few material or bank account resources, low marketable talent, and disability, and of facing unemployment and health problems as life proceeds. None of these hypothetical markets would tend to dictate equal prospects for all or equal condition for all. The upshot is that if it is so that equality of condition or prospects by any measure is not in itself morally desirable, Dworkin’s fair insurance approach has the advantage over any version of luck egalitarianism. However, as the discussion section 2.6 shows, the fair insurance approach itself attracts criticisms of unfairness. Luck egalitarianism takes the distinction between brute luck and option luck to be crucial for the determination of fair distributive policies. The more uncompromising strong and hard versions of luck egalitarianism especially invite the objection that too much is being made of too little—​that giving the distinction so much weight makes luck egalitarianism unforgiving and harsh in its recommended treatment of people who suffer grievous misfortune via their own choice or fault. Consider the example of a young person who has already received the resource share that luck egalitarianism identifies as fair and then engages in a risky sport without purchasing accident insurance, and ends up grievously injured, but could be restored to good health and good life prospects at moderate cost. Without extra help, his life prospects are grim; with help, he lives well. Luck egalitarianism in most versions says the person has suffered misfortune and is now far worse off than others through option luck not brute luck, so justice at least does not demand help for the person and (in the strong hard version) demands that no further help be given. But the luck egalitarian may respond that Dworkin’s fair insurance approach ends up paying insufficient heed to the norm of respect for personal responsibility that demands incorporating in our justice norms the distinction between brute luck and option luck. The objection is that Dworkin throws out too much of the baby with the bath water. Nothing in the fair insurance approach brings it about that hypothetical persons making their imaginary insurance distinctions will pay any attention to whether misfortunes they might suffer will arise by brute or option luck. I will, as a rationally prudent agent choosing under ignorance, be just as concerned to avoid or mitigate the misfortune of

Dworkin and Luck Egalitarianism    61 bad luck that falls on me beyond my power to control, and bad luck that I could avoid but will not in fact avoid.

2.6  Justice as Fair Insurance:  An Assessment According to Dworkin, justice requires making people equally able to insure or take action to prevent misfortune or, if that is unfeasible, compensating people for misfortune they suffer according to the coverage the member of the community with representative tastes and average income would have purchased had the opportunity to purchase such insurance been available. Is this a plausible view? Suppose luck egalitarianism is interpreted as holding that all inequality arising from brute luck should be eliminated and that all inequality arising from option luck should be left standing, so far as this is possible. This so-​far-​as-​possible qualification is necessary because brute luck and option luck can combine to produce outcomes in such a way that the aims of eliminating brute luck equality and retaining option luck inequality cannot be met together. If the luck egalitarian insistence on equality is not constrained by other values, then it can be very demanding. A  person born deaf and blind and limbless may benefit, though only slightly, from ever greater assignment of resources. The person would be worse off than others, by any standard of measurement, even if most of the world’s resources are lavished on her. Unconstrained luck egalitarianism as just characterized would insist on yet further transfer of resources to the single unfortunate so long as she remains worse off than others. The badly-​off person who is a very poor transformer of resources into welfare becomes a basin of attraction for transfers. Even if further transfer would provide no benefit at all to the recipient, if it is morally bad that some are worse off than others through no fault or choice of their own, then it is at least in one way better if better-​off persons are made worse off, even if their losses produce no benefit at all to worse-​offs. The basin-of-attraction problem and the leveling down problem clearly strike Dworkin as reasons to favor the insurance model. If I do not know whether I am disabled or not, I will be unwilling to purchase insurance that offers unlimited compensation to the disabled, channeling resources to them so long as they are unequivocally poorer in resource holdings than others. Dworkin has the hunch that the pattern of transfers recommended by hypothetical insurance will be moderately equalizing and mesh with our considered judgments. Unfortunately, sensible interpretations of the hypothetical insurance market do not yield that result (but for a vigorous defense of Dworkin’s position, see Parr forthcoming). If money will be less useful to me if I am disabled than if I am not, then given the opportunity to purchase insurance, I might well, if risk-​neutral, prefer to purchase insurance that would transfer money to me if I turn out to be able and would take money from me

62   Richard Arneson if I turn out to be disabled. The average community member might have this pattern of preference, and if so, the hypothetical insurance transfer requires reverse transfers from the disabled to the able. If we balk at the implication that justice requires reverse transfers from people who suffer the bad brute luck of disability to those who have good brute luck in this respect, we are rejecting the claim that the fair insurance model of just­ ice as articulated by Dworkin is really fair and just (Roemer 2002). If any initial willingness we might have had to accept justice as fair insurance rests on the expectation that the hypothetical insurance contraption that Dworkin constructs will yield implications for policy that fit our antecedent convictions, then it is clear the hypothetical insurance idea is window dressing. The view might just as well be stated as consisting of Dworkin’s personal judgments about how much is owed to the disabled and to the unemployed and underemployed, and those lacking marketable talent, and to those prone to sickness and injury, and to those whose parents lack wealth and resources to pass on to them, and so on. The view, if it comes to that, has no rejoinder to anyone who disagrees with Dworkin’s own personal judgments in this matter. If some huge brute luck contingency threatening widespread disability hung over our community, and insurance markets actually offered insurance against this contingency, and people’s actual decisions led to reverse transfers from unlucky to lucky when the disaster struck, we at least could appeal to people’s actual voluntary choices and their personal responsibility for their choices. We might nonetheless have strong moral reasons to disallow such an insurance market outcome, but these reasons would oppose responsibility for choices. However, the appeal to the hypothetical insurance decisions of people lacks this appeal to personal responsibility for actual choice. Moreover, the hypothetical insurance compensation is set by the average preferences of community members, and my preferences or yours might be minority preferences. So the Dworkin scheme could well end up requiring reverse transfers from unlucky to lucky, even on the part of the unlucky individuals who would never have purchased the hypothetical insurance that is imputed to them. Objecting to Dworkin’s fair insurance approach to social justice on the ground that it recommends reverse transfers from unlucky badly-​off people to lucky well-​off people might seem misguided. After all, virtually any view that gives some moral weight to gains for better-​off people will sometimes make such a recommendation. A prioritarian welfarist will hold that if a resource does hardly anything for me and would enormously benefit a better-​off person, transfer of the resource may be warranted, depending on the amounts of benefit and loss for the affected parties, even allowing for the prioritarian thumb (or perhaps elbow) on the scale for the interests of the worse-​off. I would say the reverse transfer objection to the fair insurance approach is really an objection that the factors that justify reverse transfer on Dworkin’s view are not justifying. The objection is similar to the complaint often voiced against John Rawls’s justice as fairness idea. What individuals concerned to advance their interests behind a veil of ignorance would choose simply does not capture the considerations that fix the content of social justice requirements.

Dworkin and Luck Egalitarianism    63

2.7 Conclusion Dworkin’s sophisticated fair insurance approach avoids any commitment to any version of equality of condition or prospects, and in this way sharply differs from luck egalitarianism. This is a merit of Dworkin’s position for those of us who deny that equality of condition or prospects is ever per se morally valuable. Dworkin develops sophisticated arguments supporting his idea that incorporation of any sensible personal responsibility norm in the theory of social justice rules out welfarism—​the idea that each individual’s lifetime well-​being level is the appropriate measure of her condition for the purpose of deciding what we owe to one another. This chapter has sought to rebut those arguments. Dworkin argues forcefully that the ex ante, not the ex post perspective, is the right focus for social justice principles. This chapter identifies no decisive argument for going one way or the other on this issue. Dworkin ingeniously suggests that hypothetical insurance markets offer correct guidance on the question of what we owe one another in a world of uncertainty. This suggestion merits rejection. A final comment:  Dworkin asserts that morality requires that we organize our economy on the basis of capitalist free markets, regulated as justice as fair insurance proposes. But since Dworkin’s view appeals to idealizations of market processes, he ends up asserting an independent standard of justice (what the ideal market would deliver), and it then remains a completely open question to what extent actual, feasible private-​ ownership free market institutions would be the best means to achieve this ideal.

Acknowledgments I thank Thomas Parr and Serena Olsaretti for their challenging and astute criticisms of a draft of this chapter. I also benefited from discussing the issues with Alexander Campbell and from reading his UC San Diego 2015 honors thesis.

References Anderson, E. (1999). “What Is the Point of Equality?” Ethics 109: 287–​337. Arneson, R. (1989). “Equality and Equal Opportunity for Welfare.” Philosophical Studies 56: 77–​99. Arneson, R. (2004). “Cracked Foundations of Liberal Equality,” in J. Burley (ed.) Ronald Dworkin and His Critics. Oxford: Basil Blackwell, pp. 79–​98. Arneson, R. (2007). “Desert and Equality,” in N. Holtug and K. Lippert-​Rasmussen (eds) Egalitarianism: New Essays on the Nature and Value of Equality. Oxford: Oxford University Press, pp. 262–​93. Cohen, G. (1989). “On the Currency of Egalitarian Justice.” Ethics 99: 906–​44. Cohen, G. (2004). “Expensive Taste Rides Again,” in J. Burley (ed.) Ronald Dworkin and his Critics. Oxford: Blackwell, pp. 3–​29.

64   Richard Arneson Dworkin, R. (1981). “What Is Equality? Part  1:  Equality of Welfare.” Philosophy and Public Affairs 10: 185–​246; also “What Is Equality? Part 2: Equality of Resources.” Philosophy and Public Affairs 10: 283–​345. Dworkin, R. (2000). Sovereign Virtue:  The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Dworkin, R. (2002). “Sovereign Virtue Revisited.” Ethics 113: 106–​43. Dworkin, R. (2003). “Equality, Luck, and Hierarchy.” Philosophy and Public Affairs 31: 190–​8. Dworkin, R. (2011). Justice for Hedgehogs. Cambridge, MA: Harvard University Press. Eyal, N. (2007). “Egalitarian Justice and Innocent Choice.” Journal of Ethics and Social Philosophy 2: 1–​18. Fleurbaey, F. (1995). “Equal Opportunity or Equal Social Outcome?” Economics and Philosophy 11: 25–​55. Fleurbaey, M. (2008). Fairness, Responsibility and Welfare. Oxford: Oxford University Press. Hurley, S. (2003). Justice, Luck, and Knowledge. Oxford: Oxford University Press. Kagan, S. (1998). “Equality and Desert,” in O. McLeod and L. Pojman (eds) What Do We Deserve? Oxford, Blackwell, pp. 298–​314. Kagan, S. (2012). The Geometry of Desert. Oxford: Oxford University Press. Knight, C. (2009). Luck Egalitarianism. Edinburgh: Edinburgh University Press. Lippert-​Rasmussen, K. (2001). “Egalitarianism, Option Luck, and Responsibility.” Ethics 111: 458–​579. Miller, D. (2015). “The Incoherence of Luck Egalitarianism,” in A. Kaufman (ed.) Distributive Justice and Access to Advantage. Cambridge: Cambridge University Press, pp. 131–​50. Nagel, T. (1991). Equality and Partiality. Oxford and New York: Oxford University Press. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. Parr, T. (forthcoming). “How to Identify Disadvantage:  Taking the Envy Test Seriously,” Political Studies. Rawls, J. (1982). “Social Unity and Primary Goods,” in A. Sen and B. Williams (eds) Utilitarianism and Beyond. Cambridge: Cambridge University Press, pp. 159–​85. Ripstein, A. (2007). “Liberty and Equality,” in A. Ripstein (ed.) Ronald Dworkin. Cambridge: Cambridge University Press, pp. 82–​98. Roemer, J. (1993). “A Pragmatic Theory of Responsibility for the Egalitarian Planner.” Philosophy and Public Affairs 11: 146–​66. Roemer, J. (1998). Equality of Opportunity. Cambridge, MA: Harvard University Press. Roemer, J. (2002). “Equality against the Veil of Ignorance.” Journal of Philosophy 96: 455–​72. Scheffler, S. (2003a). “What Is Egalitarianism?” Philosophy and Public Affairs 31: 5–​39. Scheffler, S. (2003b). “Justice as the Virtue of Sovereigns:  A Reply to Ronald Dworkin.” Philosophy and Public Affairs 31: 199–​206. Scheffler, S. (2005). “Choice, Circumstance, and the Value of Equality.” Philosophy, Politics, and Economics 4: 5–​28. Segall, S. (2013). Equality and Opportunity. Oxford: Oxford University Press. Sher, G. (2014). Equality for Inegalitarians. Cambridge: Cambridge University Press. Temkin, L. (1993). Inequality. Oxford and New York: Oxford University Press. Vallentyne, P. (2002). “Brute Luck, Option Luck, and Equality of Initial Opportunities.” Ethics 112: 529–​57. Williams, A. (2002). “Equality for the Ambitious.” Philosophical Quarterly 52: 377–​89.

Chapter 3

E qua lit y Versu s Pri ori t y Michael Otsuka and Alex Voorhoeve

In this chapter, we discuss two leading theories of distributive justice: egalitarianism and prioritarianism. We argue that while each has particular merits and shortcomings, egalitarian views more fully satisfy a key requirement of distributive justice:  respect for both the unity of the individual and the separateness of persons. The argument proceeds as follows. In sections 3.1 and 3.2, we introduce egalitarianism and prioritarianism, respectively, and apply these views to cases in which we are certain how things would turn out if we were to choose one way or another. We then introduce cases of risks which deprive us of such certainty. In sections 3.3–3.5, we analyze various forms of prioritarianism where decisions are made under conditions of risk and raise objections to each of them. In section 3.6, we develop the most plausible version of egalitarianism under risk and explain why it is less vulnerable to these objections than prioritarian views.

3.1  Egalitarianism under Certainty There are many forms of egalitarianism. Social and political egalitarianism holds that material and social inequalities are bad when and because they undermine individuals’ ability to live as equal citizens who are willing to offer and abide by fair terms of social co-​operation (Norman 1998; Anderson 1999; O’Neill 2008). It follows that social and political egalitarians regard inequalities as problematic when they lead to domination. One example is the political domination that arises when large inequalities in wealth lead to the control of media by an elite; another example is the power asymmetry in the workplace that occurs when the worst off are dependent on those who are better off for a minimally decent existence. Social and political egalitarians also object to stigmatizing differences in status, such as exist or have existed between men and women, aristocrat and commoner, or Brahmin and Dalit. Finally, on this view, inequalities are bad when and because they give rise to particular morally problematic attitudes. These include

66    Michael Otsuka and Alex Voorhoeve servility, envy, and a lack of self-​respect among the worst-​off, and arrogance and a jealous guarding of relative advantage among the better off. Besides often being objectionable in themselves, these attitudes undermine citizens’ attachment to an ideal of social cooperation among persons who owe each other a justification for their common institutions (Tawney 1964: 37–​8). In this chapter, we shall set aside these objections to inequality, important though they are. For they are not the topic of dispute between egalitarians and prioritarians. Instead, this dispute revolves around a further question: Is inequality objectionable beyond the ways in which it threatens the aforementioned egalitarian social and political ideals? We shall begin our attempt to answer this question with a case that is closely modelled on one that Thomas Nagel presents in his famous essay “Equality” (1979), and which has played a key role in subsequent discussions of equality and priority: Two-​Child Case with Certainty. Imagine that you are the parent of two young teenagers, Ann and Ben. Ben has recently been diagnosed with a condition that will soon give rise to a serious physical disability, but Ann has been given a clean bill of health. You have recently lost your job but have offers in two places. You must choose whether to take a city job and move your family to cramped urban accommodation in an unpleasant neighbourhood with mediocre schools or to take a job on the outskirts of town and move your family to the open spaces of a suburb with excellent schools. Either option would be equally good insofar as your own well-​being is concerned. But they would not be equally good insofar as the well-​being of each of your children is concerned. If you move to the city, able-​bodied Ann, who loves nature and sports, will be hemmed in and have only a mediocre education, but soon-​ to-​be-​disabled Ben will have access to special medical facilities that will somewhat, but far from wholly, alleviate the effects of his disability. If you move to the suburb, Ann will flourish, but Ben will not receive this treatment. Moreover, if you move to the suburb, the boost in well-​being to Ann will be greater than the boost in well-​ being to Ben would be if, instead, you moved to the city. Table 3.1 represents the relevant levels of lifetime well-​being for your two children.1

Considering only the well-​being of your two children (that is, bracketing the level of well-​being of others, yourself included), what should you choose? Utilitarianism calls for the maximization of the sum total of well-​being (utility).2 So a utilitarian would opt for the suburb in this case. However, it strikes many, including Nagel, that one should move to the city for Ben’s sake. Nagel would maintain that one has egalitarian reason to

1 

These numbers are to be interpreted as follows: an incremental improvement in well-​being of a given size always does as much good for the person (has as much prudential value for the person) no matter the level of well-​being from which this increment arises. Moreover, for every individual, 0 is a quality of life such that, from the perspective of the good of the person living that life, it is a matter of indifference that they live that life or never existed at all. (In this, we follow Adler 2012: 219–​20.) We also stipulate that 100 is an uncommonly good life. We say more on the measure of well-​being when we introduce risk in section 3.3. 2  In this chapter, we shall treat “well-​being” and “utility” as synonyms.

Equality Versus Priority    67 Table 3.1  F inal well-​being for the Two-​Child Case with Certainty Ann

Ben

City

60

39

Suburb

70

30

move to the city which overrides the utilitarian grounds one has to move to the suburb in this case. As he writes: If one chose to move to the city, it would be an egalitarian decision. It is more urgent to benefit the second child. . . . This urgency is not necessarily decisive. It may be outweighed by other considerations, for equality is not the only value. But it is a factor, and it depends on the worse off position of the second child. An improvement in his situation is more important than an equal or somewhat greater improvement in the situation of the first child. (1979: 124)

According to the principle of equality to which Nagel alludes here, it is in itself bad if some are worse off than others (through no choice or fault of theirs). This badness is above and beyond the bad effects of inequality. The badness resides in the inequality itself, so to speak. The badness is the unfairness of some being less well off than others through no fault or choice of theirs (Cohen 1989; Arneson 1997; Temkin 2001).3 (We shall assume throughout that no one is responsible for their level of well-​being.) The principle of equality doesn’t explain why it is better for all to be equally well off rather than equally badly off. It therefore requires supplementation. One such supplementation involves an appeal to a principle according to which it is in itself better if people are better off. In other words, an egalitarian should be a pluralist who combines a belief in equality with a belief in the importance of improving people’s well-​being. (We shall discuss one such form of pluralist egalitarianism in Section 3.6.) In Nagel’s Two-Child Case, it would appear that equality overrides total well-​being, since the gain in equality of a move to the city seems more important than the loss in total well-​being of such a move. One gains only one unit of total well-​being by moving to the suburb, yet one very substantially reduces the inequality between the two children. Notwithstanding his concern for improving well-​being, a pluralist egalitarian is subject to what has come to be known as the leveling down objection (McKerlie 1984: 232; Parfit 1995: 17–​18). Suppose that Ann would, due to natural causes, develop the same

3  It is instructive here to draw a contrast with the badness of having less than enough to survive. The badness of having less than enough to survive often accompanies severe inequality. But this badness is in no way constituted by the relation of inequality. It consists of the non-​comparative fact of not having enough to live on in absolute terms.

68    Michael Otsuka and Alex Voorhoeve severe disability as Ben. Pluralist egalitarians’ concern for Ann’s well-​being might lead them to regard such leveling down as bad, all things considered. Nonetheless, they are committed to the claim that it is in one way better, because fairer, if Ann and Ben are equally badly off because equally impaired. The leveling down objection holds that there is nothing good at all in such leveling down, and that egalitarianism is therefore false. Some egalitarians regard this objection as without merit (see, for example, Temkin 2003: 67–​8). We share this view. To us, it is clear that there is unfairness when some are better off than others and that this unfairness is absent when people are equally well off, so that leveling down is in one respect good. Others, however, regard it as a powerful objection. (This is true even of some once attracted to egalitarianism; see, for example, Arneson 1999: 232–​3.) Those who are so moved may be on the lookout for an alternative to egalitarianism to account for the conviction that you should move to the city in Nagel’s case. Prioritarianism is one such alternative.

3.2  Prioritarianism under Certainty In fact, as Derek Parfit points out, such an alternative to egalitarianism might be found in the very quotation of Nagel’s with which we introduced egalitarianism. In his re-​ interpretation of Nagel, Parfit writes that this passage “contains the idea that equality has value. But it gives more prominence to another idea. Nagel believes it is more important to benefit the child who is worse off. That idea can lead us to a quite different view” (1995: 19). On this different view, helping Ben is more important, not because one can improve his well-​being by a greater increment (one can’t, in this case), but because an increment in well-​being is morally more valuable the lower (in absolute terms) the level of well-​being from which this increment arises. The idea that an improvement in a person’s well-​being that arises from a lower absolute level is morally more weighty (and should therefore have priority over an equally large improvement in a person’s well-​being that arises from a higher absolute level of well-​being) is at the heart of the Priority View. Prioritarianism is not subject to the leveling down objection. On the Priority View, it is in no way better if Ann were to become impaired, since the only change in value, on this view, would be that (morally weighted) welfare is lost. Nothing is gained. Prioritarians are therefore insensitive to the elimination of the unfairness of some being less well off than others through no fault of theirs. This insensitivity to unfairness highlights the fact that the Priority View is not an egalitarian view. As Parfit explains: Egalitarians are concerned with relativities: with how each person’s level compares with the level of other people. (1995: 23)

Equality Versus Priority    69 By contrast: . . . we [prioritarians] do not believe in equality. We do not think it in itself bad, or unjust, that some people are worse off than others. That is what makes this a distinctive view. . . . On the Priority View, we are concerned only with people’s absolute levels. (1995: 22–​3)

Parfit offers the following illuminating analogy, to illustrate the non-​comparative nature of prioritarianism: People at higher altitudes find it harder to breathe. Is this because they are higher up than other people? In one sense, yes. But they would find it just as hard to breathe even if there were no other people who were lower down. In the same way, on the Priority View, benefits to the worse off matter more, but that is only because these people are at a lower absolute level. It is irrelevant that these people are worse off than others. Benefits to them would matter just as much even if there were no others who were better off. (1995: 23)

In sum, one can characterize prioritarianism as the conjunction of the following three claims: (Diminishing Marginal Value):  Each person’s well-​ being has positive and diminishing marginal moral value—​an increment that takes place from a lower level receives a higher “priority weight” than an increment that takes place from a higher level; (Separability): The moral value of (an increment in) a person’s well-​being does not depend on how anyone else fares; (Maximization): We ought to maximize the sum-​total of the moral value4 of the well-​being of persons (which is their priority-​weighted well-​being). (McKerlie 1984; Parfit 1995; Adler 2012; Broome 2015)5 4  Here and elsewhere, when we speak of “moral value” and its maximization, we do not mean to imply that prioritarianism is necessarily a form of consequentialism, according to which the fact that it would maximize moral goodness or value provides the explanation of the rightness of an act. Maximization is consistent with the affirmation of a non-​consequentialist (e.g., contractualist) explanation of the rightness of the maximization of moral value. It is also consistent with the view that the magnitude of moral value is simply a representation of the strength of our reasons for action. 5  While Parfit’s 1995 Lindley Lecture is the locus classicus for the Priority View, a view with these characteristics has a prior history in both welfare economics and moral philosophy, though it took some time for it to be recognized as distinct from egalitarianism. In economics, Serge-​Christophe Kolm (1969) and Anthony Atkinson (1970) independently formulated a social welfare function according to which each individual’s income has diminishing marginal moral value, and which sums this value across individuals (thereby respecting Separability). Their view is distinct from utilitarianism because they appeal not to the diminishing marginal individual well-​being of income but rather to its diminishing marginal social (or moral) value. However, while the Kolm–​Atkinson social welfare function can therefore be regarded as prioritarian, they refer to it

70    Michael Otsuka and Alex Voorhoeve Parfit argues that many people who may take themselves to be egalitarians because they are especially concerned with those who are worse off than others are really prioritarians. Nagel, he suggests, is one such person. Nagel says that it is “more urgent” to benefit the disabled child in the Two-​Child Case with Certainty. Parfit asks: Would it be just as urgent to benefit the handicapped child, even if he had no sibling who was better off? I suspect that, on Nagel’s view, it would. Nagel would then, though using the language of equality, really be appealing to the Priority View. (1995: 26)

Parfit’s comment raises the following general questions: When considering the fate of one person in isolation from the fate of others, what sort of benefit should we provide this person? And how, if at all, do situations in which the fate of only one person is at issue differ morally from cases in which the fates of more than one person are at issue, and in which the interests of these people may conflict? In order to answer these questions, we turn to cases involving risk.

as a form of pluralist egalitarianism. Amartya Sen (1973: 39) argued that in order to clarify the Kolm–​ Atkinson function’s special concern for those who are badly off, one should take individual well-​being (rather than income) as the object that has diminishing marginal moral value. He also argued that egalitarians need not accept Separability, because they may be concerned with how some fare relative to others. John Broome (1989) went further, arguing that egalitarians should reject Separability precisely because it implies a lack of concern for how some fare compared to others. In moral philosophy, Paul Weirich (1983) was the first to propose a view which respects Diminishing Marginal Value, Separability, and Maximization. (He did so apparently unaware of the work by Kolm, Atkinson, and Sen on social welfare functions of this kind.) Weirich described his view as “a compromise between equality and utility [well-​being],” but he failed to note that Separability is in tension with standard egalitarianism. Dennis McKerlie (1984), however, did distinguish two “different forms of egalitarianism,” one of which cares about reducing inequality (and therefore rejects Separability) and the other which accepts Separability along with Diminishing Marginal Value and Maximization. McKerlie argued that while the latter avoided the leveling down objection, it was vulnerable to a different objection: in holding that well-​being has diminishing marginal moral value even in one-​person cases, it implausibly requires individuals not to accept expectedly advantageous gambles (1984: 235). McKerlie also noted that in making the moral value of a person’s well-​being independent of whether the distributional choice one faces involves trade-​offs within a life or across different lives, the view does not “attach special importance to the difference between [separate person’s] lives” (1984: 233). (These observations lie at the heart of the criticism of the Priority View we formulate in section 3.3.) It appears that Larry Temkin (1983: 232–​4) was the first to argue unreservedly that a view which respects Diminishing Marginal Value, Separability, and Maximization (which he refers to as “extended humanitarianism”) is not an egalitarian view, given its lack of concern for how some fare relative to others. Nonetheless, he argued that it might be worth endorsing such a view in addition to egalitarianism. Parfit’s contribution was to sharpen the contrast between egalitarianism and prioritarianism and to argue that prioritarianism is superior to egalitarianism. (We are grateful to Jerod Coker for research assistance on the origins of prioritarianism in economics and philosophy.)

Equality Versus Priority    71

3.3  Final-​Well-​being Prioritarianism Consider the following one-​person case, which is a transformation of Nagel’s Two-​Child Case into a case of a single child who has an equal chance, so to speak, of turning out as either of Nagel’s two children (Otsuka and Voorhoeve 2009: 188; Otsuka 2015: §I). One-​Child Case: You have a single early-​teenage child, Cathy, who is now healthy but who has a 50% chance of developing the aforementioned disability and a 50% chance of remaining healthy. Before you know how her health will develop, you must now decide whether to take the job which necessitates a move to the city, which would enable her to receive specialist treatment if she were to develop the disability, or to take the job in a suburb, which will lead her to thrive if she is able-​bodied. Due to economic circumstances, these moves are permanent—​if, say, you have taken the job in the suburb, you will not be able to later move to the city in the event that Cathy develops the disability.

In considering this case, it is necessary to make further assumptions about the measure of well-​being we are employing. We shall assume a measure of well-​being on which a prospect has higher expected well-​being for a person just in case it would be preferred for that person’s sake after rational and calm deliberation with all pertinent information while attending to her self-​interest only. (A person’s expected well-​being is just the probability-​weighted sum of her well-​being in each possible state of the world.) One prospect has the same expected well-​being as another for a person just in case such deliberation would yield indifference between the two prospects.6 Moreover, in order to establish what the Priority View requires in this case, we must determine how it deals with risk. One prioritarian approach is to maximize expected priority-​weighted final well-​being (Rabinowicz 2002; Adler 2012). That is, one maximizes the probability-​weighted sum of priority-​weighted well-​being in each possible state of the world. We shall refer to this as “final-​well-​being prioritarianism.” On this view, you must choose the city in the One-​Child Case, for the same reason 6 

In other words, we assume that the measure of well-​being is derived from idealized preferences satisfying the Von Neumann–​Morgenstern axioms. (For discussion of how assuming this rather than another measure of well-​being determines the plausibility of prioritarianism, see Greaves (2015) and Otsuka (2015).) Note that this measure does not presuppose any particular view on what well-​being is. One might believe that two options have the same expected well-​being for a person just in case they would, if ideally rational and self-​interested, be indifferent between these options without also believing that well-​being consists of preference satisfaction. One might maintain that well-​being consists of something other than preference satisfaction and hold that the specified idealized preferences fully track the magnitude of this other thing (Otsuka and Voorhoeve 2009: 172–​3, n3). More generally, throughout, we assume that orthodox decision theory applies, according to which under risk, a decision-​maker ought to maximize the expectation of what they take to be the relevant value (so that a utilitarian ought to maximize the expected sum-​total of well-​being, a final-​well-​being prioritarian the expected sum-​total of priority-​weighted well-​being, etc.).

72    Michael Otsuka and Alex Voorhoeve you must do so in the Two-​Child Case with Certainty. In the latter, final-​well-​being prioritarianism justifies the move to the city on the grounds that improving a person’s well-​being from 30 to 39 is more morally valuable than improving a person’s well-​being from 60 to 70. The greater moral value of the improvement from 30 to 39 implies that moving to the city will have greater expected moral value in the One-​Child Case (see Table 3.2). Two features of final-​well-​being prioritarianism stand out. First, because of the way in which it distinguishes moral value from personal (or prudential) value, final-​well-​ being prioritarianism holds that one must sometimes choose an option that is contrary to the expected interests of the only person whose well-​being is at issue (McKerlie 1984; Rabinowicz 2002; McCarthy 2008; Otsuka and Voorhoeve 2009). Second, it treats some risky intrapersonal trade-​offs in which only one person’s interests are at stake as involving the very same moral calculus as interpersonal trade-​offs in which the interests of different people conflict (McKerlie 1984; Otsuka and Voorhoeve 2009). Both aspects are problematic. First, in the One-​Child Case, moving to the suburb uniquely maximizes Cathy’s expected well-​being. Given our assumptions, someone solely concerned with Cathy’s interests would therefore prefer it on her behalf. This means that no matter how things turn out, if you were to opt for the suburb, you would be able to offer her the following prudential justification: I had to balance a 50% chance of you thriving (rather than leading a cramped existence and facing mediocre schooling) if you were healthy against a 50% chance of your having access to specialist medical care (rather than not having such access) in case you developed a disability. I balanced these two potential effects on your well-​ being from the perspective of your self-​interest alone. From this perspective, the expected value of the benefits of the suburb outweighed the expected value of the benefits of the city. I therefore chose the former. In so doing, I did the best I could for you, given the information I had at the time.

We believe this gives you strong reason to choose the suburb. Moreover, when you consider Cathy’s prospects in isolation from how well off anyone else is, you have, in our view, no countervailing reason to move to the city. We conclude that when one so brackets others’ well-​being, it is at least permissible to move to the suburb. More generally, contrary to final-​well-​being prioritarianism, in risky intrapersonal trade-​offs in Table 3.2  Final and expected well-​being in the One-​Child Case Final well-​being p = 0.5 Cathy able

p = 0.5 Cathy disabled

Expected well-​being

City

60

39

49.5

Suburb

70

30

50

Equality Versus Priority    73 which inequality is not an issue, it is reasonable to accord equal moral weight to equally large increases in a person’s well-​being, independently of the baseline from which these increases take place—​that is, to maximize her expected well-​being rather than to give priority to her fate if she turns out to be worse off. 7 Indeed, when only one person’s well-​ being is at issue, it is curious to decide what to do for this person, not on the basis of the personal value of increments in her well-​being, but instead on the basis of their priority-​ weighted value, where this value is understood to be wholly impersonal in nature. Such cases highlight that a prioritarian who is committed to maximizing impersonal value is committed to choosing contrary to the prudential interests of everyone concerned, even when inequality is not at issue.8 Second, it is objectionable to apply the same moral calculus in risky intrapersonal trade-​offs and interpersonal trade-​offs. Cathy’s two potential futures are unified because they are both potential futures of hers. This unity makes it appropriate to balance the potential advantages and drawbacks of each alternative to her from the perspective of her interests.9 That is why, in making risky decisions that affect a person alone, we naturally speak of “choosing for her sake” and why, if you move to the suburb, you can offer Cathy a prudential justification for this move, no matter how things turn out. By contrast, if you move your two children to the suburb, no comparable prudential justification can be provided to the disabled child. This is because moving to the suburb was never in his interest, as only a different person (his healthy sister) could ever possibly have benefited from that. In other words, in the Two-​Child Case with Certainty, the separateness of persons makes it harder to justify foregoing the gain to Ben, who will be worse off in any case, in order to provide a slightly greater gain to Ann. In applying the same priority weights to possible increments in well-​being in both cases, final-​well-​ being prioritarianism therefore fails to respect the difference between the unity of the individual and the separateness of persons.10 7  For an extended argument for this conclusion, see Otsuka (2015). This conclusion has been much debated. For concurring opinions, see McCarthy (2008); Otsuka and Voorhoeve (2009); Williams (2012); and Rendall (2013). For dissenting views, see Rabinowicz (2002); Crisp (2011); O’Neill (2012); Parfit (2012); Porter (2012); Segall (2015); and Bovens (2015). For a review of the debate, see Weber (2014). 8  The observation that the Priority View is committed to choosing an alternative that is expectably better for no one in such cases has led some to argue that its supporters cannot wield the leveling down objection against egalitarians (Segall 2015). For further discussion of whether prioritarianism is vulnerable to some form of the leveling down objection, see Persson (2001), Brown (2003), and Porter (2011). For further discussion of the apparently impersonal nature of the Priority View, see Persson (2001). For a statement of the implausibility of a purely impersonal case for prioritarianism, see Otsuka (2015, sec. VI). 9  We assume that the children in question are in their early teens and that the disability in question is merely physical in order to ensure that Cathy can safely be regarded as the same person in both futures. 10  Empirical studies which compare people’s attitudes towards intra-​and interpersonal trade-​offs generally find a substantial difference between the two, with subjects giving a great deal more weight to improvements in well-​being for the worst-​off in interpersonal trade-​offs than in intrapersonal trade-​ offs (see, e.g., Ubel et al. 1996; Nord et al. 1999; Nord and Johansen 2014; however, see Bognar 2012 for criticism of this claim about the empirical literature). These studies therefore reveal that many people’s judgments respect the difference between the unity of the individual and the separateness of persons.

74    Michael Otsuka and Alex Voorhoeve We emphasize that this second criticism of final-​well-​being prioritarianism is distinct from the first. That is, one can disagree with our judgment that it is permissible to maximize Cathy’s expected well-​being, but still object to final-​well-​being prioritarianism’s failure to track whether a trade-​off takes place within two possible futures of the same individual or across the boundaries between people. By way of illustration, consider the following case (Voorhoeve and Fleurbaey 2012): Intra-​versus Interpersonal Case: You have two children, Denise and Edmund, who are both in their early teens. You must choose to move to the city or to a suburb. Both children will fare moderately well in the city. By contrast, the impact of moving to the suburb on your children is less clear, because there is uncertainty about the quality of the local school’s offerings in sports and art: it either, with 50% probability, has excellent sports and decent arts teaching, or, with 50% probability, has decent sports and poor arts education. In the former case, the suburb will prove especially beneficial to one “sporty” child, but leave the other child as well off as (s)he would be in the city. In the latter case, the suburb will prove somewhat burdensome to one “arty” child, but leave the other as well off as in the city. The possible benefit of the suburb to the sporty child exceeds the possible burden of the suburb to the arty child by a given amount of well-​being, d.

In order to complete the description of this case, we need to fill in which child may be benefited and which child may be burdened by a move to the suburb. Consider the following two scenarios. Intrapersonal Scenario: Denise has special interests in both sports and arts. If you moved to the suburb, she would therefore face both the chance of the benefit and the risk of the lesser burden. Edmund is unaffected by your choice. Interpersonal Scenario: Denise has special interests in sports, Edmund in arts. If the suburb would prove beneficial, Denise would reap the benefit and if it would prove burdensome, Edmund would bear the burden.

Table 3.3 represents this case (d > 0). In the intrapersonal scenario, it is the same child, Denise, who will either thrive if the suburb offers excellent training in sports and decent arts teaching, or be thwarted in her development if it offers poor arts education. Choosing the suburb therefore exposes Denise to a risk of doing less well than she might for the sake of giving her a shot at doing better than she otherwise would. (Edmund’s welfare is unaffected either way, so there is no conflict between your children’s interests.) By contrast, in the interpersonal scenario, the child who might thrive because of the suburb’s opportunities for sports (Denise) is different from the child whose interests might be thwarted there because of the lack of a decent arts education (Edmund). In the latter scenario, there is therefore a conflict of interest between Denise and Edmund. Choosing the suburb would involve imposing a risk of a burden on Edmund, which, if it materialized, would make him worse off than Denise, in order to give Denise a chance at a gain that, if it materialized, would make her better off than Edmund.

Equality Versus Priority    75 Table 3.3  F inal well-​being for the Intra-​versus Interpersonal Case One benefited in suburb (p = 0.5)

One burdened in suburb (p = 0.5)

Denise

Edmund

Denise

Edmund

City

60

60

60

60

Suburb in intrapersonal scenario

70 + d

60

50

60

Suburb in interpersonal scenario

70 + d

60

60

50

We submit that these differences make it much more difficult to justify a move to the suburb in the latter scenario. Suppose the difference d between the larger possible benefit and the smaller possible burden of the suburb is just large enough for it to be permissible to move to the suburb in the intrapersonal scenario—​the balance between the possible benefit and possible burden is such that the former outweighs the latter even after one has given any extra weight one believes is required to Denise’s situation if she is worse off, and one has compensated for the badness of inequality (if any) in this case. (We therefore do not assume that it is always permissible to maximize Denise’s expected well-​being in this scenario.) We submit that for this d, it is impermissible to move to the suburb in the interpersonal scenario, because it would not be justifiable to expose Edmund to a chance of being worse off than he might be for the sake of giving Denise this shot, thereby ensuring that she will be better off than Edmund. However, because it applies the same priority weights to increments in well-​being in intra-​and interpersonal trade-​offs, final-​well-​being prioritarianism cannot account for this difference in justifiability. It holds that for every d for which the suburb is permissible in the intrapersonal scenario, it is also permissible in the interpersonal scenario. It is interesting to compare these observations about final-​well-​being prioritarianism with the well-​known criticism of utilitarianism by David Gauthier (1963:  121–​7) and John Rawls (1999). As Rawls put it: The striking feature of the utilitarian view of justice is that it does not matter . . . how [the] sum of satisfactions is distributed among individuals any more than it matters . . . how one man distributes his satisfactions over time. . . . For just as it is rational for one man to maximize the fulfilment of his system of desires, it is right [according to the utilitarian] for a society to maximize the net balance of satisfaction taken over all of its members. The most natural way, then, of arriving at utilitarianism . . . is to adopt for society as a whole the principle of rational choice for one man . . . Utilitarianism does not take seriously the distinction between persons. (1999: 23–​4)

76    Michael Otsuka and Alex Voorhoeve Unlike utilitarianism, final-​well-​being prioritarianism is sensitive to how well lives go: it gives more weight to benefiting a person if his life will go badly than it does to benefiting a person if his life will go well. So, on this view, the boundaries between persons matter in the following sense: if one person is badly off and another well off, one has stronger moral reason to confer a benefit of a given size on the former (Otsuka 2012: 365–​6). But final-​well-​being prioritarianism fails to take seriously boundaries between persons in the following respect: it is insensitive to whether the life that goes well and the life that goes badly are possible lives of the same person or rather the lives of different people (McKerlie 1984: 233; Otsuka 2012: 368). While utilitarianism can be said to ignore the separateness of persons, final-​well-​being prioritarianism can be said to ignore the unity of the individual. The Rawlsian objection to utilitarianism is that it “adopt[s]‌for society as a whole the principle of rational choice” that is appropriate only for a single person. Final-​well-​being prioritarianism has the opposite problem: it adopts for a single person a principle that is appropriate only for a society consisting of many separate persons whose interests clash with one another (Porter 2012). Indeed, when we compare the One-​Child Case with the Two-​Child Case under Certainty, we find two key differences.11 In this Two-C​hild Case, in addition to the absence of a prudential justification to the less-​well-​off child for a move to the suburb, there is the presence of a competing-​claims-​based justification for a move to the city (Otsuka 2012:  371). To see the force of the latter, consider that one could pose the following rhetorical question about a proposed move to the suburb in the Two-​Child Case under Certainty: How . . . can [you] justify providing a benefit of a given size to someone who is already better off in order to make him better off still, when [you] could instead provide . . . [nearly as] large [a]‌benefit to someone else who is worse off? (Otsuka and Voorhoeve, 2009: 183–​4)

The competing claims approach has purchase only when there is a choice between alternatives which will either benefit one person or benefit another person. The strength of an individual’s claim is a function of how much well-​being they stand to gain and from what baseline level of well-​being, relative to another who might benefit instead (Otsuka 2012: 371; see also Lange 2017). It is again instructive to compare this observation with a distinct form of the “separateness of persons” objection first formulated by Nagel against utilitarianism. Nagel argues that utilitarianism fails to recognize the significance of competing claims:

11 

These differences are also apparent in the Intra-​versus Interpersonal Case.

Equality Versus Priority    77 [Utilitarianism] depends on an application to interpersonal conflicts of the same principles which are used to settle conflicts between reasons arising from the interests of a single person. The conditions of choice corresponding to this principle are that the chooser should treat the competing claims arising from distinct individuals as though they all arose from the interests of a single individual, himself. He is to choose on the assumption that all these lives are to be amalgamated into one life, his own. . . . But this . . . completely distorts the nature of the competing claims, for it ignores the distinction between persons. (1970: 138)

There are therefore at least two respects in which final-​well-​being prioritarianism fails to take seriously both the unity of the individual and the separateness of persons: (i) it is insensitive to the availability of prudential justifications; (ii) it is insensitive to the existence of competing claims.

3.4  Expected-​Well-​being Prioritarianism There is, however, a version of prioritarianism against which these objections have less force. On “expected-​well-​being prioritarianism,” expected well-​being is the currency of distributive justice. On this form of prioritarianism, each person’s expected well-​ being has positive and diminishing marginal moral value. Moreover, the moral value of (increments in) a person’s expected well-​being does not depend on anyone else’s expected (or final) well-​being. We ought to choose the option with the greatest sum of moral value (i.e., the greatest sum of priority-​weighted expected well-​being). Expected-​ well-​ being prioritarianism avoids the discrepancy between prudential and moral evaluation in single-​ person cases that plagues final-​ well-​ being prioritarianism. As a consequence, it treats intra-​and interpersonal trade-​offs differently: in intrapersonal trade-​offs, it requires maximizing expected well-​being; in interpersonal trade-​offs, it gives priority to whoever has lower expected well-​being. In the Two-​Child Case with Certainty, it requires choosing the city (since Ben has worse prospects), but in the One-​Child Case, it mandates choosing the suburb. In the Intra-​ versus Interpersonal Case, it holds that you ought to choose the suburb for every d > 0 in the intrapersonal scenario, but also that, for some such d, you ought not to choose the suburb in our interpersonal scenario. It therefore recognizes the difference in their justifiability. Clearly, then, expected-​well-​being prioritarianism recognizes the force of prudential justifications. It also recognizes the force of some competing claims. However, the

78    Michael Otsuka and Alex Voorhoeve following case demonstrates that there is one respect in which it does not recognize the force of competing claims: Anticorrelated Case: You have two children, Frank and Gwyneth, each of whom has a 50–​50 chance of being disabled. You know that one of them will turn out disabled, and the other healthy. You can either move them to the suburb or the city. The outcomes associated with being disabled in the city and the suburb are as in the Two-​ Child Case with Certainty; the same goes for being able-​bodied (see Table 3.4).

This Anticorrelated Case is like the One-​Child Case, and unlike the Two-​Child Case with Certainty, in this respect:  there is a prudential justification to both Frank and Gwyneth for the move to the suburb. But this Anticorrelated Case is like the Two-​Child Case with Certainty, and unlike the One-​Child Case, in this respect: if we were to learn what the outcome will be, then it would be transformed into the Two-​Child Case with Certainty (Otsuka 2012: 374). These similarities and differences can be seen in terms of our overarching theme of the difference between intra-​and interpersonal trade-​offs. In pure intrapersonal trade-​off cases such as the One-​Child Case, there are no conflicts of interest between any individuals, neither in terms of their expected well-​being nor in terms of final well-​ being. The unity of the individual then gives us reason to maximize expected well-​being. In pure interpersonal trade-​off cases such as the Two-​Child Case with Certainty, there are conflicts of interest between individuals, both in terms of their expected well-​being and their final well-​being. In such cases, the separateness of persons provides a reason to give extra weight to improvements to those who are worse off. The Anticorrelated Case is a mixed case, in which there is no conflict between individuals’ interests in expected well-​being, but there are conflicts between individuals’ final-​well-​being interests. If Gwyneth turns out to be the child with a disability, then the city is in her final-​well-​being interest while the suburb is in Frank’s. If Frank turns out to be the child with the disability, then the opposite is true. This conflict of interest gives rise to competing claims on behalf of whoever turns out to be able-​bodied and whoever turns out to be disabled. In our view, in this mixed case, you have decisive reason to respond to this conflict of interest by helping the worse-​off child and moving to the city. For how can one justify providing a benefit of a given size to someone who will already be better off in order to

Table 3.4  Final well-​being for the Anticorrelated Case Frank disabled (p = 0.5)

Gwyneth disabled (p = 0.5)

Frank

Gwyneth

Frank

Gwyneth

City

39

60

60

39

Suburb

30

70

70

30

Equality Versus Priority    79 make this person better off still, when one could instead provide nearly as large a benefit to someone else who will be worse off, and who would not even reach the (unimproved) level of the better-​off person if they (the worse-​off person) were benefited? “For he that hath, to him more shall be given, at the expense of he who hath not” is not a sound moral principle (Otsuka 2012: 376).12 However, expected-​well-​being prioritarianism ignores this conflict of interest. It responds only to the fact that the suburb is in each child’s interest ex ante and therefore mandates moving to the suburb.

3.5  A Hybrid Priority View In response to some of these objections, Parfit (2012) proposes a view that combines elements of expected well-​being and final-​well-​being prioritarianism. On this hybrid Priority View, it is true both that an increase in expected well-​being is more valuable, the lower the level of expected well-​being from which it takes place, and that an increase in final well-​being is more valuable the lower the level of final well-​being from which it takes place.13 This hybrid view avoids some of the objections raised to the other versions of prioritarianism. The expected well-​being-​prioritarian element in the view recognizes the force of prudential justifications. In the Intra-​versus Intrapersonal Case, it will therefore judge it easier to justify the suburb in the intrapersonal scenario than in the interpersonal scenario. Meanwhile, the final-​well-​being-​prioritarian element in the view ensures that one ought to look after whoever will turn out disabled in the Anticorrelated Case. Because it recognizes some differences between intra-​and interpersonal trade-​offs and gets these cases right, we consider this hybrid view more plausible than either of the other two prioritarian views surveyed here. Nonetheless, while this view boasts some of the strengths of each of its component elements, it also inherits some of their weaknesses. The final-​well-​being-​prioritarian component ensures that in risky intrapersonal trade-​off cases such as the One-​Child Case, it will sometimes recommend the alternative that is contrary to the person’s interests. It therefore remains inadequately sensitive to the presence of prudential justifications (Otsuka 2015; Voorhoeve and Fleurbaey 2016:  938–​9). Moreover, it remains inadequately sensitive to the presence of competing claims. By way of illustration of the latter point, consider the following case: Correlated Case: You have two children, Helen and Isaac, each of whom has a 50–​50 chance of being disabled. Their fates are bound together. Either both of them will turn out disabled or both will turn out healthy. You can either move them to the suburb or to the city. The outcomes associated with being disabled in the city and the 12 

For further discussion, see Fleurbaey and Voorhoeve (2013) and Frick (2013). Parfit leaves open how to balance the importance of increments in expected well-​being against increments in final well-​being, when these conflict. For an analysis of ways of doing so, see Bovens (2015) and Voorhoeve and Fleurbaey (2016). 13 

80    Michael Otsuka and Alex Voorhoeve suburb are as in the Two-​Child Case with Certainty; the same goes for being able-​ bodied (see Table 3.5).

In this Correlated Case, the expected well-​being of each of the children for each alternative is just as it is in the Anticorrelated Case. Moreover, the expected priority-​weighted final well-​being of each option is the same in both cases. On the hybrid view, you therefore have just as much reason to choose the city in the Anticorrelated Case as in the Correlated Case. We submit that this is mistaken. As we argued above, in the Anticorrelated Case, there is a conflict of interest in each state of the world. This gives rise to competing claims between the child who ends up able-​bodied and the child who will end up disabled, the latter of which, we argued, is stronger and ought to be satisfied. By contrast, in the Correlated Case, since the children’s fates are bound together, there is no conflict of interest between them. There are therefore no such competing claims. The presence of competing claims in the Anticorrelated Case and their absence in the Correlated Case entails that you have stronger reason to choose the city in the former. Because it fails to recognize this, hybrid prioritarianism fails to adequately respond to the presence or absence of competing claims. A further difference between these cases forms the basis of a final objection to all the prioritarian views we have discussed (Broome 1989; 2015). When one child’s good fortune is the flipside of the other’s ill fortune, as in the Anticorrelated Case, moving to the suburb increases inequality between them. By contrast, when the children experience good or ill fortune together, as in the Correlated Case, neither opting for the city nor choosing the suburb generates inequality. If such inequality is unfair and therefore bad, then, contrary to all prioritarian views, we have a reason against moving to the suburb in the Anticorrelated Case which we do not have in the Correlated Case. In sum, the hybrid prioritarian view proposed by Parfit (2012), while superior to both final-​well-​being prioritarianism and expected-​well-​being prioritarianism, is subject to the following objections: (i) a failure to adequately respond to the presence of prudential justifications; (ii) a failure to adequately respond to the presence of competing claims; (iii) a failure to respond to unfair inequality. We shall now examine to what extent egalitarian views can avoid these objections. Table 3.5  Final well-​being for the Correlated Case Helen and Isaac able (p = 0.5)

Helen and Isaac disabled (p = 0.5)

Helen

Isaac

Helen

Isaac

City

60

60

39

39

Suburb

70

70

30

30

Equality Versus Priority    81

3.6  Egalitarianism under Risk As mentioned, egalitarians are pluralists, because all sensible egalitarians care about improving well-​being as well as about reducing inequality. Recent work has identified one family of pluralist egalitarianism as especially attractive, because it handles the aforementioned cases well (Fleurbaey 2010). This family evaluates each distribution by its “equally-​distributed equivalent,” or EDE.14 To illustrate this idea, consider first a case of certainty, such as moving to the suburb in our original Two-​Child Case under Certainty, where Ann’s well-​being is 70 and Ben’s is 30. Suppose that you should be indifferent between moving to the suburb and a move to a hypothetical town where they each have a well-​being level of 45. Then 45 is the equally-​distributed equivalent of moving to the suburb. Next suppose that you should be indifferent between moving to the city, where Ann’s well-​being is 60 and Ben’s is 39, and moving to a hypothetical locale where each has a well-​being of 47. Then 47 is the EDE of moving to the city. Since the latter is higher than the EDE of the suburb, you should move to the city. More generally, under certainty, when an alternative leads to inequality and one ought to be averse to this inequality, then the EDE of this alternative is less than the average well-​being generated by that alternative. When an alternative contains no inequality, then the EDE is, naturally, the average well-​being generated by that alternative. In risky cases, insofar as we are concerned with final well-​being, this approach tells us to evaluate each alternative as follows. First establish the EDE for each possible distribution of final well-​being that may result from the alternative. Then take the probability-​weighted sum of these values. A risky alternative that invariably yields the same anonymized pattern of inequality in final well-​being, such as moving to the suburb in the Anticorrelated Case (with one child at 70 and one at 30), is then evaluated by the value of the EDE of this distribution, which will be less than the average expected well-​being. But a risky alternative that yields equality in each possible outcome, such as moving to the suburb in the Correlated Case, is evaluated at the expected well-​being of each person in the population. In the absence of inequality, the view is therefore utilitarian; in the presence of inequality, it gives weight to both reducing inequality and increasing well-​being. Such a final-​well-​being egalitarian view gets several of our cases right. In our One-​ Child Case, it favors moving to the suburb, because that is prudentially most valuable for the child. In the Two-​Child Case with Certainty, it mandates moving to the city, because this reduces inequality at modest cost. In the Correlated Case, it requires the

14  The EDE was introduced in welfare economics by Kolm (1969) and Atkinson (1970). However, as mentioned in fn 5, the social welfare functions they proposed were prioritarian, not egalitarian. The distinctively egalitarian credentials of an EDE social welfare function become apparent when one considers cases under risk.

82    Michael Otsuka and Alex Voorhoeve move to the suburb, because it is to the expected advantage of each and does not generate any inequality. By contrast, in the Anticorrelated Case, it mandates the city, because this reduces inequality. It therefore marks the difference between intrapersonal trade-​offs that do not generate inequality on the one hand and interpersonal trade-​offs on the other. However, our Intra-​versus Interpersonal Case shows that an exclusive concern with the distribution of final well-​being is not enough to fully respect the difference between the unity of the individual and the separateness of persons (Voorhoeve and Fleurbaey 2012). Final-​well-​being egalitarianism alone cannot draw a distinction between the suburb in the intrapersonal scenario and the interpersonal scenario. The reason is that in both scenarios, the suburb leads to the same inequality in final well-​being. This issue can be dealt with by extending egalitarianism to cover not merely the distribution of final well-​being, but also the distribution of expected well-​being.15 Such a concern for inequalities in chances meshes well with the egalitarian concern for fairness. A given inequality in final well-​being is less unfair when each person has a chance to end up better off than when the worse off have no such chance (Broome 1990; Arneson 1997; Temkin 2001). In the Intra-​versus Interpersonal Case, there is greater inequality in expected well-​being in the inter-​than in the intrapersonal scenario. Since such inequality in expected well-​being partly determines the (un)fairness of the alternatives, choosing the suburb in the intrapersonal scenario is easier to justify because it is less unfair to Edmund. More generally, this hybrid egalitarianism, which favours equality in both expected well-​being and final well-​being, deals successfully with all cases we have considered. In these cases, it is therefore duly sensitive to (i) the availability of prudential justifications; (ii) the presence of competing claims; and (iii) the unfairness of inequality. In so doing, it fully respects the difference between the unity of the individual (which gives us reason to choose in a person’s expected interest) and the separateness of persons (which gives us reason to give extra weight to the claims of those who are less well off than others with whom their claims compete and, we submit, to reduce unfair disadvantage). We therefore regard it as superior, in these respects, to prioritarian views.16 It matters that some are worse off in comparison with others, both in prospect and in final outcome.

15  For details on how to develop the EDE for hybrid egalitarianism, see Voorhoeve and Fleurbaey (2016). For an alternative egalitarian response to the Intra-​versus Interpersonal Case, see Hyams (2015). Vallentyne (2002) proposes a view which is exclusively concerned with inequalities in expected well-​ being. Such a view is implausible, in our view, for the same reason that expected utility prioritarianism was shown to be implausible in section 3.4: it does not regard choosing the suburb in the Anticorrelated Case as less justifiable than choosing the suburb in the Correlated Case. It therefore fails to respond to the presence of competing claims and of unfair outcome inequality in the Anticorrelated Case. 16  More precisely, we regard it as superior to any prioritarian view that respects Separability, which is generally regarded as a core tenet of such a view. We have not here considered versions of prioritarianism that violate Separability because they pluralistically combine a commitment to prioritarian value with a sensitivity to comparative considerations. One such version has been proposed by Andrew Williams (2012). On this view, prioritarian weights are applied to final well-​being, but such weighting is triggered only in the presence of competing claims; when there are no competing claims, a person simply has a

Equality Versus Priority    83

Acknowledgments We are grateful to Thomas Hurka, Benjamin Lange, Serena Olsaretti, and Asbjørn Schmidt for comments and to Jerod Coker for research assistance.

References Adler, M. (2012). Well-​Being and Fair Distribution. Oxford: Oxford University Press. Anderson, E. (1999). “What Is the Point of Equality?” Ethics 109: 287–​337. Arneson, R. (1997). “Postscript to ‘Equality and Equal Opportunity for Welfare’,” in L. Pojman and R. Westmoreland (eds) Equality: Selected Readings. Oxford: Oxford University Press, pp. 238–​41. Arneson, R. (1999). “Egalitarianism and Responsibility.” The Journal of Ethics 3: 225–​47. Atkinson, A. (1970). “On the Measurement of Inequality.” Journal of Economic Theory 2: 244–​63. Bognar, G. (2012). “Empirical and Armchair Ethics.” Utilitas 24: 467–​82. Bovens, L. (2015). “Concerns for the Poorly Off in Evaluating Risky Prospects.” Economics and Philosophy 31: 397–​429. Broome, J. (1989). “What is the Good of Equality?” in J. Hey (ed.) Current Issues in Microeconomics. London: Macmillan, pp. 236–​62. Broome, J. (1990). “Fairness.” Proceedings of the Aristotelian Society 91: 87–​101. Broome, J. (2015). “Equality versus Priority: A Useful Distinction.” Economics and Philosophy 31: 219–​28. Brown, C. (2003). “Giving Up on Levelling Down.” Economics and Philosophy 19: 111–​34. Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99: 906–​44. Crisp, R. (2011). “In Defence of the Priority View:  A Response to Otsuka and Voorhoeve.” Utilitas 23: 105–​8. Fleurbaey, M. (2010). “Assessing Risky Social Situations.” Journal of Political Economy 118: 649–​80. Fleurbaey, M. and Voorhoeve, A. (2013). “Decide as You Would with Full Information! An Argument against ex ante Pareto,” in N. Eyal, S. Hurst, O. Norheim, and D. Wikler (eds) Inequalities in Health: Concepts, Measures, and Ethics. Oxford: Oxford University Press, pp. 113–​28. Frick, J. (2013). “Uncertainty and Justifiability to Each Person:  A Response to Fleurbaey and Voorhoeve,” in N. Eyal, S. Hurst, O. Norheim, and D. Wikler (eds) Inequalities in Health: Concepts, Measures, and Ethics. Oxford: Oxford University Press, pp. 129–​46. claim to whatever maximizes his expected well-​being. Since the moral importance of increments in a person’s well-​being depends on whether there are others whose claims compete with his, this view violates Separability. Moreover, on our conception of the conditions under which there are competing claims, this view is not embarrassed by any of the cases so far considered. In the One-​Child Case and the Correlated Case, it favors the suburb, since this is in the expected interest of each and there are no competing claims. In the Two-​Child Case under Certainty and the Anticorrelated Case, it favors the city, since these cases involve competing claims and we therefore ought to have special concern for those who are worse off. Given its success in dealing with the cases we consider in this chapter, such a pluralistic version of prioritarianism holds promise as capturing, in a different manner from egalitarianism, the moral significance of the fact that some are worse off in comparison with others. See Weber (2014) and Lange (2017) for further discussion.

84    Michael Otsuka and Alex Voorhoeve Gauthier, D. (1963). Practical Reasoning. Oxford: Clarendon Press. Greaves, H. (2015). “Antiprioritarianism.” Utilitas 27: 1–​42. Hyams, K. (2015). “Hypothetical Choice, Egalitarianism, and the Separateness of Persons.” Utilitas 27: 217–​39. Kolm, S.-​C. (1969). “The Optimal Production of Social Justice,” in J. Margolis and H. Guitton (eds) Public Economics. London: Macmillan, pp. 145–​200. Lange, B. (2017). “Restricted Prioritarianism or Competing Claims?” Utilitas 29: 137–​52. McCarthy, D. (2008). “Utilitarianism and Prioritarianism II.” Economics and Philosophy 24: 1–​33. McKerlie, D. (1984). “Egalitarianism.” Dialogue 23: 223–​37. Nagel, T. (1970). The Possibility of Altruism. Princeton, NJ: Princeton University Press. Nagel, T. (1979). “Equality,” in Mortal Questions. Cambridge:  Cambridge University Press, pp. 106–​27. Nord, E., J.-​L. Pinto Prades, J. Richardson, P. Menzel, and P. Ubel (1999). “Incorporating Societal Concerns for Fairness in Numerical Valuations of Health Programmes.” Health Economics 8: 25–​39. Nord, E. and R. Johansen (2014). “Concerns for Severity in Priority Setting in Health Care: A Review of Trade-​Off Data in Preference Studies and Implications for Societal Willingness to Pay for a QALY.” Health Policy 116: 281–​8. Norman, R. (1998). “The Social Basis of Equality,” in A. Mason (ed.) Ideals of Equality, Oxford: Blackwell, pp. 37–​51. O’Neill, M. (2008). “What Should Egalitarians Believe?” Philosophy & Public Affairs 36: 119–​56. O’Neill, M. (2012). “Priority, Preference, and Value.” Utilitas 24: 332–​48. Otsuka, M. (2012). “Prioritarianism and the Separateness of Persons.” Utilitas 24: 365–​80. Otsuka, M. (2015). “Prioritarianism and the Measure of Utility.” Journal of Political Philosophy 23: 1–​22. Otsuka, M. and A. Voorhoeve (2009). “Why It Matters that Some are Worse Off than Others: An Argument against the Priority View.” Philosophy and Public Affairs 37: 171–​99. Parfit, D. (1995). “Equality or Priority?” The Lindley Lecture. The University of Kansas. Parfit, D. (2012). “Another Defence of the Priority View.” Utilitas 24: 399–​440. Persson, I. (2001). “Equality, Priority, and Person-​Affecting Value.” Ethical Theory and Moral Practice 4: 23–​39. Porter, T. (2011). “Prioritarianism and the Levelling Down Objection.” Ethical Theory and Moral Practice 14: 197–​206. Porter, T. (2012). “In Defence of the Priority View.” Utilitas 24: 349–​64. Rabinowicz, W. (2002). “Prioritarianism for Prospects.” Utilitas 14: 2–​21. Rawls, J. (1999). A Theory of Justice, rev. edn. Cambridge, MA: Harvard University Press. Rendall, M. (2013). “Priority and Desert.” Ethical Theory and Moral Practice 16: 939–​51. Segall, S. (2015). “In Defense of Priority (and Equality).” Philosophy, Politics and Economics 14: 343–​64. Sen, A. (1973). On Economic Inequality. Oxford: Oxford University Press. Tawney, R. (1964). Equality, 5th edn. London: Unwin. Temkin, L. (1983). Inequality. Ph.D. Thesis, Princeton University. Temkin, L. (2001). “Inequality:  A Complex, Individualistic, and Comparative Notion.” Philosophical Issues 11: 327–​53. Temkin, L. (2003). “Equality, Priority, or What?” Economics and Philosophy 19: 61–​88.

Equality Versus Priority    85 Ubel, P.  A., G. Loewenstein, D. Scanlon, and M. Kamlet (1996). “Individual Utilities are Inconsistent with Rationing Choices:  A Partial Explanation of Why Oregon’s Cost-​ Effectiveness List Failed.” Medical Decision-​Making 16: 108–​16. Vallentyne, P. (2002). “Brute Luck, Option Luck, and Equality of Initial Opportunities.” Ethics 112: 529–​57. Voorhoeve, A. and M. Fleurbaey (2012). “Egalitarianism and the Separateness of Persons.” Utilitas 24: 381–​98. Voorhoeve, A. and M. Fleurbaey (2016). “Priority or Equality for Possible People?” Ethics 126: 929–​54. Weber, M. (2014). “Prioritarianism.” Philosophy Compass 9(11): 756–​68. Weirich, P. (1983). “Utility Tempered with Equality.” Noûs 17: 423–​39. Williams, A. (2012). “The Priority View Bites the Dust?” Utilitas 24: 315–​31.

Chapter 4

Su fficien c y a nd N eeds-​B ased A pproac h e s Gillian Brock

4.1 Introduction The history of philosophical discussions about human nature and justice are replete with references to our human needs.1 Many philosophers believed that what we need to live well as human beings should inform our ideas about how to organize collective affairs and live justly with one another.2 From this perspective, the study of human needs is altogether crucial to answering the question of what justice, including distributive just­ice, requires. Just about everything concerning contemporary discussions of distributive justice is deeply contested, including what the primary question of distributive justice is, or should be. Distributive justice is, at a very general level, concerned with the fair distribution of benefits and burdens. A full account of distributive justice will need to specify which benefits and burdens are relevant, what account of fairness is being assumed, who is owed justice, and what the scope of justice ought to be. Theorists also interpret the problem of distributive justice in quite specific ways, sometimes giving different answers for different contexts. For John Rawls (the most influential contemporary political philosopher), the problem of distributive justice is “how are the institutions of the basic structure to be regulated as one unified scheme of institutions so that a fair, efficient, and productive system of social cooperation can be maintained over time, from one generation to the next?” (Rawls 2001: 50). While Rawls uses this approach for domestic justice, he adopts an entirely different perspective at the international level, where much contemporary debate is focused (Rawls 1999). For Rawls 1 

For a good overview, see Patricia Springborg (1981). To pick out just some examples, these would include Epicurus, Plato, Aristotle, Rousseau, Hegel, Marx, and Sartre. For some excellent historical treatment of the topic, see Springborg (1981). 2 

Sufficiency and Needs-Based Approaches    87 a just world order “is perhaps best seen as a society of peoples, each people maintaining a well-​ordered and decent political (domestic) regime, not necessarily democratic but fully respecting human rights” (Rawls 2001: 13). Rawls believes the domestic and international contexts are different, as in the latter, individuals belong to different societies (or peoples) and this marks an important consideration as we work out what distributive justice demands. Rawls’s view also mirrors the history of discussion about distributive justice. It is common to assume the natural order for exploring topics of distributive just­ ice is first at a state level and then at an international one. Until about the mid-​1990s, most theorists investigating the nature of distributive justice implicitly assumed the state context, and that the project of determining what distributive justice consists in was typically confined only to within one state. But over the past two decades theorists taking up a cosmopolitan perspective have challenged whether this focus is appropriate. In contemporary discussion of distributive justice, several questions are often distinguished, and answers are often deeply contested, especially to the following three questions: • To whom is justice owed? • What should we be distributing? Should we, for instance, be distributing resources, well-​being, primary goods, capabilities, social/​political status, or freedom? • Should distribution follow any patterns, such as, according to ideas of equality, priority, sufficiency, or desert? Very roughly, egalitarian accounts of justice are concerned with equality of relevant benefits. Prioritarianism holds that it is more important to provide benefits to those who have less. Sufficientarians typically claim that justice requires that everyone get an adequate (or sufficient) amount of specified goods, such as what is required to meet basic needs, and that aiming at equality beyond that sufficiency threshold is not an appropriate focus (Frankfurt 1987). While prioritarianism and sufficientarianism generally give priority to those who are worse off, desert-​based approaches, by contrast, distribute according to a criterion of desert, typically thought not to include concern for neediness. In this chapter I show how needs have a distinctive and important role to play in ideas about distributive justice and are an essential part of understanding distributive justice. We begin in section 4.2 by examining the original defense of sufficientarianism offered by Harry Frankfurt. In it Frankfurt argued that sufficientarianism is superior to egalitarianism as an account of economic justice. In section 4.3 we critically discuss the debate between egalitarians and sufficientarians in more detail. In section 4.4 I argue that both equality and sufficiency have important roles to play in an account of distributive justice, and that neither side has done an adequate job of recognizing all the important insights their interlocutors can offer. For instance, attending to equality can sometimes be required if we are concerned about sufficiency, and vice versa. I argue that there is scope for both comparative and noncomparative analyses in working out what distributive justice requires. When we elaborate on sufficiency accounts, we will have to consider comparative aspects. And there is an important role for noncomparative analysis

88   Gillian Brock as well—​a role which consideration of our human needs fills well—​in deciding which interests or dimensions of human experience we should aim to equalize. So the issue is not whether we should prefer equality to sufficiency. On this way of exploring the topic, there is a proper role for both in an account of distributive justice, and the interesting work concerns tracing the important and subtle ways in which they are connected. This, I will suggest, is at least one of the important areas in which future research might focus. While it may be the case that within the domain of social justice a fully just society is structured to enable each citizen to get what they need to enjoy a flourishing life to the best of their abilities, we often find we are quite a long way from being able to provide what a fully just society should, especially given resource constraints. We may need to realize our ideals of distributive justice progressively, conditioning them to circumstances and making gains where these are possible. Here, ensuring all have enough for a decent life will serve as an important focus for our attention, especially when there is wide-​scale remedial deprivation. Which needs are important to address in order for citizens to enjoy prospects for decent lives? If addressing needs ought to be an important focal point for public policy, we need some criteria for determining the significant ones. We turn in sections 4.5–​4.7 to a more systematic inquiry concerning which needs matter, how they do, how they can generate distinctive arguments for why we have distributive justice obligations, and how analysis of needs can provide important insights for the domain of distributive just­ ice. Section 4.5 covers the issue of which needs warrant normative attention. Section 4.6 examines some needs-​specific arguments concerning why we have responsibilities to address needs. Section 4.7 covers how to distribute according to needs. In section 4.8 I show how concern for need permeates several other prominent debates, offer some thoughts about further areas of useful inquiry, and underscore the central claims made in the chapter. We see how the concept of needs is an essential one in understanding distributive justice. There is significant work that it does well and no other concept does better. There is a role for needs to play in matters of distributive justice that does not displace concern with several other important criteria relevant to distributive justice.

4.2  The Origins of Sufficientarianism According to sufficientarian accounts of distributive justice, everyone should have enough. This position is often contrasted with positions that hold that everyone should have the same. One fairly crude way of expressing the core insight is that distributive just­ice requires that we bring it about that as many people as possible have enough (i.e., are over some threshold of adequacy). Further questions that sufficientarians must address to give an adequate account include: 1) Enough of what? 2) Enough for what? 3) How much is enough?

Sufficiency and Needs-Based Approaches    89 The origins of the contemporary debates between sufficientarians and egalitarians can be traced to an influential article by Harry Frankfurt, namely “Equality as a Moral Ideal.” Frankfurt takes as his target a particular form of egalitarianism, namely economic egalitarianism. As he understands the view, it is “the doctrine that it is desirable for everyone to have the same amounts of income and of wealth (for short, ‘money’)” (Frankfurt 1987: 21). Though it is widely believed that economic egalitarianism should be a “significant priority,” Frankfurt considers this to be a mistake. His view is that: . . . economic equality is not, as such, of particular moral importance. With respect to the distribution of economic assets, what is important from the point of view of morality is not that everyone should have the same but that each should have enough. If everyone had enough, it would be of no moral consequence whether some had more than others. I shall refer to this alternative to egalitarianism . . . as sufficiency. (Frankfurt 1987: 21–​2)

Frankfurt acknowledges that the idea of having an equal share is “more patent and accessible than the concept of having enough” (Frankfurt 1987: 24). It is not at all clear, for instance, what sufficiency means or how it should be applied. But this difficulty gives us inadequate grounds for “adopting an incorrect doctrine in preference to it” (Frankfurt 1987: 24). Indeed, one of his main aims for the article “is to suggest the importance of systematic inquiry into the analytical and theoretical issues raised by the concept of having enough, the importance of which egalitarianism has masked” (Frankfurt 1987: 24). Some of the arguments for sufficientarianism draw on the alleged failures of equality to give good advice in unusual conditions, such as extreme scarcity. Consider . . . that there is enough of a certain resource (e.g., food or medicine) to enable some but not all members of a population to survive. Let us say that the size of the population is ten, that a person needs at least five units of the resource in question to live, and that forty units are available. If any members of this population are to survive, some must have more than others. An equal distribution, which gives each person four units, leads to the worst possible outcome, namely, everyone dies. Surely in this case it would be morally grotesque to insist upon equality! (Frankfurt 1987: 30)

Whether an egalitarian is actually committed to this view is quite debatable,3 but at any rate Frankfurt’s claim is that attention to sufficiency gives better guidance here: we surely want to ensure that at least some have enough to survive. Attention to what is enough will get us a preferred outcome. Another argument Frankfurt offers for sufficientarianism is that he thinks it is often the case that we misidentify our reasons for concern in situations where some have 3 

An egalitarian might rather advise that in such cases everyone should have an equal chance to survive so a lottery mechanism might be a better way to resolve the distribution problem rather than giving everyone strictly equal amounts. The case involves extreme scarcity and so unusual rationing procedures may be advisable.

90   Gillian Brock much less, as involving concern with inequality. However, in such cases it is frequently really the importance of having enough and therefore concern for insufficiency that does the normative work. What troubles us is that some fall well short of a threshold of adequacy, while others are greatly beyond it. Unmet need often supplies much of the normative force when claims are pressed in virtue of concern for inequality. While some inequality should rightly command our normative attention, not all inequality should. We need not be troubled by inequalities between millionaires and billionaires, and the fact that they are both well beyond some threshold of having enough for a decent life operates as an important consideration explaining why the inequality is unproblematic. Appreciating that it is really insufficiency that grounds our normative attention when concern would be appropriate means we should endorse sufficientarian accounts of obligations of distributive justice in preference to egalitarian accounts. According to Frankfurt, the fundamental error that egalitarians make is that it is . . . important whether one person has less than another regardless of how much either of them has. This error is due in part to the false assumption that someone who is economically worse off has more important unsatisfied needs than someone who is better off. In fact the morally significant needs of both individuals may be fully satisfied or equally unsatisfied. Whether one person has more money than another is a wholly extrinsic matter. (Frankfurt 1987: 35)

At least two remarks are worth underscoring from this passage. First, Frankfurt helpfully directs our attention to a neglected feature: we do need to take account of what a person has already in order to know whether any remaining inequalities are ones that should trigger moral attention. Second, Frankfurt draws attention to the fact that to know whether these individuals’ lives should activate our concern, we need to know more about “the morally significant needs” of both, as they may both be quite unsatisfied or, the poor person’s might, contrary to first impressions, be better satisfied than the rich person’s. So here we see that figuring out what someone needs to enjoy a decent or reasonably good life (and how well her needs are being addressed) does play an important role: it serves as an important benchmark for knowing what to make of any residual inequalities there might be. What is the relevant benchmark? Frankfurt says: To say that a person has enough money means that he is content, or that it is reasonable for him to be content, with having no more money than he has. And to say this is, in turn, to say something like the following: the person does not (or cannot reasonably) regard whatever (if anything) is unsatisfying or distressing about his life as due to his having too little money. (Frankfurt 1987: 37–​8)

While Frankfurt makes several compelling points in “Equality as a Moral Ideal,” in another seminal article that attempts to defend the sufficientarian view, “Equality and

Sufficiency and Needs-Based Approaches    91 Respect,” Frankfurt’s position is much less persuasive. Frankfurt makes several problematic claims there, such as: “I categorically reject the presumption that egalitarianism, of whatever variety, is an ideal of any intrinsic moral importance” (Frankfurt 1997: 1). One of the major reasons he offers as to why we should reject this view is because egalitarianism invites comparisons with others, and this is entirely inessential in order for people to be content. What is important is whether one has enough to pursue one’s aims and aspirations. Resource egalitarianism encourages people to compare their lives and resource bundles with others, fostering alienation, promoting resource fetishism, and diverting attention to unimportant matters. Now, it is one thing to make a case for the importance of sufficientarianism; it is quite another to reject the importance of equality. These views can be usefully combined and are not necessarily rivals in a complete theory of distributive justice, as we see in section 4.4.

4.3  Sufficiency versus Equality:  Some Analysis Egalitarians responded to these provocative ideas.4 They marshaled support for their view by rejecting Frankfurt’s particular account of economic equality, presenting a different way to view their project and hence successfully reorienting the discussion. Egalitarians might, for instance, observe that within nations there are large gaps between the life prospects of the most well off and the least well off on many dimensions, such as income, wealth, education, health, employment, access to healthcare and many other aspects of human well-​being. It is these disparities we should work to eliminate. A better state of affairs is one in which people enjoy similar life prospects. We might call this species of egalitarianism “equality of life prospects.” Those committed to equality of life prospects might encourage us to consider a society in which there are divisions into income classes that include one group that is very rich and another that is very poor. The very poor live in fairly repugnant conditions characterized by high mortality rates from preventable disease that rarely afflict those in the very rich class; poor access to healthcare, education, or enriching social or cultural activities; exposure to high risks of crime due to an inability to buy protections that the rich regularly do; and so forth. So the rich enjoy vastly different prospects for their lives than the poor do. A resource egalitarian might then urge us to accept the principle that, other things being equal, it is desirable for people to have equal resources insofar as they promote equality of life prospects, and so mandating transfers from the rich to the poor would be justified in such cases. 4 

For excellent comprehensive treatment of the debate between sufficientarians, egalitarians, and prioritarians, see Paula Casal (2007) and Liam Shields (2012).

92   Gillian Brock The proponent of sufficiency could respond to this attempted defense by pointing out that in the account just offered, the poor manifestly do not have enough for a decent life. That is what is troubling about the situation, rather than that others in the society have more than enough for a decent life. And to add further credence to the argument, imagine the following case. All members of a society are at the level of the very rich, in the example just described, except for a class of super rich that are even better off than the very rich, and they are better off by the same relative gap as in the first example. Now, the argument goes, the super rich are obligated to transfer resources to the very rich, by the same egalitarian logic that applied in the first case. But this recommendation is both unappealing and unconvincing, many of us will believe. So it is not, after all, facts about inequality that motivate our intuition towards redistribution, but rather facts about inadequacy or insufficiency. In the case of the very and super rich, the worse off have enough for a decent life, so they do not trigger our moral concern, whereas in the first case the worst off do not have enough for a decent life so our moral concern should be rightly activated. Even if the sufficientarian is on strong ground with such examples, there are a number of other areas that are considered to be quite problematic for sufficientarians, especially for the case that Frankfurt marshals for this view canvased in section 4.2. I consider some of these next. (i) Is resource egalitarianism fetishistic, as Frankfurt contends? The egalitarian picture presented in defense of equality of life prospects would not seem to be committed to such a view. Rather, we saw that inequality of resources was abhorred because of the instrumental effects this inequality has for people’s well-​being. (ii) Are comparisons with others alienating, again as Frankfurt argues? Here we might differentiate between different levels of comparisons. Granted that an individual who is constantly comparing herself to others would probably do better not to fixate on such comparisons as this, which could be destructive on many levels, undermining her self-​esteem, a sense of her own self-​worth, possibly fostering embitteredness, not to mention being alienating. But this is not the only level of comparative analysis that might be salient to distributive justice. On some views, first-​personal or subjective comparisons may not be salient at all. Some, such as members of the government responsible for social policy, might rightly be charged with the responsibility of making such comparisons in pursuing social justice and delivering on government’s core functions. When occupying certain public roles, making comparisons might be entirely appropriate, as such comparisons can reveal important insights. They might tell us that there are important institutional failures that are giving rise to different remedial outcomes, or they might highlight what others have made possible through social, political, or technological changes, and so what else might be achievable in enabling all to live more contented lives.

Sufficiency and Needs-Based Approaches    93 (iii) Does Frankfurt employ a sliding standard? If we go along with the analysis of what matters according to sufficiency theorists, what is important is that everyone should have enough. We must then confront the issue of what it is to have enough. In some cases having enough to escape absolute poverty seems to be the relevant yardstick. In other cases a more ambitious goal is in play. Frankfurt (1987; 1997) slides between the two frequently. It seems that his most considered view is that a person has enough when she is content with what she has and would not actively seek out more. But for most people this level of contentment is only achieved when one is far above a poverty threshold, when one has had (and has pursued) numerous opportunities for enriching activities that mean a reasonable level of affluence and leisure time are available. If the standard we are to invoke constitutes a high threshold level then it is not so easy to see why this standard does indeed present as morally compelling an ideal as Frankfurt thinks (or a standard that is not shared by egalitarian aspirations). At any rate, we seem to have at least two kinds of sufficiency criteria in play in Frankfurt’s account. More generally, critics object to what they perceive as the arbitrariness of picking a particular threshold, where none can be specified in a way sufficiently determinate to do the work expected of it (Arneson 2005; Goodin 1987). (iv) Is maximizing sufficiency a normatively compelling ideal? If the aim is to move as many across the threshold of adequacy as possible, critics complain that we might be tempted to focus on those who are quite near to the threshold rather than those further away, as helping them across will involve least cost. Here we might object to the idea that we will be required to benefit those who are better off by relatively small amounts, rather than those who are worse off and need much greater help (Arneson 2005). Also, we might object to the fact that those above the threshold are of no concern (Arneson 2005; Casal 2007). So, there are some problems with the goal of trying to move people across the sufficiency threshold but not caring about them once they have done so, or indeed not caring much about them if they are too far below the threshold level ever to be able to cross it. In the former case the level of concern seems arbitrary, and in the latter case it seems insufficient, and indeed morally defective. (v) Has Frankfurt rightly captured the connection between sufficiency and equality? Sufficientarians who take the view that equality has no moral appeal, fail to consider the myriad ways in which we should care about equality which are related to our concern for sufficiency. For one thing, there are different ways to ensure everyone has enough, and some of these are more egalitarian in spirit. Suppose we could ensure everyone’s needs are met through a welfare state that is supported from taxing citizens whose needs are already all met. We have a choice between two different programs: one is a progressive tax which taxes wealthier citizens at a higher percentage to less wealthy citizens and the other is to impose a flat tax on everyone, regardless of income. The egalitarian is likely to prefer the first option. The sufficientarian would seem to be indifferent between these two: when

94   Gillian Brock everyone is above a threshold of sufficiency, they have no further guidance to offer. But we should have reasons to care about ensuring burdens are fairly distributed, even when people have enough, and an egalitarian might be right to think that a progressive tax is the fairer scheme in such cases. Is sufficiency for all a moral ideal that should have considerable normative force, whereas equality among all is not a compelling ideal? I don’t believe so. In order to see why not we need to look at the problem afresh.

4.4  Connections between Equality and Sufficiency: More Subtle and Complex than First Appreciated 4.4.1 Both Equality and Sufficiency? There are two different ways to look at what we owe others, a comparative versus a non-​comparative approach. In my view, we need both perspectives to gain proper insight into the nature of our distributive justice responsibilities towards others. Let us briefly explore each in turn, starting with the comparative perspective and, to simplify and focus attention on the way the question usually arises, let us consider the question of what we owe our compatriots. In working towards an answer we might say that all compatriots deserve equality of some kind. There are a few versions that seem like good candidates when fleshing out the relevant equality. These include viewing the equality that matters in terms of: (i) equal consideration of interests; (ii) standing in relations of equality with one another; (iii) equal positive freedom (as needed to avoid domination). Even if we accept a particular ideal of equality, this will still not tell us everything we want to know to have a determinate answer to the question of what equality itself demands. We need to supplement our analysis with more substantive views about which interests, dimensions of human experience, or activities matter and should be a focus, when we aim at equality. We need to ask questions such as: What do human beings need to function adequately? What does someone need to live well? What basic human entitlements should all people enjoy? The first question is often associated with an approach that explores basic human needs. The second question might invite us to consider what human beings need to flourish, and this kind of analysis is often associated

Sufficiency and Needs-Based Approaches    95 with a capabilities-​based approach. And the third question often invites analysis along human rights lines. In order to complete our answers concerning our distributive justice responsibilities we must address such additional questions. All these further issues add valuable dimensions to our analysis in arriving at a comprehensive view of what distributive justice requires. There is, in my view, not just one central question to answer when we try to arrive at a comprehensive account of our distributive justice obligations. Recall that our primary question of distributive justice may be phrased as something like this: What is a just basic societal structure? We will need to engage in both comparative and noncomparative investigations to address this issue. Consider a sophisticated answer to what distributive justice consists in. Rawls’s first principle of justice requires that all have a fully adequate scheme of basic liberties, while the second outlines when social and economic inequalities might be justified. Notice that even when Rawls addresses the second issue he needs the noncomparative notion of primary goods which, much like human needs, are versatile all-​purpose means—​things all people should want no matter what goals they set for themselves. Rawls’s answer to the distributive justice question incorporates both comparative and noncomparative aspects, concern for equality and concern for sufficiency. Indeed, it would seem that any sophisticated answer to the question of what distributive justice requires must include both. Rawls’s theory also includes a core insight that many needs-​ theorists embrace, namely that any reasonable view of a just society must underwrite a social minimum, and that will be informed by an account of human needs (Rawls 2001: 47–​8). Notice also that once we start exploring questions that appear to be more sufficiency-​oriented, we frequently need to look at comparative factors, because sometimes having enough of some good requires having an equal amount of it relative to others. Consider having enough freedom of political participation. What is it to have enough of this good? When we explore this question in more detail, we soon arrive at an answer that suggests having an equal level of political participation to others is the right amount. So sometimes sufficiency requires equality; to have enough of some goods (such as voting opportunities in national elections) we might need to have an equal distribution of it. There are plenty of other reasons why equality might matter to sufficiency. There is quite compelling empirical evidence that shows that we need to be concerned about intra-​societal inequalities, as groups use their positions of power to entrench advantage (Green 2008; Green 2012; Riddell 2012). In societies that are very unequal, as an empirical matter, those living in poverty will find it difficult to reliably move beyond poverty in a sustained fashion. Less appreciated might be why concern for what people need or sufficiency might matter to equality. Focusing on needs or sufficiency often gives us a focal point for deciding on the adequacy standard that should apply in particular domains. For

96   Gillian Brock instance, we might be interested in giving people equally good educations, but there are still further questions to be addressed concerned with adequacy relative to particular goals. What skill sets should they have acquired to pursue which goals? Focusing on adequacy standards to perform certain tasks must of necessity be part of our inquiry within the egalitarian project (Anderson 2007; Satz 2007). Furthermore, as we saw with Rawls’s project, even when we aim to equalize, we need to have at least a thin theory of what human beings need or what ought to matter to human beings, so we get a sense of the interests, goods, or aspects of human lives that we will care about in ensuring the relevant equality prevails. Rawls chooses the term “primary goods” for this idea. At least a proto account of what humans need is in the background of any comprehensive account of distributive justice, or so I claim. Assuming we see the value of both equality and sufficiency, how might we combine a concern for each in one theory in a coherent way? Relational egalitarianism (roughly, standing in relations of equality with one another) provides at least one coherent way to embed our concern with sufficiency within an egalitarian framework. Consider Elizabeth Anderson’s view (Anderson 1999). We should ensure that every member of a particular society has access to what they need to be able to participate fully as an equal member of a democratic society. They should have what they need to stand in relations of equality with one another, such that people are able to avoid oppression, participate in productive life, and so on. Now ensuring that we stand in relations of equality with one another does not mean that all should have all capabilities or needs met to the same extent. So long as all are sufficiently literate that they are enabled to follow discussion and debate central to being able to participate equally with others in their democracy, it does not matter if some have Ph.D.s in Philosophy while others have (say) only basic tertiary education or high-school diplomas.

4.4.2 Different Versions of Sufficientarianism While we have been focusing on Frankfurt’s form of sufficientarianism because of its dominance, there are several other more compelling versions that deserve mention. We might call Frankfurt’s version, “Maximizing Sufficientarianism.” On this view, we should aim to maximize the number of people who have sufficient (who are at or above some threshold of sufficiency). We have noted why this version is excessively vulnerable to critique as it tries to maximize the number of people who have enough. Since what matters here is maximizing the numbers who cross the sufficiency threshold, many who are too far below the threshold or too costly to bring to the threshold will be ignored or discounted. Rather, those near the threshold will be prioritized, as it is less costly to get them to cross it.

Sufficiency and Needs-Based Approaches    97 A more complex version is Crisp’s (2003), which we could call “prioritarian sufficientarianism.” On this view, absolute priority is to be given to benefiting those below the threshold rather than above it. Furthermore, for those below the threshold, priority is given to benefiting people especially when they are much worse off than others, when there are many such people, and the greater the magnitude of the benefits that can be bestowed. For cases involving only trivial benefits below a threshold or when people are above it, they are assigned no priority (Crisp 2003: 758). The prioritarian focus below the threshold has some plausibility; however, as already signaled, I find the idea that we should not care what happens to people above the threshold uncompelling. I noted already that we have good reason to care about how burdens are distributed above the threshold because of a concern for fairness, and I considered the example of different tax proposals to illustrate this. We could also make a similar point about caring how benefits above the threshold are also distributed. Imagine we have abundant resources and scarcity is not an issue. Everyone enjoys prospects for lives conducive to contentment (or whatever the preferred threshold is). The idea that we should not care about bestowing further benefits when these would add to enhanced flourishing above contentment seems unpersuasive. And we should care about distributing these benefits fairly. So, for instance, if flourishing gains are possible, we might care that people have access to these in a way that preserves standing in relations of equality with one another. A similar objection can be leveled at another influential version of sufficientarianism, namely the one Rawls holds for the international sphere. Following Rawls’s own terminology, this might be labeled “Target and cut-​off sufficientarianism”: Distributive just­ ice is concerned with the target of making sure all have enough for a decent life. Once that goal is secured, there are no particular goals of distributive justice that we need to secure. The same objection as was raised against prioritarian sufficientarianism could be marshalled against this form of sufficientarianism, namely it suggests that we should be indifferent to inequalities above the threshold once everyone has enough. But there are inequalities above the threshold that should command our attention, for instance if social policies impose greater costs on the worst off rather than the better off, even if everyone is above some sufficiency threshold. It is worth remarking that the force of the objection will be somewhat mitigated if we are attentive to empirical facts about how inequalities do translate into threats to people having enough. It seems there will be plenty of inequalities above that threshold that, as an empirical matter, should command our normative attention in virtue of our concern for sufficiency as discussed in section 4.4.1 (Green 2008, 2012; Riddell 2012). As illustrated in section 4.4.1, there are versions of sufficientarianism that are embedded in other theories, such as sufficientarianism within relational egalitarianism. On this view we should aim to ensure everyone has enough of what they need to stand

98   Gillian Brock in relations of equality with each other. This would require a fairly high level of provision and protection for all, along with other major institutional changes.5 Social relations of equality are complex and have several important dimensions which cannot be reduced to a single good or unit of analysis.6 Another way to embed sufficientarianism within a full account of distributive just­ ice is to consider how it might be applied in our actual world. When we try to implement our account of distributive justice, we might need to be sensitive to a range of factors (such as the extent of resource constraints) that we can safely put to one side in addressing the issue of what distributive justice for an ideal world consists in. When considering issues of implementation we might need to distinguish between progressive targets to aim at. A first target of a minimally just society might be to aim to secure everyone’s moral agency and basic human functioning. A second target of a more ambitiously just society might be to aim to ensure everyone’s human flourishing is promoted at higher levels beyond securing sufficient for moral agency and basic human functioning, while also ensuring relations of equality flourish in the society. What any actual society may be able to provide will be conditional on available resources. We may need to realize our ideals of distributive justice progressively, conditioning them to circumstances and making gains where these are possible within resource constraints. Here ensuring all have enough for a decent life will serve as an important focus for our attention, especially when there is wide-​scale remedial deprivation. Which needs are important to address in order for citizens to enjoy prospects for decent lives? Besides some obvious physiological needs, people often appear to make quite diverse claims about their needs. Some claim to need long periods of “alone time,” while others seem to be insatiably gregarious. Some claim that cellphones and access to the internet are basic needs, while others abhor the way these technologies interfere with what they perceive to be our authentic human needs. If addressing needs is going to be an important focal point for public policy, we will need some criteria for determining the significant ones. We turn in sections 4.5–4.8 to a more systematic inquiry concerning which needs matter, how they do, how they can generate distinctive arguments for why we have distributive justice obligations, and how analysis of needs can provide distinctively important insights for the domain of distributive justice.

5 

Some of this is argued for in Brock (2009), especially Ch. 12. In a sophisticated analysis, Paula Casal (2007) argues that sufficiency might supplement rather than replace egalitarianism or prioritarianism. She argues that sufficientarianism might well usefully solve some of the well-​known problems for egalitarianism and prioritarianism by showing that a sufficiency threshold escapes them. So, for instance, the Leveling Down Objection thought to present a problem for egalitarianism can be well met by imposing a sufficiency threshold below which equality should not be preferred. Similarly, sufficiency-​constrained prioritarianism might more adequately resist well-known challenges to prioritarianism (Casal 2007: 318–​26). 6 

Sufficiency and Needs-Based Approaches    99

4.5  Which Needs Warrant Normative Attention? What Weight Should Needs Carry? Not just any need has normative force. Which needs, if any, should rightly claim our normative attention? Coherent statements of need have the following structure: x needs y in order to z. For a coherent needs claim we can always fill in a purpose or end-​state, z, to be realized by x’s having y, no matter how obvious or mundane the end-​state, z, might be (Thompson 1987; Wiggins 1987). Statements of need are essentially instrumental claims. They express a relation that x’s having y is needed to achieve some end, z. The normative force of the needs claim will vary depending on the end-​state involved (Thompson 1987; Wiggins 1987). Clearly, not all statements of need make defensible claims on others: people may claim to need things for morally repugnant or trivial ends. It is, however, widely held in the literature that there is a group of needs that can generate normative force in a more categorical way. Our most fundamental, basic, or human needs, those central to continued human existence, can do this (Braybrooke 1987; Brock 1998; Copp 1998; Thompson 1987; Wiggins 1987). We consider next some recent accounts of how to pick out such needs. We notice there are different, but importantly overlapping ways of picking out the relevant categories. Various authors pick out central categories of needs as warranting normative attention. Harry Frankfurt (1998) suggests that there are two necessary conditions for a need’s deserving moral importance:  a need is morally important if harm typically results when the need is not met and that harm is outside the person’s voluntary control. David Wiggins (1987; 1998) also believes harm arising from factors beyond one’s control is central to why needs matter, when they do. To develop this idea, he introduces several terms to reflect important differences. He distinguishes the gravity of the harm that would ensue if the need is not met (“badness”), from the urgency with which harm would ensue. A need is basic if it results from a law of nature, an unalterable and invariable environmental fact, or a fact about human constitution. Needs can be entrenched when they are inflexible to modification, or substitutable when they are not. Using this terminology, Wiggins defines vital needs as ones that are badly needed in a way that is entrenched and scarcely substitutable, and it is these vital needs that matter morally according to him (Wiggins 1987; 1998). David Braybrooke’s (1987) account is one of the most influential in the philosophical literature. He develops an account of basic needs in terms of what is necessary for social functioning. Something is a need if, without its satisfaction, one would be unable to carry out four basic social roles, namely, those of citizen, parent, householder, and worker. By examining several lists proposed by the United Nations and other influential public bodies, he extracts their common elements and offers a systematic account of the needs one would have over the course of a life. The list consists of needs for a

100   Gillian Brock life-​supporting relation to the environment; for whatever is indispensable to preserving the body intact in important respects (including food, water, exercise, and periodic rest); for companionship; for education; for social acceptance and recognition; for sexual activity; for recreation; and for freedom from harassment, including not being continually frightened. Focusing on what humans typically do (through consideration of roles) provides Braybrooke with a good reference point for compiling this list. Someone might claim that not all these needs apply to everyone (e.g., the need for sexual activity for a nun, or the need for companionship for a hermit), but it is important to note that Braybrooke is interested in deriving a list of items that are plausibly needed in order to carry out the four roles he identifies. In order to have the genuine choice to perform the role (such as that of being a parent), one typically needs items on the list. Another highly influential account which bridges philosophical and empirical disciplines is that of Len Doyal and Ian Gough (1991). Doyal and Gough’s view is that needs are universalizable preconditions that enable non-​impaired participation in any form of (human) life. Chief among these preconditions will be physical health and the mental competence to deliberate and choose, or autonomy. They recognize a class of “intermediate needs,” which aim to connect the two basic needs with knowledge available about basic needs in the social sciences. These are nutritious food and clean water, protective housing, a non-​hazardous work environment, a non-​hazardous physical environment, appropriate healthcare, security in childhood, significant primary relationships, physical security, economic security, appropriate education, safe birth control, and safe child-​bearing. Braybrooke’s and Doyal and Gough’s accounts highlight important features of recent influential accounts of basic needs:  the importance of social (not just physical) functioning in particular communities; the relevance of information about human needs collected by the natural and social sciences; and the importance of cross-​cultural comparison. More generally, there are several common elements to these and other recent accounts of normatively salient needs canvassed above. The needs that matter morally are those that are necessary, indispensable, or inescapable, at least with respect to human functioning in social groups (Brock 1998; Thompson 1987; Wiggins 1998). Moreover, if such needs are not met, we are unable to do anything much at all, let alone lead a recognizably human life (Copp 1998; Nussbaum 1998). Having needs satisfied is essential to our ability to function as human agents (Copp 1998; Gewirth 1978; O’Neill 1998; Shue 1980). While one dominant approach in the needs literature is to emphasize the link between needs and human agency, other approaches can be discerned, such as, connecting our basic needs with what is required for human flourishing (Nussbaum 1998; Reader 2005)  rather than the perceived more basic goal of functioning as human agents. Theorists often adopt one or other of these standards depending on different tasks. If the task at hand is to define what the minimally just state ought to do, then theorists often construct their answers using the goal of basic functioning. However, if we are inquiring as to what the fully just state ought to do, suggesting that a goal of enabling all to address what is needed for flourishing might be more appropriate. Ensuring citizens’

Sufficiency and Needs-Based Approaches    101 central human needs are satisfied to an adequate level is necessary for the minimally just (or legitimate state), but a fully just state might need to do more. Many states simply do not have the luxury of aiming at being fully just societies, given their histories and available resources. It is quite plausible that in our nonideal world, claims of need might have variable importance and might need to be especially sensitive to the resources available and the burdens that meeting them would place on others (e.g., Copp 1998; also Brock 2009). I should also comment briefly on what the focus for “attention to needs” should rightly be. The responsibility to meet needs is often thought to entail directly meeting needs, for instance, supplying food for the hungry. This may or may not be the direct object of the responsibility. My preferred way of expressing the content of duties with respect to needs is that the responsibilities are to enable others to meet their needs themselves (ceteris paribus). Enabling is a process that involves a number of different elements depending, for instance, on the capacities of the person. If Diane enables Edith to eat fish, this might involve a point at which Edith may need to be simply given fish directly, but another point at which Edith might need to be taught certain skills to catch fish, and to be helped with opportunities to exercise her fishing skills. Addressing needs might focus on provision for all the stages involved in enabling one to meet one’s needs oneself.

4.6  Do We Have Responsibilities to Meet Needs? Some Needs-​Specific Arguments Along with the usual array of arguments concerning why we have responsibilities to attend to concerns of distributive justice, several kinds of arguments can be discerned in the literature which aim to show that we have strong responsibilities to meet needs as a special focus for our distributive justice responsibilities. Next we discuss some common argument strategies. One strategy starts from showing that having needs met is a necessary condition for human agency, autonomy, or good human functioning, and then proceeds to argue that these special necessary conditions should be protected or underwritten by communities of co-​operators (Baker and Jones 1998; Copp 1998; Gewirth 1978; Wiggins 1987). Insofar as we care about the benefits of co-​operation, we have reasons to protect its bases. Another strategy commonly pursued in the literature is to show how people are especially vulnerable to coercion or having their rights violated if they do not have their needs met (Braybrooke 1987; Gewirth 1978; Goodin 1998; O’ Neill 1998; Shue 1980). Being needy in certain circumstances (especially ones in which there are vast power differences between the needy and the affluent) can lead to opportunities for coercion. But whatever else we expect of a state, we minimally expect it to protect us from rights’

102   Gillian Brock violations and coercion, so responsibilities to remove our vulnerabilities to force and injustice and so to meet certain kinds of needs, are hereby generated. Related to this line of argument is another common strategy—​one that examines the task of government. At the very least, governments should protect vital interests and enable citizens to meet their needs for themselves (insofar as this is possible). Meeting needs is therefore a crucial part of governments’ core function, according to this line of argument (Braybrooke 1987; Wiggins 1987; Nussbaum 1998. Others suggest that in addition, part of government’s job is to act in certain kinds of cases as a co-​ordinator of our moral responsibilities (Goodin 1998). If we all have some responsibilities to help others with their needs (as we do), an efficient way to co-​ordinate and discharge these responsibilities may well be through state-​organized assistance. According to libertarians, by contrast, the job of government is quite minimal: to ensure there are no rights’ violations or liberty infringements. There are many arguments that specifically target libertarian interlocutors and their conception of justice with the intention of showing how making space for meeting needs is necessary for the coherence of their own accounts (e.g., Baker and Jones 1998; Brock 1995; Sterba 1998). These arguments also often aim to show that cherished libertarian beliefs about the nature of rights to life, liberty, or property, when examined, commit one to meeting needs, given a detailed analysis of the nature of the rights and liberties at issue (Shue 1980; Gewirth 1978; Brock 1995; Sterba 1998; Baker and Jones 1998). Others argue that the ideals of self-​ reliance with respect to meeting needs, when examined, are untenable without considerable concession in the direction of support for needs. Some needs are for services that no one can seriously be expected to provide for oneself (e.g., education or healthcare), some needs are intrinsically social (such as the need for companionship), and some people (such as very young children) are incapable of meeting their own needs for themselves. Baker and Jones (1998) argue that the common libertarian ideal of self-​reliance—​that of strict individual responsibilities with respect to needs satisfaction—​entails collective responsibilities to meet needs and provide social structures that are organized to ensure that individuals have real opportunities to satisfy their own basic needs. We see, then, that several theorists argue that enabling citizens to meet needs is, or should be, a legitimate and core function of governments, across a wide ideological spectrum of political views.

4.7  When and How are Needs Salient in Distribution? How much priority should the activity of enabling people to meet needs have relative to other activities or demands for resources? What normative significance should needs have? It is commonly held that needs have, or should have, a certain lexicographical priority over desires, preferences, and wants (Wiggins 1987; Frankfurt 1998; Braybrooke

Sufficiency and Needs-Based Approaches    103 1987). Having a need for something makes a more compelling claim than having a desire or preference for the same thing (Wiggins 1987; Frankfurt 1998; Braybrooke 1987). Harry Frankfurt describes a Principle of Precedence which captures this insight (1998). He notices, for instance, that people are: . . . widely disposed to accept the proposition that a need for something preempts a desire for that thing. This proposition, which I shall call the Principle of Precedence, attributes to needs only a quite minimal moral superiority over desires. It maintains no more than that when there is competition between a desire and a need for the same thing, the need starts with a certain moral edge. That is, when A needs something that B wants but does not need, then meeting A’s need is prima facie morally preferable to satisfying B’s desire. (Frankfurt 1998: 20)

Claims in virtue of needs typically enjoy a prima facie advantage over claims pressed using the discourse of desires. Needs sometimes also enjoy a similar advantage when in competition with other claims. In matters of distributive justice there is the question of how needs should relate to claims pressed in virtue of desert or equality, which also have a role to play in distributive contexts. David Miller (1999) argues, for instance, that all three of these important concepts (needs, desert, and equality) should play basic roles in a theory of social justice, with their importance varying depending on what is being distributed to whom and by whom. In some contexts need seems to matter greatly, for instance when considering the distribution of medical resources. In other contexts, equality or desert might be a more salient consideration. While there certainly are connections between each of these central concepts (for instance, meeting needs may be a way of pursuing equality in certain cases), each of these concepts often engages with a different core concern in matters of distributive justice, and so each should play an important and distinct role in a full account of distributive justice. There are a couple of important distribution problems when deciding on the importance to give a particular claim in virtue of need. First, there is the problem of mediating between needs claims and non-​needs claims. This problem involves the weight claims needs should enjoy compared with claims made in virtue of competing considerations such as individual desert, productive contribution, rights, or other important projects which we, as a society, might deem worthy of resource allocation, such as protecting the environment or investing in infrastructure. Once we have decided that distribution according to need should trump, there are several problems concerning how to proceed. There are problems of commensurability, and problems even when commensurability can be assumed. First, problems associated with lack of commensurability. Various different needs claims compete for normative attention and there is the problem of mediating among these claims, such as in the case of the need for medical attention versus the need for housing. There may be competition among different types of needs claims, and the claims may vary in intensity, extent, urgency, kind, and the like. Alice needs medical attention; Bob needs housing. Who is in greater need or to which need

104   Gillian Brock should we direct our attention if we cannot do both? Putting problems of commensurability to one side, let us consider cases of only one type of need and assume commensurability so that we can place people all on one scale. Within the domain where we recognize that needs should operate as salient in distribution and can assume commensurability, what is it to distribute according to needs? There are at least four plausible ways to negotiate situations of competition among claims of need: 1) strict priority to the most needy; 2) priority to those whose needs can be efficiently or effectively met with scarce resources; 3) equality in provision for needs even though all remain needy; 4) meeting needs via lotteries in cases of extreme scarcity. Sometimes the person who is most needy should get resources in proportion to need. This is especially the case when resources needed to meet needs are fairly abundant. So, if Andrew is severely dehydrated, and suffering unmet need for hydration at, say –​5 units, and Betty is also dehydrated but suffering unmet need at only –​2 units, Andrew should get 5 units of water and Betty 2 to bring them up to hydration satiation. But such a principle of distribution proportionate to unmet needs is not always the best principle to follow. Whether or not distributing in proportion to unmet needs is advisable depends on the extent of scarcity, the types of resources we are distributing, the ways in which recipients are able to convert resources into needs-​fulfillment, what is needed to address needs, and the like. So, if there is a high level of scarcity, we might give more attention to those who are neediest, prioritizing them above all others. But sometimes this strategy would be imprudent, for instance in cases which suggest triage policies would be more appropriate. Triage is a well-​known practice for using scarce resources most efficiently, for instance in battle situations. Casualties are divided into three categories: those who are badly wounded but whose chance of recovery is small even with extensive medical resources, those who are in need of medical attention and, if and only if they receive it, have good prospects for recovery, and those who will recover eventually without treatment or resource use. Under a triage policy, the second group is prioritized, even though they have less urgent needs than the first group, because scarce resources allocated here are expected to have the biggest impact on favorable outcomes. Devoting all our resources to the most needy might mean there are none left for the less needy, but those who might have greater prospects of recovery. Sometimes distributing equally in the face of severe scarcity is more plausible, for instance in the case of trapped miners, where giving each of ten trapped miners an equal distribution of the one can of tuna available is appropriate, even though an equal distribution will leave all of them still quite hungry and no needs will be remotely well satisfied. Furthermore, if the good needed is not divisible, then a lottery in which each has an equal chance of securing the good might be the fairest way to distribute the one heart that four patients all need equally urgently in order to survive.

Sufficiency and Needs-Based Approaches    105 So, when goods are fully divisible and transferable, we might employ a strategy of distributing goods as closely as possible in proportion to neediness, but when resources are not fully divisible, following equality guidelines (either giving each an equal share or an equal chance to receive the scarce good) might be more plausible. This range of ways of addressing unmet needs is not necessarily any kind of problem; rather, it points to complexity in the face of (i) diverse needs; (ii) diverse ways of satisfying needs; (iii) needs that have different urgencies; (iv) differences in whether or not needs can be partially satisfied; (v) diverse capacities to convert resources into states of need satisfaction, and the like. Similar difficulties and complexity arise with other distributive justice criteria, such as desert. The obligation to address needs is often thought to be subject to something like a threshold criterion of taking responsibility for meeting needs. (Of course it is no easy matter to decide for which activities people can properly be held responsible. However, there will sometimes be available generally plausible views about what is a matter of fate and what a matter of individual responsibility that can assist in making such determinations.) The idea is that if we supply resources for someone to meet needs which are then squandered, the obligation to meet needs in virtue of justice has been discharged, notwithstanding the fact that the resources have not succeeded in meeting needs. In such cases we may not have an obligation to supply further resources, not at any rate in virtue of justice, though we may still have a duty in virtue of humanitarian considerations.

4.8  Needs, the Scope-ofJustice Question, and some Concluding Remarks The debate concerning meeting needs extends in various directions. Importantly, there is the issue of the scope of justice: whose needs deserve consideration when we are distributing according to need? Are we implicitly talking about compatriots only or does the concern extend across borders to include non-​compatriots? And what about future generations? Should their needs act as constraints on how we may permissibly distribute resources now? (For extended treatment of these issues, see Chapter 29 Justice across Borders by Michael Blake and Chapter 32 Future Generations by Rahul Kumar in this volume.) Here I make some brief comments on the case of the extension across state borders, typically thought to limit our responsibilities in important respects. The debate on distributive justice originated in a context where a state-​centric model of the primary agents responsible for doing the distribution dominated. But over the past few decades, the domain of distributive justice has gone global. For some, this changed nothing. The set of agents among whom distribution should justly occur might be larger,

106   Gillian Brock but the principles that should govern what counts as a just distribution are unaltered. For others, the global extension changed everything. John Rawls is a good example of the latter. In the global domain, we have to make special allowance for other features. Individual human beings are members of peoples and those peoples have aspirations to self-​determination and a desire to be considered equal to other peoples. We must modify our accounts of distributive justice to make allowances for these important group memberships, and also the fact that we must acknowledge the reality that not all peoples in the world are liberal, or can reasonably be expected to conform their behavior towards liberal values. Future research might usefully invert this order of investigating what distributive justice consists in. How might starting with an account of what global distributive just­ ice consists in affect our views about what social justice should be? How might starting with a more global perspective on distributive justice change the debates surrounding sufficientarianism versus egalitarianism? One plausible way in which starting with the global perspective might alter our accounts is in inviting consideration of a range of prior concepts before we can focus on issues of distributive justice more narrowly construed. These might include analysis of the nature of property entitlements and authority, which might usefully inform alternative accounts concerning the meaning and value of appropriate sovereignty, the weight human rights ought to carry in accounts of distributive justice, how to weigh our commitment to the importance of honoring human rights over property rights, and so on. In this chapter we have addressed several debates concerning needs and distributive justice. I argued that the debate between egalitarians and sufficientarians has not been as productive as it could be, especially when we acknowledge the insights both sides can learn from each other. The issue is not whether we should prefer equality or sufficiency, on this way of exploring the topic, but rather what is the proper role for each in an account of distributive justice. There is further work to do in tracing the important and subtle ways in which they are connected. In this chapter we also addressed several prominent debates concerning needs and distributive justice. We see that among needs-​theorists there is convergence on what counts as a need and the needs claims that give rise to appropriate claims of justice. Needs can be a source of powerful responsibilities grounded in concern for justice, according to a wide range of theories of justice. There is a role for needs to play in distributive justice that does not displace concern with several neighboring concepts such as desert or equality. Though there are several ways of interpreting the advice to distribute “to each according to his needs,” some important guidance on how to establish priorities among different needs claims is also available. The concept of needs is an essential one in understanding the distributive justice and normative landscape. There is essential work that it does well and no other concept does better. There is a role for needs to play in matters of distributive justice that does not displace concern with several other criteria that are salient in a comprehensive account of distributive justice.

Sufficiency and Needs-Based Approaches    107

References Anderson, E. (1999). “What is the Point of Equality?” Ethics 109: 287–​337. Anderson, E. (2007). “Fair Opportunity in Education: A Democratic Equality Perspective.” Ethics 117: 595–​622. Arneson, R. (2005). “Distributive Ethics and Basic Capability Equality: ‘Good Enough’ is not Good Enough,” in A. Kaufman (ed.) Capabilities Equality: Basic Issues and Problems. London: Routledge, pp. 17–​43. Baker, J. and Jones, C. (1998). “Responsibility for Needs,” in G. Brock (ed.) Necessary Goods: Our Responsibilities to Meet Others’ Needs. Oxford: Rowman and Littlefield, pp. 219–​32. Blake, M. (2018). “Justice across Borders,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 619–39. Braybrooke, D. (1987). Meeting Needs. Princeton, NJ: Princeton University Press. Brock, G. (1995). “Is Redistribution to Help the Needy Unjust?” Analysis 55: 50–​60. Brock, G. (ed.) (1998). “Introduction,” in G. Brock (ed.) Necessary Goods: Our Responsibilities to Meet Others’ Needs, Oxford: Rowman and Littlefield, pp. 1–​18. Brock, G. (2009). Global Justice: A Cosmopolitan Account. Oxford: Oxford University Press. Casal, P. (2007). “Why Sufficiency is not Enough.” Ethics 117: 296–​326. Copp, D. (1998). “Equality, Justice, and the Basic Needs,” in G. Brock (ed.), Necessary Goods: Our Responsibilities to Meet Others’ Needs. Oxford: Rowman and Littlefield, pp. 113–​33. Crisp, R. (2003). “Equality, Priority, and Compassion.” Ethics 113: 745–​63. Doyal, L. and Gough, I. (1991). A Theory of Human Need. Houndsmills: Macmillan. Frankfurt, H. (1987). “Equality as a Moral Ideal.” Ethics 98: 21–​43. Frankfurt, H. (1997). “Equality and Respect.” Social Research 64: 3–​15. Frankfurt, H. (1998). “Necessity and Desire,” in G. Brock (ed.) Necessary Goods:  Our Responsibilities to Meet Others’ Needs. Oxford: Rowman and Littlefield, pp. 19–​32. Frankfurt, H. (2000). “The Moral Irrelevance of Equality.” Public Affairs Quarterly 14: 87–​103. Gewirth, A. (1978). Reason and Morality. Chicago, IL: University of Chicago Press. Goodin, R. (1987). “Egalitarianism, Fetishistic or Otherwise.” Ethics 98: 44–​9. Goodin, R. (1998). “Vulnerabilities and Responsibilities: An Ethical Defense of the Welfare State,” in G. Brock (ed.) Necessary Goods:  Our Responsibilities to Meet Others’ Needs. Oxford: Rowman and Littlefield, pp. 73–98. Green, Duncan (2008). From Poverty to Power: How Active Citizens and Effective States can Change the World. Oxford: Oxfam International. Green, Duncan (2012). From Poverty to Power: How Active Citizens and Effective States can Change the World, 2nd edn. Rugby, UK: Practical Action Publishing. Kumar, R. (2018). “Future Generations,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 689–710. Miller, D. (1999). Principles of Social Justice. Cambridge, MA: Harvard University Press. Nussbaum, M. (1998). “Aristotelian Social Democracy,” in G. Brock (ed.) Necessary Goods: Our Responsibilities to Meet Others’ Needs. Oxford: Rowman and Littlefield, pp. 135–​56. O’Neill, O. (1998). “Rights, Obligations, and Needs,” in G. Brock (ed.) Necessary Goods: Our Responsibilities to Meet Others’ Needs, Oxford: Rowman and Littlefield, pp. 95–​112. Otsuka, M. and Voorhoeve, A. (2018). “Equality versus Priority,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 65–85. Rawls, J. (1999). Law of Peoples. Cambridge, MA: Harvard University Press. Rawls, J. (2001). Justice as Fairness: A Restatement. Cambridge, MA: Harvard University Press.

108   Gillian Brock Reader, S. (ed.) (2005). The Philosophy of Need. Cambridge University Press, Cambridge. Riddell, R. C. (2012). “Navigating between Extremes: Academics Helping to Eradicate Global Poverty.” Ethics and International Affairs 26: 217–​43. Satz, D. (2007). “Equality, Adequacy, and Education for Citizenship.” Ethics 117: 623–​48. Shields, L. (2012). “The Prospects for Sufficientarianism.” Utilitas 24(1): 101–​17. Shue, H. (1980). Basic Rights:  Subsistence, Affluence and US Foreign Policy. Princeton, NJ: Princeton University Press. Springborg, P. (1981). The Problem of Human Needs and the Critique of Civilization. London: Allen and Unwin. Sterba, J. (1998). “From Liberty to Universal Welfare,” in G. Brock (ed.) Necessary Goods: Our Responsibilities to Meet Others’ Needs. Oxford: Rowman and Littlefield, pp. 185–​217. Thomson, G. (1987). Needs. London: Routledge. Wiggins, D. (1987). “Claims of Need,” in D. Wiggins (ed.) Needs, Values, Truth. Oxford; Oxford University Press, pp. 1–​57. Wiggins, D. (1998). “What is the Force of the Claim that One Needs Something?” in G. Brock (ed.) Necessary Goods:  Our Responsibilities to Meet Others’ Needs. Oxford:  Rowman and Littlefield, pp. 33–​55.

Additional Reading Crisp, R. (2004). “Egalitarianism and Compassion.” Ethics 114: 119–​26. Frankfurt, H. (2000). “The Moral Irrelevance of Equality.” Public Affairs Quarterly 14: 87–​103.

Chapter 5

The Capabilit y A pproac h Ingrid Robeyns

5.1 Introduction The capability approach forms one of the most recent additions to the landscape of theories of distributive justice. Although some of the central notions of the capability approach can be traced back to Aristotle, Adam Smith, and Karl Marx, in the contemporary literature on distributive justice the notion of capabilities was first suggested in Amartya Sen’s 1979 Tanner Lecture (Sen 1980). In that lecture Sen asked the famous “Equality of what?” question:  assuming we advocate a form of equality, what kind of good should be equalized? In other words, what should the “metric” or “currency” (Cohen 1989) of egalitarianism be? While Sen coined the idea of capabilities in the specific discussion on equality, it soon became understood as a proposal within theories of distributive justice more generally. Every account of distributive justice has to put forward a metric of justice—​the distribuendum with which the account of distributive justice will be concerned; the claim of the capability approach to distributive justice is that functionings and/​or capabilities should be the metric of justice. Thirty-​five years later, the capability approach is widely regarded as offering us an alternative position on the metric of distributive justice, and a large literature has emerged examining that view. This chapter will introduce that literature, point to some of the main challenges that capabilitarian theories of distributive justice face, and outline what needs to be done if we want to further this literature.1 Yet before embarking on these tasks, it is important to pre-empt a common misunderstanding. On its most comprehensive interpretation, the capability approach is not just a 1 

For this chapter, I partly extend some earlier attempts at explaining and commenting upon the literature on the capability approach and on capabilitarian theories of justice in particular (Robeyns 2013, 2016).

110   Ingrid Robeyns contribution to theories of distributive justice where it proposes a metric of justice. Instead, its scope is much wider, both in terms of the debates in which it intervenes, and the disciplines in which it plays a role. If one asks the question what the capability approach precisely is, the answer often depends on the perspective from which that question is asked. If one looks at the capability approach starting off from the philosophical literature on distributive justice, one is tempted to understand the capability approach as a theory of distributive justice—​yet that is only one of the debates or literatures in which the capability approach has been put forward (Robeyns 2017). In order to avoid this common misunderstanding, it may perhaps be helpful to make a distinction between the capability approach, which is the general framework that is used in a number of different disciplines for various purposes, and capability theories or capability accounts, which use this approach for a particular purpose, such as to develop a theory of well-​being, poverty, or justice. This distinction is important, because a claim or an argument made by a capability scholar could be either a claim about the capability approach or about particular capability theories which are developed in a particular debate (e.g., theories of distributive justice, or development ethics, or welfare economics). In order to properly understand a particular capabilitarian claim or argument, one needs to know what the exact scope of the claim is. The distinction between the capability approach and capability theories or accounts is also important because without understanding this distinction one cannot understand how it can be consistent for a particular capability theorist to make claim A in one article, and claim the denial of A in another article. Yet this can be consistent if claim A is only valid for capability theory X, while one rejects claim A for capability theory Y, or for the capability approach in general. For example, Amartya Sen has endorsed particular selections of capabilities for particular empirical applications, but denies that there can be a single selection of relevant capabilities for the capability approach in general. In contrast to what some may think, Sen’s endorsement of both claims need not be inconsistent. In the remainder of this chapter, however, we will focus only on the capability approach as an intervention in the literature on social and distributive justice. After a brief introduction to various metrics of justice and the role a metric of just­ ice plays in a theory of justice (section 5.2), we turn to the metric of justice in the capability approach (section 5.3). Next we compare the capability approach with John Rawls’s Justice as Fairness (section 5.4). In section 5.5 we examine whether the capability approach is genuinely different from a welfarist metric. The answer to that question is in part dependent on which the relevant capabilities are taken to be (section 5.6) and how one judges a person’s overall level of advantage, which requires aggregation (section 5.7). Finally, in section 5.8 I will conclude that the theoretical resources of the capability approach as it currently stands are not sufficient for a complete capability theory of justice, but that additional aspects of a theory of justice need to be added. As a consequence, various capability theories of justice are conceivable.

The Capability Approach    111

5.2  Distributive Justice and the Metric of Justice Richard Arneson (2010: 103) has presented the following metaphor that is helpful for understanding theories of distributive justice in general, and for the specific claims that the capability approach makes. A theory of distributive justice can be compared with an engine, in which the capability approach provides only one part, or “module.” Not all theories of justice have exactly the same modules, yet theories of distributive justice must specify at least two things: a metric and a distributive rule. The “metric of justice” is the good possession of which by individuals is to be compared when making statements of distributive injustices; hence, the metric of justice identifies the dimensions in which the interpersonal comparisons that are an integral part of claims of justice are made. The distributive rule specifies what justice requires in terms of the distribution of that good between people. Typical distributive rules are equality, sufficiency (a level that everyone should be situated above), or priority to the worst off. Moreover, there are other modules that make up the full theory of distributive justice, such as the grounds for its principles of justice, or a specification of the duties of justice, rather than merely the rights of justice (see section 5.7). As Arneson (2010:  103–​ 4) point outs, one of the theoretical difficulties of comparing the capability approach to justice with other theories is that one limits the comparison to only one module, whereas the comparison really requires an assessment of the entire engine. We can’t make an overall comparison of, say, cars, by only comparing their wheels or their maximum speed. Thus, while some political philosophers mistakenly think the capability approach is a theory of equality or a theory of justice, the capability approach to distributive justice only specifies a metric of justice, and hence a variety of capability theories are possible, depending on the other modules that are integrated with it. Even if we grant that we are just comparing different metrics, the question is whether that can be done in a sound way if we don’t know what the other characteristics of the theory of distributive justice are, for example what its scope, domain, and meta-​ethical commitments are. It may well be the case that we find one metric the most plausible for a theory of distributive justice with scope A, domain B, and meta-​ethical commitments C and D, whereas another metric is more convincing for a theory with characteristics scope E, domain F, and meta-​ethical commitments G and H. This potential complication is often ignored in debates on the metric of distributive justice. In this chapter these issues will only be mentioned briefly. Nevertheless, more research on which modules and which meta-​ ethical commitments can be put together coherently and plausibly would be a worthwhile line of further research.

112   Ingrid Robeyns

5.3  Functionings, Capabilities, and the Metric of Justice Capability theories of justice claim that when making interpersonal comparisons of “advantage” as part of a judgment of justice, we should focus on people’s functionings and/​or capabilities. Let us first look at how functionings and capabilities are defined, then look at the reasons for choosing functionings and capabilities as the metric of justice, and finally address the question whether the capabilitarian metric of justice should be functionings or capabilities or both. Functionings are “beings and doings,” that is, various states of human beings and activities that a person can undertake. Examples are being well nourished, being undernourished, being educated, being illiterate, being part of a supportive social network, being part of a criminal network, being depressed, travelling, caring for a child, voting in an election, taking part in a debate, taking drugs, killing animals, and eating animals. To every functioning corresponds a capability, which is the real opportunity one has to achieve that functioning. If we say that person A has a capability to obtain functioning X, we are saying that if A chooses to (be or do) X, then A will succeed to (do or be) X. Hence, capabilities stand to functionings as an opportunity stands to outcome, or the potential stands to the realized. Capabilities can also be understood as freedoms or opportunities—​yet when they are described as freedoms, this almost always comes with the qualifiers real or genuine or effective. By using the term effective, two aspects of the freedom notion of capabilities is stressed (Olsaretti 2005). First, describing capabilities as effective freedoms contrasts capabilities with merely formal freedoms. Second, “freedom is ‘effective’” in the sense that a person is said to have their “freedom increased ‘in effect’ when someone does something to her or for her which she would choose to have done to her or for her if given the chance, even if she does not actually choose that” (Olsaretti 2005: 91). In the earlier literature on the capability approach, capabilities have often been defined as valuable doings or beings. For example, Sen described the capability approach as “a particular approach to well-​being and advantage in terms of a person’s ability to do valuable acts or reach valuable states of being” (1993: 31, emphasis added). This value-​ laden interpretation of the notion of functionings and capabilities is widespread among capability scholars, especially outside philosophy. Yet there is also a value-​neutral sense, which defines functionings as beings and doings—​without thereby restricting them to valuable beings and doings. One could be tempted to think that this is merely a semantic difference: either we define functionings as beings and doings, or else as valuable beings and doings. However, there are reasons to think that the choice for the value-​ neutral versus the value-​laden interpretation of the notion “functioning,” and hence also of “capability,” is more than merely a semantic issue. It can be argued that a value-​ neutral interpretation is superior, since it makes much more explicit the need to decide which functionings are relevant, and which are not. A value-​neutral conceptualization

The Capability Approach    113 of “functioning” and “capability” can also allow for functionings that have a negative value, which would certainly be the case for many forms of harms that we can suffer through injustices. Moreover, a value-​laden interpretation would make it very hard to know what to call a being or a doing that is arguably not valuable, such as, say, counting blades of grass. In terms of the various metrics of justice, which will be discussed at greater length below, it is clear that “counting blades of grass” cannot be understood as a resource, nor is it reducible to a mental state (as in the case of welfare and happiness). But what is it, then? On a value-​neutral understanding of functionings, we have no difficulty in describing “counting blades of grass” as a functioning, albeit a functioning which may not be worth protecting, or which most likely will not be a concern for social justice. So while it is important, when reading across the capability literature, to take note of the fact that “functioning” and hence also “capability” are used in both value-​laden and value-​neutral ways, this chapter will discuss the concepts as being value-​neutral.2 What reasons are there to focus on the evaluative space of functionings and capabilities, rather than other possible metrics of justice, such as resources or welfare? The capability approach is motivated by a number of commitments, which lead to its rejection of resources and welfare as the metric of justice. The first commitment is what Carter (2014: 81–​2) calls the anti-​fetishist move. This is the concern that judgments of justice should focus on what is intrinsically valuable to people, rather than what is instrumentally valuable. For the capability approach, functionings and/​or capabilities are intrinsically valuable, whereas resources are merely instrumentally valuable. Rephrased in terminology that is often used in the capability literature and by Sen in particular, we should focus on the ends rather than the means of well-​being, and the capability approach postulates that these ends are what people are able to be and to do. The anti-​ fetishism commitment leads to the rejection of resourcist theories such as Ronald Dworkin’s (1981b), and, to some extent, also John Rawls’s theory of justice (see section 5.4). Still, it is important to acknowledge that in those theories of justice the accounts of resources are still relatively broad, while in philosophical work on justice in a particular domain or the corresponding institutional analysis (think of recent work on basic income or on property-​owning democracy), the definition of resources often further narrows to financial resources. The anti-​fetishist commitment of the capability approach is certainly a rejection of the latter as the relevant metric of justice, without necessarily denying the instrumental value of those resources. A second commitment of the capability approach is to human diversity. For the capability approach, the problem with resources as a metric of justice is not only that those are the means to living, but also that people differ in their ability to convert means into valuable opportunities (capabilities) or outcomes (functionings) (Sen 1992: 26–​8, 36–​8). Since ends are what ultimately matters when thinking about well-​being and the quality of life, means can only work as reliable proxies of people’s opportunities to achieve those

2  Ian Carter (2014) discusses some implications of the value-​laden interpretation of the notion ‘functioning’ for the ability of a capability theory of justice to avoid paternalism.

114   Ingrid Robeyns ends if all people have the same capacities or powers to convert those means into equal capability sets. Capability scholars believe that these inter​personal differences are far-​ reaching and significant, and that theories that focus on means tend to downplay their normative relevance. The sources of inter-​individual differences to convert means into ends can be personal (e.g., impairments), social (e.g., social norms), or environmental (e.g., living in an area affected by malaria). However, as we shall see in sections 5.4–5.8, the core normative commitment of the capability approach to take into account (at least to some extent) people’s interpersonal differences results in certain redistributive effects that are not beyond criticism. A final point we need to address is whether the metric of justice is functionings, or the genuine access to those functionings (i.e., capabilities), or both. Functionings are achievements, whereas capabilities are opportunities; hence, the question can be rephrased in terms of an old debate in political theory; namely, whether we want the capability account of distributive justice to be an outcome or an opportunity theory— that is, whether we think that we should assess injustices in terms of functionings, or in terms of capabilities, or some combination of the two. At the level of theory and principles, most theorists of justice endorse the view that justice is done if all have equal genuine opportunities, or if all reach a minimal threshold of capability levels. Translated to the capability language, this would imply that at the level of theory and principles, capabilities are the relevant metric of justice, and not functionings. However, while most theorists defend opportunities rather than outcomes, the focus on opportunities is not entirely uncontested (e.g., Fleurbaey 2002; Phillips 2004; Wolff and De-​Shalit 2007). Moreover, Claassen (2014) has argued that many capabilities presuppose certain levels of achieved functionings (e.g., the capability to run for a political office presupposes having achieved a minimal level of education), and hence a capability theory that is worked out in sufficient detail will often include claims that some minimum levels of achieved functionings are part of what justice requires.

5.4  Rawlsian Justice and the Capability Approach One common methodological practice in the theoretical literature on distributive just­ ice (and normative political theory/​philosophy more generally) is that the strength of a theory is tested by comparing it with other theories. This comparison of different theories is part and parcel of what happens in this literature. Yet there is seldom a systematic justification for why theory X is compared with theory Y and not with some other theories. In the case of the capability approach, we can observe that philosophers defending the capability approach have most explicitly targeted John Rawls’s metric of justice, the “social primary goods” (Nussbaum 2006; Sen 1980, 2009). I will first describe what arguments have been put forward in the debate between Rawlsian theorists and

The Capability Approach    115 capabilitarian theorists, and then return to the question of whether this comparison is the most important one to be made if one wants to test the capability approach to justice. In his 1979 Tanner lecture entitled “Equality of What?” Sen (1980) argued that “the primary goods approach seems to take little note of the diversity of human beings.” He continued: If people were basically very similar, then an index of primary goods might be quite a good way of judging advantage. But, in fact, people seem to have very different needs varying with health, longevity, climatic conditions, location, work conditions, temperament, and even body size. . . . So what is being involved is not merely ignoring a few hard cases, but overlooking very widespread and real differences. (Sen 1980: 215–1​6)

Sen argues that Rawls’s difference principle would not justify any redistribution to the disabled on grounds of disability. Rawls’s strategy has been to postpone the question of our obligations towards the disabled, and exclude them from the scope of his theory. Rawls certainly does not want to deny our moral duties towards the people that fall outside the scope of his theory, but he thinks that we should first work out a robust and convincing theory of justice for the “normal” cases and only then try to extend it to the “more extreme cases” (Rawls 2001:  176). Rawls stressed, especially in his later work, that in his theory “everyone has physical needs and psychological capacities within the normal range,” and therefore he excludes people with severe physical or mental disabilities from the scope of justice as fairness (2001: 170–​6). In his earlier work (Rawls 1971), Rawls justified the restriction by arguing that a theory of justice should in any case apply for “normal cases”—​if the theory is inconsistent or implausible for such cases, then it will certainly not be an attractive theory for the more challenging cases, such as people with severe disabilities. We could postpone the question of how to treat people with disabilities to one of the later (legislative) stages of the design of the basic structure of society, though, of course, even in his earliest discussions of this issue Rawls thinks that a complete theory of justice must deal adequately with the claims of people whose abilities fall outside the normal range, and that any theory that cannot do so should be rejected on those grounds. In later work Rawls (2001: 176) no longer argues that the case of justice towards the disabled had to be postponed to the legislative phase, but rather that we have to try to extend justice as fairness to include those cases. Rawls has not pursued this task systematically himself, though he has emphasized the role that his conception of the person possessed of the capacities for a sense of justice and a conception of the good plays in developing that part of a theory of justice, and has argued that this conception enables him to deflect accusations of “fetishism” about the primary goods (Rawls 2001: 176–​8). However, this doesn’t seem to provide a full response to Sen’s concerns in his Tanner lecture, which was not only about the case of the severely disabled. Sen’s more general critique concerned what he saw as the inflexibility of primary goods as a metric of just­ ice. Sen believes that the more general problem with the use of primary goods is that it

116   Ingrid Robeyns cannot adequately deal with the pervasive differences among people. Primary goods, he argues, cannot adequately account for differences among individuals in their abilities to convert these primary goods into what people are able to be and to do in their lives. Primary goods are among the valuable means to pursue one’s life plan. However, the real opportunities or possibilities that people have to pursue their own life plan are not only influenced by the primary goods that they have at their disposal, but also by a range of factors that determine to what extent they can use these primary goods to generate valuable states of being and doing. Hence, Sen claims that we should focus on the extent of substantive freedom that a person effectively has, that is, their capabilities. Sen sometimes adds that functionings, too, may need to be taken into account. The vagueness in his defense for a particular metric of justice (in this case: is the metric of justice comprised of capabilities, functionings or both?) reflects Sen’s meta-​theoretical commitment to let questions of justice be decided by social choice procedures, rather than by theorizing only (Sen 2009). A more recent wave of philosophical enquiry has highlighted how complex the comparisons between Rawls’s theory of justice and the capability view are (Pogge 2002; Richardson 2006; Robeyns 2009; Brighouse and Robeyns 2010). One reason is that the capability metric is a general metric of well-​being freedom, whereas the social primary goods metric emerges as one element of an integral and complex theory of institutional justice (rather than social justice more broadly, let alone the even wider category of moral evaluations). Also, Rawls’s theory of justice is an ideal theory of justice since it tries to outline the conditions of a completely just (yet “realistic”) utopia, which the currently developed capability accounts do not aspire to do. This means that it is very hard to compare Rawls’s work on justice with the existing philosophical work on the capability approach, since their scope and theoretical aims are not the same (Robeyns 2009, 2016). Regarding scope, Rawls’s theory of justice is limited to (1) the basic structure of society (that is, the set of most important social institutions); (2) liberal democratic societies rather than also to nondemocratic and illiberal societies; and (3) to the principles of justice insofar as they apply to people in their capacity as citizens. The scope of the capability approach can be summarized as “justice applies everywhere”—​that is, it applies to all human beings independently of their country of birth or residence, and not only to public institutions but also to the social ethos and to social practices. Regarding theoretical aims, one can safely say that most capability scholars tend to disfavor top-​down theorizing, and prefer to find out how theory or philosophy can help us make the actual world, a social institution, or a practice, more just, rather than to work more abstractly on the principles of justice and their justification. This last difference has been a main point of attention in Sen’s more recent work (Sen 2009). Still, it does not follow that this methodological preference for bottom-​up theorizing is a necessary or inevitable characteristic of a capabilitarian theory of justice. Rawlsians have criticized the capability approach too, and not all of their critiques have been sufficiently rebutted (Pogge 2002; Kelly 2010). First, it is claimed that the capability approach is endorsing a particular comprehensive moral view, which Rawlsians

The Capability Approach    117 find objectionable. Rawls aims to stay away from a perfectionist account of justice, and the question is whether this is possible for a capability theory of justice. This is an important area of dispute, to which we will return briefly in section 5.6. Another main Rawlsian objection to the capability approach concerns the publicity criterion, which stipulates that the conception of justice must be public and the necessary information to make a claim of injustice must be verifiable by all and be easily accessible. Rawlsians argue that a theory of justice needs a public standard of interpersonal comparisons, as otherwise the obtained principles of justice among citizens with diverse conceptions of the good life will not prove stable. The suggestion is that as capabilities are very hard to measure or assess in such a public fashion, and as assessing whether one has them would require very large amounts of difficult-​to-​assess information, the capability approach is unworkable as a theory of justice. Not everyone agrees that this is a valid complaint. For example, Elizabeth Anderson (2010: 85) has argued that the capabilities metric does meet the publicity criterion, while Richard Arneson (2010: 114) has argued that concerns of justice overrule concerns of publicity: if social justice can only be achieved by relying on measures that violate the publicity criterion, then that is a price worth paying. This brief overview of the debate between Rawls’s defense of social primary goods as the metric of justice and capabilitarian justice shows that a large literature has emerged between these two accounts of justice. There are plausible historical reasons to explain this, related to the fact that Rawls’s theory of justice is often seen as the most influential contemporary account, in addition to the fact that the two most prominent capabilitarian theorists—​Sen and Nussbaum—​picked out Rawlsian justice as their target for comparison. Yet one could wonder whether this is really where the action should be. Rawls’s theory of justice has much in common with the capability approach, and quite plausibly much more than with utilitarian accounts of justice, or with purely material​resource accounts or accounts that defend subjective welfare as the metric. Rawlsian just­ ice and the capability approach both see persons as agents, and as human beings living in a relational context (in the capability approach, this is captured by including a range of ‘social capabilities’; in justice as fairness, by arguing for the social basis of self-​respect as the most important social primary good). This is not to deny that there are significant disagreements between those two theories, which have been highlighted already—​such as the degree to which interpersonal diversity has to be taken into account (e.g., as in the case of disabilities). Yet it is also instructive that some philosophers have argued that the conceptual differences between social primary goods and capabilities are not that big, and that one way to see them is as complementary metrics of justice (Richardson 2006). Social primary goods are not necessarily incompatible with capabilities, and there is even some degree of overlap, especially in those social primary goods that are opportunities. Capabilitarian accounts of justice and Rawls’s theory of justice are thus much less theoretical opponents than the intensity of this debate has suggested. Social primary goods are only one possible metric of justice with which capabilities and functionings can be compared. One surprising observation is that relatively little attention has been paid to comparing functionings and capabilities as the metric of

118   Ingrid Robeyns justice with subjectivist metrics, such as happiness or “welfare.” One may wonder: can the capability approach escape the criticism of welfarism, or is it, upon closer consideration, a form of welfarism in disguise, as Dworkin (2002) has claimed?3

5.5  Capabilities, Welfare, and Happiness Welfarist metrics of justice use mental states or subjective judgments as the relevant metric. They come in a variety of forms, from classical hedonism (with its focus on the balance of pleasures and pains) to views of happiness understood as overall satisfaction with life, to views that take welfare to be preference fulfillment. In all those cases, a person’s level of ‘advantage’ cannot be assessed from the position of an outsider, but depends on the mental states of the person. In his highly influential paper “Equality of Welfare,” Dworkin (1981a) argues against welfare as a metric of justice whereby one of the grounds is that it does not hold people responsible for their “ambitions.” If Lisa wants to be a professional piano player and needs an expensive education and an expensive piano to practice and perform, whereas Karen wants to be a childminder for which she can train with a much cheaper education, then why would Lisa, on grounds of justice, be entitled to more resources than Karen in order to fulfill her ambitions? This should be worrying for the capability approach, since “being a piano player” and “being a childminder” are functionings (capabilities that a person has realized). If the capability approach advocates equality of capabilities, then the person with the expensive preference (becoming a piano player) will need many more resources than the person who does not have such expensive preferences (becoming a childminder). Interestingly enough, many capabilitarians have argued against subjective metrics based on arguments that are to some extent similar to the expensive preference objection: the “Adaptive Preference Objection.” This objection states that welfare (happiness, satisfaction) as the metric of justice legitimizes a distribution whereby very little is given to those persons who have learnt to become satisfied with little—​such as the lower class in class societies, the women in sexist societies, or the Dalits in caste societies. If we equalize welfare, these oppressed groups will get less than their fair share because they have been conditioned to be satisfied with less, so the argument goes.

3  Dworkin’s criticism of Sen’s capability approach was phrased differently, namely as the claim that the capability approach has to boil down to either equality of (Dworkinian) resources, or else as equality of welfare. For arguments that capabilities do not necessarily collapse into either resources or welfare, see Williams (2002) and Pierik and Robeyns (2007). In this chapter we will proceed on the assumption that capabilities do not necessarily have to collapse into either resources or welfare, but rather investigate the conditions under which they can avoid becoming welfarist.

The Capability Approach    119 Sen and Nussbaum have said very little on the problem of expensive taste—​possibly since their concerns have primarily been with questions of justice for the poorest and guaranteeing minimal levels of capabilities for all. Anderson (2010: 85–​7) has addressed the question of both adaptive preferences and expensive taste, and used these as arguments against welfarist metrics of justice. For Anderson, metrics of justice should be objective, and she claims that one strength of the capability approach is that it is an objective metric of justice. But the question is whether that is true:  is the capabilitarian metric of justice an objective metric? That depends on which capabilities are selected as the relevant dimensions. It is possible to select only capabilities to obtain objectively valuable states of being and doing other than mental states. If one were to add “the capability of feeling happy” as a capability that is relevant for matters of justice, one would go beyond an objective metric. Sen has repeatedly stated that the capability to be happy is “a major aspect of the freedom that we have good reason to value” (Sen 2009: 276). However, this is best interpreted as a statement about the capability of feeling happy being a relevant capability for capability accounts and theories that are not about justice (e.g., capability accounts about the decisions we make that affect our own well-​being) (Sen 2009: 282–​ 4). In other words, one plausible reading of Sen’s view is that happiness matters, but not for matters of justice. This underscores the importance, which was highlighted in section 5.1, of the distinction between the capability approach and the (more specific) capability accounts and theories. Happiness is a functioning, and its corresponding capability may be relevant, but not as a matter of justice. However, this doesn’t fully respond to Dworkin’s worry that some accounts of distributive justice are not sensitive to differences in people’s ambitions. Can the capability approach answer Dworkin’s worry? As will be explained in section 5.7, this depends on what the distributive rule is that a particular capability theory of justice defends. The upshot is that the capability approach can be developed in a range of capabilitarian theories of justice, and that whether that more specific theory is prone to Dworkin’s objections depends on the selection of capabilities, as well as on the distributive rule and method of aggregating. We will discuss these two challenges in turn.

5.6  Which Capabilities are Relevant for Justice? A major challenge for a capability theory of justice is the question of which capabilities matter. In answering this question, philosophers have had two different notions of justice in mind. One sees the question of justice as a question about truth, sharply distinguished from questions about implementation, justice-​enhancing policies, feasibility, and other practical concerns. G. A. Cohen (2008) is an important representative of this line of work. However, very few articles analyzing or defending the capability

120   Ingrid Robeyns approach to justice take this line. An exception is Peter Vallentyne (2005), who argues that all functionings should be included when considering issues of justice. According to Vallentyne (2005: 362), “given that any functioning could, under some circumstances, enhance (or otherwise affect) the quality of someone’s life, it is a mistake to exclude some functionings from consideration. To do so would leave out something that is relevant for justice.” However, Vallentyne adds that this is not to deny that when designing policies, we need to select the most important capabilities. Since virtually all capability theorists implicitly or explicitly understand “justice” as a practical concept, that is, as a concept that will help us determine what we ought to do or how we ought to shape social institutions, they take some (minimal) feasibility constraints into account. Those working in a more practical line of political philosophy have argued that considerations of justice require that we demarcate morally relevant from morally irrelevant and morally bad capabilities (Nussbaum 2003; Pogge 2002; Pierik and Robeyns 2007). Put differently, any capability account of justice will have to tell us which capabilities are relevant and which are not for purposes of justice. Amartya Sen (2004, 2009: 242–​3) has notoriously refused to answer this question, claiming that processes of public reasons and democratic deliberation should lead to the selection of relevant capabilities. Yet this “democratic route” to selecting the relevant capabilities requires a specific account of the deliberative processes that are needed, and that specific account has not been provided by Sen. While several capability theorists have debated issues of democratic deliberation in the context of development questions or other policy decisions, within the context of distributive justice this work remains to be done. The second way to select the relevant capabilities for the purpose of justice is the “criteria route,” whereby the criteria that the selected capabilities should meet are proposed and defended. A  prominent example of the “criteria route” is Elizabeth Anderson’s (1999) theory of democratic equality. Anderson (1999: 316), who aims to develop the outline of a political theory of justice (rather than a theory of social justice that encompasses all spheres of life), argues that people should be entitled “to whatever capabilities are necessary to enable them to avoid or escape entanglement in oppressive social relationships” and “to the capabilities necessary for functioning as an equal citizen in a democratic state,” without giving a complete list of which capabilities meet these criteria. The third way to select the relevant capabilities for the purpose of justice is the “objective-​list route.” Following Derek Parfit’s notion of objective list theories, which are theories claiming that “certain things are good or bad for us, whether or not we want to have the good things, or to avoid the bad things” (Parfit 1984: 493), the capability theorist proposes an objective list of well-​being which will be the concern of distributive justice. Richard Arneson (2010) defends this version of the capability approach, which he dubs the “perfectionist capability theory”—​without, however, specifying an account of well-​being in terms of an objective list. Martha Nussbaum’s (2000, 2006, 2011) minimal theory of justice is the best-​known version of the capabilities approach that relies on an objective list of well-​being.

The Capability Approach    121 Nussbaum’s theory of social justice is comprehensive, in the sense that it is not limited to an account of political justice, or to liberal democracies. Rather, her account holds for all human beings, independently of where they are living or what their particular needs are. The main demarcation of Nussbaum’s account is that it provides only “a partial and minimal account of social justice” (Nussbaum 2006:  71) by specifying thresholds of a list of capabilities that governments in all nations should guarantee to their citizens. Nussbaum’s theory focuses on thresholds, but this does not imply that reaching these thresholds is all that matters for social justice; rather, her theory is partial and simply leaves unaddressed the question of what social justice requires once those thresholds are met. Nussbaum’s well-​known list contains capabilities that are grouped together under ten “central human capabilities”: life; bodily health; bodily integrity; senses, imagination, and thought; emotions; practical reason; affiliation; other species; play; and control over one’s environment (Nussbaum 2006: 76–​8; 2011: 33–​4). Nussbaum (2000: 70–​7; 2006: 78–​81) justifies her list by arguing that each of these capabilities is needed in order for a human life to be “not so impoverished that it is not worthy of the dignity of a human being” (2000: 72). She defends these capabilities as being the moral entitlements of every human being on earth. She formulates the list at an abstract level and argues that the translation to implementation and policies should be done at a local level, taking into account local differences. Nussbaum argues that this list can be derived from a Rawlsian overlapping consensus and stresses that her list remains open-​ended and always open for revision (Nussbaum, 2000: 77), yet other philosophers have taken issue with her claim that this would result in a form of political liberalism, claiming that she is a perfectionist liberal and cannot avoid paternalism after all (Barclay 2003). The question of perfectionism and paternalism is one that does not only affect Nussbaum, since many critics have raised doubts whether the capability approach can avoid perfectionism and paternalism at all (e.g., Nelson 2008; Carter 2014; Claassen 2014). Even if we grant that capability theories of justice should only be concerned with capabilities rather than functionings, and we bracket—​for the sake of the argument—​Claassen’s (2014) point that many capabilities require minimal levels of other functionings, we are still left with the fact that, on the capability view, some citizens will be required to support the capabilities of their fellow citizens which go against their own ideas of the good life (Nelson 2008). The usual response by capability theorists that citizens are not forced into functioning but are merely offered the capability will not do. Why would Andrea need to fund access to university education of some of her co-​citizens, if Andrea endorses a notion of the good in which the good life requires merely basic education combined with a lot of hands-​on learning on the job, and moreover higher education is considered a “bad,” both for individuals and for society (e.g., because it threatens the innocent souls of students and corrupts their character)? Yet if Andrea’s notion of the good is not regarded as unreasonable, and the capability account of justice requires the state to provide the effective freedom to higher education, than Andrea will have to contribute, via the coercive nature of taxation, to the funding needed for making

122   Ingrid Robeyns higher education accessible. Nelson (2008) concludes that there is no way in which the capability approach can be neutral.

5.7  Aggregation and the Distributive Rule The selection of relevant capabilities is one major challenge for the capability approach to justice. Another is the question of how to aggregate the different capabilities that are judged to be relevant for justice. If judgments of justice require us to make interpersonal comparisons of people’s overall freedom to achieve well-​being, don’t we need a way to aggregate the value of the different capabilities into an overall value? In order for us to be able to judge whether person A is better off than person B, and to decide whether this constitutes a matter of injustice, we do need to make interpersonal comparisons at an aggregated level. For such interpersonal comparisons we need to choose a metric, which we discussed in section 5.6. In the case of multidimensional metrics of justice, such as Rawls’s social primary goods, Dworkinian resources, or capabilities, we also need a mechanism to aggregate the various dimensions, as well as a distributive rule to decide which distributions of the aggregated metric of justice constitute injustice. Clarity on these matters is needed if we want to compare theories of distributive justice, and assess their plausibility. Unfortunately, the literature on capabilitarian justice is remarkably silent on these questions. There are some proposals for aggregating capabilities either by using social choice procedures (Chakraborty 1996)  or else by equating the value of a capability by its contribution to a person’s happiness (Schokkaert 2007), but these are made in the context of welfare economics, rather than as part of assessing justice. Amartya Sen deliberately refuses to investigate the question how such aggregation should or could be done, since he believes that the quest for aggregation is driven by a concern for complete rankings. To Sen’s mind, striving for complete rankings is a mistake (Sen 2009). The capability literature thus seems to leave us with empty hands. Nevertheless, the problem of aggregating the dimensions of the metric of justice is of course relevant for all multidimensional metrics of justice, including Rawls’s social primary goods metric and Dworkin’s resource egalitarianism. Could the capability approach adopt the aggregation mechanisms available in Dworkin’s or in Rawls’s theory? On one interpretation of Rawls, the aggregation problem is an unsolved problem in his theory of justice, since the social primary goods are noncommensurable. On another interpretation, the social primary goods of basic liberties, opportunities, and the social basis of self-​respect, will be distributed equally by Rawls’s first principle of justice (the principle of equal liberties), implying that for distributive questions the relevant social primary goods reduce to income and wealth (assuming the social basis of self-​respect is

The Capability Approach    123 best taken care of by the realization of the principle of equal liberties). If it is problematic to reduce the full range of social primary goods to only income and wealth, then there is no satisfactory solution to the aggregation problem in Rawls’s theory of justice either. So the problem of aggregating would then be a problem not only for capability theories of justice, but also for the social primary goods metric. Dworkin (1981b) proposes the so-​called “envy-​test” to make comparisons between the “resources” that people hold, which in Dworkin’s conceptualization includes not only their material possessions and leisure time, but also their skills, talents, and handicaps. The distributive rule that Dworkin defends is equality: we should all have an equal amount of resources, and the envy-test can tell us whether that ideal of equality is met or not. If one is willing to take a pill to trade with the place of another person, taking the entire “package deal” of their life, one has a justified complaint that one has not received one’s fair share in life. However, while Dworkin’s envy-test may be a useful heuristic device, the envy-test entails impossible epistemological requirements, and operates against a number of background assumptions that make Dworkin’s theory highly idealized (Pierik and Robeyns 2007). Those background assumptions put Dworkin’s theory in highly idealized circumstances, in which there are no preference formation mechanisms, unjust social norms, or legacies of racism, sexism, and other forms of systematic bias against groups. In those circumstances, it is possible to adopt the Dworkinian envy-test for a capability theory of justice, although so far it hasn’t been analyzed as to whether that would provide us with a plausible capabilitarian account of justice. In any case, the point which capability theorists of justice should take home is that a principle of fairness, such as the Dworkinian envy-test, will have to be developed if one wants the capability metric to do the work of a complete theory of distributive justice. A capability theory of justice that endorses sufficiency as the distributive rule, such as Martha Nussbaum’s theory (2000, 2006, 2011), can avoid the problem of aggregation, since such a theory specifies that justice requires only that all people meet a certain threshold level for each capability. Nevertheless, even such a theory has two major problems to solve. First, we will need to know where the thresholds are set, or by which procedure, or which foundational principles determine the thresholds. Ultimately, the normative decisions that confront the selection of relevant capabilities re-​emerge here, albeit in a slightly different guise. Second, in an unjust world, the theory of transitional justice or the non​ideal theory of justice will need to tell us which capability to prioritize if not all people are above the thresholds of all relevant capabilities. Should we prioritize education, health, or being able to hold a decent job? Nussbaum (2011: 37–​8) has argued that the impossibility to get all people above the thresholds for all capabilities involves a tragic choice, which should prompt us to ask the question of how we can work towards a future where this is no longer the case. For Nussbaum, this strategy will be sufficient, since “If the whole list has been wisely crafted and the thresholds set at a reasonable level, there usually will be some answer to that question” (Nussbaum 2011: 38). One could wonder whether this response really solves the problem: many of the one billion most deprived people on earth are below reasonable thresholds on most of the

124   Ingrid Robeyns capabilities on Nussbaum’s list. Which capabilities should justice-​seeking organizations or governments prioritize?

5.8  A Family of Capabilitarian Theories of Justice The capability approach is often taken to be an egalitarian theory or a theory of social or distributive justice. While it is true that the capability approach can be developed into a theory or account of justice, it only provides one aspect of such an account (or a “module,” to use Arneson’s terminology introduced in section 5.2). The account provided is the metric of justice, hence a clear claim on what should count for interpersonal evaluations for the purpose of justice. However, additional modules are needed before one can speak of a theory of justice. Nussbaum (2000, 2006, 2011) offers us a capability theory of justice, but her theory too doesn’t amount to a full theory of social justice. Moreover, it would be a mistake to think that there can be only one capability theory of justice; on the contrary, the open nature of the capability approach allows for the development of a family of capability theories of justice. But this prompts the question: what is needed to develop a full capability theory of justice, and what other aspects of a theory of justice have already been developed by capability theorists? Assuming that the capability theory of justice has addressed the issues we have outlined—​that is, the choice for functionings, capabilities or both, the selection of the relevant capabilities, and the question of aggregation—​several other theoretical choices remain to be made and defended (Robeyns 2016). First, a theory of justice needs to explain on what basis it justifies its principles or claims of justice. For example, in Rawls’s theory of justice the two principles of justice are justified by the thought-​experiment of the original position and the more general social contract framework on which this is based. Dworkin’s egalitarian justice theory starts from the meta-​principle of equal respect and concern, which Dworkin then argues supports the principles that the distribution of burdens and benefits should be sensitive to the ambitions that people have, but should not reflect the unequal natural endowments with which individuals are born. One could also develop a capability theory of justice, arguing that the ultimate driving force is a concern with autonomy or with human dignity, or with human vulnerability, or perhaps a combination of these. If capability scholars want to develop a full theory of justice, they will also need to explain on what bases they will justify their principles or claims. As mentioned earlier, Nussbaum starts from a notion of human dignity, whereas the Senian strand in the capability approach stresses the importance of what people have reason to value. Those reasons are then put together in a collective (or public) account of public reasoning; this is, ultimately, an account of justice that at its core has some strong procedural elements (or, to put it in Sen’s preferred terminology, it is a social choice perspective on theorizing

The Capability Approach    125 about justice). However, little work has been done so far to flesh out Sen’s embryonic idea of “having reason to value,” and it therefore remains unclear whether the capability approach has a solid unified rationale on the basis of which a full account of justice could be developed. Second, a capability theory of justice needs to take a position on the “distributive rule” that it will endorse: will it argue for plain equality, or for sufficiency, or for prioritarianism, or for some other (mixed) distributive rule? Both Martha Nussbaum’s and Elizabeth Anderson’s theories are sufficiency accounts (Anderson 1999, 2010; Nussbaum 2006), but from this it does not follow, as one sometimes reads in the secondary literature, that the capability approach entails a sufficiency rule. Sen may have given the (wrong) impression of defending straight equality as a distributive rule, by asking the question, “Equality of what?” (Sen 1980), though a careful reading shows that he was merely asking the question “If we want to be defending equality of something, then what would that be?” In fact, Sen has remained uncommitted to one single distributive rule, which probably can be explained by the fact that he is averse to building a well-​defined theory of justice, but prefers rather to investigate how real-​life unjust situations can be turned into more just situations, even if perfect justice is unattainable (Sen 2009). The capability approach clearly plays a role in Sen’s work on justice since, when assessing a situation, he will investigate inequalities in people’s capabilities and analyze the processes that led to those inequalities. However, Sen has an eclectic approach to theorizing, and hence other notions and theories (such as human rights or more formal analyses of freedoms from social choice theory) also play a role in his work on justice. The presence and importance of the capability approach in Sen’s work on just­ice is thus undeniable, but should not be seen as the only defining feature. Third, a capability theory of justice needs to specify where the line between individual and collective responsibility is drawn, how it will be decided, and by whom. There is a remarkable absence of much discussion about issues of responsibility in the capability literature, in sharp contrast to political philosophy and welfare economics, where this is one of the most important lines of debate, certainly since the publication of Ronald Dworkin’s (1981b) work on justice and equality, which led to what Elizabeth Anderson (1999) has called “luck  egalitarianism.” An exception is Peter Vallentyne (2005:  365), who has argued that the relevant metric of justice is “brute luck capabilities,” these being the capabilities that one has as a matter of brute luck only; those that one has as a matter of option luck (that is, because of one’s own choices) are not a concern of justice. While Vallentyne endorses a strict separation between theorizing about justice on the one hand, and policy and institutional design on the other, the question of responsibility also has important effects for the more practical approach to justice. Indeed, whether one wants to discuss it explicitly or not, any concrete capability-​based policy proposal can be analyzed in terms of the division between personal and collective responsibility; but this terminology is largely absent from the capability literature. There may be plausible explanations for why this issue is not discussed in the capability literature, but if a capability theory of justice wants to be applicable to questions of justice, then it cannot but confront the question about

126   Ingrid Robeyns the just division between personal and collective responsibility (Pierik and Robeyns 2007: 148–​9). This brings us to a related issue: a theory of justice generally specifies not only rights, but also duties. However, capability theorists have remained largely silent on the question of who should bear the duties for securing the selected capabilities. Nussbaum passionately argues that all people all over the world should be entitled, as a matter of justice, to threshold levels of all the capabilities on her list; but apart from mentioning that it is the governments’ duties to guarantee these entitlements (2006: 70), she remains silent on the question of who precisely should bear the burdens and responsibilities for realizing these capabilities. Yet as Onora O’Neill (1996: 141–​6) has argued, questions of obligations should be central to any account of justice. This short and presumably incomplete list of the “modules” which a complete capability theory of justice would need to comprise makes clear that a capability theory of justice is theoretically much more demanding than the basic presupposition of the capability approach to distributive justice, namely its claim that functionings and/​or capabilities are the best metric of justice. While much has been written on the capability approach in recent years, by an increasing number of scholars, including philosophers, much of the philosophical work needed for turning the open-​ended capability approach into a capability theory of justice remains to be done.

Acknowledgments I am grateful to Sem de Maagt and Serena Olsaretti for helpful comments on an earlier draft of this chapter, and to the Netherlands Organization of Scientific Research (NWO) for research funding.

References Anderson, E. (1999). “What is the Point of Equality?” Ethics 109(2): 287–​337. Anderson, E. (2010). “Justifying the Capabilities Approach to Justice,” in H. Brighouse and I. Robeyns (eds) Measuring Justice: Primary Goods and Capabilities. Cambridge: Cambridge University Press: pp. 81–​100. Arneson, R. (2010) “Two Cheers for Capabilities,” in H. Brighouse and I. Robeyns (eds) Measuring Justice:  Primary Goods and Capabilities. Cambridge:  Cambridge University Press, pp. 101–​27. Barclay, L. (2003). “What Kind of Liberal is Martha Nussbaum?” SATS:  Nordic Journal of Philosophy 4(3): 5–​24. Brighouse, H. and Robeyns, I. (eds) (2010). Measuring Justice: Primary Goods and Capabilities. Cambridge: Cambridge University Press. Carter, I. (2014). “Is the Capability Approach Paternalistic?” Economics and Philosophy, 30: 75–​98. Chakraborty, A. (1996). “On the Possibility of a Weighting System for Functionings.” Indian Economic Review 31: 241–​50.

The Capability Approach    127 Claassen, R. (2014). “Capability Paternalism.” Economics and Philosophy 30: 57–​73. Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99(4): 906–​44. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Dworkin, R. (1981a). “What is Equality? Part 1: Equality of Welfare.” Philosophy and Public Affairs 10: 198–​246. Dworkin, R. (1981b). “What is Equality? Part 2: Equality of Resources.” Philosophy and Public Affairs 10: 283–​345. Dworkin, R. (2002). Sovereign Virtue:  The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Fleurbaey, M. (2002). “Development, Capabilities and Freedom.” Studies in Comparative International Development 37: 71–​7. Kelly, E. (2010). “Equal Opportunity, Unequal Capability,” in H. Brighouse and I. Robeyns (eds) Measuring Justice: Primary Goods and Capabilities. Cambridge: Cambridge University Press, pp. 61–​80. Nelson, E. (2008). “From Primary Goods to Capabilities: Distributive Justice and the Problem of Neutrality.” Political Theory 36(1): 93–​122. Nussbaum, M. (2000). Women and Human Development:  The Capabilities Approach. Cambridge: Cambridge University Press. Nussbaum, M. (2003). “Capabilities as Fundamental Entitlements:  Sen and Social Justice.” Feminist Economics 9(2/​3): 33–​59. Nussbaum, M. (2006). Frontiers of Justice:  Disability, Nationality, Species Membership. Cambridge, MA: Harvard University Press. Nussbaum, M. (2011). Creating Capabilities: The Human Development Approach. Cambridge, MA: Harvard University Press. Olsaretti, S. (2005). “Endorsement and Freedom in Amartya Sen’s Capability Approach.” Economics and Philosophy, 21: 89–​108. O’Neill, O. (1996). Towards Justice and Virtue. Cambridge: Cambridge University Press. Parfit, D. (1984). Reasons and Persons. Oxford: Oxford University Press. Phillips, A. (2004). “Defending Equality of Outcome.” Journal of Political Philosophy 12(1): 1–​19. Pierik, R. and I. Robeyns (2007). “Resources versus Capabilities:  Social Endowments in Egalitarian Theory.” Political Studies 55(1): 133–​52. Pogge, T. (2002). “Can the Capability Approach be Justified?” Philosophical Topics 30(2): 167–​228. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (2001). Justice as Fairness: A Restatement. Cambridge, MA: Harvard University Press. Richardon, H. S. (2006). “Rawlsian Social Contract Theory and the Severely Disabled.” Journal of Ethics, 10: 419–​62. Robeyns, I. (2009). “Justice as Fairness and the Capability Approach,” in Kaushik Basu and Ravi Kanbur (eds) Arguments for a Better World. Essays for Amartya Sen’s 75th Birthday. Oxford: Oxford University Press, pp. 397–​413. Robeyns, I. (2013). “The Capability Approach (to Social Justice),” in F. D’Agostino and G. Gaus (eds) Routledge Companion to Social and Political Philosophy. New  York:  Routledge, pp. 456–​66. Robeyns, I. (2016). “The Capability Approach,” in Edward N. Zalta (ed.) Stanford Encyclopedia of Philosophy. http://​plato.stanford.edu/​archives/​sum2011/​entries/​capability-​approach/​ (accessed 8 November 2017).

128   Ingrid Robeyns Robeyns, I. (2017). Wellbeing, Freedom and Social Justice:  The Capability Approach Re-​ Examined. Cambridge: Open Book Publishers. Schokkaert, E. (2007). “Capabilities and Satisfaction with Life.” Journal of Human Development 8(3): 415–​30. Sen, A. (1980). “Equality of What?,” in S. McMurrin (ed.) Tanner Lectures on Human Values. Cambridge: Cambridge University Press, pp. 195–​220. Sen, A. (1992). Inequality Re-​Examined. Oxford: Clarendon Press. Sen, A. (1993). “Capability and Well-​Being,” in M. Nussbaum and A. Sen (eds) The Quality of Life. Oxford: Clarendon Press, pp. 30–​53. Sen, A. (2004). “Capabilities, Lists and Public Reasons: Continuing the Conversation.” Feminist Economics 10(3): 77–​80. Sen, A. (2009). The Idea of Justice. London: Allen Lane. Vallentyne, P. (2005). “Debate:  Capabilities versus Opportunities for Wellbeing.” Journal of Political Philosophy 13: 359–​7 1. Williams, A. (2002). “Dworkin on Capability.” Ethics 113: 23–​39. Wolff, J. and de-​Shalit, A. (2007). Disadvantage. Oxford: Oxford University Press.

Chapter 6

Libertarian i sm, L e ft and Rig h t Joseph Mazor and Peter Vallentyne

Libertarianism is a school of moral/​political thought that is committed to full or near-​ full individual self-​ownership. In the realm of distributive justice, libertarianism is probably most famous (or infamous) for its rejection of redistributive taxation. However, the libertarian view of distributive justice is less monolithic and more complex than is often recognized, and in this chapter we aim to clarify (and in in places tentatively defend) this view.1 We begin by briefly defining “distributive justice” and “libertarianism” (section 6.1). We then turn to considering self-​ownership (section 6.2), property rights in natural resources (section 6.3), and property rights in human-​made goods (section 6.4). Finally, we consider directions for future research (section 6.5) and conclude.

6.1 Preliminaries 6.1.1 Justice and Distributive Justice While there are many ways of understanding “justice” (Vallentyne 2003a), for the purposes of this chapter, we take justice to be concerned with enforceable moral duties, that is, those duties whose compliance may be permissibly induced using force or threat thereof.2 We define distributive justice as that area of justice relating to the distribution of economic benefits and burdens. 1  One of us, Peter Vallentyne, is a left libertarian. The other, Joseph Mazor, is a sympathetic critic of libertarianism. However, in this chapter we adopt the perspective of defenders of libertarianism. 2  There are coherent pacifist versions of libertarianism that recognize various libertarian duties but no enforceable duties. Such libertarian theories would have nothing to say about justice as we define it here. Since every prominent libertarian thinks that coercive force may be used to protect property rights, we will not consider these pacifist libertarian theories here.

130    Joseph Mazor and Peter Vallentyne

6.1.2 Libertarianism As the term suggests, libertarianism entails a commitment to respect for liberty. However, this is a commitment to a very particular conception of liberty: liberty as full or near-​full self-​ownership. The commitment to this type of liberty can be justified on a variety of grounds,3 including consequentialism (e.g., Epstein 1998), social contract (e.g., Narveson 1998), autonomy (e.g., Grunebaum 2000), natural rights (e.g., Mack 1990), and the value of freedom from interference.4 We will not explore the foundations of the commitment to self-​ownership in this chapter. We will focus instead on explaining this commitment and its implications for distributive justice.

6.2  Full and Near-​Full Self-​O wnership Self-​ownership is simply a special case of ownership where the owner and the thing owned are one and the same. We therefore begin our discussion of self-​ownership by considering the more general concept of ownership.

6.2.1 Ownership Ownership can be best understood as a collection of certain Hohfeldian liberties, claim-​ rights, powers, and immunities that some person has with respect to some entity.5 We will refer to these rights as property rights. These include:6 (1) control rights over the entity (liberty-​rights to use, a power to authorize use or encroachment by others, and a claim-​right that others not use or encroach upon the entity without one’s authorization); (2) rights to transfer one’s property rights over the entity to others (by sale, rental, gift, or loan); (3) enforcement rights (e.g., rights of prior restraint, punishment) with respect to others’ violations or potential violations of one’s property rights over the entity; (4) rights to compensation when someone violates one’s property rights over the entity without one’s permission; (5) immunities to the non-​consensual loss of one’s property rights over the entity. 3 

For a critique of the libertarian commitment to self-​ownership, see Lippert-​Rasmussen 2008. For a debate on whether maximal equal negative liberty leads to traditional libertarianism conclusions, see Narveson and Sterba 2010. For a discussion, see Vallentyne 2011. 5  For a more detailed account of the nature of these rights, see Hohfeld 1919. 6  This list is enumerated and briefly discussed in Vallentyne, Steiner, and Otsuka 2005: 203–​4. 4 

Libertarianism, Left and Right    131 We will not take a position on which property rights are precisely necessary and sufficient for ownership. However, we will assume that an entity’s owner must at least have liberty-​rights to use the entity and either a claim-​right against others’ encroachment on the entity or a right to appropriate compensation if others use the entity without the owner’s permission.7 We take it that these are the minimal requirements for ownership. The more property rights an owner has with respect to an entity, the stronger their ownership over the entity is. It is worth emphasizing straightaway that even the strongest form of ownership over an entity does not guarantee a person the effective moral freedom to make use of that entity in any practical sense. After all, there are basically no actions that someone takes with respect to an entity that use or encroach upon only that entity. For example, when I ride my motorcycle, I occupy space, discharge pollution into the air, and generate sound waves that encroach on other individuals. Ownership of my motorcycle only grants me a liberty-​right to use the motorcycle itself. Before we can conclude that I have moral permission to ride my motorcycle somewhere, we need to know about my property rights with respect to all the entities that my motorcycle ride would use or encroach upon. This is a general point that applies to all types of ownership, including self-​ ownership, which we turn to next.

6.2.2 Self-​Ownership Self-​ownership is ownership of one’s own person.8 Setting aside a variety of complex issues about personal identity, we will make the simplifying assumption that a person is their body. Thus, one can think of self-​ownership intuitively as giving an individual (moral) rights over their body that are similar to the legal rights that slave-​owners had (and in some parts of the world unfortunately still have) over their slaves (Cohen 1995: 68). A wide variety of theories are committed to granting individuals some type of self-​ ownership (e.g., Thomson 1992: Ch. 8). What distinguishes libertarianism from these other theories is the commitment to the following key moral claim (or something similar to it): Every rational agent initially possesses full self-​ownership.9 We will refer to this as the full self-​ownership thesis.

7 

We set aside here the important question of what constitutes appropriate compensation. As Cohen 1995: 68–​9 argues, the term “self ” in self-​ownership has a purely reflexive significance. 9  More accurately, the thesis requires that every agent possess the same set of self-​ownership rights. By “initially,” we mean before the agent contracts away any of these rights or loses these rights as a result of engaging in rights violations. 8 

132    Joseph Mazor and Peter Vallentyne

6.2.3 Full Self-​Ownership Full self-​ownership assigns each person the logically strongest control rights (1) and transfer rights (2) over themselves possible.10 By “logically strongest” we mean both maximally stringent and maximally extensive. “Maximally stringent” means that the ownership rights can never be overridden by other moral considerations (e.g., those having to do with human welfare). “Maximally extensive” means the most complete list of Hohfeldian rights within the relevant categories possible.11 However, full self-​ownership does not assign a person the logically strongest self-​ ownership rights possible in categories (3), (4), and (5). The reason is this: It would be impossible to grant every person (or indeed, even more than one person) full self-​ownership were it defined this way. After all, the stronger we make one person’s rights of compensation and enforcement (i.e., what they may do to others in self-​defense), the weaker we have to make others’ immunity to the non-​consensual loss of self-​ownership rights, and vice versa. Thus, libertarians define full self-​ownership instead as granting a person a set of self-​ownership enforcement rights (3), compensation rights (4), and immunity to loss of self-​ownership rights (5) that are as strong as possible subject to the constraint of being compatible with others having the same set of rights in categories (3), (4), and (5) over themselves (Vallentyne, Steiner, and Otsuka 2005: 203–​5). Having defined the concept of full self-​ownership, we now turn to considering three prominent criticisms of the full self-​ownership thesis having to do with indeterminacy, small encroachments or risks of encroachments, and enforceable duties to assist.

6.2.4 Is Full Self-​Ownership Indeterminate? Some theorists have criticized libertarianism by arguing that the concept of full self-​ ownership is indeterminate (Fried 2004: 78–​80). And indeed, there is some truth to this criticism. Full self-​ownership is indeterminate when it comes to the question of what a person may do in self-​defense. Since it is the case that the stronger we make one person’s enforcement and compensation rights (3  and 4), the weaker we have to make another’s immunity to non-​consensual loss of self-​ownership rights (5) and vice versa, there are admittedly many possible bundles of rights in categories (3), (4), and (5) that would satisfy the definition of full self-​ownership.12 10  We also assume that full self-​owners cannot be deprived of their other property rights simply because they exercise their self-​ownership rights. So depriving individuals of the right to breathe air (assuming that they already have that right) simply because they smiled would violate their full self-​ownership. 11  Our discussion of extensiveness and stringency follows Wall 2009: 400. 12  However, full self-​ownership is not completely indeterminate in the realms of rights in categories (3), (4), and (5). For example, full self-​ownership cannot include both strong rights in (3) and (4) and strong rights in (5) (since this would not be compatible with granting others the same set of rights). It also cannot include both weak rights in (3) and (4) and weak rights in (5) (since it would be possible

Libertarianism, Left and Right    133 One conception of full self-​ownership could grant me full rights to defend my body, including the right to kill you if you are threatening to punch me. Another could deny me the right to do anything to you in response to your attack on my body. Another might provide me with enforcement rights that are proportional in some way to the threat you pose. Each of these is an admissible conception of full self-​ ownership. Thus, full-​self ownership is indeed indeterminate when it comes to the question of self-​defense. However, some theorists (e.g., Fried) have claimed that full self-​ownership is also indeterminate when it comes to the question of whether one person may exercise their self-​ownership control rights in ways that violate others’ self-​ownership control rights (Fried 2004: 79). Libertarians, on the other hand, argue that there is no conflict between self-​ownership control rights. For example, though you have full control rights over your fist, you cannot use your fist to punch my nose. Fried is unconvinced. She argues that such intuitively obvious examples simply obscure the indeterminacy regarding the conflict between different self-​ownership control rights. To evaluate Fried’s criticism, let us consider the following example:  Realizing that I am running late to an important appointment, I start jogging while others are walking nearby, thereby introducing a small risk that my hand will accidentally hit someone else’s body. Is such jogging permissible? Fried’s argument suggests that the full self-​ownership thesis does not provide a determinate answer to this question. However, Fried is mistaken. She seems to think that libertarians must evaluate the permissibility of jogging by somehow balancing the control rights I  have over my body against the control rights others have over their bodies to see which one would “win” (Fried 2004: 79). This is wrong. The property rights I have over my body, even in their logically strongest form, do not include a right to take whatever actions I want with my body (just as full ownership over my motorcycle does not give me a right to ride the motorcycle wherever I want). To evaluate the permissibility of my jogging, we need to know what rights I have with regards to any other entities that my jogging might encroach upon or use. Since my jogging generates a risk of incursion on others’ bodies, and since granting every person the logically strongest possible control-​rights over their body gives them rights against even small risks of incursions, the full self-​ ownership thesis condemns my decision to start jogging, and does so in a fully determinate way.13 Although the prohibition on my jogging is admittedly problematic, the problem here is not indeterminacy.

to give a person stronger property rights in one of the categories in this case without violating the compatibility constraint). 13 

Note that, while others’ self-​ownership rights place moral restrictions on my actions in this case, these moral restrictions do not conflict with my self-​ownership rights in any way.

134    Joseph Mazor and Peter Vallentyne

6.2.5 Full Self-​Ownership and Small Incursions or Risks of Incursions The problem illustrated by the jogging case is this: libertarianism seems committed to an implausibly strong version of self-​ownership. The full self-​ownership thesis implies that any action that causes small incursions or risks of incursions on others is impermissible. Yet this seems to generate unacceptable restrictions on human activity. Libertarians have considered this problem in the context of pollution and driving a car (Nozick 1974: 78–​81). However, as the jogging example suggests, the problem is more ubiquitous than is commonly recognized. In fact, almost any action (e.g., even carefully walking to the grocery store) will impose some additional risk of incursion on another’s body relative to the option of not doing anything. And since full self-​ownership does not grant any weight to positive freedoms or human welfare in a contest with the rights that people have against (even tiny risks of) bodily incursions, the troubling upshot of the full self-​ownership thesis seems to be universal, morally required paralysis (Mack 2011: 112–​13). However, this problem does not necessitate abandoning a commitment to strong self-​ownership rights altogether. Instead, libertarians can respond to this problem by slightly stepping back from the full self-​ownership thesis to endorse some version of the near-​full self-​ownership thesis: the thesis that every rational agent initially has a set of self-​ownership rights that is “close” to full self-​ownership rights (where “close” entails minor, circumscribed deviations, either in terms of stringency or in terms of extensiveness).14 For example, in order to permit actions like jogging while others are around, libertarians might weaken individuals’ control rights over themselves to allow sufficiently small (i.e., “trivial”) incursions or risks of incursions by others. Needless to say, this theoretical move would require accounts of what counts as a trivial incursion and why such sufficiently small incursions are permissible. We will not explore these important issues further here.15

6.2.6 Full Self-​Ownership and Duties of Physical Assistance We turn instead to a third criticism of the full self-​ownership thesis relating to its rejection of enforceable duties to assist others. Even if a child is drowning in a shallow puddle

14  Admittedly, libertarianism is sometimes understood as requiring full self-​ownership. On this view, any move to merely near-​full self-​ownership would be an abandonment of libertarianism in the strictest sense. However, such an understanding of libertarianism seems to us to be overly restrictive (it would exclude a large number of theorists commonly referred to as “libertarian”). 15  For a discussion of the different libertarian approaches to address the problem of small incursions, see Sobel 2012.

Libertarianism, Left and Right    135 next to the only person who is able to help, a commitment to the potential rescuer’s full self-​ownership prohibits anyone (including the state) from forcing the person to help (even if doing so would be quite easy for them) (Arneson 2010: 183). Most people find this implication of the full self-​ownership thesis to be unacceptable. Libertarians have several responses to this important objection. First, they can point out that their theory does not reject all duties to physically assist. The core libertarian claim is only that individuals have no general, enforceable duties to physically aid others. This is compatible with their having special enforceable duties (e.g., from contract or past wronging) to aid specific others. It is also compatible with individuals having non-​ enforceable general moral duties to aid others. Second, libertarians can highlight the unappealing implications of principles that allow for blanket violations of self-​ownership in order to assist the disadvantaged. For example, Nozick (1974: 206) points out that such principles might grant some entity (e.g., state) the right to forcibly remove non-​essential organs (e.g., an eye from a person with two) in order to help those who are suffering (e.g., the congenitally blind). Such a forced transfer of eyes seems morally unacceptable. Third, some libertarians have pointed out that it is often possible to alleviate suffering without violating anyone’s self-​ownership. A  great deal of human suffering could be alleviated if only the badly off owned more resources. Indeed, some libertarians (left ​libertarians) have been particularly interested in the role that the distribution of natural resources can play in increasing the welfare of the less fortunate. We turn now to discussing the libertarian debate over natural resource property rights.

6.3  Property Rights in  Natural Resources Natural resources can be defined (roughly) as objects of value16 that (1) have not been produced by any rational agent and (2) are not part of a being with moral standing.17 The debate over the ownership of natural resources defines the left/​right libertarian divide. Left libertarians endorse some type of egalitarian ownership of natural resources. Right libertarians do not. Note that this debate cannot be settled by an appeal to the full or near-​full self-​ ownership thesis. Natural resources, by definition, are not part of, and have not been 16  By objects of value, we mean objects that serve someone’s needs, desires, wants, interests, or values (Narveson 1998: 15). There are also natural objects without value (e.g., a worthless rock) and objects that are bad for humans in some way (e.g., toxic natural chemicals). These are not resources given our definition and, for simplicity, we do not address the libertarian view on these objects here. 17  A being has moral standing just in case its will or interests matter morally for their own sake. We leave open here the possibility that there may be beings with moral standing other than rational agents.

136    Joseph Mazor and Peter Vallentyne produced by, any rational agent. Moreover, since self-​ownership rights are not rights to take action, full self-​ownership rights do not include rights to use or encroach upon any natural resources. Having full self-​ownership does not even guarantee a person the right to breathe air! Since the central moral principle that unites libertarians says nothing about natural resource property rights, it is perhaps unsurprising that natural resources have been a locus of significant disagreement among libertarians. We will attempt to clarify the different basic positions in this contentious debate by considering three interrelated questions: (1) Are natural resources initially owned? (2) If not, are there any moral restrictions on appropriation of natural resources? (3) If so, what are these restrictions? In considering these questions, we will assume for simplicity a single generation of adults living in a single nation-state. We will briefly consider issues relating to minors, intergenerational, and international justice at the end of this chapter.

6.3.1 Are Natural Resources Initially Owned? The first question that divides libertarians is this: Are natural resources initially owned? If they are initially owned, then no one may use the owned natural resources without either obtaining the owner’s permission or providing the owner with appropriate compensation. A variety of thinkers (including some libertarians) have endorsed the claim that natural resources are initially owned. Some have suggested that natural resources should be seen as jointly ​owned in the sense that their utilization requires a collective decision-​ making process such as majority decision-​making (Grunebaum 2000: 54–​9) or unanimous consent (Cohen 1995: 94–​5). However, the initial joint-​ownership idea is open to criticism. As both Narveson (1998: 12) and Rothbard (2000: 224) point out, Cohen’s joint-​ownership proposal would leave individuals unable to enjoy any substantive liberty (since they would need others’ permission to even breathe).18 Moreover, Rothbard (2000: 224) argues that it is simply impractical for billions of people to jointly exercise any sort of collective ownership rights over the world’s natural resources.

18 

Rothbard’s and Narveson’s appeals to substantive liberty to dismiss Cohen’s collective ownership proposal open them up to the charge of inconsistency. After all, in dismissing the claims of the desperately poor to assistance, Rothbard and Narveson reject the value of substantive liberty and defend self-​ownership instead. Indeed, Cohen’s (1995: 94–​102) purpose in introducing his joint natural resource ownership proposal is precisely to force libertarians to recognize the importance of substantive liberty.

Libertarianism, Left and Right    137 Cohen (1995: 102–​3) also considers a different type of initial ownership: Namely, he suggests that each person might be seen as having initial ownership of an equal share of all available natural resources.19 However, the equal-​ share initial ownership proposal is also problematic. For one thing, it is not clear who should be granted ownership of which share of natural resources. And while we could rely on some central authority to allocate ownership of particular shares to particular individuals,20 such an authority did not always exist, and it is unclear how equal-​initial-​ownership proposals would work in its absence. Due to these problems, the vast majority of libertarians reject the initial ownership view. They hold instead that natural resources are initially unowned and morally available for anyone’s use. We will refer to this as the common-​use view and the initial situation as the common-​use state. On this view, actions like picking apples from a tree or bathing in a stream do not require anyone’s permission, nor do they generate any duties of compensation (Roark 2012: 689). Though quite popular among libertarians, the common-​use view is also open to criticism. One key problem is that, intuitively, there seem to be significant moral restrictions even on mere natural resource use when there is scarcity and when a person uses a large portion of the available natural resources (Roark 2012: 695). For example, it seems wrong for a person in the common-​use state to bathe in the only source of potable water (thus fowling it for everyone else) or to pick one of the three remaining apples in society without providing any compensation to others. Yet the proponents of the traditional common-​use view seem committed to permitting these actions. Despite this criticism, the majority of contemporary libertarians continue to endorse the common-​use view. However, they do not insist that natural resources must remain in the common-​use state forever. Instead, they hold that individuals have a moral power to appropriate (i.e., acquire ownership of) previously unowned natural resources, at least under certain conditions.21 The key philosophical question, which we turn to next, concerns the conditions under which an individual may appropriate natural resources.

6.3.2 No Moral Restrictions on Appropriation One answer to this question, endorsed by libertarians like Rothbard (2000) and Narveson (1998), is that there are no moral restrictions on natural resource appropriation (libertarians who endorse this position are sometimes called radical-​right libertarians). 19 

Cohen mistakenly attributes this view to Steiner. Although Steiner holds that every person has an initial moral claim to an equal share of natural resources, he holds that no one initially has ownership of any particular share (Steiner 1994: 235 n211). The appropriation process determines which person obtains ownership rights over which particular natural resources. 20  An example of this is Dworkin’s (1981: 283–​90) natural resource auction. 21  An interesting question, which we cannot address here, concerns the boundaries of the appropriated property. When I appropriate land, do I also obtain rights to the space above the land? If so, how far up? Do I obtain rights to what is below the land? How far below?

138    Joseph Mazor and Peter Vallentyne On this view, individuals not only have a liberty-​right to use natural resources in the common-​use state, they also have an unconditional moral power to acquire permanent ownership of currently unowned natural resources by taking some action (e.g., laboring on the resource).22 However, many theorists have criticized this radical-​ right libertarian position. Primarily, critics have questioned whether merely being a first-​comer should entitle a person to the enormous value of undeveloped natural resources.23 And while some radical-​right libertarians (e.g., Narveson 1998; Rothbard 2000: 225–​6) have argued that human activity is responsible for the entire value of natural resources, this position has been criticized (Mazor 2009: 43–​52). If natural objects indeed have a value that has not been created by anyone, it is not clear why first-​comers should be able to appropriate this entire value for themselves.

6.3.3 The Equal Claims View and Restrictions on Appropriation In fact, many libertarians endorse restrictions on natural resource appropriation. They hold (implicitly or explicitly) that all individuals have some type of initial moral claim24 to natural resources—​claims that fall short of initial ownership, but which nevertheless ground, not only liberty-​rights to use natural resources, but also conditional immunities against the loss of these liberty-​rights. In this section, we will consider theories that hold that individuals have equal initial moral claims to natural resources. We will refer to this as the equal claims view. Under what conditions can individuals lose their liberty-​rights to use particular natural resources? Several prominent libertarians have answered this question by appealing to Locke’s (Locke and MacPherson 1980: §33) famous contention that one person’s appropriation of natural resources may be permissible (i.e., does not run afoul of others’ initial moral claims to natural resources) as long as it leaves others with “enough and as good.” This is known as the Lockean Proviso. The key debate among these libertarians has been about the interpretation of this proviso. Nozick (1974: 176–​7) defends the following interpretation of the proviso: He suggests that one person’s appropriation must leave “enough and as good” only in the sense that others are no worse off than they would have been had all natural resources remained in the common-​use state.

22 

There is significant debate over the precise actions that are necessary to appropriate natural resources. Some libertarians argue that appropriation requires mixing one’s labor with a natural resource (Rothbard 2000: 223–​7). Others insist that first possession is all that is necessary for appropriation (Narveson 1998: 11). Still others suggest that it is only necessary to publically stake a claim (Vallentyne 2007: 273). We will not explore this debate here. 23  For a version of this criticism, see Otsuka 2003: 23–​4. 24  These are moral claims in the broad sense of the term, rather than Hohfeldian claim-​rights.

Libertarianism, Left and Right    139 Nozick’s account of permissible natural resource appropriation is open to a wide variety of objections (Kymlicka 2001: 111–​2 1), and here we will focus on the one we take to be the most important:  Like the natural resource appropriation theories of Rothbard and Narveson, Nozick’s theory condones enormous, seemingly arbitrary inequality in the distribution of natural resource wealth.25 As G. A. Cohen (1995: 79–​80) points out, Nozick’s theory would allow one person to appropriate all available natural resources as long as that person pays others a wage to work those natural resources that is just high enough so that they are no worse off than they would have been in the common-​use state. And since the common-​use state is likely to be quite poor (due to the lack of sufficient incentive to develop or forebear from overusing natural resources), the morally required wages would be fairly low. As Otsuka (2003: 23–​4) argues, this outcome seems both blatantly unfair and implausible as an interpretation of the requirement to leave “enough and as good” for others. Hillel Steiner offers a different account of the Lockean Proviso. Steiner claims that a person has left “enough and as good” if and only if they have appropriated no more than an equally valuable share of pristine natural resources (1994: 235–​6). Since the first appropriator is not morally permitted to appropriate a share that is any greater in value than the share taken by the last appropriator, Steiner’s proviso is not subject to the criticism that it grants a significant unfair advantage to first-​appropriators.26 However, the equal share interpretation of the Lockean Proviso is subject to a different criticism:  Namely, it is insensitive to unfair inequalities in individual circumstances (Quong 2011: 68–​70). To see the problem, assume a world with only two people and a single natural resource called manna. Imagine that one of the people has an illness (through no fault of their own) that makes it the case that they need two-​ thirds of the available manna to survive while the other person only needs one-third to survive. Would it be just to permit the healthy person to appropriate half of the manna? Otsuka argues that it would not. He explicitly rejects the position that all individuals have equal initial moral claims to natural resources. Instead, he holds that those who are unfairly disadvantaged have greater initial moral claims to natural resources than do the unfairly advantaged (Otsuka 2003: 29). Otsuka endorses conditions on natural resource appropriation that aim to mitigate these unfair inequalities (as does Vallentyne). We now turn to considering this type of view on permissible appropriation in greater detail.

25 

While Nozick (unlike the radical-​right libertarians) accepts that there are restrictions on natural resource appropriation, he ultimately condones very unequal distributions of natural resource ownership (which is why he is classified as a right libertarian). 26  Steiner (1994: 268) does, however, like most libertarians, condone individuals’ appropriation of more than the benchmark share as long as they pay redress for the excess share equal to the market value of the over-​appropriated natural resources to those who have under-​appropriated. We do not address this part of Steiner’s theory here.

140    Joseph Mazor and Peter Vallentyne

6.3.4 Equality-​Aimed Restrictions on Appropriation To understand Otsuka’s and Vallentyne’s accounts of permissible natural resource appropriation, it is first necessary to understand their conception of unfair inequality. Vallentyne (2002) and Otsuka 2003: 25 hold that it is unfair that some have less opportunity for welfare than others. Although it is impermissible on their view to violate anyone’s self-​ownership to correct for inequalities in opportunities for welfare, Otsuka and Vallentyne endorse restrictions on natural resource appropriation that go at least part of the way towards correcting these inequalities. That is, they endorse equality-​ of-​opportunity-​for-​welfare-​aimed restrictions on natural resource appropriation (or equality-​aimed restrictions for short). While Otsuka and Vallentyne’s theories (like Steiner’s) do not significantly advantage first-​comers, they are subject to a variety of other potential criticisms. One potential criticism (from the left) is that these theorists give too much weight to self-​ownership relative to equality of opportunity for welfare. Once it is conceded that equality of opportunity for welfare is an important moral value, it seems strange (though not philosophically incoherent) to effectively grant self-​ownership absolute or near-​absolute priority over this egalitarian value. Why are fairly minor violations of self-​ownership (e.g., painlessly taking a small amount of blood) more important than very large gains in equality of opportunity for welfare (Arneson 2010: 184)? A second criticism is that these theorists do not go far enough in their pursuit of equality of opportunity for welfare, even given an unwavering commitment to self-​ ownership. After all, it is possible to have a regime of natural resource ownership that is wholly consistent with full self-​ownership while at the same time fully achieving equality of opportunity for welfare (e.g., by requiring the advantaged to transfer sufficient resources to the disadvantaged as a condition of breathing air). Yet neither Vallentyne nor Otsuka calls for such a regime.27 A third criticism, this time from the right, is that this type of libertarian theory appeals to the wrong notion of fairness. Many theorists, including several right-​leaning liberal theorists, argue that, rather than requiring any type of equality, fairness requires respecting individuals’ claims to the fruits of their labor (including natural talents).28 If this is what fairness requires, then it would be unfair to adjust downward the amount 27 

Vallentyne (Tideman and Vallentyne 2001: 451–​2) holds that appropriation makes the appropriator morally liable for transferring only the competitive market value of the natural resource and no more to those with lower opportunity for welfare. Yet this limit seems somewhat ad hoc. Otsuka (2003: 32) argues that the natural resource appropriation regime should be structured to promote equality of opportunity for welfare subject to the caveat that the resulting natural resource property rights regime not excessively curtail the substantive liberty (i.e., what Otsuka calls the “robust self-​ownership”) of the more advantaged. However, it is unclear why the value of “robust self-​ownership” is appealed to here, whereas elsewhere Otsuka finds non-​robust (i.e., traditional) self-​ownership to be a sufficient conception of liberty. 28  See, for example,  Munzer 1990: 254–​91. See also Chs 7 and 8 of Miller 1999. For criticism, see Vallentyne 2012.

Libertarianism, Left and Right    141 of natural resources a person can appropriate simply because her labor has been (or can be anticipated to be) particularly productive (e.g., as a result of more economically valuable natural talents). A fuller discussion of these criticisms and Vallentyne and Otsuka’s potential responses is unfortunately beyond the scope of this chapter. To sum up, most contemporary libertarians agree that natural resources are initially unowned. But they disagree about the restrictions on natural resource appropriation. Radical-​right libertarians like Narveson and Rothbard hold that there are no restrictions on appropriation. Nozick and Steiner, on the other hand, endorse restrictions on appropriation that are based on the view that individuals have equal initial moral claims to natural resources. Vallentyne and Otsuka also endorse restrictions on appropriation. But they hold that those who are unfairly disadvantaged have greater moral claims to natural resources than do others. They endorse conditions on appropriation that aim to mitigate unfair inequalities. These different libertarian views on natural resource appropriation and their relationship to the traditional left/​right libertarian divide are summarized in Fig. 6.1. Left libertarians Equality-aimed appropriation restrictions Otsuka Vallentyne

Right libertarians Equal-claim-based appropriation restrictions Steiner

Nozick

No restrictions on appropriation

Narveson Rothbard

fig. 6.1  Libertarian views on natural resource appropriation

6.4  Property Rights in Artifacts and Distributive Justice The final major category of property rights that we have not yet discussed is property rights in artifacts (natural resources that have been improved through human action). If we accept a libertarian account of self-​ownership and natural resource property rights, the libertarian account of property rights in artifacts is not at all complex. These property rights simply arise as a result of voluntary exercises of pre-​existing property rights (self-​ownership rights, natural resource property rights, and property rights in pre-​ existing artifacts). Here is a simple example: Albert owns land with some apple trees on it. He makes Bella the following offer: If she would pick ninety of his apples, she can have thirty of the apples she picks. We can describe this offer more formally as follows: Albert conditionally transfers all of his property rights to thirty of his unpicked apples to Bella (while also authorizing Bella to encroach upon his land and apple trees) in exchange for Bella transferring certain limited, time-​sensitive self-​ownership control rights to Albert.

142    Joseph Mazor and Peter Vallentyne Once Bella has picked the ninety apples, she gains ownership of thirty of them while Albert now owns sixty picked apples. The picked apples are examples of artifacts. Having discussed self-​ownership, property rights in natural resources, and property rights in artifacts (i.e., the vast majority of property rights in society),29 surprisingly little remains to be explained about the libertarian view of distributive justice (at least as a matter of ideal theory).30 This is because libertarians generally do not appeal to particular end-​state principles to determine the proper distribution of economic benefits and burdens in society.31 The just distribution of benefits and burdens is simply that which arises from the voluntary exercise of property rights. Once we understand the libertarian account of these property rights (and particularly self-​ownership and property rights in natural resources), the libertarian account of distributive justice follows straightforwardly. However, while this account of distributive justice may be parsimonious, it is certainly not uncontroversial. We turn now to briefly consider several prominent objections to this account having to do with the role of luck, the moral limits of markets, discrimination, paternalism, unjust prices, the libertarian notion of freedom, the plight of the desperately poor, and the conflict between economic prosperity and property rights.

6.4.1 Brute Luck Some theorists criticize libertarianism by pointing out that many of the differences in the economic benefits and burdens enjoyed by individuals can be traced back to factors for which the individuals themselves are not morally responsible. They are thus unjust (because unfair). For example, no one can be said to be responsible for the natural talents that they are born with (they are the result of an unchosen genetic lottery and thus a matter of brute luck). Many theorists who make this argument call for the rejection of the full or near-​full self-ownership thesis and advocate redistribution of wealth to correct for unfair inequalities in economic benefits (Arneson 2006). As we have seen, some libertarians have been willing to accommodate this concern to some extent by structuring natural resource property rights to address these types of unfair inequalities (see Section s.3.4). However, they have not been willing to abandon the commitment to full or near-​full self-​ownership in response to this criticism.

29  We have already explicitly mentioned certain entities (e.g., animals) that are not necessarily included in this discussion. We will raise some other exceptions (e.g., minors) in section 6.5. 30  We will briefly return to the issue of rectificatory justice (what should be done when property rights have been violated in the past) in section 6.5. 31  The exceptions are those libertarians who endorse equality-​aimed restrictions on natural resource property rights. These libertarians still view distributive justice solely in terms of exercise of property rights. But the conditions on individuals’ natural resources property rights (and, by extension on the artifacts made with natural resources) can change depending on how the owner fares in terms of equality of opportunity for welfare relative to others.

Libertarianism, Left and Right    143

6.4.2 The Moral Limits of Markets Libertarians generally do not place any moral constraints on individuals’ transfer rights. This means that in theory at least, everything is available for sale. This includes organs, life preservers, and sex. A person can even sell themselves into slavery if they so choose. A variety of theorists have criticized this aspect of libertarianism. They have advocated limits to what can be bought and sold on the market based on protecting human dignity (Sandel 2012), preserving the social meaning of the goods in question (Anderson 1990), and respecting fellow citizens as moral equals (Satz 2010). However, libertarians have generally refused to condone restrictions on transfer rights on the basis of these values.

6.4.3 Discrimination Another troubling implication of the libertarian commitment to unrestricted transfer rights is the permissibility (as a matter of justice) of all forms of discrimination (Freeman 2001: 135–​6). Imagine that, although Carol is a better apple picker than Betsy, Albert refuses to hire her to pick his apples because she is black and Albert believes that blacks are “dirty.” Although libertarians can condemn Albert’s discriminatory action as immoral, they cannot condemn such action as unjust (Vallentyne 2006).

6.4.4 Paternalism Another problem with viewing voluntary choice as sacrosanct is this. People make mistakes, and sometimes their mistakes have very serious consequences for their well-​ being. Is it always impermissible to encroach on individuals or their property for their own good? If a person is about to cross a bridge that unbeknownst to him is rotten (and there is no time to inform him), is it really impermissible to physically prevent him from crossing the bridge? Some libertarians have entertained the possibility of allowing paternalism in cases in which the relevant interference does not violate a person’s will. However, most libertarians have refused to endorse any type of encroachment on individuals or their property, even for their own good.32

6.4.5 Just Prices Another important challenge to the libertarian account of distributive justice has to do with the distribution of the benefits from individual transactions (a critique 32 

For an excellent discussion, see Wall 2009.

144    Joseph Mazor and Peter Vallentyne that some have raised under the heading of just prices). A  wide variety of theorists (e.g.,Wertheimer 1996) have argued that the distribution of the benefits of voluntary exchange might in fact be unjust (e.g., because unfair or exploitative) in certain cases (e.g., when one of the parties holds a monopoly on the relevant good).33 Libertarians generally reject such criticisms. On the libertarian view, as long as a bargain was agreed to voluntarily, there can be no legitimate claim that the terms of the bargain are unjust.34

6.4.6 An Overly Narrow Notion of Freedom? Others have criticized libertarianism for endorsing an implausibly narrow notion of freedom (Olsaretti 2009b:  101–​61). Although libertarians sometimes cast themselves as defending freedom very broadly, understood as a “right to decide what would become of [oneself] and what [one] would do” (Nozick 1974: 171), libertarians are seemingly unconcerned (at least as a matter of justice) with individuals who are forced by their circumstances into certain courses of actions. A well-​known example is the worker who has no acceptable choice (e.g., they would starve otherwise) but to work for the capitalist for whatever wage the capitalist offers (Cohen 1995: 34–​7). Some libertarians have responded by pointing out that, while the worker may be forced to work, they are not coerced into working. Unjust restrictions on freedom arise, on the libertarian view, when one person causes another person to be forced to choose a certain path by violating (or threatening to violate) her rights. Since there is no rights-​ violating coercion in the case of the worker, it is perfectly coherent for the libertarian to resist seeing her plight as a matter of justice (Barnes 2012). Critics of libertarianism have conceded that coercion may raise special moral concerns. However, these critics continue to argue that the lack of acceptable alternatives by some individuals (e.g., the worker) can call into question the justice of the arrangements into which those individuals enter (Olsaretti 2013).

6.4.7 Desperate Poverty Another objection to the libertarian account of distributive justice is that it implausibly privileges individuals’ property rights in artifacts over the moral claims of the desperately poor. Imagine that someone (e.g., David) is suffering from serious malnutrition

33 

Libertarians sometimes suggest that monopoly prices will not be a problem as long as unjust legal barriers to entry in markets are removed (Nozick 1974: 17). When profits are very high, new firms will have incentive to enter the relevant market, and this will drive down prices. However, as economists (e.g., Stiglitz 2000: 78) have long recognized, there are a variety of non-​legal barriers to entry in many cases (e.g., very high fixed costs in certain industries). 34  See, for example, Nozick 1974: 63–​5. An important exception is Steiner 1984, who calls into question the justness of agreements whose terms were affected by a previous rights violation.

Libertarianism, Left and Right    145 even though he owns his rightful share of natural resource wealth.35 Why should David not be given property rights to, say, some of Bella’s apples? Some readers may view the moral claims of the desperately poor as obviously more weighty than the artifact property rights of the well off. Libertarians can offer two responses that should at least give such readers some pause. First (and most famously), libertarians argue that redistribution of human-​made wealth is on a par with forced labor (Nozick 1974: 169–​70). Redistributing ten apples from Bella to David is morally akin to forcing Bella to work for David for the time it took Bella to pick those apples. There have admittedly been many criticisms of the analogy between redistribution of wealth and forced labor (e.g., Fried 2004: 80). However, even if the analogy is not perfect, there is at least some sense in which redistribution is morally akin to forced labor. When rights to the apples that Bella picked are transferred from Bella to David without Bella’s consent, there is some sense in which Bella is used as a means for David’s ends. After all, Bella’s labor has benefited David rather than Bella, against Bella’s will. Insofar as there is a problem with some individuals being used as a means for others’ ends, there does seem to be at least a weighty argument against redistribution of human-​ made wealth. Libertarians can also point out that an account of distributive justice that always allows desperate poverty (or even blameless desperate poverty) to trump respect for property rights in human-​made wealth would have implications that few would be willing to accept. Such an account would, for example, require a massive redistribution of human-​made wealth from individuals in wealthy countries to poor individuals in the developing world.36 Some (e.g., Nagel 2005) have tried to resist this implication by appealing to the moral relevance of state borders. However, for those unconvinced that political boundaries have this type of normative force, the libertarian insistence on the moral importance of property rights provides a promising alternative avenue for resisting the radical demands for massive global redistribution.

6.4.8 Economic Prosperity vs Property Rights Libertarian society is often pictured as highly unequal but also wealthy overall due to the power of free enterprise, well-​defined property rights, and the lack of stifling government regulation. However, in reality, an unwavering commitment to property rights can (counterintuitively perhaps) generate very serious economic difficulties. Indeed, as economists have long recognized in their discussion of market failures, a system of voluntary exchanges can fail to realize significant economic benefits in a wide variety of circumstances (Stiglitz 2000: 77–​90). 35  For example, under Vallentyne’s theory, if natural resources are perfectly plentiful so that they lack a market value, David would not be due any transfers from Bella. 36  Of course, not everyone finds this implication to be unattractive. Some (e.g., Unger 1996) have even made the positive case for such redistribution.

146    Joseph Mazor and Peter Vallentyne It is difficult to overstate the economic problems that such market failures could potentially cause in a libertarian society. Who would build the roads? Who would have an incentive to undertake technological innovation in the absence of at least temporary protection from competition? How would money work? Libertarian society might not only have pockets of desperate poverty. It could also be fairly poor overall. Libertarians have several responses to this objection. Some (e.g., Mack 1986) have argued that critics underestimate the potential of private market solutions to so-​ called market failures. Others (Nozick 1974:  79) have responded to this problem by stepping back from their absolute commitment to property rights, often in contexts in which unaddressed market failures would have the most economically debilitating consequences.37 However, libertarians can also point out that other theories of distributive just­ ice either fail to address the conflict between respect for property rights and aggregate economic benefits or, in the case of utilitarianism, choose to sacrifice property rights for the sake of aggregate benefits in ways that are unattractive. For example, utilitarianism could easily condone policies that force a few people from their homes for the sake of a private redevelopment project that is only moderately beneficial for society as a whole. The unattractive implications of these alternative theories at least raise the possibility that, despite its problems, some version of libertarianism might be the most plausible account of distributive justice on offer, all things considered.

6.5  Directions for Future Research Before concluding, we would like to briefly highlight five topics for future research that have particularly important implications for libertarian accounts of distributive justice.38 These involve the moral status of minors, future people, and people in other nation-​states, property rights in non-​divisible natural resources, and rectificatory justice. In laying out these topics for future research, our focus will be on raising important questions and on providing the reader with references to further reading.

37  Nozick (1974: 79) does not insist that a polluter must obtain the explicit consent of every person whose rights might be encroached upon by her air pollution (since insisting an explicit consent would almost surely lead to a society in which no pollution at all could occur—​a society which would be very poor indeed). 38  Note that there are many other important areas for further research, including the moral status of animals and preventative justice (i.e., what a person may do to prevent potential rights violations), which we do not discuss here because they have relatively few implications for distributive justice.

Libertarianism, Left and Right    147

6.5.1 Minors Throughout this chapter, we have assumed that the only beings with moral standing are autonomous agents. This leaves out sufficiently young children as well as severely cognitively impaired adults (i.e., minors). The treatment of minors raises important questions for libertarianism. Can minors be owned by others? Do minors have any rights?39 Should the proviso on appropriation of natural resources take minors into account? If so, how? Who has a duty to ensure that the rights of minors are respected?40 Who, if anyone, has the right, and perhaps duty, to be the minors’ custodians (Vallentyne 2003b)? Since so much of what matters for distributive justice happens before we reach majority, the questions regarding the status of minors are clearly a pressing topic for future research.

6.5.2 Intergenerational Justice We have assumed throughout that there is only a single generation of people. What is the moral status of past people and their wishes? Do future people who will definitely exist have moral standing? Do future people who may exist, but also may not exist, have standing? If they have standing, do they have full self-​ownership (e.g., that can be violated by burying a toxin that will be released in 100 years)? Does the proviso give them a right to some share of the value of natural resources?41 The moral status of future people has important implications for a variety of issues in distributive justice such as obligations of environmental protection/​conservation.

6.5.3 International Justice We have assumed above a single nation-​state. However, there are important questions of international distributive justice that arise once we relax this assumption. Libertarians uniformly agree that national boundaries make no difference with respect to self-​ ownership. But do individuals in a particular nation-​state have special moral claims

39 

Those who endorse the choice-​protecting conception of rights deny that minors have any rights (since they lack any autonomous agency to protect). Those who endorse an interest-​protecting conception of rights (or a hybrid conception) can hold that minors have rights (since they have interests to protect). For a superb analysis of choice-​protecting vs interest-​protecting rights, see Kramer, Simmonds, and Steiner 1998. For discussion of whether children have a form of self-​ownership, see Vallentyne, 2003b; Andersson, 2007. 40  For discussion of who has the duty to provide children with their fair share of wealth, see Rakowski 1991: Ch. 7; Casal and Williams 1995; Vallentyne 2002; Olsaretti 2009a; Steiner and Vallentyne 2009. 41  For a discussion, see Fabre 2009; Steiner and Vallentyne 2009; Mazor 2010.

148    Joseph Mazor and Peter Vallentyne to the natural resources within their borders?42 Who has property rights to natural resources that span national borders?

6.5.4 Non-​Divisible Natural Resources Another important area for future research concerns property rights in natural resources that cannot be straightforwardly divided into individually appropriatable portions. Consider, for example, rights to the atmosphere. Although we could try to grant individuals the power to appropriate specific portions of the atmosphere, the molecules will move around and thus the portions will not be stable. How, then, should rights to the atmosphere be assigned on the libertarian view? 43

6.5.5 Rectificatory Justice The final pressing area for future research is rectificatory justice (the rights that individuals have as a result of previous rights violation). Given the centrality of historical transactions in legitimating the current distribution of property rights in libertarianism, questions of rectificatory justice have enormous importance for the libertarian account of distributive justice. Is there a right to punitive damages, or is rectification limited to victim restoration/​compensation? 44 Is the duty to compensate based on strict liability or is it sensitive to agent-​responsibility for the harm imposed (e.g., how foreseeable it was)? Is there a moral statute of limitations on the rights of rectification (after which a right ceases to be valid)? If so, what? If not, how should one proceed, given that we have very little knowledge of what rectification is required for the massive and systematic injustice of the distant past?45

6.6 Conclusion Our goal in this chapter has been to explore the libertarian account of distributive justice, including internal disagreements, external criticisms, and areas for future research. We began with a discussion of self-​ownership and considered criticisms relating

42 

For an argument that natural resources are commonly owned, see Risse 2012: Ch. 6. For general discussion, see Steiner 1999; Tideman and Vallentyne 2001. 43  For one Lockean view on rights to the atmosphere, see Bovens 2011. 44  See Barnett 1977, 1998 for a defense of the claim that there is a right only to victim restoration. 45  Nozick 1974: 231 indicates an openness to a one-​time Rawlsian redistribution of wealth to rectify massive past injustices. Narveson 2009, on the other hand, argues that, in the absence of specific compelling evidence, one should act as if the status quo is just.

Libertarianism, Left and Right    149 to indeterminacy, small incursions (or risks of incursions), and enforceable duties of assistance. We then turned to natural resource property rights and considered the debate over whether natural resources are initially owned and, if not, the conditions under which they may be appropriated. Finally, we discussed the libertarian account of just economic activity, surveyed a variety of criticisms of this account, and concluded by raising important topics for future research. As we recognized throughout this chapter, libertarianism is subject to many criticisms. Moreover, there are many gaps and unresolved issues. However, no theory is beyond reproach, and we believe that libertarianism has much to offer philosophers interested in questions relating to the just distribution of economic benefits and burdens.

References Anderson, E. (1990). “The Ethical Limitations of the Market.” Economics and Philosophy 6(2): 179–​205. Andersson, A.  ​K. (2007). Libertarianism and Potential Agents:  A Libertarian View of the Moral Rights of Foetuses and Children. Ph.D. dissertation, Stockholm University. Arneson, R. J. (2006). “Luck Egalitarianism Interpreted and Defended.” Philosophical Topics 32(1): 1–​20. Arneson, R. J. (2010). “Self-​Ownership and World Ownership: Against Left-​Libertarianism.” Social Philosophy and Policy 27(1): 168–​94. Barnes, G. (2012). “Why Is Coercion Unjust?: Olsaretti Vs. The Libertarian.” Analysis 72(3): 457–​65. Barnett, R. (1977). “Victim’s Rights: Restitution or Vengence.” Ethics 87: 279–​301. Barnett, R. (1998). The Structure of Liberty: Justice and the Rule of Law. Oxford: Clarendon Press. Bovens, L. (2011). “A Lockean Defence of Grandfathering Emission Rights,” in D.G. Arnold (ed.) The Ethics of Global Climate Change. Cambridge:  Cambridge University Press, pp. 122–​44. Casal, P. and Williams, A. (1995). “Rights, Equality and Procreation.” Analyse & Kritik 17(1): 93–​116. Cohen, G. A. (1995). Self-​Ownership, Freedom, and Equality. New York: Cambridge University Press. Dworkin, R. (1981). “What Is Equality? Part 2: Equality of Resources.” Philosophy and Public Affairs 10(4): 283–​345. Epstein, R. (1998). Principles for a Free Society. Reading, MA: Perseus Books. Fabre, C. (2009). “Preconception Rights,” in S. de-​Wijze, M. Kramer, and L. Carter (eds) Hillel Steiner and the Anatomy of Justice: Themes and Challenges. New York: Routledge, pp. 53–​69. Freeman, S. (2001). “Illiberal Libertarians:  Why Libertarianism is not a Liberal View.” Philosophy & Public Affairs 30(2): 105–​51. Fried, B. (2004). “Left-​ Libertarianism:  A Review Essay.” Philosophy & Public Affairs 32(1): 66–​92. Grunebaum, J. (2000). “Autonomous Ownership,” in P. Vallentyne and H. Steiner (eds) Left-​ Libertarianism and Its Critics: The Contemporary Debate New York: Palgrave, pp. 48–​73. Hohfeld, W. (1919). Fundamental Legal Conceptions. New Haven, CT: Yale University Press.

150    Joseph Mazor and Peter Vallentyne Kramer, M. H., Simmonds, N. E., and Steiner, H. (1998). A Debate over Rights: Philosophical Enquiries. Oxford: Clarendon Press. Kymlicka, W. (2001). Contemporary Political Philosophy: An Introduction. New York: Oxford University Press. Lippert-​ Rasmussen, K. (2008). “Against Self-​ Ownership:  There are no Fact-​ Insensitive Ownership Rights over One’s Body.” Philosophy & Public Affairs 36(1): 86–​118. Locke, J. and Macpherson, C. B. (1980). Second Treatise of Government. Indianapolis, IN: Hackett Publishing. Mack, E. (1986). “The Ethics of Taxation:  Rights Versus Public Goods,” in D. R. Lee (ed.) Taxation and the Deficit Economy. San Francisco, NY: Pacific Research Institute for Public Policy, pp. 487–​514. Mack, E. (1990). “Self-​Ownership and the Right of Property.” The Monist 73(4): 519–​43. Mack, E. (2011). “Nozickian Arguments for the More than Minimal State,” in R. M. Bader and J. Meadowcroft (eds) The Cambridge Companion to Nozick’s Anarchy, State, and Utopia. Cambridge: Cambridge University Press, pp. 89–​115. Mazor, J. (2009). A Liberal Theory of Natural Resource Property Rights. Ph.D. dissertation, Harvard University. Mazor, J. (2010). “Liberal Justice, Future People, and Natural Resource Conservation.” Philosophy & Public Affairs 38(4): 380–​408. Miller, D. (1999). Principles of Social Justice. Cambridge, MA: Harvard University Press. Munzer, S. R. (1990). A Theory of Property. Cambridge: Cambridge University Press. Nagel, T. (2005). “The Problem of Global Justice.” Philosophy & Public Affairs 33(2): 113–​47. Narveson, J. (1998). “Libertarianism vs. Marxism: Reflections on G. A. Cohen’s Self-​Ownership, Freedom, and Equality.” The Journal of Ethics 2(1): 1–​26. Narveson, J. (2009). “Present Payments, Past Wrongs: Correcting Loose Talk about Nozick and Rectification.” Libertarian Papers 1: 1–​17. Narveson, J. and Sterba, J. P. (2010). Are Liberty and Equality Compatible? New York: Cambridge University Press. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. Olsaretti, S. (2009a). “Choice, Circumstance, and the Cost of Children,” in S. de-​Wijze, M.  Kramer, and I. Carter (eds)  Hillel Steiner and the Anatomy of Justice:  Themes and Challenges. New York: Routledge, pp. 70–​84. Olsaretti, S. (2009b). Liberty, Desert and the Market. Cambridge: Cambridge University Press. Olsaretti, S. (2013). “Coercion and Libertarianism:  A Reply to Gordon Barnes.” Analysis 73(2): 295–​99. Otsuka, M. (2003). Libertarianism without Inequality. New York: Oxford University Press. Quong, J. (2011). “Left-​Libertarianism:  Rawlsian not Luck Egalitarian.” Journal of Political Philosophy 19(1): 64–​89. Rakowski, E. (1991). Equal Justice. Oxford: Oxford University Press. Risse, M. (2012). On Global Justice. Princeton, NJ: Princeton University Press. Roark, E. (2012). “Applying Locke’s Proviso to Unappropriated Natural Resources.” Political Studies 60(3): 687–​702. Rothbard, M. (2000). “Entrepreneurship, Entitlement, and Economic Justice,” in P. Vallentyne, and H. Steiner (eds) Left-​ Libertarianism and its Critics:  The Contemporary Debate. New York: Palgrave, pp. 214–​27. Sandel, M. (2012). What Money can’t Buy: The Moral Limits of Markets. New York: Penguin.

Libertarianism, Left and Right    151 Satz, D. (2010). Why Some Things Should not be for Sale:  The Moral Limits of Markets. New York: Oxford University Press. Sobel, D. (2012). “Backing Away from Libertarian Self-​Ownership.” Ethics 123(1): 32–​60. Steiner, H. (1984). “A Liberal Theory of Exploitation.” Ethics 94(2): 225–​41. Steiner, H. (1994). An Essay on Rights. Cambridge, MA: Blackwell. Steiner, H. (1999). “Just Taxation and International Redistribution,” in I. Shapiro and L. Brilmayer (eds) Global Justice. New York: New York University Press, pp. 171–​91. Steiner, H. and Vallentyne, P. (2009). “Libertarian Theories of Intergenerational Justice,” in A. Gosseries and L. Meyer (eds) Justice between Generations. Oxford: Oxford University Press, pp. 50–​76. Stiglitz, J. E. (2000). Economics of the Public Sector. New York: W. W. Norton. Thomson, J. J. (1992). The Realm of Rights. Cambridge, MA: Harvard University Press. Tideman, N. and Vallentyne, P. (2001). “Left-​Libertarianism and Global Justice,” in B. M. Leiser and T. D. Campbell (eds) Human Rights in Philosophy and Practice. Aldershot, UK: Ashgate, pp. 443–​57. Unger, P. (1996). Living High and Letting Die. New York: Oxford University Press. Vallentyne, P. (2002). “Brute Luck, Option Luck, and Equality of Initial Opportunities.” Ethics 112(3): 529–​57. Vallentyne, P. (2003a). “Justice in General,” in P. Vallentyne (ed.) Equality and Justice. New York: Routlege, pp. xi–​xviii. Vallentyne, P. (2003b). “Rights and Duties of Childbearing.” William and Mary Bill of Rights Journal 11: 991–​1010. Vallentyne, P. (2006). “Left Libertarianism and Private Discrimination.” San Diego Law Review 43: 981–​92. Vallentyne, P. (2007). “On Original Appropriation,” in M. Murray (ed.) Liberty, Games and Contracts: Jan Narveson and the Defence of Libertarianism. Aldershot, UK: Ashgate Press, pp. 173–​8. Vallentyne, P. (2011). “Equal Negative Liberty and Welfare Rights.” International Journal of Applied Philosophy 25: 237–​41. Vallentyne, P. (2012). “Taxation, Redistribution, and Property Rights,” in A. Marmor (ed.) Routledge Companion to Philosophy of Law. New York: Routledge, pp. 291–​301. Vallentyne, P., Steiner, H., and Otsuka, M. (2005). “Why Left-​Libertarianism is not Incoherent, Indeterminate, or Irrelevant: A Reply to Fried.” Philosophy & Public Affairs 33: 201–​15. Wall, S. (2009). “Self-​ Ownership and Paternalism.” Journal of Political Philosophy 17(4): 399–​417. Wertheimer, A. (1996). Exploitation. Princeton, NJ: Princeton University Press.

Chapter 7

Desert-​B ased J u st i c e Jeffrey Moriarty

Justice requires giving people what they deserve. Or so many philosophers—​and according to many of those philosophers, everyone else—​ thought for centuries. Aristotle says that “all agree that the just in distributions must accord with some sort of worth [axia]” (1999/​350 bce, 1131a1125–​1127). And Mill says that “it is universally considered just that each person should obtain that (whether good or evil) which he deserves” (1998/​1863: 98; see also Ross 2002/​1930 and Sidgwick 1981/​1907). In the 1970s and 1980s, however, perhaps under the influence of Rawls’s (1971) desert-​less theory, desert was largely cast out of discussions of distributive justice. Now it is making a comeback. In this chapter I consider recent research on the concept of desert, arguments for and against its requital, and connections between desert and other distributive ideals. I suggest that desert-​sensitive theories of distributive justice, though they face obstacles, have a promising future.

7.1  What is Desert? Desert: a three-​place property uniting a subject, a thing, or treatment, and a fact about the subject (Feinberg 1999/​1963; Olsaretti 2004). To say that person P deserves thing T is to say that P has the property of being deserving of T. P has this property in virtue of some fact F about him, for example that he has performed a certain action. F is called the “desert-​base.”1 Desert-​claims are normatively significant. If P deserves T, then necessarily, there is a reason that P should have T (Feinberg 1999/​1963). It does not follow, however, that P should have T all things considered. Desert provides a reason for giving T to P, but that reason can be overridden by other, more powerful reasons. As this implies, not all cases

1 

This section draws on Moriarty (2009).

Desert-Based Justice   153 in which there is a reason for someone to have something is a case in which they deserve it. It is not easy to say, however, what makes desert special. Most, if not all, normative concepts admit of some degree of vagueness and indeterminacy. But philosophers’ intuitions about desert seem especially variable. To clarify desert’s nature, writers have sought to distinguish it from the related concept of entitlement. Perhaps the most important difference is that entitlement can be a wholly “institutional” artifact, while desert must always be in some sense “natural.” Desert is natural both in the sense of being pre-​institutional and, to use Scheffler’s (2000) phrase, pre-​justicial. A person can become entitled to a thing or treatment by satisfying rules or criteria for its distribution. By contrast, to deserve something, as Feinberg says, “one must satisfy certain conditions of worthiness which are written down in no legal or official regulation” (Feinberg 1999/​1963: 71)—​and, we might add, in no theory of justice. Desert is thus a critical notion: it can tell us what the rules of institutions and theories of justice should be (Miller 1999). Any linguistically competent person can produce a range of apparently valid desert-​ claims. Heroes deserve reward. Criminals deserve punishment. Hard workers deserve success. But why are these responses the deserved responses? Desert is often understood in terms of “fittingness” or “appropriateness” (Cupit 1996). So we might say, for example, that criminals deserve punishment, and not reward, because punishment is a fitting or appropriate response to crime. But this raises a new question. Why is punishment fitting or appropriate for crime, and not for heroism? Several writers link deserved treatments with people’s natural responses (Miller 1999; Olsaretti 2004; Scheffler 2003). Thus Feinberg says that people “naturally entertain certain responsive attitudes toward various actions, qualities, and achievements” (1999/​1963: 81). These “responsive attitudes”—​attitudes like admiration, resentment, and gratitude—​he says, are “the basic things persons deserve,” while the concrete things or treatments people deserve are “the natural or conventional means of expressing the morally fitting attitudes” (1999/​ 1963: 82). There is considerable debate about the conditions for desert:  about what subjects can be deserving, what things and treatments can be deserved, and what facts can serve as desert-​bases. However, two conditions for desert are widely, but not universally, accepted. I call the first the “responsibility condition” and the second the “value condition.” The responsibility condition is often expressed as a requirement that if P deserves T in virtue of F, then P is responsible for F. Thus Rachels says: “the concept of desert serves to signify ways of treating people that are appropriate responses to them, given that they are responsible for [certain] actions or states of affairs” (1997: 180; see also Miller 1999). But other, more elaborate, versions of this condition have been proposed (Smilansky 1996). The type of responsibility at issue here is moral responsibility. While there are many different accounts of this concept, most claim that, for P to be morally responsible for F is for P to a be a proper subject for moral appraisal, as might be expressed by praise or blame, for having done F (Levy 2005). (Hereafter, when I use ‘responsibility’, I mean moral responsibility.) In addition to constraining the range of facts that can be

154   Jeffrey Moriarty desert-​bases, the responsibility condition constrains the range of subjects that can be deserving. It implies that only subjects that can be responsible can be deserving. The value condition has two parts. First, if P deserves T in virtue of F, then F has value, or is appropriately the subject of an appraising or evaluative attitude. As Schmidtz says, “to judge a person deserving is to respond to features of the person that we judge to be of value” (2002: 775). The feature could be positively valued, as when we say that a person who has performed a heroic deed deserves a reward. Or it could be negatively valued, as when we say that a person who has committed a crime deserves punishment. Second, the things and treatments subjects are said to deserve have value. As Feinberg says, they are “ ‘affective’ in character . . . favored or disfavored, pursued or avoided, pleasant or unpleasant” (Feinberg 1999/​1963: 61). In desert-​claims, these values line up. Those who do good deserve good, while those who do bad deserve bad. Not all writers endorse the responsibility and value conditions. The standard counterexample to the former involves desert of compensation (Feldman 1995; see also Vilhauer 2009). People are said to deserve compensation for bad things that happen to them, even if—​or rather, precisely because—​they are not responsible for them. And while there is less disagreement about the value condition, some writers hold a view of desert that is incompatible with it. They think need is a desert-​base, so that, for example, people deserve food simply in virtue of needing it (McLeod 1996). Defenders of the responsibility and value conditions have, in turn, tried to defuse these apparent counterexamples (for a defense of the responsibility condition, see Olsaretti 2004; for a defense of the value condition, see Miller 1999). I will not try to settle these disputes here. However, I will restrict my attention in this chapter to desert-​claims that satisfy the responsibility and value conditions. This is because it is this kind of desert that has been making a comeback in recent years in discussions of distributive justice (Arneson 2007; Miller 1999). And, as we will see, it is this kind of desert that figures prominently in arguments for desert’s requital. It is to these arguments that we now turn.

7.2  Why Should People Get What They Deserve? As mentioned, to say that P deserves T is to say that there is a reason for P to have T. But this is a pro tanto and not a conclusive reason. It may be outweighed by other, competing reasons—​reasons for P not to have T, or for someone else to have T. For example, we might claim that teachers deserve higher salaries, but their having higher salaries might require infringing taxpayers’ rights to private property, or failing to satisfy poor children’s medical needs. To decide what to do in cases of conflict, minimally, we need to say something about why people should have what they deserve. A variety of accounts of desert’s normative significance have been offered. Below I review and make some critical observations about the most prominent ones.

Desert-Based Justice   155

7.2.1 Desert as Basic According to some writers, the reason that people should have what they deserve is simply that they deserve it. As Miller puts it, the reason is “basic”: “[t]‌he performance has taken place, and [P’s] being put in a position to enjoy [T] is the fitting or appropriate response on our part to that fact” (Miller 1999: 136). If we understand desert as a relationship of fittingness between a subject and an object, however, this is tantamount to saying that people should have what they deserve because they deserve it, that is, because the subject, in virtue of performing some action, now deserves the object. Miller confirms this interpretation a few sentences later when he says that “desert judgments” imply that “[P’s] doing [F] at one moment simply is a reason for his being given [T] at some later moment” (1999: 136, emphasis in original). Schmidtz (2006) also offers a basic argument for giving people what they deserve. He considers a person named Bob who deserves a pay raise, and says: (1) “To be sure, it would be a misuse of terms to say that Bob deserves a pay raise on the grounds that giving him a raise would have utility” (2006: 58–​9). He then adds: (2) “if we ask why we should acknowledge that Bob is a great worker, a big part of what makes Bob’s efforts worthy of recognition is that his efforts are of a kind that make us all better off ” (p. 59, emphasis in original). It might look like Schmidtz is offering an instrumental argument for giving people what they deserve (for more on this type of argument, see section 7.2.3.). That is, it might look like he is saying: (1) while Bob doesn’t deserve a pay raise because his getting it will have utility; (2) the reason that Bob should get the pay raise he deserves is because his getting it will have utility (e.g., by incentivizing others to perform similar acts). But while the foregoing sentence accurately characterizes Schmidtz’s (1), it does not accurately characterize his (2). Schmidtz appeals to “utility” in (2), not as a forward-​looking justification for giving Bob what he deserves (giving T to P will increase utility), but as the past-​looking reason for, or basis of, Bob’s desert (P deserves T because he increased utility). And why should Bob get the pay raise he deserves? According to Schmidtz, because “Bob’s efforts  . . .  are of a kind that make us all better off,” that is, simply because Bob has done the thing that makes him deserving.2

2 

Utility plays a similar role in Miller’s “basic” argument. Miller says that the performances in virtue of which people are deserving are those that are “positively appraised or valued by the surrounding community” (1999: 135). “No doubt,” he says, “in the background there often stands some idea of social utility” (1999: 135). We appraise positively what we think “enriches our lives in one way or another” (1999: 135). Here also, a person’s past generation of social utility is appealed to both as a desert-​maker and as the reason to give people what they deserve. For Miller, people should be given what they deserve because they deserve it, and one way they can come to be deserving is by performing actions that have social utility.

156   Jeffrey Moriarty

7.2.2 Desert and Non-​Instrumental Value Another argument for giving people what they deserve is just a bit less basic than the previous one. It claims that the reason that people should have what they deserve is that it is good, other things equal, that they have it. (The basic argument does not appeal to goodness in this way.) Ross gives the most famous version of this argument. He says: If we compare two imaginary states of the universe, alike in the total amounts of virtue and vice and of pleasure and pain present in the two, but in one of which the virtuous were all happy and the vicious miserable, while in the other the virtuous were miserable and the vicious happy, very few people would hesitate to say that the first was a much better state of the universe than the second. (Ross, 2002/​1930: 138)

What explains our belief that the first world is better than the second cannot be the amounts of pleasure, pain, virtue, and vice in each world, because these are identical. The only difference, according to Ross, is the distributions of pleasure and pain in each. Thus, it must matter who is happy and who is miserable. Since we think that the first world is better than the second, we think that the state of affairs in which the virtuous are happy and the vicious are miserable is better than the state of affairs in which the virtuous are miserable and the vicious are happy. It follows that “the apportionment of pleasure and pain to the virtuous and the vicious respectively”—​that is, the requital of desert—​is an “independent” or “intrinsic” good (2002/​1930: 138). Even if we grant that Ross’s argument succeeds on its own terms, it may seem incapable of justifying the wide range of desert-​claims that we typically make. In one way, Ross’s argument may seem too narrow. It draws conclusions about the apportionment of happiness and misery to the virtuous and vicious, respectively. But people are said to deserve other kinds of treatments (e.g., rewards, punishments) for performing other types of actions (e.g., heroic deeds, crimes). In another way, Ross’s argument may seem not narrow enough. It is naturally suited to justifying claims of “global desert,” that is, claims that it is good, other things equal, for those who have led virtuous lives as a whole to be happy overall, and for those who have led vicious lives as a whole to be unhappy overall (Hurka 2003; Tadros 2011). But many desert-​claims are “local”; for example, Jones deserves ten years in prison for armed robbery, or Smith deserves a bonus for making a big sale. This impression is mistaken. With respect to the “too narrow” problem, we could insert more general value concepts in the place of “happiness,” “misery,” “virtue,” and “vice” in Ross’s thought experiment, and still reach the same conclusion. That is, if we compare the world in which good and bad treatment of an unspecified nature is proportionate, respectively, to good and bad behavior of an unspecified nature, with the world in which unspecified good and bad treatment is proportionate, respectively, to unspecified bad and good behavior, we still think the former world is better than the latter world. In fact, I suggest Ross’s experiment supports primarily the more general conclusion that

Desert-Based Justice   157 treatments of a certain value should be proportionate to behavior of the same value. He simply makes a further claim about the sources of value in people’s behavior (namely, virtue and vice) and the treatments they receive (namely, happiness and misery). Ross’s thought experiment also seems capable, understood in the right way, of surmounting the “not narrow enough” problem. When we make local desert-​claims, we abstract away from the other good and bad behavior and treatments—​that is, the other potential desert-​bases and deserved treatments—​in a person’s life, and focus exclusively on the putative desert-​base and deserved treatment implicated in that claim. As far as our local desert-​judgment is concerned, that action and treatment pair is the person’s whole life. We say that it is good, for example, that a person receives punishment in virtue of committing a particular crime, ignoring the other actions they have performed and the other treatments they have received. And insofar as these actions and treatments are ignored, our judgment is justified. Once we take off our blinders and take these factors into account, however, the desert-​based goodness of the state of affairs in which a criminal receives punishment may be diminished, if it would make him, as Ross says, more “sinned against than sinning” (2002/​1930: 59). But this does not undermine the initial desert-​claim we make, given its implicit limitations.3 Both the “basic” and the “non-​instrumental” arguments capture the idea that there is something fundamental about desert’s normative significance. But, for this reason, both may fail to convince the widest range of writers. Some do not feel the force of the intuitions on which these arguments rely (Scanlon 1998); even their adherents acknowledge that they will seem “mysterious” to some (Miller 1999). More importantly, neither argument is particularly well suited to proving that theories of justice should be desert-​ sensitive. Justice does not require the realization of every normatively significant state of affairs. Perhaps for this reason, some writers have sought to link the requital of desert with the promotion of other values, at least some of which might be regarded as clearly relevant to theories of distributive justice. Linking desert to justice in turn strengthens the case for giving people what they deserve.

7.2.3 Desert and Instrumental Value Rachels says that a social practice of requiting desert promotes a variety of values that are relevant to theories of distributive justice. First, it “permits people . . . to control their own fates by allowing them to earn good treatment at the hands of others” (1997: 190; see also Schmidtz 2006). Second, it makes “the distribution of burdens and benefits more nearly equal” (Rachels 1997: 190), since people who deserve good treatment typically do so because they have taken on some extra burden (e.g., hard work). Finally, requiting desert 3  There is another way that Ross’s thought experiment may seem “not narrow enough,” namely, it does not help us to identify the specific treatments people deserve for performing specific actions. For example, it does not tell us how many years in prison people deserve for armed robbery. But there may be no precise answer to that type of question (Miller 1999).

158   Jeffrey Moriarty encourages moral behavior and discourages immoral behavior, by attaching rewards and penalties, respectively, to them. It is unclear, however, whether instrumental arguments can justify all of the desert-​claims that, intuitively, we might like to make. For example, while rewarding hard workers (with, e.g., money) will usually promote hard work, it may not always. Suppose P has worked hard but is widely perceived to have slacked off. Rewarding him may incentivize slacking off. Yet it seems right to say that, in this case, P deserves to be rewarded, and that this is a reason to reward him. Defenders of instrumental arguments might respond by saying that their arguments should be given a “rule” rather than an “act” interpretation. That is, they might claim not that each and every act of rewarding hard work makes society better off (with respect to the promotion of a particular value), but that society is better off if it accepts a rule that requires rewarding hard work than if it does not. Society’s acceptance of this rule then provides the reason to give hard worker P in our example a reward (Hooker 2000). But, in making this move, instrumentalists about desert will face the familiar problems of rule-​consequentialism. Society might do better if it accepts the rule “always reward hard work” than if it accepts the rule “Never reward hard work,” but it might do even better if it accepts the rule “Reward hard work except when the worker is widely perceived to have slacked off.” This would imply, however, that there is no desert-​based reason to reward P in our example. How instrumentalists respond to this depends on whether they think that all or only some desert-​claims can be justified on instrumental grounds. Those who think that all desert-​claims can be justified on instrumental grounds would have to concede that, since rewarding P does not make society better off (or is not required by a rule the acceptance of which would make society better off), it is not the case that P deserves to be rewarded. But this seems extreme. It is intuitively plausible to suppose that people who work hard deserve to be rewarded, and that this is a reason to reward them, even if rewarding them would not promote any other value. It makes more sense, I think, to concede that the reach of instrumentalist arguments is limited—​that they are capable of justifying some but not all desert-​claims. Since all desert-​claims have normative significance, this in turn suggests that instrumental considerations do not provide the only justification of desert-​claims, but rather sometimes provide additional reasons to give people what they deserve.

7.2.4 Desert, Responsibility, and Respect A number of writers have linked giving people what they deserve with respecting them, and in particular, with respecting them as responsible or autonomous agents. This may appear to be a promising avenue for establishing that theories of distributive justice should be desert-​sensitive, since it is plausible to assume that people are treated unjustly to the extent that they are treated disrespectfully. But, as I will show, this promise has not been fulfilled. The connection between desert, respect, and responsibility provides at

Desert-Based Justice   159 best a reason for theories of justice to be desert-​sensitive, not a requirement that they be desert-​sensitive. Writers who link desert, respect, and responsibility are not always clear about what the link is. A passage from Rachels is representative: Treating people as they deserve is one way of treating them as autonomous beings, responsible for their own conduct. A person who is punished for his misdeeds is held responsible for them in a concrete way. He is not treated as a mindless automaton, whose defective performances must be ‘corrected’, or whose good performance promoted, but as a responsible agent whose actions merit approval or resentment. The recognition of deserts is bound up with this way of regarding people. (1991: 144, emphasis in original)

Rachels begins by saying that treating people as they deserve is “one way of ”—​or sufficient for—​treating them as responsible beings. Similarly, Moore says that “[w]‌e respect someone’s autonomy when we punish him only because he deserves it, not because his punishment is useful to others’ welfare and not because his punishment is a therapy to cure his ‘sickness’ ” (1997: 150). But Rachels concludes by suggesting that treating people as they deserve is “bound up with”—​or necessary for—​treating them as responsible beings. Similarly, Lucas says, “[i]f we deny people their deserts . . . [t]hey are not in our eyes autonomous agents who had it in their power to act or not to act, but merely natural phenomena which we have been manipulating at our will” (1980: 202). There are two arguments here. Both start with: (1) We should treat people as responsible beings, out of respect for them. Then we have: (2) Treating people as they deserve is either (a) a way of treating them as responsible beings or (b) necessary for treating them as responsible beings. The arguments are (1) + (2a) and (1) + (2b). Do either, or both, succeed?4 To begin, we must assume that people are responsible beings. There would be no problem with treating people as non-​responsible beings if they were non-​responsible beings. There is no problem, for example, with treating houseplants as non-​responsible beings. Given this assumption, it does seem right to say that people should be treated as responsible beings, out of respect for them (Lucas 1980; Moore 1997). That is, (1) seems correct. It is disrespectful to treat normal human adults as if they were small children, or even worse, as “mindless automata” or “merely natural phenomena.” In normal circumstances, we should see their acts as robustly their own, and regard them as proper subjects of moral appraisal in light of them. In his famous discussion of punishment,

4  Schmidtz says the point of requiting desert is in part “to show respect for customs and institutions and characters that make people better off ” (2006: 59). But he does not say why treating people as they deserve is a way of respecting them. As a result, this statement has more in common with the “basic” arguments we considered in section 7.2.1 than with the arguments considered in this section.

160   Jeffrey Moriarty Morris says when we fail to see people this way, “we display a lack of respect for the moral status of individuals” (1968: 487). (2a) also seems right. That is, it seems right to suppose that treating people as they deserve is a way of treating them as responsible beings, given how we have understood the concept of desert. Desert, we have assumed, requires responsibility. When we treat a person as they deserve, we suppose that they are deserving. And when we do this, we suppose that they are responsible for what makes them deserving. Thus the argument (1) + (2a) succeeds. This argument should be taken seriously. It provides a reason to treat people as they deserve that is grounded in respect for persons, a powerful moral ideal. But this argument also has potentially serious limitations. If treating people as they deserve is not the only way to treat people as responsible beings, and if there are other, better ways to do so, then the respect-​based case for treating people as they deserve loses much of its force. (2b) attempts to block this possibility. It says that treating people as they deserve is required if we are to treat them as responsible beings. The writers who make this claim—​ Lucas and possibly Rachels—​do not put forward an argument for this position. Perhaps they think that it is intuitively obvious. But upon inspection, (2b) seems to be false. Given the truth of (2a), which says that that treating people as they deserve is a way of treating them as responsible beings, (2b) can be established by showing that there is no other way to treat people as responsible beings. But it seems that there is. Suppose that P commits a crime. One thing we might do, as Kelly (2002) says, is judge that P is responsible for her crime, and then communicate that judgment to her. This is a way of treating P as a responsible being. After all, when our houseplants drop leaves onto our kitchen counters, we do not judge them responsible for the mess and communicate that judgment to them. But we might go no further than this. That is, we might not go on to give P the treatments that criminals are commonly said to deserve, namely, we might not resent P for what she has done or inflict punishment on her. If this is right, then treating people as responsible beings does not require treating them as they deserve, at least in the sense of giving them the things and treatments that they are commonly said to deserve. It might be wondered whether (2b) can be defended by appealing to a Strawsonian conception of responsibility (Scheffler 2003). On Strawson’s view, treating people as responsible beings requires being susceptible to the reactive attitudes in our dealings with them. And these attitudes, according to Scheffler, “always present themselves as merited or deserved responses to the individuals who are their targets” (2003: 71; see also Miller 1999). This might seem to show that if we—​at least in normal circumstances—​do not treat people as they deserve by having and expressing our reactive attitudes, then we are not treating them as responsible beings. This is false, for reasons alluded to above. First, one can be susceptible to—​in the sense of being disposed to have—​the reactive attitudes towards a person without actually having them. To be sure, if a person is susceptible to the reactive attitudes, then she will be inclined to have them, and other things equal, she will have them. But this inclination can be overcome. If Q perceives that his resenting P for her wrongdoing will

Desert-Based Justice   161 prevent him from getting on with his life, he may resist his inclination to resent her, and possibly succeed. Second, it does not follow from the fact that one has a reactive attitude that one should express it, either verbally or in the form of a concrete treatment. Q may resent P for her wrongdoing, but this does not mean Q should express his resentment, to P or anyone else, let alone that Q should express it in the form of a concrete treatment. Of course, there often are good reasons to express our reactive attitudes. Expressing our resentment to wrongdoers in the form of a punishment can motivate them and others not to perform wrongful acts in the future (Feinberg 1999/​1963). But this is an instrumental reason to give people what they deserve. (2b) says that treating people as responsible beings requires treating them as they deserve. I have argued that this is false. It follows that the argument (1) + (2b) fails. Respecting persons does not require treating them as they deserve.

7.3  The Asymmetry of Desert It is interesting to observe that, while desert was for a time absent from many theories of distributive justice, it never disappeared from theories of retributive justice (Moriarty 2003). And even those who now think desert should play little or no role in distributive justice do not think that desert should be excised from retributive justice. Rawls (1971) famously endorsed this “asymmetric” treatment of desert, but in recent years, Scheffler (2000, 2003) has been its most prominent defender. This asymmetry is typically thought to require justification, for the simple reason that people can deserve the distribuenda of both distributive and retributive justice—​social benefits and burdens in the first case, and punishment in the second. In a (2000) article, Scheffler argues that desert is individualistic, which makes it an appropriate criterion in theories of retributive justice, which are also individualistic, but not in theories of distributive justice, which are holistic. This argument has been the subject of extensive criticism (Hurka 2003; Husak 2000), and in his (2003), Scheffler acknowledges its weaknesses.5 But in that article, he advances a new argument which can be understood as a partial defense of the asymmetry. He claims that a desert-​free

5  For Scheffler, distributive justice is holistic in the sense that, before we can determine whether a given person has her “just share” of social benefits and burdens, we need to compare her claims to others’ claims and to the total supply. This is not the case, Scheffler says, in retributive justice. Because the supply of the distribuenda (namely, punishment) is effectively unlimited, we can determine whether a person has their “just share” by looking at their claims alone. One problem with this argument is, as Husak (2000) says, that the supply of punishment is not actually unlimited. Punishment costs money—​ money which could otherwise be spent on social benefits. If social benefits are limited, so is punishment. Second, as Hurka (2003) says, Scheffler seems to assume that all desert is “non-​comparative,” that is, people’s deserts are determined solely by what they have done. But desert is often “comparative,” that is, people’s deserts are determined by what they have done compared to others, and so desert is especially suitable for use in “holistic” distributive theories.

162   Jeffrey Moriarty retributive theory would be “intolerably revisionist” (2003: 69), but a desert-​free distributive theory would be tolerable. This argument has received no critical attention. In this section I examine it and show that it fails. Scheffler says that judgments of retributive desert “are related in a particularly intimate way” to our practices of responsibility (2003: 75). (1) First, our reactive responses to criminals are due to their having violated a normative standard, namely one embodied in the relevant criminal law. (2) Second, the punishment the criminal is said to deserve is the “public, institutional vehicle for the expression of just this type of reactive response” (2003: 75). Neither point, Scheffler says, is true in the case of distributive desert, and in particular, the desert of economic benefits. With respect to the first point: while judgments of economic desert may be legitimate, and may express reactive attitudes, they do not arise, Scheffler says, from a “perception that the individual has committed a breach of expectations” (2003: 75). They likely arise, instead, from a belief that he has received “unfair advantages” (2003: 75). With respect to the second point: the income a person is said to deserve “does not itself constitute an institutional vehicle whose function it is to give public expression to [a reactive attitude]” (2003: 75). Since our practices of responsibility are closely linked with retributive desert but not with distributive desert, Scheffler concludes, a desert-​free retributive theory is intolerably revisionist, while a desert-​free distributive theory is not. As mentioned, this is not, and is not intended by Scheffler to be, a complete defense of the asymmetry. Even if eliminating desert from theories of distributive justice is “tolerable,” it may still be wrong. Still, it is an attempt to show that desert matters less in distributive justice than in retributive justice. I will argue that Scheffler fails to establish even this much. Consider first Scheffler’s point (2): his claim that punishment is a “public, institutional” vehicle that gives expression to the attitude that we think the offender deserves, while income is not a public, institutional vehicle that gives expression to the attitude that we think the economic agent deserves. This is true. But this fact by itself cannot be used to justify the asymmetry. It is, in effect, a statement of it. Punishment and income could play different roles in a just society. In particular, income could be distributed by a central authority, much as punishment now is, to give expression to attitudes that we think economic agents deserve. The fact that punishment and income in fact play certain roles in society is not evidence that they should play them. If Scheffler’s argument succeeds, it is because of the other difference between distributive and retributive desert he identifies. This is his claim that (1) judgments of retributive desert express reactive responses to breaches of normative expectations, while judgments of distributive desert do not. Scheffler further believes that (1a) only reactive responses to breaches of normative expectations are “importantly constitutive of treating people as responsible agents” (2003: 75). The argument must be that, given the importance of treating people as responsible agents, and given (1) and (1a), it is more important to express (through responses and treatments that can be understood as “giving people what they deserve”) judgments of retributive desert than judgments of distributive desert. I will show that both (1) and (1a) are problematic. I begin with (1).

Desert-Based Justice   163 It is plausible to regard the criminal law as codifying normative expectations society has for its members. When a person violates such a law, it is natural to say that he deserves a negative reactive response such as resentment, and that punishment expresses that response. In this sense, Scheffler is right to say that judgments of retributive desert express reactive responses to breaches of normative expectations. In support of his claim that judgments of distributive desert do not express reactive responses to breaches of normative expectations, Scheffler considers a person who is paid a high salary, but who we judge deserves a lower one. Our desert-​judgment in this case, he says, does not reflect a belief that the high-​earner has breached a normative expectation. For we do not expect people to make the (e.g.) large economic contributions that would make them deserving of high salaries.6 It is permissible for people to make more modest contributions. Scheffler is right that this kind of desert-​judgment does not express a response to a breach of normative expectations. But this is not surprising. It is a judgment about the amount of economic benefits a person should receive. In the same way, a judgment J about the amount of punishment an offender should receive does not express a response to a breach of normative expectations. It may presuppose a distinct judgment J1 that the offender has breached an expectation, but J itself embodies an assessment of the relative seriousness of her offense. The question becomes:  are there any normative expectations relevant to theories of distributive justice to which desert-​judgments are, or might be, keyed? Scheffler appears to think not, but he offers no argument for this conclusion. And it is unjustified. A number of writers have offered theories of distributive justice which incorporate normative expectations, and in particular, the expectation that all members of society must make a contribution to it. White argues that “each citizen has an obligation to make a decent productive contribution . . . to the community” (2003: 91). And Rawls says that his justice as fairness “expresses the idea that all citizens are to do their part in society’s cooperative work” (2001: 179). While neither of these theorists link desert-​judgments to this expectation, it is natural to do so. It is natural to say, that is, that people who culpably fail to live up to this expectation deserve no economic benefits. This intuition likely lies behind the popular view that the receipt of welfare benefits should be tied to work. To be clear, I am not arguing that theories of distributive justice should give weight to this or any other desert-​claim tied to the violation of a normative expectation. Whether such theories appropriately recognize normative expectations and desert-​claims tied to them is a difficult matter that cannot be resolved here. My point is simply that, in dismissing the possibility of such expectations and claims in the case of distributive justice—​while granting their reality in the case of retributive justice—​Scheffler again assumes to be true what he must prove to be true.

6  This assumes that contribution is the basis of economic desert. But nothing hangs on this. We reach a similar conclusion if the basis of economic desert is something else, such as effort. For a discussion of this issue, see McLeod (1996).

164   Jeffrey Moriarty It might be objected that, even if what I  have said so far is right, it barely dents Scheffler’s defense of the asymmetry. Even if economic benefits are denied, for desert-​ based reasons, to people who do not contribute to society, it does not follow that people’s shares of economic benefits should be proportionate to their deserts. That is true, but it is not what is at issue. Scheffler does not argue that criminals’ punishments should be proportionate to their crimes. He argues that claims that people deserve punishment—​as opposed to not deserving it—​have a special significance, because they are tied to whether people breach—​as opposed to don’t breach—​certain normative expectations. Similarly, I have suggested that claims that people deserve economic benefits—​as opposed to not deserving them—​may have a special significance, because they may be tied to whether people breach—​as opposed to not breach—​certain normative expectations. In any event, there is a deeper problem with Scheffler’s argument. Remember that his goal is to show that treating people as responsible beings requires requiting their retributive, but not their distributive, deserts. He does this by combining the claim that (1) judgments of retributive desert express reactive responses to breaches of normative expectations, while judgments of distributive desert do not, with the claim that (1a) only reactive responses to breaches of normative expectations are “importantly constitutive of treating people as responsible agents” (2003: 75). Having argued that (1) is unjustified, I will now give reason for thinking that (1a) is false. Consider two claims: “P deserves blame for robbing a bank,” and “Q deserves praise for discovering a new vaccine.” Scheffler’s (1a) implies that seeing P as a legitimate target of blame is “importantly constitutive” of treating him as a responsible agent, while seeing Q as a legitimate target of praise is not. (Reward and punishment are assumed by Scheffler to be the concrete expressions of praise and blame, respectively.) This is because P breaches a normative standard by robbing a bank, while Q does not by discovering a new vaccine. If anything, Q goes beyond what morality requires of him. Intuitively, however, I detect no difference between them. It seems plausible to suppose that, if we fail to see P as a legitimate target of blame, then we are failing to see him as a responsible agent. We do not see creatures that we regard as incapable of responsibility (e.g., young children, animals, plants) as legitimate targets of reactive attitudes. But it seems just as plausible to suppose that, if we fail to see Q as a legitimate target of praise, then we are failing to see him as a responsible agent. Put another way, it is implausible to suppose that seeing people as responsible beings requires only that we see them as deserving of blame for violating normative expectations, and not as deserving of other, positive reactions for performing beneficial acts. To use Rachels’s (1991) phrase, praise seems no less “bound up” with the recognition of persons as responsible agents than blame. If so, then assuming that judgments of retributive desert express reactive attitudes such as blame, and that judgments of distributive desert express reactive attitudes such as praise, the former seem no more central to our practices of responsibility than the latter. In sum, Scheffler believes that, compared to judgments of distributive desert, judgments of retributive desert are more closely tied to our practices of responsibility, and so are harder to eliminate from our theories of justice. In response, I denied that reactive responses to breaches of expectations are more central to our treatment of

Desert-Based Justice   165 persons as responsible agents than reactive responses to the performance of beneficial acts. Even if they were, I claimed, some judgments of distributive desert may have the former structure. I conclude that Scheffler’s proposed distinction between retributive and distributive desert fails. By his logic, rejecting distributive desert should be no more tolerable than rejecting retributive desert.

7.4  Desert and Luck Egalitarianism Given desert’s connection to responsibility, giving desert a place in a theory of distributive justice is a way of making people’s distributive shares dependent on what they are responsible for, and hence their choices (cf. Lippert-​Rasmussen 2009). Luck egalitarians also think that people’s distributive shares should depend on their choices. But standard formulations of this view make no mention of desert (Eyal 2007; Cohen 1989; Temkin 1993). While luck egalitarians are aware of connection between choice and desert (see especially Temkin 1993), their focus is on the concept of choice, not on the concept of desert. As I will show, however, the connections between luck egalitarianism and desert-​ theory are strong, and have become clearer as (some) luck egalitarians have developed or modified their view in response to questions and challenges.7 As the name suggests, early formulations of luck egalitarianism combine a “pro” equality view with an “anti” luck view. Thus Temkin says that the luck egalitarian’s central belief is that “it is bad—​unjust and unfair—​for some to be worse off than others through no fault or choice of their own” (1993: 13; see also Cohen 1989). Suppose that the “currency” of justice—​to use Cohen’s (1989) phrase—​is welfare, and suppose that A has ten units and B has five units. According to early formulations of luck egalitarianism, if this inequality is not due to A’s or B choice (i.e., if it is due to luck), then it is bad, but if it is due to choice, then it is not bad. Notice that this formulation of luck egalitarianism implies nothing about unchosen (and chosen) equalities. Suppose that both A and B have ten units of welfare, and this is due to luck. Perhaps B made a choice that should have led him to have five units of welfare, but by luck ended up with ten. Suppose also that B’s extra five units were not extracted from A or anyone else, but descended upon B like manna from heaven. On the one hand, there is no inequality to worry about. On the other, A’s and B’s relative position is due to luck. This formulation of luck egalitarianism does not tell us how (i) the state in which A has ten and B has five compares to (ii) the state in which both A and B have ten. This represents a possible gap between luck egalitarianism and desert-​theory. The desert-​theorist—​the person who thinks that it is good (or just or fair—​hereafter these

7 

As this suggests, there are many different versions of luck egalitarianism; not all of them have been developed or modified in the same way.

166   Jeffrey Moriarty alternatives will be assumed) for people to get what they deserve—​thinks that it is bad if some are worse off than others through no choice of their own and if some are equal to others through no choice of their own. In both cases, people fail to get what they deserve. But early formulations of luck egalitarianism leave it open for luck egalitarians to believe that unchosen equalities—​scenarios in which some are equal to others through no choice of their own—​are not bad (perhaps because they are equalities). But later formulations of luck egalitarianism close this gap. Cohen says that “there is a symmetry in the luck egalitarian’s attitude to plain, ordinary equality, and plain, ordinary inequality—​both are bad if and only if they are in disaccord with choice” (2006: 44; see also Temkin 2011). On this revised formulation, luck egalitarianism holds that, in the above example, (i) the state in which A has ten and B has five is preferable to (ii), the state in which both A and B have ten. Unchosen equalities are bad, at least relative to chosen inequalities. This has the effect of adding a “pro” choice element to the “pro” equality and “anti” luck elements of their view (Stemplowska 2013). Now for luck egalitarians, there is a reason for people to suffer (or enjoy) the inequality-​generating consequences of their choices, not just a reason against their suffering (or enjoying) the inequality-​generating consequences of unlucky (or lucky) breaks. Desert-​theorists agree. Even if we accept Cohen’s revised formulation of luck egalitarianism, there is still much potential space between it and desert-​theory. Both Temkin’s early and Cohen’s revised formulation of luck egalitarianism place no limitations on the consequences of choices that people should suffer (or enjoy). The desert-​theorist, by contrast, thinks that people should suffer (or enjoy) all and only those consequences that they deserve to suffer. But, as we will see, in response to objections, some luck egalitarians have qualified their accounts of the consequences of choice, and in ways that further reduce the space between luck egalitarianism and desert-​theory (Arneson 2001). One problem that has been raised for luck egalitarianism concerns praiseworthy but costly choices (Eyal 2007; Temkin 2011). Suppose that A and B are neighbors who each have ten units of welfare. Suppose that their mutual neighbor C’s house catches on fire with C asleep inside. B is home for the weekend, while A is away. B sees that he can save C, but that if he does, he will be burned himself. B decides to save C and, as expected, gets burned. B’s welfare level is reduced to seven by this event, while A’s remains at ten. Luck egalitarianism seems to suggest that it is not bad—​even good—​that B is worse off than A. For this inequality is due to B’s choice. While we have not articulated a complete account of desert, it is unlikely that desert-​ theorists will agree. They will say that B does not deserve to be worse off than A because B chose to save C. If anything, B deserves to be better off than A. In fact, many luck egalitarians want to resist the conclusion that it is not bad—​even good—​that B is worse off than A. Some do so by appealing to values external to their theory. They say that, while luck egalitarianism itself says that it is not bad—​even good—​ that B is worse off than A, other values (e.g., utility) tell against this result (Tan 2012). But this seems ad hoc. Moreover, we might doubt that there is anything good about B’s being worse off than A. Perhaps in response to this, other luck egalitarians reinterpret or modify their view so that it does not imply that it is not bad—​even good—​that B is

Desert-Based Justice   167 worse off than A. In doing so, they bring luck egalitarianism closer to desert-​theory in its implications. Moreover, some justify this modification by appealing to desert and its cognates. Temkin, for example, says B should not suffer the full cost of his choice to save C because B does not deserve to suffer this cost. He says that what egalitarians really care about is comparative fairness. As a result, they also “care about moral desert, because they . . . believe it is . . . unfair if a less morally deserving person fares better . . . than a more morally deserving person does” (2011:  66). Similarly, Eyal says that luck egalitarians should believe that people should suffer only the consequences of “culpable choices” (2007: 6), that is, choices that are free and “at least somewhat morally wrong” (2007: 6). In our example, B’s choice to save C, while free, was not at all morally wrong, so B should not be made to suffer its consequences. Eyal says that his adjustment to luck egalitarianism is not desert-​based. But this is dubious. Culpability is often treated by legal theorists as a basis of negative desert (Alexander, Ferzan, and Morse 2009). A determination of whether a person is culpable for an offense just is a determination of whether they deserve punishment for it (cf. Tadros 2011). Another problem that has been raised for luck egalitarianism concerns merely imprudent choices with truly disastrous consequences. Suppose that the reason C’s house catches on fire in the above example is that C, after a long and difficult day, and resting comfortably in her favorite armchair, decides to smoke a cigarette. C falls asleep and the cigarette drops from her hand, igniting the carpet. Being a sound sleeper, C wakes up only after her house is fully engulfed in flames. Critics have suggested that an implication of luck egalitarianism is that C should be “abandoned” to her fate, since the fire is due to C’s choice to smoke while sleepy (Anderson 1999). Again, desert-​theorists are unlikely to agree. They will say that, despite C’s poor choice to smoke while sleepy, C does not deserve to suffer the full consequences of doing so. Suppose that, without B’s intervention, this is the loss of all her possessions and life-​ threatening burns. The “punishment,” desert-​theorists will say, does not fit the “crime.” And again, many luck egalitarians want to resist the conclusion that it is not bad—​ even good—​for C to be badly burned and lose all her possessions. While some appeal to values external to their theory (Tan 2012), others seek to reinterpret or modify luck egalitarianism itself so that it does not have this implication. Temkin says that “full responsibility for one’s choices doesn’t entail full responsibility for one’s predicament,” and that egalitarians should be concerned instead that people enjoy or suffer the “expected value of their choices” (2011: 65). Dworkin (2003) says that, in an ideal world, people would purchase insurance against the kind of devastating losses C suffers in the fire. He therefore understands society’s refusal to abandon C to her fate as an implication of, rather than a deviation from, equality. Both Temkin’s and Dworkin’s qualifications have the effect of bringing luck egalitarianism closer in its implications to desert-​theory. The challenges raised by “praiseworthy but costly choices” and “merely imprudent choices with truly disastrous consequences” press luck egalitarians to answer a more general question. That is, what are—​or should be taken to be—​the consequences of people’s choices (Olsaretti 2009)? Is B’s suffering nagging injuries a (properly understood) consequence of his choice to rush into a burning building? Is C’s being badly

168   Jeffrey Moriarty burned and losing all of her possessions a (properly understood) consequence of her choice to smoke while sleepy? Does it matter if B has a moral obligation to rescue C, or if C has a moral obligation to compensate B? Many writers have claimed that luck egalitarians implicitly, or in some cases explicitly, endorse a “natural rewards” approach to consequence identification (Anderson 1999; Fleurbaey 2008; Vallentyne 2002). On this view, the consequences of people’s choices are whatever actually results from them, provided that others involved act within their (negative) rights. As seen, however, this approach commits luck egalitarians to some implausible conclusions, which they have been at pains to avoid. But egalitarians may be able to avoid them by adopting a different theory of consequences. We have seen already that, in certain cases, it seems more plausible that people should suffer (or enjoy) the consequences that they deserve to suffer than that they should suffer (or enjoy) whatever consequences happen to flow from their choices. And some writers, notably Temkin (2011), have sought to move luck egalitarianism in this direction (see also Arneson 2007). It may well be, as Dekker (2009) suggests, that luck egalitarians should want for people to suffer or enjoy all and only the consequences that they deserve to suffer or enjoy (cf. Anderson 2008; Olsaretti 2009). This is an important topic of future research.

7.5  Desert, Value, and Neutrality There is widespread agreement that, if a person deserves something, it is in virtue of something about him that has value. This is sometimes expressed as the claim that, for P to be deserving in virtue of F, F must be (appropriately) the subject of an appraising or evaluative attitude. Some political philosophers—​namely, neutralists—​have argued that this poses a problem for incorporating desert into theories of distributive justice. They claim that states should not aim to promote, or justify their policies by appealing to, any controversial conception of the good. Since desert-​claims imply controversial value judgments, the argument goes, desert should be rejected as a distributive criterion. Let us call this the “neutrality argument” against desert. Rawls is the most prominent proponent of this argument. He says that “a conception of moral desert as moral worth of character and actions cannot be incorporated into a political conception of justice in view of the fact of reasonable pluralism” (2001: 73).8 Because citizens have different views about what is valuable, they have different views about what makes people deserving. So, in Rawlsian terms, desert cannot be part of a theory of justice that is the subject of an overlapping consensus of persons with reasonable comprehensive doctrines. Hurley also endorses the neutrality argument. 8  This is not Rawls’s only anti-​desert argument. In A Theory of Justice, he suggests that rewarding desert is “impracticable,” because it is impossible to separate that part of a person’s achievement (or effort) which is due to his autonomous choice from that part which is due to (dis)advantageous natural and social factors (1971: 311–​12). For a critical discussion of this argument, see Moriarty (2005).

Desert-Based Justice   169 She says that “it is not for society to decide which efforts are more, less, or equally deserving” (2003: 201). Doing so requires making “objectionably paternalist or perfectionist” judgments, both of which are unacceptable in a “pluralistic liberal democracy” (2003: 200, 202).9 It is no coincidence that desert was largely cast out of theories of justice during the 1970s and 1980s, as these were the decades in which neutralism was ascendant in liberal political philosophy. It is also no coincidence that the re-emergence of desert has coincided with the rise of perfectionism, a view that says that states can “take sides” in debates about value. To be sure, a commitment to perfectionism does not entail a commitment to desert. But it is open to perfectionists to sanction the recognition by the state of the types of value-​claims embedded in desert-​claims, and many perfectionists are in fact friends of desert (Arneson 2003, 2007; Sher 1987, 1997). It might seem, then, that the neutrality argument against desert hangs on whether neutralism or perfectionism is correct. To some extent this is right. The more perfectionist a political theory is, the more room it leaves for desert. We cannot enter into this debate here. What I will do instead is suggest that there is room for desert even on a more neutral political theory. The belief that there isn’t, I suggest, may stem from mistaken beliefs about desert or neutralism. First, consider desert. Rawls seems to think that a desert-​sensitive theory of distributive justice is one that gives a role to “moral desert,” that is, desert whose basis is “the moral worth of character and actions” (2001: 73). Similarly, Temkin understands “proportional justice” as the view that “people should fare well precisely to the extent that they are morally deserving, where this is a function of their virtue or moral character” (2011: 54, emphasis in original). Assuming that the state’s actions ought to be guided by considerations of justice, any theory of justice that gives a role to desert—​understood in this way—​will be deeply unattractive. Only the most illiberal societies have state agencies concerned with the promotion of moral virtue. Desert, however, should not be understood in this way. All desert is moral in the sense that desert is a moralized concept. To say that P deserves T in virtue of F is to say that there is a reason for P to have T. But not all desert is moral in the sense that it is based on the “moral worth of character and actions” or “moral virtue.” In fact, most desert is not moral in this sense (Miller 1999; Sher 1987). People can come to be deserving in virtue of performing certain actions without having moral motives for those actions, and without being morally virtuous. A scientist deserves to win the Nobel Prize in virtue of their important discovery; a soldier deserves the Congressional Medal of Honor in virtue of their “conspicuous acts of gallantry and intrepidity”; an entrepreneur deserves

9  We might see luck egalitarianism in its early formulations as attempting to capture the less controversial aspect of desert—​its connection to responsibility—​while leaving aside the more controversial aspect—​its connection to value. In reinterpreting or modifying their theories to meet the questions and criticisms we have discussed, however, luck egalitarians likely commit themselves to the sorts of controversial value-​claims that neutralists are skeptical about. For an incisive discussion of this issue, see Hurley (2003).

170   Jeffrey Moriarty the profits from their firm in virtue of their shrewd business decisions. These agents may be deserving of these things even if they have self-​serving motives and lack moral virtue. Next, consider neutralism. Neutralists do not think that states should never aim to promote, or justify their actions by appealing to, claims about value. They should only refrain from doing so when there is reasonable disagreement about what is valuable (Arneson 2003). So if there is no reasonable disagreement about a certain value, then even a neutralist state can recognize desert-​claims based on this value. And we have reason to think that such values exist. The standard treatment of punishment provides a useful starting point. It is widely thought that offenders should receive the amount of punishment that, or at least no more punishment than, they deserve (Moore 1997). As this is often put: the severity of the punishment should match, or at least not exceed, the seriousness of the offense. This means that determining how much punishment an offender deserves requires assessing how bad his offense is. I am aware of no writer who rejects the principle that punishments should be proportionate to crimes on neutralist grounds. If so, then at least some value judgments are legitimate, even for neutralists. The question is whether any such judgments are legitimate on the distributive “side” of justice. It would be surprising if none were. It would be surprising, that is, if there were no reasonable disagreement about the badness of the things people do, but always reasonable disagreement about the goodness of the things people do. For goodness and badness are, in some sense, two sides of the same coin. And indeed some critics of state neutrality have suggested that there is no reasonable disagreement about the desirability of many goods, including “aesthetic experience, human relationships, amusement and play, [and] knowledge” (Chan 2000). We might add to this list of goods the conditions for their achievement, including health, wealth, and security. Thus if, for example, aesthetic experience is valuable, then a person does something valuable when they produce a series of beautiful paintings. Likewise, if health is valuable, then a person does something valuable when they discover new medicines. These actions, in turn, can be the bases of desert-​claims that even a neutralist state, in a “pluralistic liberal democracy,” can recognize.

7.6 Conclusion This chapter covered a lot of ground but left many issues unexplored. It discussed the nature of desert, arguments for its requital, the asymmetric role of desert in distributive and retributive justice, the connections between desert-​theory and luck egalitarianism, and the neutralist case against desert. Among the issues it did not explore, but which are currently the subjects of lively debate, are the connections between non-​comparative and comparative desert (Kagan 2012), and the relative weight of claims of desert and welfare (Skow 2012). The latter issue is especially important to the development of a comprehensive desert-​sensitive theory of distributive justice. The incomes that people

Desert-Based Justice   171 receive in the market do not track any reasonable conception of their economic deserts (Olsaretti 2004). Ensuring that they do, however, may introduce significant welfare-​ diminishing inefficiencies into the economic system. The best theory of distributive just­ice will balance the value of desert against the value of welfare, and perhaps many other values. Whatever the final shape of this theory, it seems clear that desert will play some role in it. After a long hiatus, desert is back in discussions of distributive justice, and is likely to remain there for some time to come.

Acknowledgments For extremely detailed and valuable comments on drafts of this chapter, I  thank Serena Olsaretti. A  portion of this chapter was presented at the second annual New Orleans Invitational Seminar on Ethics (NOISE). I thank members of that audience and especially my commentator, Simon Cabulea May, for helpful discussion.

References Alexander, L., Ferzan, K. K., and Morse, S. J. (2009). Crime and Culpability:  A Theory of Criminal Law. New York: Cambridge University Press. Anderson, E. (1999). “What is the Point of Equality?” Ethics 109(2): 287–​337. Anderson, E. (2008). “How should Egalitarians Cope with Market Risks?” Theoretical Inquiries in Law 9(1): 239–​70. Aristotle. (1999/​350 bce). Nicomachean Ethics. T. Irwin (ed. and trans.). Indianapolis, IN: Hackett. Arneson, R. J. (2001). “Luck and Equality.” Aristotelian Society Supplementary Volume 75(1): 73–​90. Arneson, R. J. (2003). “Liberal Neutrality on the Good: An Autopsy,” in G. Klosko and S. Wall (eds) Perfectionism and Neutrality. Lanham, MD: Rowman & Littlefield, pp. 191–​218. Arneson, R. J. (2007). “Desert and Equality,” in N. Holtug and K. Lippert-​Rasmussen (eds) Egalitarianism: New Essays on the Nature and Value of Equality. New York: Oxford University Press, pp. 262–​93. Chan, J. (2000). Legitimacy, Unanimity, and Perfectionism. Philosophy and Public Affairs 29(1): 5–​42. Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99(4): 906–​44. Cohen, G. A. (2006). “Luck and Equality: A Reply to Hurley.” Philosophy and Phenomenological Research 72(2): 439–​46. Cupit, G. (1996). Justice as Fittingness. New York: Oxford University Press. Dekker, T. J. (2009). “Choices, Consequences and Desert.” Inquiry 52(2): 109–​26. Dworkin, R. (2003). “Equality, Luck, and Hierarchy.” Philosophy & Public Affairs 31(2): 190–​8. Eyal, N. (2007). “Egalitarian Justice and Innocent Choice.” Journal of Ethics and Social Philosophy 2(1): 1–​18. Feinberg, J. (1999/​1963). “Justice and Personal Desert,” in L.P. Pojman and O. McLeod (eds) What do We Deserve? A Reader on Justice and Desert. New York: Oxford University Press, pp. 70–​83.

172   Jeffrey Moriarty Feldman, F. (1995). “Desert: Reconsideration of Some Received Wisdom.” Mind 104(413): 63–​77. Fleurbaey, M. (2008). Fairness, Responsibility, and Welfare. New York: Oxford University Press. Hooker, B. (2000). Ideal Code, Real World:  A Rule-​Consequentialist Theory of Morality. New York: Oxford University Press. Hurka, T. (2003). “Desert: Individualistic and Holistic,” in S. Olsaretti (ed.) Desert and Justice. New York: Oxford University Press, pp. 45–​68. Hurley, S. L. (2003). Justice, Luck, and Knowledge. Cambridge, MA: Harvard University Press. Husak, D. N. (2000). “Holistic Retributivism.” California Law Review 88(3): 991–​1000. Kagan, S. (2012). The Geometry of Desert. New York: Oxford University Press. Kelly, E. (2002). “Doing without Desert.” Pacific Philosophical Quarterly 83(2): 180–​205. Lippert-​Rasmussen, K. (2009). “Justice and Bad Luck,” in E. N. Zalta (ed.) The Stanford Encyclopedia of Philosophy. http://​plato.stanford.edu/​archives/​fall2009/​entries/​justice-​bad-​ luck/​(accessed 8 November 2017). Levy, N. (2005). “The Good, the Bad, and the Blameworthy.” Journal of Ethics & Social Philosophy 1(2): 1–​16. Lucas, J. R. (1980). On Justice. Oxford: Clarendon Press. McLeod, O. (1996). “Desert and Wages.” Utilitas 8(2): 205–​21. Mill, J. S. (1998/​1863). Utilitarianism. R. Crisp (ed.). New York: Oxford University Press. Miller, D. (1999). Principles of Social Justice. Cambridge, MA: Harvard University Press. Moore, M. (1997). Placing Blame: A General Theory of the Criminal Law. Oxford: Clarendon Press. Moriarty, J. (2003). “Against the Asymmetry of Desert.” Noûs 37(3): 518–​36. Moriarty, J. (2005). “The Epistemological Argument against Desert.” Utilitas 17(2): 205–​21. Moriarty, J. (2009). “Deserving Jobs, Deserving Wages,” in J. Smith (ed.) Normative Theory and Business Ethics. Lanham, MD: Rowman & Littlefield, pp. 119–​46. Morris, H. (1968). “Persons and Punishment.” The Monist 52(4): 475–​501. Olsaretti, S. (2004). Liberty, Desert and the Market. Cambridge: Cambridge University Press. Olsaretti, S. (2009). “Responsibility and the Consequences of Choice.” Proceedings of the Aristotelian Society 109(2): 165–​88. Rachels, J. (1991). “What People Deserve,” in J. Arthur and W. H. Shaw (eds) Justice and Economic Distribution, 2nd edn. Englewood Cliffs, NJ: Prentice Hall, pp. 136–​48. Rachels, J. (1997). Can Ethics Provide Answers? And Other Essays in Moral Philosophy. Lanham, MD: Rowman & Littlefield. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (2001). Justice as Fairness: A Restatement. Erin Kelly (ed.). Cambridge, MA: Harvard University Press. Ross, W. D. (2002/​1930). The Right and the Good. P. Stratton-​Lake (ed.). New York: Oxford University Press. Scanlon, T. M. (1998). What We Owe to Each Other. Cambridge, MA: Harvard University Press. Scheffler, S. (2000). “Justice and Desert in Liberal Theory.” California Law Review 88(3): 965–​90. Scheffler, S. (2003). “Distributive Justice and Economic Desert,” in S. Olsaretti (ed.) Desert and Justice. New York: Oxford University Press, pp. 69–​92. Schmidtz, D. (2002). “How to Deserve.” Political Theory 30(6): 774–​99. Schmidtz, D. (2006). Elements of Justice. New York: Cambridge University Press. Sher, G. (1987). Desert. Princeton, NJ: Princeton University Press. Sher, G. (1997). Beyond Neutrality: Perfectionism and Politics. New York: Cambridge University Press.

Desert-Based Justice   173 Sidgwick, H. (1981/​1907). The Methods of Ethics. Indianapolis, IN: Hackett. Skow, B. (2012). “How to Adjust Utility for Desert.” Australasian Journal of Philosophy 90(2): 235–​57. Smilansky, S. (1996). “Responsibility and Desert: Defending the Connection.” Mind 105(417): 157–​63. Stemplowska, Z. (2013). “Luck Egalitarianism,” in G. Gaus and F. D’Agostino (eds) The Routledge Companion to Social and Political Philosophy. New York: Taylor & Francis, pp. 389–​400. Tadros, V. (2011). The Ends of Harm: The Moral Foundations of Criminal Law. New York: Oxford University Press. Tan, K.-​C. (2012). Justice, Institutions, and Luck. Oxford: Oxford University Press. Temkin, L. S. (1993). Inequality. New York: Oxford University Press. Temkin, L. S. (2011). “Justice, Equality, Fairness, Desert, Rights, Free Will, Responsibility, and Luck,” in C. Knight and Z. Stemplowska (eds) Responsibility and Distributive Justice. New York: Oxford University Press, pp. 51–​76. Vallentyne, P. (2002). “Brute Luck, Option Luck, and Equality of Initial Opportunities.” Ethics 112(3): 529–​57. Vilhauer, B. (2009). “Free Will Skepticism and Personhood as a Desert Base.” Canadian Journal of Philosophy 39(3): 489–​511. White, S. G. (2003). The Civic Minimum. New York: Oxford University Press.

Pa rt  I I

DI ST R I BU T I V E J U ST IC E A N D OT H E R V I RT U E S

Chapter 8

Retribu tive J u st i c e Larry Alexander

8.1 Introduction The topic of this chapter is the relationship between retributive justice and distributive justice. All versions of retributive justice rely on the reality of desert and specifically negative desert. Accounts of distributive justice, however, frequently do not treat desert as a basis of the just distributive share of goods or welfare. Or, if they do make desert a basis of just distributive distribution, the notion of desert they rely upon is different from that of the retributivists. So there is an apparent asymmetry in the treatment of desert as we move from retributive justice to distributive justice. Some of the asymmetry is due to retributivists having as their desert basis the viciousness of will displayed by offenders, whereas some theorists of distributive justice, if they even make desert a relevant consideration, focus on contribution and effort as desert bases, rather than on virtuous will.1 A second source of asymmetry between the roles of desert in retributive and distributive justice is that many (but surely not all) retributivists treat desert as noncomparative—​the justice of the punishment A receives can be assessed independently of the punishment B receives. On the other hand, many (but not all) theories of distributive justice treat it as essentially comparative, so that the justice of A’s share of goods is dependent on the magnitude of B’s share. If, because of these asymmetries, neither retributive nor distributive justice can be reduced to the other, how can they be reconciled? Or, if they cannot be reconciled because of these asymmetries, is there a position on distributive justice that eliminates 1 

Thomas Hurka, for example, places retributive desert within the domain of what he calls “moral desert,” which holds that the virtuous deserve pleasure and the vicious deserve to suffer. For economic desert, however, Hurka requires that income be a function solely of contribution and effort regardless of the virtue or viciousness of the recipient. See T. Hurka (2003). “Desert: Individualistic and Holistic,” in S. Olsaretti (ed.) Desert and Justice (Oxford: Oxford University Press), pp. 45–​68.

178   Larry Alexander the barriers to their reconciliation? Can we be retributivists and still have a compatible account of distributive justice, and if so, what account or accounts will qualify? These are the questions this chapter addresses.

8.2  Retributive Justice One naturally thinks of retributive justice in connection with crime and punishment. Retributivists—​and I confess to being one of them—​believe that punishment should be measured by the ill desert of the offender. The offender should receive no more punishment than he deserves for his offense—​desert is a ceiling on permissible punishment. That is why it is wrong to punish one who is innocent of any wrongdoing, which is but an instance of punishment greater than deserved. On the other hand, most retributivists believe that not only is ill desert necessary for permissible punishment but also that it is sufficient for permissible punishment. No other good, such as deterrence, rehabilitation, or incapacitation, need be achieved beyond requiting ill desert in order for punishment to be permissible. Having the guilty suffer to the extent they deserve is a good worthy of individual and social pursuit and resources (see Moore 1997: 153–​88; Alexander and Ferzan 2009: 7–​10). Retributivists divide over how great a good meting out retributive desert is and thus how it compares to those goods that must be sacrificed to achieve it. The strongest retributivists believe retributive justice is a great good and that we are obligated to sacrifice almost all other goods in order to achieve it. The weakest retributivists believe the good of retributive justice is outweighed by almost all other goods and is almost always permissibly forgone. Most actual retributivists are moderates about the weight of retributive justice (see, e.g., Husak 1987: 224–​6; Husak 2008: 196–​206). Retributivists also divide over other matters. Some believe that retributive justice is comparative—​what B deserves for a given wrong is measured by the punishment A was given for the same wrong.2 Most retributivists, including me, believe that retributive desert is noncomparative, and that punishing one who noncomparatively deserves that punishment is not a violation of retributive justice even if a similar wrongdoer receives less or no punishment. Retributivists also divide over whether the results of wrongful acts affect retributive desert—​for example, whether a murderer deserves more punishment than one who attempts to murder but fails, or whether a reckless driver who injures someone deserves more punishment than an equally reckless driver who, solely due to luck, does not. (I am in the camp that answers that question “No” [Alexander and Ferzan, 2009: Ch. 5].) 2  In a recent article, Ronen Avraham and Daniel Statman take the position that retributive justice has both comparative and noncomparative dimensions, and that the justice of the punishment one receives is a function of both. See R. Avraham and D. Statman (2013). “More on the Comparative Nature of Desert: Can a Deserved Punishment be Unjust?” Utilitas 25(3): 316–​33.

Retributive Justice   179 Although I  shall not have the space to defend adequately the positions I  take on matters that divide retributivists, allow me at least to give you a glimpse of my reasons for holding those positions. Perhaps the most important of my positions in terms of how it affects the arguments in the rest of this chapter is that which holds retributive desert to be noncomparative.3 What one deserves for one’s wrongdoing is not affected by how others of similar desert are treated. It does not turn on where one lives or when one lives. It is invariant through time and space. A wrongful killing by a caveman deserves the same punishment as a similarly wrongful killing of a citizen of the United States in 2014. Retributive desert is cosmic, the same on the planets of Alpha Centauri as on Earth.4 One consequence of the noncomparative position is that when two people commit identical wrongs, and A receives the punishment he deserves and B receives less or no punishment, A cannot complain that his punishment is unjust and that he has been wronged by the failure to punish B. Perhaps there was some societal good to be achieved by remitting B’s punishment. Or perhaps B got off lightly due to improper partiality by the sentencing judge—​perhaps B is the judge’s son-​in-​law—​or perhaps it was due to naked bigotry towards A because of A’s race or religion. In the latter cases, the judge has violated duties towards the citizenry, duties that require judges to mete out retributive punishment on proper bases. Even so, A himself, having received the punishment he deserves, has no valid complaint about his punishment.5 One difficulty with the noncomparative view that I favor is that when the wrongdoer is already suffering undeservedly, or is living in a Hobbesian society in which almost everyone, himself included, is suffering, retributive punishment may seem unwarranted. If the miscreant’s level of suffering is already exceeding what he deserves for his wrongdoing, then punishing him would appear unjust. Yet at the same time, not punishing him would leave him no worse off than his properly behaving neighbors, which also appears unjust, given that unequally deserving people would then be being treated equally. I shall return to this problem of undeserved suffering in section 8.4. The comparative view of retributive desert, however, has problems that dwarf the problem of undeserved suffering. One problem is that it provides no standard for determining how much to punish the first wrongdoer, or the first wrongdoers of specific wrongs. If no one has ever been punished for, say, armed robbery, then what the first armed robber deserves has no answer.6 Does he deserve death, a long prison term, or a 3   Of course, we may try to determine the noncomparative desert for committing offense X by comparing offense X to offense Y, the noncomparative desert for which we are more certain. 4  Of course, people can be more or less sensitive or squeamish within and across cultures and historical periods with respect to how they are affected by particular kinds of punishment. Put differently, the same term of imprisonment or amount of corporal punishment will cause different amounts of suffering in different people, in different places, and in different times. That is not only a point I acknowledge, but it is also a point I exploit. See section 8.4.1. 5  For a somewhat different view, see Avraham and Statman 2013 (n. 2). 6  Merely comparing armed robbery to the punishment meted out for a different crime will not help. Even if, on a comparative view of retributive desert, we can determine that petty theft deserves less punishment than armed robbery, it will not tell us how much less. If the armed robber got five years’ imprisonment, any amount less than that would be consistent with comparative desert. Only

180   Larry Alexander mild rebuke? That is a difficult matter on the noncomparative view of retributive desert, but at least that view assumes that there is an answer. On the comparative view, however, there is no answer because there is no desert. Without a comparator, comparative desert is an empty notion. The second problem with the comparative conception of desert is related to the first. One’s just desert is, on this view, hostage to the punishments that have been meted out in the past. If the first caveman was stoned to death for taking more than his share of mastodon meat, then petty thieves now deserve stoning. On the other hand, if a murderer was fined some trivial amount, small fines are all that murderers now deserve. And if widely disparate punishments have been handed out for rape, there is no answer to what rapists now deserve.7 I believe these problems with the comparative conception of desert are fatal. So I will stick with the noncomparative conception, aware of its difficulties—​the basic difficulty of determining what is deserved for the various kinds of wrongdoing, and the more specific difficulty of determining what wrongdoers deserve who have previously suffered undeservedly, or who are presently so suffering or will in the future. Aside from the controversy over whether retributive desert is comparative or noncomparative, another matter that divides retributivists is that of the currency of retributive desert. I believe that currency to be suffering. If a wrongdoer’s treatment does not achieve his suffering to a requisite extent, then he has not received the punishment he deserves. Others disagree, however. Some cash out retributive desert in terms such as loss of liberty irrespective of how much the wrongdoer suffers (see, e.g., Markel and Flanders 2010). A  ten-​year term of imprisonment for each of two equally deserving wrongdoers is justly deserved even if one suffers intensely and the other enjoys prison  life. Others view retributive punishment as a matter of expressing to the wrongdoer society’s condemnation of his act (Duff 2001). On this view, the wrongdoer’s suffering is again not an essential part of retributive punishment, though it is a normal consequence of the hard treatment that is the medium of the expression of condemnation. The first view seems to me to be obviously false. There are, of course, considerable and quite obvious practical difficulties that attend trying to ensure that equally culpable wrongdoers suffer equally. Suffering is impossible to measure in practice even if a noncomparative view of retributive desert can determine whether four years’ or four minutes’ imprisonment is deserved. 7 

One might believe that one could devise a complex desert function that takes account of both noncomparative desert and comparative desert (see, e.g., Hurka 2003, n. 1). But note that the comparators for comparative desert can range across all time and space. If the noncomparative desert for committing crime X is five years’ imprisonment, but the Soviet Union punished crime X with death (as did cave men), does overall retributive justice demand that we punish crime X with more than five years? Likewise, if a lot of historical societies have let violators of crime X off with mere verbal rebukes, should we say that therefore violators of crime X deserve a punishment a good deal less than five years? Both of those questions seem to me easy to answer: “No.”

Retributive Justice   181 we can give a theoretical account of it such that we then know what facts in the world count as suffering and its magnitude. And even were we able to gauge suffering accurately, the administrative costs of attempting to achieve suffering commensurate with desert would be enormous and foreclose the achievement of goods other than retributive justice. Nevertheless, putting aside those practical considerations, it seems odd to deny that negative desert should turn on suffering.8 Suppose that most people hate brussels sprouts. They suffer if forced to eat them. Thus, forcing wrongdoers to eat brussels sprouts is adopted as a form of punishment. Sue, however, loves brussels sprouts. She craves them. It would be strange, then, to say that Sue had been “punished” by being forced to eat brussels sprouts. And the same goes for any form of punishment. If it does not cause a given wrongdoer to suffer, it is not a punishment of that wrongdoer even if it is a punishment of other wrongdoers. 9 The second, expressive, view gets something right, but not the currency of retributive punishment. There is a connection between punishment and an expression of condemnation, but the connection, as I  see it, goes like this:  In teaching someone—​say, a child—​the meaning of the term “wrong,” one necessarily associates that meaning with the propriety of the negative reactive attitudes. When one has committed a wrong, one should experience guilt. When another has committed a wrong, one should experience indignation and believe that the wrongdoer should be the target of expressions, verbal or behavioral, of blame. To experience guilt is to suffer psychologically, and the expression of blame is meant to induce, at a minimum, the psychological suffering of guilt. I say “at a minimum” because the blamer may intend to produce suffering beyond the pangs of guilt, such as alienation of one’s esteem, affection, and companionship. The reactive attitudes, deserved suffering, and the notion of “wrongness” are all bound up together. When we impose upon someone the punishment he deserves, we are both intending deserved suffering and expressing reactive attitudes. Deserved suffering cannot be deleted without neutering the reactive attitudes. On the other hand, given the existence of deserved suffering, a wrongdoer can experience that suffering without anyone’s imposing it on him, much less imposing it on him with an expressive intention. Finally, a word on that other great divide within retributivism, the divide over whether the results of wrongful acts (and omissions) affect the magnitude of retributive 8 

By negative desert I mean retributive desert, i.e., deservingness of suffering. Positive desert is the mirror opposite, the deservingness of benefits due to above-​average virtuousness. 9  Of course, one might respond that by forcing Sue to eat brussels sprouts, Sue has lost something valuable, namely, the freedom to choose whether to do so. Is the loss of something valuable rather than the suffering such a loss likely to cause the proper currency of retributive justice? I do not believe so. For even if everyone loses something of value by being forced to eat brussels sprouts, the value of what is lost will vary tremendously from person to person. In Sue’s case, the value to her of the freedom not to eat brussels sprouts is minimal. For others, it is much greater. Their punishments are materially unequal even if formally equal. The same is true of loss of liberty: The agoraphobe suffers from it a good deal less than does an outdoorsman.

182   Larry Alexander desert. Does a successful murderer deserve more punishment than an attempted murderer who acts with the same level of culpability? Does a reckless driver who kills a pedestrian deserve more punishment than an equally reckless driver who luckily avoids causing death? Does a pickpocket who comes away with the victim’s wallet deserve more punishment than one who discovers the victim’s pocket to be empty? As I  said, I  hold with those who say results do not matter (Alexander and Ferzan 2009:  Ch. 5). Put differently, result luck does not affect one’s negative desert. There are two components of my and others’ case against results affecting desert. One component is based on hypotheticals that make distinguishing among defendants based on who caused the harm in question appear silly or even pathological (Alexander and Ferzan 2009:  172–​80). The other component is based on the weakness of the various accounts of causation that are offered to link the wrongdoer’s act with its resultant harm (Alexander and Ferzan 2009:  180–​8; Alexander and Ferzan 2012a; Alexander and Ferzan 2012b; Alexander 2011). I shall not repeat those components here. That is all I need to say about retributive justice. In what follows I shall be using the version of it that I  find most persuasive, namely, that retributive justice is a good worthy of pursuit, but one that does not trump all other goods; that it is noncomparative; that its currency is suffering; and that it turns solely on what the actor is trying to do or the risks he believes he is imposing, not on the results of his act. Most of the problems that I shall raise, however, will affect versions of retributive just­ ice other than mine.10

8.3  How Great a Good? If one believes that retributive justice is an intrinsic good, but not the only intrinsic good, then the resulting picture of distributive justice must be pluralistic about the intrinsic goods to be distributed. Take even the simplest picture of distributive just­ ice, one in which there is, aside from the intrinsic good of retributive justice, only one other intrinsic good that matters for purposes of just distribution—​pleasure, welfare, or resources. Meting out retributive justice will most likely come at a cost of that good. I do not mean the cost to the offender, for retributive justice regards his loss of pleasure, welfare, or resources as a good. Rather, I mean the cost to those who must capture the 10 

There are a number of other problems I do not address here that proponents of retributive justice must deal with: the free will/​determinism debate and its relevance to desert; the desert of those who violate deontological constraints for altruistic reasons; whether there is a “volume discount” of desert for those who commit mass genocide (e.g., Mao, Stalin, Hitler, and Pol Pot) or who commit many attempted crimes in a short time period (e.g., one who fires several shots within a few seconds attempting to kill his intended victim); and how to treat changes in personal identity between the time of the commission of the crime and the time of imposition of punishment. For a recent attempt to deal with this last issue, see T. Pummer (2013). “Does Division Multiply Desert?” Philosophical Review 123(1): 43–​77.

Retributive Justice   183 offender, try and convict him, and then punish him. These things can be achieved only by sacrificing time, energy, and tangible resources, and these sacrifices can be translated into sacrifices of pleasure, welfare, or resources. The question then becomes, how do we weigh the gain in one good (retributive just­ ice) against the loss in the other good, whatever that is? That question arises whether we view retributive justice to be subsumed under distributive justice as a good up for just distribution, or whether we view it as a competitor of distributive justice and need a way to balance gains or losses of retributive justice against gains or losses of other goods. There is no obvious common currency other than the abstract currency of intrinsic goodness that these can be reduced to, for the goods of pleasure, welfare, or resources were supposed to be the basic currency of distributive justice, and nothing appears to be more basic than these. Yet it is hard to imagine how the good of retributive justice can be reduced to pleasure, welfare, resources, or any other candidate for the currency of distributive justice. Of course, if one is already a pluralist about the currency of distributive justice and believes that there are several incommensurable goods up for distribution, having to trade off these goods against another incommensurable good—​meting out retributive justice—​might be met with indifference. Dealing with incommensurable goods is difficult, one might say, whether there are two, three, or twenty such goods. Adding retributive justice to the mix does not make matters worse than they already are. For again, pursuing retributive justice is costly in terms of pleasure, welfare, or resources. I think, then, that if one believes in retributive justice, one has to be a pluralist about the goods which are the subject of distributive justice, as the good of retributive justice must somehow mesh with the goods of distributive justice. Again, this is true whether retributive justice is a good subsumed under distributive just­ ice or a competitor that must be weighed against the claims of distributive justice. Pluralism makes matters difficult for any theory of distributive justice, at least if it is supposed to provide a policy guide. I happen to be one who thinks that even apart from the problem of retributive justice, the most plausible version of distributive justice will posit a plurality of goods, rather than a unitary good such as pleasure, welfare, or resources. (I actually believe that any plausible account of how “resources” should be measured will inevitably reduce resources to either welfare or pleasure or else itself be pluralistic [Alexander and Schwarzschild 1986].) But even if one rejects this, introducing retributive justice as a good will perforce make one a pluralist.11

11 

There is one theory of distributive justice that could mesh with retributive justice without endorsing pluralism beyond that introduced by overly strong retributive justice, and that is a theory that treats distributive justice as the mirror image of retributive justice. On such a theory, retributive justice distributes pain in proportion to viciousness, and distributive justice distributes pleasure in proportion to virtue. I discuss such a theory below (see section 8.4.3).

184   Larry Alexander

8.4  Meshing Retributive and Distributive Justice Can we achieve retributive justice without sacrificing distributive justice? In other words, are retributive and distributive justice jointly realizable? That is the question that will occupy me for the remainder of this chapter. Here is how I will proceed. I will first assume that the correct version of distributive justice is an entitlement version, such as Nozick’s (1974), with no particular pattern of distribution of goods required. I will argue that there are real problems reconciling retributive justice with such an entitlement version of distributive justice. That is because retributive justice works with deserved suffering, whereas an entitlement theory allows for undeserved suffering to occur without remediation. I will then turn to patterned theories of distributive justice and point out the problems reconciling them with retributive justice. I conclude inconclusively, except for one point about which I am certain: reconciling retributive and distributive justice is difficult.

8.4.1 Retributive Justice in an Entitlement World Suppose an entitlement theory of distributive justice is correct. Our holdings, great or small, can be just whether they are equal, unequal, or conform to any other pattern, and whether we have been lucky or unlucky, or virtuous or vicious. I need not examine the details beyond this. What is important is that no injustice is shown merely because the poor have been unlucky and the wealthy lucky, or because the vicious, not the virtuous, are prospering. The question I wish to raise here is whether retributive justice based on deserved suffering can fit into a scheme of distributive justice that is based on entitlements that are not based on desert, and that allows for great disparities of fortune and misfortune among the otherwise equally deserving. Suppose then that Tom and Tina both commit offenses that merit fifty units of suffering as their retributive desert. Tom is a wealthy celebrity, and one month in jail, with the humiliation and degradation that Tom will suffer along with the loss of freedom, is sufficient to produce the fifty units. Tina, on the other hand, has been dealt a tough hand by life. She is used to squalid living conditions, and she is an unknown to the public and will suffer no public humiliation by going to jail. It will take six months in jail to cause her fifty units of suffering. A short sentence for Tom and a lengthy one for Tina will bring about the retributively deserved suffering for each. Yet, it seems perverse to let the already very fortunate Tom out of jail five months sooner than hard-luck Tina. Have Tina’s undeserved hardships hardened her and thereby, ironically, made her liable to a longer jail term than Tom’s? Has Tom’s undeserved good fortune softened him such that even a short stint in jail, which will be very painful, is the most punishment he should receive?

Retributive Justice   185 Now one response to this apparent perversity is to attribute it to the use of jail time as the method of punishment. Suppose that instead of being sent to jail, offenders were subjected to corporal punishment. And suppose further that we could measure the suffering that corporal punishment was causing each offender, so that we could tell exactly when Tom and Tina had suffered fifty units. Would we not then have meted out to each his or her desert-​based punishment without producing perverse results and despite their otherwise quite unequal and non-​desert-​based entitlements? Here is why I think the problem remains. Let us suppose that Tom and Tina suffer the same amount from the same amount of corporal punishment. Fifty lashes produces fifty units of suffering for both. Now suppose that Tina had previously been convicted of a similar offense and had been given fifty lashes. It then is discovered that she was innocent of that offense. In other words, she received fifty units of undeserved suffering through a botched attempt at retributive justice. Now, however, she is guilty of a similar offense and deserves fifty units of suffering for it. Yet it would only be fitting—​just—​ if she were fully compensated for the previously undeserved suffering inflicted on her. And the compensation that would be “full” would be to cancel the present punishment. For Tina has already been punished the amount she now deserves (Alexander 2013).12 Now let me alter this hypothetical. Let us suppose that Tina had not been erroneously convicted and punished previously. Instead, some vigilantes believed she had committed the offense in question, and they captured her and subjected her to fifty units of suffering. The vigilantes have disappeared, so Tina cannot be compensated by them. She has again already been “punished” for the crime of which she is now convicted, though this time the punishment was not inflicted by the state. Does it matter that the state that now seeks to punish Tina was not the entity that previously punished her undeservedly? I find it difficult to believe that it does. For if Tina does not now deserve fifty units of suffering when the state previously inflicted an undeserved fifty units of suffering on her, that must be because she has already undeservedly suffered and not because it was the state that caused it. And the state is just us acting collectively. Does it matter that Tina not only did not deserve the suffering inflicted by the vigilantes and by the state, but that she was entitled to better treatment by them? One might think so. But here is why I doubt that Tina’s entitlements and not her desert are what matters. For suppose Tina has been correctly convicted of the offense and now deserves fifty units of suffering. On the way to receive her fifty lashes, a tree branch falls on her, causing fifty units of suffering. Indeed, it causes exactly the same kind of suffering as the fifty lashes would have caused. Should Tina still receive the fifty lashes, or has she already been punished enough? Is it really necessary for an official of the state to administer the suffering if nature intervenes to administer the same amount? What 12  For further discussion of the problems raised in this section, see G. Ezorsky (1972). “The Ethics of Punishment,” in G. Ezorsky (ed.) Philosophical Perspectives on Punishment (Albany: State University of New York Press): xi–​xxvii; A. Kolber (2009). “The Subjective Experience of Punishment.” Columbia Law Review 109(1): 182–​236; A. Kolber (2012). “Unintentional Punishment.” Legal Theory 18(1): 1–​29.

186   Larry Alexander would be the point? The state can say to Tina, “You got what you deserved, and Nature has spared us the necessity of doing it ourselves.” Or suppose the state has a machine that administers the fifty lashes. Normally, a state official turns on the machine. But in Tina’s case, a tree branch fell on the “on” switch and activated the machine, which gave Tina fifty lashes. Would Tina still deserve another fifty lashes because Nature, not an official, activated the machine? You can see where this is going. For if the falling tree branch can justify remitting Tina’s sentence, why does not her life of undeserved hardships do the same? If in her life, Tina has accumulated a large credit balance with respect to desert, why does this not offset the negative desert of her offense? It seems to me that it should. Or at least I can think of no good reason why the negative desert one has incurred from a particular act should be viewed in isolation from all other desert bases in one’s life. The implications, however, are surely unsettling. Past undeserved suffering amounts to a “get-out-of-jail-free” card. On the other hand, in order to reject that implication, one must reject retributive desert or non-​desert-​based entitlement theories or both, or so it appears to me.

8.4.2 Retributive Justice in Non-​Desert-​Based Patterned Theories of Distributive Justice 8.4.2.1 Non-​Responsibility-​Based Patterned Theories Let us assume that the correct theory of distributive justice is a patterned one, the criteria of which make no reference to responsibility or desert. Utilitarian, total or average, egalitarian, prioritarian, maximin, and other similar theories of distributive justice that dictate distributing goods so as to achieve the requisite end-​state without regard to people’s choices or their positive or negative desert fall into this category. I can be brief here. These theories cannot accommodate retributive justice. If giving an offender the punishment he deserves will not maximize utility, will result in inequality, or will fail to maximize the position of the worst off, then these theories will rule it out. Desert-​based punishment has no place in non-​responsibility-​based patterned theories of distributive justice. This should be obvious in the case of the utilitarian. Utilitarians prescribe punishment and its severity, as with every other policy, based on what will maximize utility. Where utility conflicts with desert, utility trumps. And the same is true for egalitarians: equality trumps desert. Rawlsians have written little about criminal punishment, and Rawls himself argues against desert-​based distributive justice and possibly against the reality of desert (Rawls 1971: 103–​4, 310–​15). Samuel Scheffler interprets Rawls as denying that distributive just­ ice can be based on desert but not denying the reality of desert or its relevance to distributive justice (Scheffler 2000). Scheffler himself endorses this asymmetry between the role of desert in retributive justice and its role in distributive justice. Distributive justice,

Retributive Justice   187 unlike retributive justice, cannot be based on individuals’ personal characteristics, for this would be to ignore such relevant factors as the level of material wealth and others’ desires and contributions (Scheffler 2000: 984–​5). There cannot be prejusticial distributive desert, meaning that just distributions are logically prior to questions of distributive deservingness (Scheffler 2000: 985–​6). Distributive justice must be comparative or what Scheffler calls “holistic,” unlike retributive justice (Scheffler 2000: 986–​7). Distributive justice operates in a condition of scarcity, whereas retributive justice does not (Scheffler 2000: 986). This asymmetric position, however, does not elide the problem of determining the retributive desert of those whose overall condition is not desert-​based. If the positions of the least and most advantaged members of society are just but are underserved—​because there is no prejusticial desert—​then should the bad fortune of the least advantaged and the good fortune of the most advantaged count in determining the punishments they deserve? The asymmetry of retributive and distributive desert seems just as daunting on Rawls’s and Scheffler’s accounts as it does on pure entitlement accounts such as Nozicks’s. 13

8.4.2.2 Responsibility-​Based Patterned Theories of Distributive Justice: Luck ​Egalitarianism Luck-​egalitarian theories of distributive justice would seem to be more hospitable to retributive justice. These theories seek to equalize the effect of “brute luck” or chance on people’s distributive shares but allow “option luck” or choice to enhance or diminish those shares. Retributive justice surely focuses on the choice side of the ledger and thus, at first glance, would seem compatible with luck ​egalitarianism. Now luck  egalitarianism is beset by various difficulties having to do with how to divide up one’s fortune between that which is due to chance or brute luck and that which is due to choice or option luck. What if, due to one’s genes and environment, one is a bad chooser? What if one takes what appears to be a tiny, prudent risk and things turn out horribly bad? Choice or chance? What if one takes what appears to be a highly imprudent risk and comes out smelling like roses? Choice or chance? The difficulties are dizzying.14 13  Scheffler’s asymmetric account is criticized by Jeffrey Moriarty. F. Moriarty (2003). “Against the Asymmetry of Desert.” Nous 37(3): 518–​36. Scheffler later softened his denial of prejusticial distributive desert. S. Scheffler (2003). “Distributive Justice and Economic Desert,” in S. Olsaretti (ed.) Desert and Justice (Oxford: Oxford University Press), pp. 69–​91. Both Scheffler and Moriarity believe that economic desert must be comparative, as does Hurka (see Hurka 2003, n. 1, at 56–​68). However, I will argue below that comparative desert is equally problematic in the distributive realm as in the retributive realm (see section 8.4.3). 14  I take the position that, apart perhaps from quantum mechanics, and leaving aside the issue of whether human choices are determined, when we speak of “risks” that a person chooses, we are using an epistemic notion, not an ontic one. Risk is an assessment of the probability of an outcome based on a limited amount of information. With full information—​that possessed by God or Laplace’s demon—​ the risks of an event are either one or zero, depending on whether the event will or will not happen. Frequency accounts of risk are not opposed to this view. They are based on arbitrary selections of reference classes, and they do not pertain to single events. If one takes as a reference class fair flips of a

188   Larry Alexander For our purposes here, however, these difficulties are somewhat beside the point. For what is crucial is that although choice is necessary for retributive desert, the results of choice are not coextensive with desert. The implications of this for fitting retributive just­ice into a luck-​egalitarian scheme will become clear in a moment. Assume, for example, that Al attempts to kill Betty but fails. Indeed, not only is Betty not harmed or even frightened, but Al’s misaimed bullet unearths a treasure buried on Al’s property. Al’s choice was lucky for both Al and Betty. If, as luck ​egalitarianism would have it, Al should both suffer the consequences and reap the rewards of his choices, then Al should get to keep his riches. The consequences of his vicious choice were good for all concerned. Yet in terms of retributive desert, Al should be made to suffer greatly. Likewise, suppose Betty slaps Al, intending only to cause him a little pain. Unbeknownst to Betty, Al has an “eggshell skull” and is hospitalized for months as a result of her slap. Her retributive desert is slight. But the luck egalitarian might require her to compensate Al, which would cause her to suffer in excess of her retributive desert. Now the luck egalitarian might contend that the results of Al’s and Betty’s choices are to be put on the chance side of the choice–​chance divide. Although Al chose to fire at Betty and Betty chose to slap Al, the consequences themselves were not chosen. Therefore, giving Al and Betty their retributive deserts can be consistent with luck ​egalitarianism, rightly conceived. I will not argue the point, except to note that this argument reveals the previously noted difficulty of separating choice and chance.15 Unless the choice–​chance distinction is reduced to a question of what the chooser deserves—​in which case responsibility-​ based versions of luck egalitarianism collapse into the desert-​based versions of distributive justice that I take up next—​there will be a gap between what the chooser deserves and what distributive share luck egalitarianism will allocate to him. Retributive and distributive justice will dictate different answers to what people are due. coin, then the risk of heads will be fifty percent if one has no information other than it’s a fair flip. On the other hand, a physicist who knows the weight distribution of the coin, its orientation and height above ground when flipped, the force and angle of the flip, the barometric pressure, etc., etc., may assign a much higher or lower probability to heads than fifty percent. And God, with total information, would assign “probabilities” of one or zero. Therefore, a “prudent” risk is one that appears so to the actor, given what the actor knows. A different person, possessing different information, might not deem the risk to be prudent. The only truly prudent risks to take are those that turn out well. Quite obviously, however, the luck egalitarian cannot operate with the latter notion of prudent risk-​taking. Nor can he avoid assigning risks that were assessed to be prudent by the actor, but that turn out to be imprudent, to the chance or brute luck side of the ledger, rather than to the choice or option luck side. And ex post, because the outcome will either be higher (one) or lower (zero) than the risk he estimated, the actor will either have suffered brute bad luck if the outcome is bad or brute good luck if it is good. No outcome will ever correspond to the “risk” the actor perceived. For related objections to the distinction between option luck and brute luck, see L. Temkin (2011). “Justice, Equality, Fairness, Desert, Rights, Free Will, Responsibility, and Luck,” in C. Knight and Z. Stemplowksa (eds) Responsibility and Distributive Justice (Oxford: Oxford University Press), pp. 51–​76, 64–​5; G. Sher (2003). “Effort and Imagination,” in S. Olsaretti (ed.) Desert and Justice (Oxford: Oxford University Press), pp. 205–​17; R. Arneson (2003). “The Smart Theory of Moral Responsibility,” in S. Olsaretti (ed.) Desert and Justice (Oxford: Oxford University Press), pp. 233, 252–​3. 15 

See arguments in n. 14.

Retributive Justice   189 Luck ​egalitarianism’s choice–​chance distinction does not mesh with retributive desert because the distinction between choice and chance is not a distinction in terms of desert (Knight 2011: 152–​73). What we need is a desert-​based patterned theory of distributive justice.

8.4.3 Retributive Justice in Desert-​Sensitive Patterned Theories of Distributive Justice The easiest way to make retributive and distributive justice compatible would appear to be making one’s distributive shares reflect one’s desert. Those who deserve punishment will, if they receive that punishment, have their overall level of distributive goods reduced to the level they deserve. Distributive and retributive justice reconciled! Things are not quite so neat, however. For if people’s negative desert should reduce their distributive shares, should not their positive desert increase their distributive shares? (Matravers 2011: 136–​51). Is desert a continuum from the terrible suffering that horrendous wrongs merit to the vast riches that heroic supererogatory acts merit? Or are positive and negative desert asymmetrical, so that wrongful acts merit suffering but virtuous acts do not merit rewards, at least as a matter of justice? Consider, first, the problems of treating positive and negative desert as symmetrical and on a continuum.16 As I have asserted, though without a complete supporting argument, retributive desert is best seen as noncomparative. The amount of suffering murderers and thieves deserve does not turn on the wealth of the society or other contingencies of time and place. So if positive and negative desert are symmetrical, the rewards merited by virtuous acts are also noncomparative and independent of the contingency of the society’s wealth. Now, meting out noncomparative negative desert, while not cost-​free, is not terribly costly in terms of resources. A lot of suffering can be inflicted for very little cost. (Corporal punishment is a good example.) On the other hand, noncomparative positive desert can be costly—​far costlier perhaps than a society can afford. Think of what a person deserves, positively, for, say, falling on a hand grenade to save several people. If, fortunately, the grenade fails to explode, the heroic act surely appears to merit a huge reward. But if the hero’s society is relatively poor—​or there are several such heroes to reward—​the reward justice dictates may exceed society’s wealth or at least result in a severe diminution of the wealth of those whose overall desert is at a neutral point on the desert continuum—​either they have done little to merit either punishment or reward, or the punishment they deserve 16 

Although some might argue that retributive desert is based on vicious choices, distributive desert need not be based on virtuous choices. Therefore, negative and positive desert need not be symmetrical and on a continuum. It is natural, however, to assume that if vicious choices are the desert base for punishment, virtuous choices are the desert base for rewards. I will nonetheless discuss the possibility that positive and negative desert are asymmetrical.

190   Larry Alexander and the rewards they deserve balance out. Indeed, even if there are no saints or heroes—​ everyone is moderately virtuous—​if the society is poor, everyone’s distributive share may be less than he or she noncomparatively deserves. Now it just may be the case that in virtuous but poor societies, no one has the distributive share that he or she deserves. And in wealthy but nonvirtuous societies, everyone has more in terms of distributive shares than they deserve. But contemplating these possibilities, as well as the enormous difficulty of determining what someone deserves absolutely (noncomparatively) for virtuous acts, has led most theorists to reject noncomparative distributive desert in favor of comparative distributive desert.17 Comparative distributive desert faces the same daunting problems that comparative retributive desert faces, however. For example, take two people of equal native talent who put in identical effort in order to contribute to their society’s wealth. They would appear to be equal in distributive desert. But one lives in an advanced society and one in a primitive one. Their resulting contributions will vary considerably, as will the wealth available to reward them. If they are equal in terms of desert but unequal in terms of reward, which one, if either, is getting what he deserves? Do we deserve to live better than our cavemen ancestors? The destitute in North Korea? Do we deserve to have less than the fabulous wealthy inhabitants of a distant planet, or than what we hope our descendants will have? There is no coherent rationale for confining comparative desert to comparisons within one society at one moment in time. (My armchair diagnosis for why champions of comparative distributive desert confine it to a given society at a given historical moment is that they conflate desert and fairness; but people can be treated fairly or unfairly with respect to the division of undeserved gains and losses.) If noncomparative distributive desert is the frying pan, comparative distributive desert is the fire. But if we jettison distributive desert altogether, we are back to the problem of how to insert retributive desert into a system in which people’s fortunes are not determined by their desert. So we must continue to investigate the possibility of distributive desert that can be integrated with retributive desert. Beyond the general problem of determining one’s distributive desert lies the problem of circumstantial luck. Suppose the society contains several would-​be heroes, people who would have jumped on a grenade to save their comrades or who would have made other great sacrifices to benefit others had the opportunity arisen. But no such opportunities for great supererogation ever arose for these would-​be heroes. Nevertheless, given that luck should not affect desert, do these people not deserve large distributive shares because of their willingness to engage in acts of great supererogation?

17 

Hurka and Shelly Kagan both accept the idea of noncomparative (absolute) overall desert (Hurka, n. 1, at 45–​57); S. Kagan (1999). “Equality and Desert,” in L. P. Pojman and O. McLeod (eds) What Do We Deserve? A Reader on Justice and Desert (New York: Oxford University Press), pp. 298–​314, 300. Scheffler rejects noncomparative distributive desert. Scheffler, note 14, at 83. I noted the difficulty of determining noncomparative retributive desert—​what a given culpable act deserves everywhere and at all times. But compared to determining noncomparative distributive desert—​what a given virtuous act deserves everywhere and at all times—​determining noncomparative retributive desert seems at least manageable.

Retributive Justice   191 Circumstantial luck should not affect negative desert. We may think Joe, had he not been born to great wealth, would have been a thief. We may think that Flo, had her husband not been faithful, would have murdered him. But Joe does not deserve to be punished as a thief if he has not, in fact, stolen anything. And Flo does not deserve to be punished as a murderer if she has not killed anyone. Perhaps we reject the relevance of circumstantial luck to negative desert because we do not believe there is a fact of the matter about what someone would have done in circumstances that never arose. Does this reason for rejecting the relevance of circumstantial luck to negative desert apply equally to positive desert? Perhaps. But if Al’s wealth is going to be reduced in order to fund a reward for Alice for her supererogatory acts, is there not force to Al’s complaint that had he been in Alice’s circumstances, he would have acted similarly?18 No one suffers if would-​be malefactors are spared punishment because of their circumstantial luck. Those who commit malevolent acts can be given their noncomparative negative desert, which, given that the desert is noncomparative, does not depend upon how others are treated. And no one else has a complaint about would-​be, but not actual, malefactors escaping punishment. The situation is different with positive desert. Would-​be but not actual committers of supererogatory acts will have their wealth reduced to reward those whose circumstantial luck presented them with opportunities to commit those supererogatory acts. Perhaps, then, we should regard negative and positive desert as asymmetrical. Wrongdoers deserve punishment based on their wrongful choices, but saints and heroes do not deserve rewards based on their saintly and heroic choices if those rewards would make their distributive shares larger than the shares of those of similar virtuous character, but who have not had the opportunities to display their virtue. Such a view might eliminate the problems of the symmetrical, continuum view of positive and negative desert: the limited resources for rewarding supererogatory acts that, as a noncomparative matter, might deserve a huge amount, and the possible unfairness of having circumstantial luck provide only some of those of virtuous character with the opportunities to achieve positive desert. But that would depend on whether the claims of all those with virtuous character would or would not outstrip the available resources. Is, however, an asymmetrical view of positive and negative desert a tenable one? Is having the vicious suffer a matter of justice but having the virtuous prosper beyond justice’s province? A point in favor of asymmetry is that symmetry suggests a moral ledger notion of desert, with supererogatory acts as credits and wrongful acts as debits. And this in turn

18  Larry Temkin appears to answer this question affirmatively. For him, it is one’s character, one’s willingness to do good deeds, that determines one’s deservingness (Temkin 2011: 54). See also P. Vallentyne (2003). “Brute Luck Equality and Desert,” in S. Olsaretti (ed.) Desert and Justice (Oxford: Oxford University), pp. 169–​86, 173–​4. Perhaps the asymmetry can be explained by the fact that negative desert typically can be avoided by omitting to act and does not depend upon opportunities, whereas positive desert does depend on opportunities if it depends upon acts rather than the willingness to act.

192   Larry Alexander suggests, counterintuitively, that supererogatory acts can function as “get-out-of-jail-free” cards. Commit enough of them, or commit particularly heroic ones, and one can go on a crime spree with impunity. To reject this is to favor asymmetry. Another point in favor of asymmetry is that rewarding supererogatory acts might threaten to undermine their supererogatory quality. At least this is true for those supererogatory acts that do not risk death or serious bodily injury—​harms that no ex post rewards can compensate. It is difficult to be an altruist if one’s altruism is commensurately rewarded. Still, the symmetry view, that virtue and vice are on a continuum and commensurable, has some force. One point against the asymmetrical view of positive and negative desert and in favor of symmetry is that we commonly think that a record of altruistic acts should mitigate an otherwise deserved punishment. This suggests that positive and negative desert are at least comparable and perhaps continuous. Additionally, it seems natural to suppose that if justice requires that the vicious suffer, it must also require that the virtuous prosper. It may be that the problems of limited resources and of circumstantial luck can be surmounted, in theory at least, if not in practice.

8.5 Summing Up Here is where we are. I have shown that meting out deserved punishment, which is what retributive justice consists in, meshes uneasily with standard accounts of distributive justice. If the latter is an entitlement account, then the problem is how to insert deserved suffering into a system in which undeserved suffering does not itself amount to an injustice. If the account of distributive justice is a non-​responsibility-​based patterned one, such as utilitarianism, egalitarianism, or prioritarianism, then retributive justice has no place in the account. There may be good consequentialist reasons to punish on such accounts, for punishing some people might help realize the theory-​dictated end state. Nevertheless, such punishment would not be premised on the person’s deserving it. It might be harsher than the person deserves or less than he deserves. Or the person may not deserve punishment at all. Consequentialist punishments are not retributive and thus not within retributive justice. They are more akin to protective measures, such as electric fences, razor wire, guard dogs, and moats, that make attempting criminal acts more difficult or dangerous. That is the aim of purely consequentialist punishment. On responsibility-​ based patterned theories of distributive justice, the concern is whether the person’s option luck (choice) or brute luck (chance) is in play. But although those who deserve punishment have chosen to offend, what they deserve is independent of the results of their choice. Seriously wrongful choices, those that deserve severe measures of comeuppance, may end up producing beneficial consequences for all, whereas minimally wrongful choices may produce disastrous results.

Retributive Justice   193 Responsibility-​based patterned theories of distributive justice, therefore, are at odds with punishments based on desert. The only account of distributive justice that appears to mesh with retributive justice is one that is at least in part desert-​based. But there are problems if not only negative desert is to be punished but positive desert is to be rewarded, problems of resource limits and of circumstantial luck. Yet treating positive and negative desert asymmetrically is itself difficult to justify. My conclusion, therefore, is that hard theoretical work needs to be undertaken if we are to have a plausible account of the relation between retributive and distributive justice. I surely have not given such an account in this chapter. What I hope I have done is to have  identified the various minefields that a satisfactory account must avoid and the issues it must resolve.

References Alexander, L. (2011). “Michael Moore and the Mysteries of Causation in the Law.” Rutgers Law Journal 42(2): 301–​14. Alexander, L. (2013). “You Got What You Deserved.” Criminal Law and Philosophy 7(2): 309–​19. Alexander, L. and Ferzan, K. K. (2009). Crime and Culpability:  A Theory of Criminal Law. New York: Cambridge University Press. Alexander, L. and Ferzan, K. K. (2012a). “Ferzander’s Surrebuttal.” Criminal Law and Philosophy 6(3): 463–​5. Alexander, L. and Ferzan, K. K. (2012b). “‘Moore or Less’:  Causation and Responsibility.” Criminal Law and Philosophy (6)(1): 81–​92. Alexander, L. and Schwarzschild, M. (1986). “Liberalism, Neutrality, and Equality of Welfare versus Equality of Resources.” Philosophy & Public Affairs 16(1): 85–​110. Arneson, R. (2003). “The Smart Theory of Moral Responsibility,” in S. Olsaretti (ed.) Desert and Justice. Oxford: Oxford University Press, pp. 233, 252–​3. Avraham, R. and statman, D. (2013). “More on the Comparative Nature of Desert: Can a Deserved Punishment be Unjust?” Utilitas 25(3): 316–33. Duff, A. (2001). Punishment, Communication, and Community. New York: Oxford University Press. Ezorsky, G. (1972). “The Ethics of Punishment,” in G. Ezorsky (ed.) Philosophical Perspectives on Punishment. Albany, NY: State University of New York Press, pp. xi–​xxvii. Hurka, T. (2003). “Desert: Individualistic and Holistic,” in S. Olsaretti (ed.) Desert and Justice. Oxford: Oxford University Press, pp. 45–​68. Husak, D. (1987). Philosophy of Criminal Law. New York: Oxford University Press. Husak, D. (2008). Overcriminalization. New York: Oxford University Press. Kagan, S. (1999). “Equality and Desert,” in L. P. Pojman and O. McLeod (eds) What Do We Deserve? A  Reader on Justice and Desert. New  York:  Oxford University Press, pp. 298–​314, 300. Knight, C. (2011). “Responsibility, Desert, and Justice,” in C. Knight and Z. Stemplowksa (eds) Responsibility and Distributive Justice. Oxford: Oxford University Press, pp. 152–​73. Kolber, A. (2009). “The Subjective Experience of Punishment.” Columbia Law Review 109(1): 182–​236.

194   Larry Alexander Kolber, A. (2012). “Unintentional Punishment.” Legal Theory 18(1): 1–​29. Markel, D. and Flanders, C. (2010). “Bentham on Stilts: The Bare Relevance of Subjectivity to Retributive Justice.” California Law Review 98(3): 907–​88. Matravers, M. (2011). “Mad, Bad, or Faulty?” in C. Knight and Z. Stemplowksa (eds) Responsibility and Distributive Justice. Oxford: Oxford University Press, pp. 136–​51. Moore, M. (1997). Placing Blame. New York: Oxford University Press. Moriarty, F. (2003). “Against the Assymmetry of Desert.” Nous 37(3): 518–36. Nozick, R. (1974). Anarchy, State, and Utopia. New York; Basic Books. Pummer, T. (2013). “Does Division Multiply Desert?” Philosophical Review 123(1): 43–​77. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Scheffler, S. (2000). “Justice and Desert in Liberal Theory.” California Law Review 88(3): 965–​90. Scheffler, S. (2003). “Distributive Justice and Economic Desert,” in S. Olsaretti (ed.) Desert and Justice. Oxford: Oxford University Press, pp. 69–91. Sher, G. (2003). “Effort and Imagination,” in S. Olsaretti (ed.) Desert and Justice. Oxford: Oxford University Press, pp. 205–​17. Temkin, L. (2011). “Justice, Equality, Fairness, Desert, Rights, Free Will, Responsibility, and Luck,” in C. Knight and Z. Stemplowksa (eds) Responsibility and Distributive Justice. Oxford: Oxford University Press, pp. 51–​76. Vallentyne, P. (2003). “Brute Luck Equality and Desert,” in S. Olsaretti (ed.) Desert and Justice. Oxford: Oxford University Press, pp. 169–86.

Chapter 9

The Go od S o c i et y Steven Wall

In a good society people pursue worthwhile goals, engage in valuable activities, realize virtue, support excellence, and have fitting attitudes toward objects of intrinsic value. A good society has just political and social institutions, but its goodness includes more than its justice. The question that motivates this chapter is, how should considerations of the good, including both the human good and the impersonal good, inform our thinking about distributive justice in political societies of the sort we are familiar with in the modern world? My discussion of this question takes for granted several background assumptions. First, I  assume that modern societies, and their political institutions, should promote, actively and intentionally, the good of their members. Second, I assume that there is a plurality of goods and a plurality of ways of living a good human life that can be realized in these societies. And, third, I assume that the goodness of a human life depends, in part, on whether it is the product of autonomous engagement with the goods that it participates in or realizes. The first assumption sets a general goal for politics, whereas the second and third assumptions have implications for how the goal should be pursued. Just political societies, I am assuming, aim to promote the good of their members. If this is correct, then considerations about good human lives are integral to determining the content of justice. Discussions of justice—​or, more precisely, distributive justice—​ often1 focus narrowly on the goods and resources that are produced and distributed by social cooperation within a political society. These goods and resources, it is often said, should be distributed in a manner that gives each member of the society a fair chance to pursue his own conception of the good life. In light of its background assumptions, this chapter departs from these common views. It argues that people do not have a claim to a fair chance to pursue their own conception of the good life, but rather a claim to have a fair chance to pursue a sound or worthwhile conception of the good life. It also

1 

Often, but not always. Those who take welfare, or opportunity for welfare, as the currency of distributive justice take a broader view of its scope.

196   Steven Wall rejects the narrow focus on goods and resources and argues that environmental goods, including the makeup of the moral and cultural environment of the society, are also elements of advantage for distributive justice. The realization of justice contributes to the goodness of a society. But there are other goods to consider as well. A good society not only promotes good human lives, but also supports impersonal goods. By impersonal goods I mean to refer to goods, such as the achievement of excellence in artistic and scientific endeavors, that have a value that is not exhausted by their contribution to the goodness of human lives.2 This chapter argues that the effective promotion of impersonal goods often requires that resources and opportunities be distributed in a manner that favors those with special talents. This kind of favoritism is not unjust, since all members of a political society have reason to promote the impersonal good. But while it is not unjust, it is not required by justice either. A good society realizes goods beyond the good of justice. The background assumptions of this chapter are controversial. No real effort is made to defend them here. They constitute a limitation in the scope of the argument I advance. Nevertheless, in sections 9.1 and 9.2, I clarify the assumptions and try to put them in an attractive light. The argument of the chapter overall is cumulative. Earlier sections raise problems and concerns that are addressed in later sections.

9.1  Pluralism and the Human Good Just political societies favor sound conceptions of the good and promote virtues that are important to living well.3 To be sure, there are limits to politics. A society can have well-​ designed political institutions and yet its members, or many of them, can fail to lead good lives. They may pursue misguided conceptions of the good or lack virtues that are necessary for living well, for example. That is one reason why a fully just society could fail to be a fully good society. Over and above its concern with the good of its members, a good society respects and promotes impersonal goods. But justice does not require that the political institutions of a society honor this further commitment. That is a second reason why a fully just society could fail to be a fully good society. A just society promotes the good of its members, or at least strives to do so. How to understand the good of human beings is, of course, one of the deepest questions in philosophy. There is no single plan or way of life that uniquely realizes the human good for people in modern societies—​or so this chapter assumes. But, if so, what explains this pluralism? 2 

In characterizing impersonal goods in these terms, I don’t mean to deny that human life has impersonal value. My characterization is stipulative. 3  A political society could aim to promote the good of its members indirectly. In some circumstances, the best way to promote the good might be for the state to refrain from doing so directly. I mention this possibility to put it to one side.

The Good Society    197 Many philosophers answer this question by appealing to desires. The good for a human being, they say, is a function of what he desires, or would desire if he had adequate information about the different possible objects of his desire. Since human beings desire many different things, this view can explain how there can be a wide plurality of good ways of living. But this response is not the only possible answer to our question. The fact of pluralism can be explained more directly by appeal to the variety of goods that can enrich a human life. Goods such as pleasure, knowledge, friendship, achievement, and political engagement can be realized in myriad ways, and different combinations of these goods can yield lives of roughly equal value. Moreover, since different people have different talents and temperaments, the same pursuits will not contribute to the goodness of their lives in the same way or to the same extent. Objective facts about value, and objective facts about people, in short, can explain the pluralism about the human good. No appeal to the subjectivity of desire is needed. Yet even on the objective view I  have just sketched, it remains an open question whether the human good depends in some way on human desires. Much will turn on the nature of the purported dependence. Since the topic is important for the present discussion, it will be helpful to try to get clearer on it. (In what follows I understand ‘desire’ very broadly as encompassing a range of pro attitudes including valuings and carings as well as what are more naturally thought of as desires.) A strong desire-​dependence condition holds that if a person desires something, the satisfaction of the desire promotes his good.4 This condition is inconsistent with the objective view I have sketched. But a weaker desire-​dependence condition is available. Even if the fact that a human being desires something does not make it true that its attainment would make his life go better, it may be true nonetheless that nothing can contribute to the good of a human life unless that person desires it. This weak desire-​dependence condition is consistent with an objective account of the human good. Thus an (otherwise) objective account of the human good can be combined with a weak version of the desire-​dependence condition. At the conclusion of his well-​known discussion of what makes a life go better, Parfit hints at a view of this kind: We might claim, for example, that what is good or bad for someone is to have knowledge, to be engaged in rational activity, to experience mutual love, and to be aware of beauty, while strongly wanting just these things. On this view, each side in this disagreement saw only half of the truth. Each put forward as sufficient something that was only necessary. (Parfit 1984: 502)

This hybrid view has clear attractions. It captures the appealing thought that when we engage with valuable objects we respond to the good-​making properties of the objects 4  For the satisfaction of a desire to contribute to the good of a human being certain further conditions would need to be satisfied, for example, the desire must relate to the person’s life in an appropriate way. I ignore this complication here.

198   Steven Wall rather than project our desires and sentiments on to them. At the same time, it excludes the unappealing thought that we could lead good lives by being forced to take up objectively valuable pursuits for which we could see no value or that left us cold. Parfit merely proposes the hybrid view, but others affirm it. Dworkin, for example, claims that there are objectively better and worse ways of leading a life, but he insists that no life can be made better against the grain of a person’s convictions and desires (Dworkin 2011: 195–​ 213). If we follow Dworkin and accept the hybrid view, then we will need to say more about the desire-​dependence that is thereby affirmed. Consider two cases. In case one, a person willingly engages with an objectively valuable activity, but he would not desire to do so under ideal informational conditions. In case two, a person unwillingly engages in the same activity, but he would desire to do so under ideal informational conditions.5 Dworkin claims that we cannot benefit from engaging in an activity if we see no value in it. For example, if we can see no point to religious practice, then we cannot benefit from it, even if engaging in the practice were objectively valuable. Our actual attitudes would bar engagement in the activity in a way that would contribute to our good. If this is right, then the fact that we would desire to engage in the activity under different conditions—​conditions under which we had more relevant information—​would not change the fact that our actual engagement with the activity does not contribute to our good. Generalizing on this point, a plausible version of the hybrid view holds that the good for a human being consists in both engaging with objective value and in actually desiring to do so. The human good, on this view, is dependent on actual human desires, but only insofar as these desires enable people to engage with objectively valuable goods in a manner that allows them to benefit from them. So understood, the appeal to desire plays an ancillary role in explaining pluralism about the human good. Dworkin’s hybrid view affirms weak desire-​ dependence. But weak desire-​ dependence, as so far formulated, may still be too strong. Dworkin advances an uncompromising claim. No activity or pursuit, no matter how objectively valuable, can contribute to the goodness of a human life unless that person desires it or sees value in it. A less austere view is available. Some—​perhaps many—​objectively valuable activities are such that they contribute to the good of those who participate in them only if these people desire to participate in them, or at least believe them to be valuable. On this view, the desire condition is not a strict necessary condition. It allows that an activity could contribute to one’s good, even if one did not desire to engage in it. A still less austere view holds that, while the appropriate desires are never necessary for an objectively valuable pursuit to add value to a human life, they enhance, or often enhance, the contribution that these pursuits make to the goodness of such a life.6 5  The hypothetical desires in these examples could be desires, or endorsements of desires, or both. These different possibilities do not matter for the purposes of the examples. 6  The less austere formulations of the weak desire-​dependence condition allow the hybrid view to dodge some of the well-​considered criticisms pressed against it in Arneson 2006.

The Good Society    199 I have been discussing the dependence of the human good on human desires. Rejecting a strong desire-​dependence condition, I have suggested that a weak desire-​ dependence condition is attractive, and that it is compatible with an objective account of the human good. We need not here decide which formulation of the weak desire-​ dependence condition is best for a hybrid view of the sort that Parfit introduced. If some such condition is sound, then we can see two ways by which a human being can fail to lead a good life. He may fail to realize objectively valuable goods or he may fail to have the desires toward these goods that enable him to engage with them well. Accordingly, in promoting the good, a good society must do more than ensure that its members have access to objectively valuable goods. It must also promote, at least to the extent that it can do so, the appropriate desires towards these goods (at least when the goods are such that desiring them appropriately is necessary to engaging with them well). Expressed differently, a good society will strive to promote sound conceptions of the good life among its members. A couple of final points will help to clarify the understanding of the human good that I am here proposing. First, a good life for a person is one in which he engages with objective goods in the right spirit, but there is no implication that such a life is self-​ centered. The goodness of a life for a person can be augmented by actions that set back his welfare for the sake of others or for impersonal goods.7 Second, an unsound conception of the good for a human being need not be one that involves treating others as inferior. Racist and sexist conceptions of the good are defective because they exclude living with others on equal terms, but one can pursue a false conception of the good, even if one is committed to treating others as free and equal. Like Rawls’ grass counter, one may simply spend too much time engrossed in worthless pursuits. This point deserves emphasis, since many who reject the broad claim that political institutions should promote the human good accept the narrow claim that these institutions should discourage conceptions of the good that are inconsistent with viewing others as free and equal.

9.2  Autonomy and Reason I have claimed that a good society favors sound conceptions of the good, at least to the extent that it can do so. But should not human beings be left free to decide for themselves what is valuable and worth doing? A good society does not coerce or manipulate its members into pursuing the good. Addressing this concern requires us to investigate the nature and value of autonomy. 7  To some ears, this claim will sound odd. The goodness of a life for a person, they will think, just consists in facts about his well-​being. In contrast, I think the goodness of a life for a person consists in facts about how choiceworthy it is (to adopt a term from Scanlon) and these facts include, but are not limited to, facts about the person’s well-​being.

200   Steven Wall This chapter assumes that autonomy is an aspect of the human good. Part of what makes a human life go well is how one pursues the projects that make up one’s conception of the good. If two lives consist of valuable pursuits and experiences, and if these pursuits and experiences are of roughly equal value, but one life realizes autonomy while the other does not, then the former life will be better. For this reason, the good society—​ and its political institutions—​will be autonomy promoting. Some writers characterize autonomy in unappealing terms.8 The autonomous life is presented as highly individualistic, overly rationalistic, or inherently hostile to traditional ways of life. The autonomy I have in mind does not have these implications. It is not an inflamed conception of the ideal. Autonomy requires that one have access to a range of valuable options, that one have the basic capacities necessary to engage with these options, and that one not be coerced or manipulated into pursuing them. It also requires that one take charge of one’s life and not simply submit to the will of others or drift aimlessly from one moment to the next. So understood, autonomy is compatible with deference to authority; for in governing oneself one often will need to take advantage of the expertise of others. Still, a person who lets others make all or most of his major life decisions would not be autonomous. The requirement that one take charge of one’s life gives rise to an interesting challenge to autonomy’s claim to be an aspect of the human good.9 Suppose a person would make better decisions by letting others make his most important life decisions for him. Would it not be better for him to do so? If so, then autonomy may not have the value I am assuming it has. Perhaps reason responsiveness, not autonomous self-​government, is what really matters. On this view, we should care about autonomy, but only to the extent that being autonomous helps us to respond well to the reasons we have. It is worth pausing to consider what might be wrong with this view. If pluralism about the human good is correct, then for each of us there is more than one way of leading a fully good human life. There is no single best path for each of us to choose. This implies that if you defer to others to make all of your major life decisions, then, even if these others choose well for you, they would not simply be enabling you to do what you have most reason to do: they would be exercising discretion in directing you to lead one kind of life rather than another. In submitting to them you would be submitting to their wills and not simply to the discipline of reason. The challenge to autonomy that we are considering—​keep in mind—​contrasts autonomy with reason responsiveness. To realize autonomy we must direct our lives by our own will and not by the will of another. That is why coercion and manipulation pose a special threat to autonomy. But if achieving this kind of independence made us poor at responding to reason, then its value would be questionable. However, assuming pluralism about the human good, there is a role for our will to play in our lives that does not 8  See, for example, the discussion of communitarian, feminist, and post-​modern criticisms of autonomy in Christman and Anderson 2005: 3–​4. 9  The challenge is skillfully pressed in Valdman 2010. For a fuller response to it than is provided here see Wall 2016.

The Good Society    201 consist in its simply accepting or rejecting the demands of reason.10 Imagine a perfectly reason-​responsive agent who confronts a number of valuable career options that are equally good, and for which he is equally well suited. In selecting one over the others, the agent would, at least in this area of his life, lead his life on his own terms. The claim that autonomy is valuable implies that this agent would realize a good that he would not have realized had he simply deferred to the will of another in making this choice. This claim is plausible. Holding other things constant, autonomy adds value to a life. Those who challenge autonomy’s value might concede that living one’s life on one’s own terms contributes to its value. There is, they might say, “intrinsic value in living an acceptable life, where one’s life is acceptable, roughly, if it is guided or shaped by one’s deepest commitments” (Valdman 2010:  774). But then, so the challenge would run, this fact does not yet establish that it is valuable for people to make their own decisions about how to lead their lives. One could delegate decision-​making to an individual or group who was committed to making decisions that were guided by one’s deepest commitments. Here, in deferring to the will of another, one would be deferring to a will that was constrained by one’s own will. This concessive response, I believe, concedes more than it realizes. It helps itself to the thought that people have deep commitments that others must respect in making decisions for them. But how exactly do these deep commitments become a person’s own? The natural answer is that they are the product of that person’s own past decision-​making. If her own deep commitments were implanted in her by others, then we would be hard pressed to explain why others must be constrained by them in making decisions for her now. Why, in short, would living an acceptable life add value to a person’s life now if it were not true that her deepest commitments were the product of her own past autonomous decision-making? In responding to the challenge to autonomy’s claim to be an aspect of the human good, I have emphasized the compatibility of the exercise of autonomy with reason responsiveness. I also have claimed that if we hold constant the objective goods realized in a life, as well as the desires toward these goods, then a life that realizes autonomy is better than a life that does not. Can more be said? Might one life go better than another even though it scored less well in terms of its realization of objective goods (and the appropriate desires toward those goods) if it realized autonomy, whereas the other life did not?11 If it is accepted that autonomy adds value to a life, then this possibility is plausible. If autonomy adds value to a life, it is likely more than a mere tiebreaker between lives that are otherwise equally valuable. The view that has emerged from this discussion is that the goodness of a human life is a function of the objective goods it realizes or participates in (as well as the appropriate

10  On this point, see Raz’s discussion of value and incommensurability  (Raz 1999: Ch. 3). The connection between autonomy and pluralism that I am relying on here also draws on Raz’s discussion (Raz 1986: 395–​9). 11  Since the realization of autonomy itself is an objective good, the formulation of this question is not ideal. By objective goods I here mean to refer to objective goods other than autonomy. This qualification should be borne in mind in the ensuing discussion.

202   Steven Wall desires toward those goods), and the degree to which the life realizes autonomy. Fig. 9.1 depicts how the two variables relate to each other.

Life 1

Goodness Life 2

|Sufficient autonomy Autonomy

fig. 9.1  Autonomy and good lives

The x axis represents the degree of autonomy realized in the life and the y axis represents the goodness realized in the life, which is a function of both the objective goods realized and the desires toward those goods. The curves represent two lives. The height of curve is a strict function of the objective good it realizes. Life 1 realizes more objective good than Life 2, but it is a nonautonomous life. Life 2 realizes less objective good than Life 1, but it realizes autonomy to a sufficient degree. It is possible, I am suggesting, for Life 2 to be overall better than Life 1, even though Life 1 realizes more objective good than Life 2. The graph only very roughly depicts the relative contributions that objective goods and autonomy make to the goodness of a human life. But it conveys several important ideas. First, the realization of autonomy comes in degrees. More autonomy is better, but there is no reason for people to maximize autonomy, assuming that we could make sense of that demand. Second, the graph allows for the possibility that nonautonomos lives are better than autonomous lives. This would be the case if the nonautonomous life contained substantially more objective good than the autonomous life. Autonomy augments the goodness of a human life; it is not a condition of its goodness. Third, the graph represents only objective goods. It does not consider how objective bads—​ engaging in evil or degrading pursuits, for example—​affect the goodness of a life. This simplification allows me to leave to one side the question of whether objective bads diminish the goodness of a life or whether they merely fail to add to its value. The relationship between autonomy and objectively bad pursuits does raise an important issue, however, and it is an issue that brings us back to the concern expressed at the beginning of this section. That concern, to recall, is that a society that aims to promote sound conceptions of the good will fail to respect the autonomy of its members. It runs the danger of manipulating or coercing them into pursuing the good. This concern cannot be dismissed simply by pointing out that the members of a society can promote the good of one another in ways that do not involve coercion and manipulation. This is true, but I have claimed that the political institutions of a good society should promote the good of their members. And many have thought that political action is, at bottom, coercive.

The Good Society    203 Our discussion of autonomy and its value point the way toward a response to the concern. Political action that aims to promote sound conceptions of the good will favor valuable pursuits and discourage worthless ones. In supporting valuable pursuits, political measures promote autonomy by seeking to ensure that all members of the society have access to a wide range of valuable options. In discouraging base pursuits, political measures may set back autonomy, but not the autonomy that contributes value to a human life. For the value of autonomy, on the view I have proposed, is not found in rebellion against reason, but rather in making choices from among a set of reasonable options. The claims just advanced will meet with resistance. Three objections in particular will be pressed. First, I have characterized autonomy as an ideal that adds value to a human life. But autonomy is often understood to be a right, and if it is a right, then it may be wrong to infringe it, even if doing so would advance the good of the person whose autonomy is set back. Second, autonomy, as I have emphasized, is reduced if a person is subjected to the will of another. But political measures that favor some pursuits and discourage others may subject citizens to the will of those in political power. Third, even if political measures can promote valuable pursuits and discourage base ones without setting back the valuable autonomy of anyone, it does not follow that there is no danger in adopting these measures. In particular, given the pluralism about the human good that I have endorsed, there is the danger that these measures will become sectarian, with some valuable pursuits favored over others that have an equal claim to support. The first objection raises large issues about the nature and value of autonomy that cannot be discussed here.12 The second and third objections, however, pinpoint concerns that we need to address. Responding to them requires us to turn our attention now to the topic of distributive justice.

9.3 Advantage A society can fail to be just, and therefore not good in one respect, because of the way that advantages get distributed within it. It is possible that a just distribution of advantage is one that results from voluntary exchange between persons who have justly acquired their goods (Nozick 1974). If so, then a just society respects the entitlements of its members to goods that they have acquired in the right way, and it does not aim to bring about any particular distribution of advantage. But if we put this view to one side—​as I propose to do here—​and consider the possibility that a just society must ensure, or take steps to ensure, that certain distributive outcomes obtain, then we will need to think harder about the nature of advantage for the purposes of distributive justice. If advantage must be distributed in a certain way for the distribution of advantage to be just, then how should it be characterized? 12 

For a response to this objection see Wall 2013.

204   Steven Wall This is the currency question in the philosophical literature on distributive justice. It is distinguished from the question of which distributive principle(s) are correct. I will have nothing to say about this latter question. Obviously, it is important, but my focus will be squarely on the currency question. The term “advantage” is not a neutral one in the currency debate. It contrasts with “resources” and “welfare” in that it is broader than both of them. Two people can be equal in either resources or in welfare or both and yet be unequal in advantage. Cohen, who introduced the term into the debate, claimed that it encompasses an “unlovely heterogeneity” of vectors or components. The pressure to move in the direction of this “unlovely heterogeneity” is that alternative, and more unified, accounts of the currency of distributive justice are undone by compelling counterexamples. Resources are inadequate as a currency, since people who are differently situated need different amounts of resources to achieve the same level of advantage (Sen 1980). Welfare, understood either in hedonic or preference satisfaction terms, is inadequate as a currency, since people who are equal to others in terms of it may not be equally advantaged. One person, for example, may suffer from a disability that warrants compensation or accommodation, even if he is equal in welfare to others (Cohen 1989). Neither the hedonic nor the preference satisfaction conception of welfare is particularly plausible. So counterexamples to welfare construed in these terms may only establish that these are inadequate conceptions of welfare. Moving to a more objective conception of welfare, one more in line with the view of the human good advanced in section 9.1, may enable the welfarist to avoid the objections standardly pressed against the welfare metric. If a person is paralyzed and unable to move around, for example, then even if she enjoys her life or has most of her preferences satisfied, she may still suffer a significant welfare deficit if mobility is an objective component of welfare. To use Sen’s language, she may lack an important capability, and if we care about her welfare, then we will need to address this incapacity. There are two problems with this proposal, however. First, given pluralism about the human good, any objective conception of welfare will need to encompass a wide range of objective goods, and these goods may not be such that they can be ranked or ordered. So, if we are inclined to accept welfare as a currency metric because we seek to avoid the “unlovely heterogeneity of advantage” that Cohen called attention to, then we should not be happy with this proposal, since it, in effect, reproduces the heterogeneity of advantage within the category of welfare. But, second and more interestingly, the proposal should not be accepted, since there is more to advantage than welfare, even on the broadened objectivist conception we are now considering. Two people can be equal in welfare, but unequal in their capacity to realize their rational non-​welfare aims. They can differ in what Sen calls their agency freedom.13 And 13  To clarify: Sen construes agency freedom in terms of the freedom to pursue one’s rational aims, where rational aims include not only aims that further one’s well-​being, but also other aims as well. “A person’s agency freedom refers to what the person is free to do and achieve in pursuit of whatever goals or values he or she regards as important” (Sen 1985: 203). Feeding agency freedom into the metric of

The Good Society    205 agency freedom is plausibly a component of advantage. If this is right, then the notion of advantage should not be identified simply with what is prudentially good. The best life for a person, as I have noted, may be one in which he sacrifices his own welfare, at least to some extent, for the sake of others, or for impersonal goods. A good society promotes good human lives. For this reason, its conception of advantage will reflect this concern.14 The most plausible conception of advantage is not unified. It is sensitive to, among other things, resources, welfare, and people’s prospects for realizing their rational aims. The sources of disadvantage, on this conception, are likewise diverse. People can be disadvantaged because they lack resources [resource disadvantage], or suffer from a welfare-diminishing handicap, affliction, or experience [welfare disadvantage], or live in an environment that makes it impossible, difficult, or costly for them to pursue their rational aims [environmental disadvantage].15 In the remainder of this section I  highlight this third source of disadvantage—​ environmental disadvantage. Doing so will put me in a position to respond to the objections raised above to my discussion of autonomy. I start with an example from Cohen. Paul loves photography, while Fred loves fishing. Prices are such that Fred pursues his pastime with ease while Paul cannot afford to. Paul’s life is a lot less pleasant as a result: it might even be true that it has less meaning than Fred’s does. I think the egalitarian thing to do is to subsidize Paul’s photography. (Cohen 1989: 20)

Paul fares less well than Fred because his tastes are expensive. His tastes are expensive because of the economic environment in which he pursues his pastime. This environment makes photography more expensive than fishing. Let us add that photography and fishing are both valuable pursuits and that neither can be said to be more valuable than the other. Cohen claims that Paul suffers a disadvantage. I agree. Cohen also claims that Paul’s disadvantage entitles him to compensation. On this issue, I do not here take a stand. To determine whether Paul is entitled to a subsidy we must decide whether it is fair to hold him responsible for the disadvantage he confronts. The only point I wish to extract from the example is the modest point that Paul suffers a setback in interests (relative to Fred) in virtue of the fact that his hobby is more costly to pursue than Fred’s hobby, given the advantage thus yields a view that diverges from opportunity for welfare views. Following Vallentyne, we might call such a view the “opportunity for value view” (Vallentyne 2005). 14 

There are complications here that I pass over. The conception of advantage relevant to distributive justice is, in part, a moralized notion. It can diverge from an account of advantage that was concerned only with how human beings fare. That is, in thinking about advantage for distributive justice purposes we are thinking in part about what kinds of claims people can make on one another. For discussion of this point, see Scanlon 1998: 109–​11. 15  The three sources of disadvantage are cross-cutting in various ways. Environmental disadvantage impacts both rational aim satisfaction and welfare, for example.

206   Steven Wall economic environment that both confront. An account of advantage must determine whether this setback registers in its calculation of advantage.16 Environmental disadvantages are not only economic, but also moral and cultural. Each person has an interest in living in an environment that supports valuable pursuits and discourages base ones. This interest follows from the account of the human good discussed earlier. Recall that, on the view there advanced, a good human life consists of objectively valuable pursuits and appropriate desires toward those pursuits. We do not have an interest in pursuing whatever we take an interest in, for we might take an interest in pursuits that have no value at all. Our interest in leading a good life is an interest in pursuing a sound conception of the good, not in pursuing just any conception of the good that we happen to adopt. This fact explains why each of us is advantaged by living in an environment that favors the good over the bad. Consider now a variant on Cohen’s example. In addition to Paul and Fred, we introduce David, whose pastime is centered on a valueless activity, the nature of which can be left to the reader’s imagination. On my view, David is not disadvantaged if the environment in which he lives makes his pastime costly or difficult to pursue relative to Paul and Fred’s pastime. Although he may experience some frustration, which is a cost, he has no genuine interest in pursuing the valueless activity in question. His interest lies in abandoning his pastime and taking up a valuable one, and an environment that makes the pursuit of his pastime more costly will serve this interest. Even if one thinks that Paul deserves compensation (in Cohen’s example), one should not conclude that David does. Suppose, however, that the state has intervened to make David’s pastime more costly than others. In this scenario, David may be able to object that the state disadvantages him by setting back his autonomy. By subjecting him to the will of others, the state thereby diminishes his autonomy, which is, I have assumed, an aspect of the human good. Is this judgment correct? Not without qualification. David’s pastime might involve activities that risk physical harm to others. He might like to hunt game in areas populated by people, for example. If so, then the state should stop him from doing so. True, its action will subject David to the will of others, but the subjection is justified, since the state has a duty to enforce justice. David’s autonomy is not infringed by state action that enforces justice. If we say this, as I think we should, then we need to distinguish subjection to the will of others that compromises autonomy from that which does not. The appeal to duties of justice allows us to do so. Consider a final case. David’s pastime involves activities that threaten no physical harm to anyone, but that degrade the moral and cultural environment in which he lives. Since, as argued above, environmental disadvantages engage justice, and since each of us has an interest in living in a moral environment that supports valuable pursuits and discourages valueless ones, it follows that justice could require the state to intervene so

16 

This is the issue that divides Dworkin from Cohen.

The Good Society    207 as to make David’s pastime, and others like it, more costly or difficult to pursue. When and if this were the case, David’s autonomy would not be compromised.17 He would not be subjected to the will of others in a way that compromised his claim to lead his life on his own terms.18 The argument of the foregoing paragraph relies on the freighted claim that the state has a duty of justice (owed to those who are subject to its authority) to protect the moral and cultural environment from activities that degrade it. This claim requires unpacking. Work must be done to clarify how an activity could degrade the moral environment in a way that would impose an unjust environmental disadvantage on others.19 I will not undertake that work here, since I have wanted only to present the outline of a response to the objection that I raised at the end of section 9.2. That objection, to recall, held that when the state intentionally favors valuable pursuits over worthless pursuits it thereby invades the autonomy of some by subjecting them to the will of others. The response to the objection is that autonomy is bounded by justice, and that if the state has a justice-​based duty to sustain or promote the right kind of moral environment, then its efforts to favor some pursuits over others need not set back the autonomy of its members. An objection remains from section 9.2. As noted, even if political measures can promote valuable pursuits without setting back the autonomy of anyone, it does not follow that these measures are compatible with justice. The measures may arbitrarily favor some valuable pursuits over others that have an equal claim to support. This, in turn, could contribute to an unjust distribution of advantage. People have an interest in pursuing their own conception of the good, provided that it consists of valuable activities and ends. Let us introduce two final characters, Mark and Sally. Mark is committed to a conception of the good that is centered on sports, while Sally is committed to one that is centered on art. If the state favors art over sports, then

17 

It may be objected that we should distinguish duties of justice broadly construed from duties of justice that concern the protection of autonomy. Only the latter duties, it may be thought, are such that when they are enforced no one’s autonomy is set back. On this view, A’s autonomy is not limited if he is forced to respect B’s autonomy. But when A is forced to comply with some other non-​autonomycentered duty of justice, then his autonomy is set back. I do not find this view attractive, but if it were adopted, then one would still face the question of whether people have an autonomy-​based interest in committing injustice that is strong enough to outweigh the wrong they do to others. If one answers no to this question, then one must grant that the state can act rightly when it enforces justice, even when its doing so sets back the autonomy of the wrongdoers. This view concurs with the view I have advanced in the text. There is not a compelling (general) autonomy-​based objection to state efforts to promote sound conceptions of the good. 18  Matters are different if the state is not required to interfere with David’s pursuits. If justice is not engaged, then David may have an autonomy-​based claim to be left free to engage in his worthless pastime. 19  With reference to the specific issue of hard paternalism, Richard Arneson has provided a nice account of how the argument can go. A regime of strict anti-​paternalism creates a moral and cultural environment that predictably advantages those with good choice-​making abilities over those with poor choice-​making abilities. This fact itself presents an issue of distributive justice (see Arneson 1989).

208   Steven Wall Mark will be disadvantaged relative to Sally. To insist that Mark has the same opportunity as Sally to pursue a valuable conception of the good, whether centered on art or sports, would miss the point. The same thought can be expressed differently. If the state sustains a moral environment that favors Sally’s conception of the good relative to Mark’s, then Mark may have a complaint that he has not been given a fair opportunity to pursue his conception of the good, since he suffers an environmental disadvantage that Sally does not. I say he may have a complaint, but he may not. The state could have justice-​based reasons to favor art over sports in the context in question. Environments are public goods (or public bads). And moral and cultural environments, I have been stressing, are not equally hospitable to all conceptions of the good. Perforce some interests will need to be favored over others. The fact that a large number of its members would benefit from one environment may be a sufficient justice-​based reason for the state to favor it, even if some other environment would be better for some of its members. Fairness in this context requires balancing interests, and (plausibly) the numbers matter in determining the fair balance. In discussing Mark and Sally’s rival conceptions of the good, I have been assuming that these conceptions are equally valuable. Let me now drop that assumption. Sally’s conception, let us stipulate, is more valuable than Mark’s. If the state supports Sally’s conception of the good over Mark’s, then it is not arbitrarily advantaging Sally relative to Mark, since it has a reason to favor her conception of the good over his (it is more valuable). Would this reason—​namely, the fact that more valuable pursuits have a stronger claim to support than less valuable pursuits—​silence Mark’s complaint that he is unjustly disadvantaged by the state’s action? Not necessarily. Once again, interests are in competition here. Mark has an interest in pursuing his valuable conception of the good, but the state may have a duty to sustain an environment that favors more valuable over less valuable pursuits. This duty is grounded in the general interest of the members of the state to live in an environment that helps them to lead the best lives that they can lead. Mark’s interest can diverge from this general interest, given Mark’s investment in and commitment to his current conception of the good. When this is the case, Mark’s interest and the general interest must be balanced to determine when, and if, Mark has a complaint. The concern that, if the state aims to promote sound conceptions of the good then it will end up unfairly disadvantaging some who pursue valuable conceptions of the good, is thus an important one, and it raises very difficult issues—​issues that I have only just touched upon. (The vague appeal to balancing leaves all the hard work to be done!) But it should now be clear that the concern does not really challenge the line of argument I am advancing in this chapter. I have been arguing that a good society will promote sound conceptions of the good and that the currency of advantage that it relies on for purposes of distributive justice will reflect this concern. The balancing of competing claims to advantage, including competing claims to environmental advantage, require us to move beyond the currency question and to consider the plausibility

The Good Society    209 of various distributive principles. This task, important as it is, is not one that I have addressed.

9.4 Impersonal Goods Just societies, I have been claiming, promote the human good. But human goods may not be the only goods. The preservation of an exotic species may be good, even if the species does not contribute to the good of any person. Likewise, the value of an artistic masterpiece may exceed the contribution it makes to the good of those who can experience and appreciate it. Goods of this kind are impersonal.20 Perhaps there are no such goods. The value of purported impersonal goods may be fully exhausted by their contribution, or their potential to contribute, to the human good. If so, then a good society need only concern itself with the human good. Putting this skepticism aside, I  ask, how do impersonal goods—​on the assumption that there are such goods—​affect distributive justice in a good society? My answer is that justice does not require a society to spend resources on impersonal goods, but justice does not rule it out either. Moreover, a good society, as contrasted with a merely just society, will encourage its members to respect and promote impersonal goods. An important category of impersonal goods is perfectionist goods. These goods are human achievements. Discussing them, Thomas Nagel writes: Examples are provided by scientific discovery, of artistic creation, of space exploration, perhaps. These pursuits do of course serve the interests of the individuals directly involved in them, and of certain spectators. But typically the pursuit of such ends is not justified solely in terms of those interests. (Nagel 1979: 129–​30)

Nagel continues:  it is “important to achieve fundamental advances in mathematics or astronomy,” even if doing so has no practical value. “The mere existence of such understanding, somewhere in the species, is regarded by many as worth substantial sacrifices.” I agree. But support for perfectionist goods has been thought to conflict with distributive justice. To see how it could do so, suppose, just for simplicity’s sake, that justice requires an equal distribution of money. At t1 this equal distribution has been realized in a society. At t2 the society imposes a tax on its citizens and uses the funds to promote 20 

As these claims suggest, impersonal goods often contribute to the good of human beings, but this contribution does not exhaust their value. In a strict sense, impersonal goods refer to the remainder of goodness that is not due to this contribution.

210   Steven Wall perfectionist goods. A consequence of this state action is that money is transferred to those persons who have the talent and motivation to achieve the goods. The just distribution has now been upset, as inequalities have been introduced. Rawls expressed a version of this worry: The perfectionist idea is that some persons have special claims because their greater gifts enable them to engage in the higher activities that realize perfectionist values. (Rawls 1999: 152)

Rawls thought that this idea called into question whether a just society could allocate public resources to support excellence in science, art, and culture. In presenting this example, I invited us to suppose that money is the currency of advantage. But the same concern arises when we adopt a more adequate metric of advantage that includes welfare and rational aim satisfaction. Starting from a stipulated just distribution of advantage, a talented artist capable of great achievement may get an additional increase in advantage from state support for perfectionist goods. (In a similar way, a person who is strongly committed to preserving the coral reefs may get an additional increase in advantage from state support for the preservation of this natural impersonal good.) Is this unjust? The question is difficult to answer in the abstract, admittedly. Different distributive principles likely will support different answers. But some general comments can be ventured. Begin by noticing that Rawls’s worry, as it is expressed in the passage above, is misleading in one key respect. Those with greater gifts, we might agree, do not have a special claim to support. It is not unjust to them if they are denied support. This could be true, while it also could be true that it is not unjust to others to support them. In devoting resources to the preservation and promotion of perfectionist goods, the state favors these goods. It need not show favoritism toward any of its members. However, it would be unjust, arguably, if it favored some over others simply on the grounds that they have greater gifts. The state need not treat its members unjustly when it supports impersonal goods. The question remains whether the distribution of advantage that results from its action would be unfair. One response is to grant that the resulting distribution would be unfair, but hold that the state is justified in its action, nonetheless. This would be a case of justified unfairness. But a different response, I believe, is more plausible. To explain it, I proceed circuitously. Consider a man who has a very unfitting attitude towards impersonal goods. He desires to destroy them. If given the chance, he will deface a painting in the Louvre or despoil the Grand Canyon. Fortunately, we can prevent him from acting on his desires. As a result, he will experience frustration, and his level of advantage will be diminished. Does this man have a claim in justice to be compensated for this disadvantage? Plausibly no. His desires are perverse. He desires what is intrinsically bad, that is, the destruction of intrinsic goods. As such, his desires merit condemnation. From the standpoint of distributive justice, deficits in advantage due to

The Good Society    211 perverse desires should not count. Call this the perverse tastes amendment to the advantage metric.21 Now consider someone who desires to respect or promote impersonal goods. She is the mirror image of the man with perverse tastes. Her desires merit commendation. From the standpoint of distributive justice, gains in advantage due to commendable desires should not be held against people. Call this the commendable tastes amendment to the advantage metric. The perverse tastes amendment is very plausible. But if one accepts it, then one should accept the commendable tastes amendment as well. Both amendments are underwritten by the same idea.22 The idea, roughly, is that justice does not penalize people for having fitting attitudes towards objects of intrinsic value. If those with perverse tastes merit compensation, then this compensation will come at the expense of those with nonperverse tastes. Likewise, if commendable tastes are factored into the metric of advantage, then those with commendable tastes will be penalized relative to a distribution of advantage that did not include commendable tastes in its metric. By excluding perverse and commendable tastes from the currency of advantage, we can say that while a just distribution of advantage does not direct a society to respect and promote impersonal goods, it does not present an obstacle to its doing so either. We then can add that, in addition to being just, a good society will preserve and promote impersonal goods. By doing so, it advantages those whose own good is thereby advanced. This consequence, in turn, will encourage its members to have fitting attitudes toward these goods. The point of justice is the human good, but a just society is perfected by its concern for the impersonal good.

Acknowledgments Thanks to Tom Christiano, and especially Serena Olsaretti, for helpful comments. This chapter was made possible through the support of a grant from the John Templeton Foundation. The opinions expressed in this publication are those of the author and do not necessarily reflect the views of the John Templeton Foundation.

References Arneson, R. (1989). “Paternalism, Utility and Fairness.” Revue Internationale de Philosophie 170: 409–​23. 21  Compare this claim with Cohen’s response to offensive tastes. Offensive tastes concern the pleasure a person derives from discriminating against others or subjecting them to a lesser liberty. Cohen comments: “From the point of view of justice, such pleasures deserve condemnation, and the corresponding preferences have no claim to be satisfied, even if they would have to be satisfied for welfare equality to prevail” (Cohen 1989: 9–​10.) 22  The line of argument here presupposes that the metric of advantage is not adequately captured by a resourcist view. Resourcists are under no pressure to accept either amendment.

212   Steven Wall Arneson, R. (2006). “Desire Formation and the Human Good.” Royal Institute of Philosophy Supplement 59: 9–​32. Christman, J. and Anderson, J. (eds) (2005). Autonomy and the Challenge to Liberalism. Cambridge: Cambridge University Press. Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99(4): 906–​44. Dworkin, R. (2011). Justice for Hedgehogs. Oxford: Oxford University Press. Nagel, T. (1979). “The Fragmentation of Value,” in T. Nagel, Mortal Questions. Cambridge: Cambridge University Press, pp. 128–​41. Nozick, R. (1974). Anarchy, State and Utopia. NewYork: Basic Books. Parfit, D. (1984). Reasons and Persons. Oxford: Oxford University Press. Rawls, J. (1999/1971). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press. Raz, J. (1986). The Morality of Freedom. Oxford: Oxford University Press. Raz, J. (1999). Engaging Reason. Oxford: Oxford University Press. Scanlon, T. M. (1998). What We Owe to Each Other. Cambridge, MA: Harvard University Press. Sen, A. (1980). “Equality of What?,” in S. McCurrin (ed.) The Tanner Lectures on Human Values. Vol. 1. Salt Lake City: University of Utah Press, pp. 197–220. Sen, A. (1985). “Well-​Being, Agency and Freedom.” The Journal of Philosophy 829(4): 169–​221. Valdman, M. (2010). “Outsourcing Self-​Government.” Ethics 120(4): 761–​90. Vallentyne, P. (2005). “Debate: Capabilities versus Opportunities for Well-​Being.” Journal of Political Philosophy 13(3): 359–​7 1. Wall, S. (2013). “Moral Environmentalism,” in C. Coons and M. Weber (eds) Paternalism: Theory and Practice. Cambridge: Cambridge University Press, pp. 93–114. Wall, S. (2016). “Autonomy as a Perfection.” American Journal of Jurisprudence 61(2):  175–​94.

Chapter 10

T he Ethics of C a re Virginia Held

What makes a society fair? The focus of contemporary political theory has been overwhelmingly on distributive justice at least since John Rawls asserted in 1971 that “justice is the first virtue of social institutions” and directed his attention to distributive just­ ice (Rawls 1971: 3). This outlook and its assumptions and values have primarily been developed to evaluate social and political institutions within societies: what does justice require of and for citizens and the states whose governments they establish? Questions have then been asked about whether and how the principles of justice for states apply to the international context. Consideration of the moral issues of international affairs and global developments have, if anything, been even more heavily concentrated on questions of justice, to the extent that they have moved beyond issues of war and peace. As Sarah Clark Miller shows, in dealing with the global problems that morality ought to address, a “justice-​dominated discourse” has been apparent. “Cosmopolitan theories of justice,” she writes, “have dominated contemporary philosophical discussions of global responsibility” (Miller 2012: 121). In these same recent decades, however, an alternative moral outlook to the one behind the focus on justice has been developing: the ethics of care. It, too, can be applied to societies, and to the globe. It focuses on the values of care seen most clearly in contexts of family and friendship but then extended to politics and society and international relations. It attends especially to actual practices of care and to caring relations between actual persons and groups, and to the values involved. It asks whether and how societies are caring, and looks at how well, or not, the values of care are reflected in a society’s or the world’s practices and institutions. The questions it asks may be even more important and fundamental than those about justice, though they are seldom asked in the mainstream literature of political and social theory. This is beginning to change. It can be understood that without many years of a care whose values conflict with those assumed by the dominant theories of justice, no child would become the self-​interested rational individual of these theories that then repudiate the outlook of care. The nearly exclusive emphasis on justice may be understandable given the dominant moral theories of the period. Or perhaps the dominance of these moral theories can

214   Virginia Held itself be best understood in light of the political theories they reflect. The dominant moral theories can be interpreted as generalizations, to the moral level, of the prevailing liberal political theories emphasizing impartial, universal rules of law, rights, and just­ice for independent individuals. In any case, the political and the moral reinforce each other. The dominant moral theories have been Kantian morality and utilitarianism, and both, together with their legal and political implications, can be understood as moralities of justice. They can be contrasted with the alternative outlook that developed in the final quarter of the twentieth century and the beginning of the twenty-​first and is still in its infancy: the ethics of care. It may well be that distributive justice is not the best focus to give to the moral considerations to which we ought especially to be directing our attention. It may be that the dominance of theories of justice, law, and rights at the expense of many other moral values and approaches is less helpful than moral efforts ought to be. We might think more about other ways than those of justice and law to address various contemporary and future moral problems, and develop different and more fruitful emphases in our moral thinking. To do so would not be a rejection of the concerns of distributive justice, which would remain important in their appropriate domain. What would be questioned is the nearly exclusive focus of attention on the approach of justice through law and rights, and the neglect of the vast areas of society, and the values most important in and for them, that lie beyond. When we look at a society we can see that most of its rich patterns of relatedness and interdependence and mutual pursuit of mutual interests proceed without the intervention of the law. And yet they can be and often are guided by moral considerations. If we consider the kind of morality suitable for these sorts of activities and relations, we see how they are not especially focused on justice or law or rights. They especially concern themselves with meeting needs, building trust, and doing so with sensitivity in ways that reflect mutuality. If they do not do this, we can argue that they ought to. Rights and law belong to a limited domain within the social. The social and civil society must be composed of relations that are at least somewhat trusting and mutually considerate (see Putnam 1994; Keane 2003). Enough of us must care sufficiently that our fellow members of society are treated decently for there to be the cooperation and support that enable legal or political systems to function. The implementation of just­ ice is dependent on a background or underlying network of social relations that are at their core caring relations. In the other sectors of society than the contexts of family and friendship, caring relations may be relatively weak, not the strong attachments that characterize family ties, but they are nevertheless essential. To focus our moral attention and efforts on care rather than on justice might be a promising alternative for many problems. Theories of distributive justice are theories for persons imagined as independent, autonomous, and relatively self-​sufficient individuals. Justice concerns the principles of law and rights to which they could agree from a purely hypothetical position in which all are free and equal. But many issues are not best conceptualized as questions of law

The Ethics of Care    215 and rights. They more appropriately and fundamentally concern other aspects of society: the activities of the economy, the provision of services, the practices of civil society groups, for instance. They occur within legal constraints, but are not themselves primarily legal matters. Consider education and its plethora of institutions and activities. Law provides certain bounds within which its practices are conducted, but education is certainly not primarily a legal activity and its institutions are not primarily legal institutions. It should be guided by moral considerations, but not primarily by moral theories of justice. Or consider healthcare. Law can importantly influence the practices of health insurance and healthcare delivery. But the vast amount of activity composing the healthcare sector of society is not primarily an activity of law. Society is vastly more than its legal system. Or, especially, consider the economy. Issues of distributive justice concern the fair division of goods after the economy has worked its production, but most of what goes on in an economy is not primarily a matter of law and its institutions. Yet how we conduct our economic activities and how the enormous sector of society that is its economy is organized is of crucial moral significance. In thinking about morality and our societies and about global society, we would do well to overcome the routine tendency to imagine the issues of justice and rights and law (whether seen as moral or legal) as being the ones to which nearly all our moral concern and attention should be devoted. In the period during which there has been such a concentration of attention on issues of distributive justice, feminist theorists have developed the major new approach to morality that is the ethics of care, and from this perspective the nearly exclusive emphasis on distributive justice appears limited at best. We can see the promise of alternative approaches, and the need for their development. Perhaps a focus on the moral approaches of care would be fruitful for dealing with many of the most serious problems we face as members of communities and of global society. A focus on care might better motivate or more effectively guide or perhaps provide a superior understanding of what we ought to do and how we ought to live than the approach of distributive justice.

10.1  The Ethics of Care The ethics of care is built on a feminist understanding of the values incorporated into actual practices of care, and the standards that can be articulated by which to evaluate and improve these practices. When it was first developed, care was formulated as an outlook contrasting with that of justice. The first virtue of justice is impartiality between independent individuals. Justice rests on universal principles recognizable by reason and valid for all. Such principles are applied to particular cases. Care appreciates the moral emotions of

216   Virginia Held empathy for particular others. It asks us to respond to and to meet the needs of the actual persons for whom we take responsibility, and to do so effectively and with sensitivity and respect. It calls for trust and mutuality, recognizing persons as dependent and interdependent. Its focus is on caring relations. The ethics of care does not start with the fully formed individual agent considering how he ought to act, or to which rules or agreements he ought to consent. It is built on the awareness of the enormous amount of valuable and unrecognized labor involved in care, especially the care of children, without which there would be no persons. It acknowledges rather than overlooks the reality that human beings are dependent on other human beings and are essentially vulnerable and in need of care for substantial parts of their lives. It understands that human beings in society are interdependent, as capitalists are dependent on workers, consumers and producers on each other, and as we all carry out our parts of the interconnected activities that compose a society. Fiona Robinson is one of the most influential care theorists writing about care in international relations. She emphasizes the need for moral theory, of whatever kind, to attend to the vast amount of work involved in care, including at a global level. She argues that: . . . relationality and interdependence, and the responsibilities and practices of care that arise therewith, are fundamental aspects of moral life and sites of political contestation that have been systematically denied and obfuscated . . . A political theory of care brings into view the responsibilities and practices of care that sustain not just ‘bare life’ but all social life, from nuclear and extended families to local, national and transnational communities. . . . It emphasizes an ontology of relationality and interdependence that accepts the existence of vulnerability. (Robinson 2010: 1)

A care perspective arose from the awareness of feminists considering the enormous moral relevance of the practices of care in which so many women are engulfed for so much of their lives. Work such as housework, childcare, and care of the ill or elderly had previously been dismissed as belonging to the realm of ‘nature’ rather than being a distinctively human activity. It had been interpreted as being governed by ‘instinct’ rather than by conscientious thought, and it had been considered by almost all moral philosophers to be irrelevant to morality. Feminists, however, started paying attention to this work and practice and concluded that it was of huge relevance to morality (Ruddick 1980 and 1989; Noddings 1984). It incorporated moral values, was conducted and evaluated according to moral standards, and could be judged to need improvement in accord with moral aspirations. The values illuminated by attending to practices of care are quite different from those propounded by the dominant moral theories, but are no less important, compelling, and morally persuasive. As development of the ethical insights involved in attending to care proceeded, what came to be understood as the ethics of care came into being. It has become a recognizable alternative approach to the ethics of justice. In the view of

The Ethics of Care    217 many, it is more fundamental, more comprehensive, and potentially more universally appealing than the ethics of justice that have been dominant in the West for the past few centuries. Once the moral values in practices of care are identified, we can recognize how they ought to be extended beyond their immediate contexts. Empathy, responsiveness to needs, and trust, for instance, can be seen as general values of relations for other contexts as well as for those of caring labor for particular others. When something is a moral value, it becomes of general relevance. When literature or a revered text portray an admirable act of courage or exemplary case of compassion, we understand it to be an example that can lead us to the valuing of such qualities in general and in other persons and contexts. When we comparably understand the values of care and its focus on relations, these are not merely for the particular cases we experience, but for social contexts generally. The ethics of care had its beginnings in the work of Sara Ruddick on the thinking involved in the practice of mothering (Ruddick 1980, 1989), and the work of Nel Noddings in exploring the phenomenology of care work (Noddings 1984). Initial suggestions of a contrast between justice and care emerged in the psychological inquiries of Carol Gilligan, who found gender differences in the ways people interpreted moral problems (Gilligan 1982). Most of the men and boys studied tended to construe the problems presented as calling for the application of abstract moral principles, such as principles of justice, to situations of conflicting rights. Many of the women and girls studied, on the other hand, interpreted the problems as particular and concrete, as problems of maintaining human relationships and meeting the actual needs of those for whom they felt responsible. Gilligan thus identified a “different voice,” a voice of care that differed from the more traditional moral approaches of justice. In subsequent decades the ethics of care was further developed by feminist moral philosophers (see e.g., Held 1993 and 2006; Tronto 1993; Bowden 1997; Kittay 1999; and Slote 2007), and it became more clearly contrasted with the ethics of justice of the dominant Kantian and utilitarian approaches (Held 1995). This work can clearly lead us to wonder whether the overwhelming focus on justice that I have been describing in thinking about moral issues in society and the world is based on gender bias rather than on anything more defensible. Gilligan has continued her inquiries, especially with the study of adolescents (Gilligan 2013). She shows how both boys and girls when they are young tend towards caring, and how in adolescence boys are led by patriarchal influences to turn away from caring and girls to suppress their abilities to express themselves. She finds these developments to be experienced as trauma. Her inquiries show that the psychological assumptions of the ethics of care, that human beings are capable of what it recommends, are not unrealistic. Her inquiries support the view that we need not accept the assumptions of much political and moral thought that human beings seek always or primarily to promote their own individual interests. The ethics of care is a moral approach advocated for men as well as women, applicable to caring practices within the family, and, like other moral theories, to the full range of human activity.

218   Virginia Held Practices and values of childcare, eldercare, and other care of course vary by class and race and culture in a society, and in the same society at different times. Even greater variations can be found around the globe and throughout history. However, we can find the common aims sought and deeper values served in the variety of care practices. They all intend the survival and adequate well-​being of those cared for, except in very special circumstances. They have in common that they intend to be nonviolent practices. Practices of care must all include such values as effectively responding to need. They all require empathetic understanding, mutuality, and trust. They depend for their success on the relation between recipient of care and provider, and such relations are valued and appreciated. Standards of good care are part of the various actual practices and can be used to evaluate and improve them. For instance, in a practice aimed at restoring the health of a patient, if a particular treatment worsens the patient’s condition, it should be changed (other things equal). In bringing up children, if a kind of punishment intended to correct a child’s behavior makes the behavior worse, the practice of using it should be corrected. Care should not be understood as altruism, asking for endless giving on the part of selfless providers. It understands needs for care of the self. Self-​sacrifice vs self-​interest is not the right framework within which to interpret the issues. For the most part, in contexts of care, our own interests are not simply pitted against those of others, though they may certainly at times conflict. Underlying contexts of care, we want what will be good for both or all of us together. Seeing the alternatives as egoism vs altruism is the wrong way to think about what is at stake. We want those we care for, our children for instance and especially, to do well along with ourselves. As recipients of care, we want our caregivers to do well along with us. As providers or receivers of care, we want the relations between us to be characterized at their most basic by mutuality and cooperation, not competition. When caring relations are good ones, the whole framework of self versus other needs to be rethought, and then rethought for other contexts. Practices of care include, most clearly, childcare, eldercare, and nursing, and also much education and healthcare. Citizenship and peacekeeping and many other practices can be interpreted as practices of care (Bowden 1997; Sevenhuijsen 1998; Tronto 2008). Care involves labor guided by the objectives and standards of the various practices. Some of the actual labor involved may be done without caring emotions or intentions, as when a parent feeds a child but feels nothing but resentment at the burden, or when a healthcare worker hates the patients being tended and works only for the pay involved. But we can understand that, although it is better that children be fed than go hungry, and better that patients be tended than not, care is better care when done with the appropriate feelings and motives. Care is often work that ought to be paid for, but when the primary motive of a hospital’s healthcare is profit, such care is not as good as it should be (Held 2006: Ch. 7). Practices of care should include considerations of justice: childcare responsibilities should be divided fairly between parents, multiple patients should be treated fairly, childcare and healthcare workers should receive the compensation they deserve, and the like. But the highest priorities of such practices should be the values of care rather

The Ethics of Care    219 than of justice. Before questions of just distribution even become relevant, there must be persons cared for adequately enough to live and work, and goods produced sufficiently to be distributed. Not all feminists are advocates of the ethics of care, but the ethics of care and attention to care activities are central and important parts of feminist thought. Almost all feminists want to revise existing conceptions of law, justice, and rights to better assure justice for women (Gould 2004; Jaggar 2009a and c) and many have examined issues of global gender justice. Alison Jaggar has shown, for instance, how “women everywhere are disproportionately vulnerable to poverty, abuse, and political marginalization” (Jaggar, 2009c: 33). Feminists want to add requirements for women’s equality and protections against violence where laws and institutions fail to provide for them. Jaggar considers whether the definition of genocide should be “expanded to include female infanticide, the systematic withholding of food, medical care, and education from girls, and the battery, starvation, mutilation, and even murder of adult women,” and whether rights against genocide should accordingly be revised (Jaggar 2009b: 11). However, many feminists want to go even further than assuring that women are fairly represented in existing theories and institutions. They see the need to reconceptualize the dominant moral and political theories themselves because of the bias they see in these theories (Held 1993). The development of the ethics of care has been an expression of this aim (see e.g., Held 2006; Tronto 1993; Bowden 1997; Kittay 1999; Slote 2007). We want to overcome the gender bias built into moral theory, not only in the way this theory has been applied. The ethics of care of course rejects the gender hierarchy in existing institutions—​it is a feminist ethic. But it goes further: it rejects the gender hierarchy in the theorizing that prioritizes justice over care.

10.2  Care and the Individual At the heart of the dominant moral theories and theories of justice is the concept of “liberal individual.” Despite many critiques, for instance from Marxists and more recently feminists, these theories are still influenced to a remarkable degree by the Hobbesian and Lockean conceptions of man in the “state of nature,” contracting with other men to establish a basic “social contract” for society. Hobbes imagined these men “as if but even now sprung out of the earth, and suddenly, like mushrooms, come to full maturity without all kind of engagement with each other” (Hobbes 1972 [1642]: 205). Locke imaged these men being always able to go off into unoccupied territory (to America, Locke suggested) and make their own way, using their labor to acquire property. Hence any exchange they entered into in an actual market was imagined to be voluntary (Held 1976). However wildly unrealistic such conceptions are, critics are answered with the assertion that such images are only hypothetical. They are said to be theoretical constructs such as Rawls’ “original position” from which, imagining ourselves as free

220   Virginia Held and equal and ignorant of our actual advantages or disadvantages, we can think about the basic principles of justice for society to which we could agree. However, since the principles of justice are supposed to serve real persons in actual circumstances in a world carved up into states with enormous disparities of power and advantage, within them and between them, the problems of whether such hypothetical principles should apply to the world we live in remain formidable, even within their appropriate spheres of law and politics and economics. When these assumptions about persons are extended, as they often are, to persons everywhere, even in contexts of family and friendship, they are most clearly unsatisfactory. To the ethics of care, persons are not just imagined to be relational and interdependent. This is the way they are seen to actually be. They start out in need of vast amounts of actual caring labor and continue to need it at various times. Some persons need it continuously. The ethics of care appreciates that care is not only an empirical necessity. Moral values are embedded in practices of care. The social arrangements that structure existing practices of care are often in need of reform, especially in the way the overwhelming proportion of care is performed by women and ill-​paid members of minority or immigrant or migrant groups who have little choice but to do this work. Nevertheless, we can discern the values in existing practices of care and we can understand them even more clearly in such practices as they ought to be. Care is both practice and value. The value of care, within the ethics of care, has been developed to be of at least as much importance as the value of justice. It is thought by some to be even more fundamental: We can understand that, although persons can exist without justice, they cannot exist without the care essential for every child to survive. This care contains within its practices the values that the ethics of care makes visible. Attending to the practices of care illuminates such values as responsiveness to needs, empathetic understanding, and sensitivity. It especially promotes trust and mutuality between care givers and recipients, and then in wider contexts. It calls for understanding caring relations from the point of view of recipients of care as fully as from the point of view of providers.1 In this way it offers the moral resources for avoiding the paternalism and patronizing attitudes that often accompany compassion and charity, within societies and in global efforts. The ethics of care shows the way to achieve good relations of care, and it is caring relations, not merely virtuous dispositions, that are especially valued. Morality, for the ethics of care, does not resemble a contract voluntarily entered into between independent, equal, and self-​interested individuals, or an agreement between rational and autonomous persons making laws for themselves and all others. The ethics of care starts with an understanding that no person would have survived without the care of others for many years of their lives. It understands all persons as interdependent embodied persons in actual historical and social and personal circumstances. 1  For discussion of practices that might adequately bring the perspectives of recipients of care into the making of policies, and the provision and delivery of social services, see Marian Barnes. (2012). Care In Everyday Life: An Ethic of Care in Practice. (Bristol. UK: The Policy Press).

The Ethics of Care    221 Care occurs between persons of very unequal power who did not choose each other. Parents have vast power over their helpless infants and morality requires them to not misuse it. We cannot choose our parents or the particular children of our procreation, yet morality guides our interactions with them. Persons only come to resemble, to some extent if at all, the independent autonomous individuals of the dominant political and moral theories as a result of extensive interactions with interdependent others. It is a great advantage of the ethics of care that, unlike various other theories, it applies to persons of unequal power in relationships they did not choose, as well as to persons imagined as free and equal. And it is much more suitable than are more individualistic theories for dealing with groups, and such issues as nationalism or ethnic conflict, issues so important in global contexts. In fostering trust and mutual concern rather than confrontation and violence, the approach of care lends itself to reducing hostility between groups.2 The ethics of care is often more appropriate than traditional theories for a world of states and persons of very unequal power in situations and relations to which they did not voluntarily agree, which is the world as it is in historical reality. Care’s values of empathetic understanding, responsiveness to needs, and building of trust could go far in confronting actual rather than merely hypothetical conflicts.

10.3  Care and Moral Theory The ethics of care is concerned with meeting the needs of actual and embodied persons located in actual contexts, and doing so effectively. It is not, however, a kind of consequentialism, evaluating actions only in terms of their consequences. Activities involved in the practice of care should also be evaluated on more deontological grounds in terms of the caring intentions and motives they express. Care seeks to meet, with sensitivity and respect, actual needs. It rejects the view of the dominant theories that morality must be based only on ideal, abstract, impartial principles recognizable by reason. To the ethics of care, empathy and caring emotions not only help us carry out the dictates of reason but help to enable us to understand what morality recommends. Some theorists see care theory as a kind of virtue theory, and there are certainly affinities between them. There are also, however, important differences (see McLaren 2001; Saunder-​Staudt 2006; Slote 2007). The ethics of care is a feminist approach to morality that builds on experience not previously focused on in the development of moral theories. And it is caring relations, rather than the virtuous dispositions of individuals, that are valued by the ethics of care. Some elements of the moralities of the British sentimentalists and especially of Hume’s ethics can be helpful in developing care theory (see Baier 1994, and Tronto 1993), but the ethics of care sees persons as relational, not the independent individuals of the British moralists.

2 

See section 10.7.

222   Virginia Held Care is based especially on actual experience. Every person has received care without which they would not have survived childhood, and such care has incorporated moral values. Many persons have provided a great deal of care of others and almost everyone has provided some care. Every person can thus reflect on his or her experience of giving and receiving care, and can evaluate the practices involved. Existing practices are often inadequate, insensitive, or paternalistic, and should be changed. The recipients and providers of care that was not good care can reflect from their own experience on how it should be improved. Marian Barnes writes on the practice of care in the social services. She describes the emphasis of care ethics on lived experience. “A key argument of an ethic of care,” she writes, is that care as practice and as moral and political value cannot be understood in the abstract. We need to consider the lived experience of giving and receiving care, and how context, conflicts and power impact the difficult moral decisions as well as the practical tasks of care. (Barnes 2012: 40)

The ethics of care is thus based on experience that truly is universal: the experience of care. It does not employ the religious foundations of many other moral theories, foundations that are often divisive and at best unpersuasive to those who do not share the religion in question. Care ethics can effectively appeal to conscientious persons across very different cultures.

10.4  Critiques of Care A criticism of the ethics of care that is sometimes made, including by feminists, is that it is too conservative, that it may support reactionary views about the appropriate roles of women. Some fear it may reinforce the exploitation of women as caregivers, rather than promoting major social change. Onora O’Neill, for instance, worried that “a stress on caring and relationships . . . may endorse relegation to the nursery and the kitchen, to purdah and to poverty. In rejecting ‘abstract liberalism,’ such feminists converge with traditions that have excluded women from economic and public life” (O’Neill 1992: 55). And Claudia Card expressed comparable doubts, suggesting that “more likely mistaken for a caring virtue is women’s misplaced gratitude to men who take less than full advantage of their power to abuse or who offer women the privilege of service in exchange for ‘protection.’ Women have assumed caretaking responsibilities as a debt of gratitude for such ‘benefactions.’ ” But, Card wrote, “misplaced gratitude is a kind of moral damage women have suffered” (Card 1995: 94).

The Ethics of Care    223 As many of us defending the ethics of care have been saying for some time, however, the view of the ethics of care as supporting the traditional subordination of women is seriously mistaken (Held 1993, 2006; Sevenhuijsen 1998; Robinson 1999). Carol Gilligan’s recent work, and a recent book by David Richards with whom she has taught, gives further support to the rejection of these fears (Gilligan 2013; Richards 2013). Beyond insisting on the equal rights of women in existing social institutions, the ethics of care overthrows the gender hierarchy of morality. What could be more revolutionary than overthrowing patriarchy in the most basic ways we think about how we ought to live and what we ought to do? The ethics of care is clearly revolutionary in the most important way: it changes how people think and feel about the most basic questions they face. It asks for the transformations and restructurings of society, politics, law, economic activity, social services, the family, and personal relations, away from the assumptions of patriarchy and toward the world of caring and the kind of justice caring calls for. As Gilligan expresses it in her book Joining the Resistance, our exploration . . . has led us to see the ethic of care, grounded in voice and relationship, as an ethic of resistance both to injustice and to self-​silencing. It is a human ethic, integral to the practice of democracy and to the functioning of a global society. More controversially, it is a feminist ethic, an ethic that guides the historic struggle to free democracy from patriarchy. (Gilligan 2013: 175; see also Tronto 2013)

Other critiques of the approach of care deny that it is significantly different from standard moral theorizing. George Sher, for instance, claimed that there is no special distinction between the way standard morality and the ethics of care would handle a moral problem. Concerning whether we should attend to abstract rules and principles or rather to the contextual particulars of the problem, he wrote, “the proper question is not so much whether context is relevant, but rather how many, and which aspects of it are pertinent to our moral decisions, and how these interact to generate moral duties” (Sher 1987: 180). Every moral decision, he asserted, “requires both sensitivity to context and abstraction from it . . . ” and decisions about care are no different (Sher 1987: 181). Similarly, in Sher’s view, if we look at the other ways in which the ethics of care is thought to be different, we find that it is not. “The oppositions of concrete and abstract, personal and impersonal, duty and care,” he wrote, are “generic determinants of the moral problematic. We have always known that an adequate theory must assign each its proper place” (Sher 1987: 187–​8). Sher was writing largely about Carol Gilligan’s empirical findings. Since then, the ethics of care has been developed considerably further as a moral theory. Much depends on what one takes the “proper place” to be for the contextual, the personal, and the values of care, and the importance accorded them by the ethics of care is clearly quite different from that in the standard moral theories. It is not as if the traditional, universalistic moral theories have no problems with the full range of moral concerns. They have difficulties making room for our particular

224   Virginia Held responsibilities to our families and friends, yet their defenders believe that reasonable lines can be drawn. Care ethicists, comparably, think they are not limited to the particularities of the contexts of care but can reasonably extend their understandings of them to general concerns. Sher’s criticisms of the ethics of care were echoed by Will Kymlicka in his Contemporary Political Philosophy:  An Introduction (Kymlicka 1990). He wrote that one way to try to distinguish care and justice “is to say that justice aims at universality or impartiality whereas care aims at preserving the ‘web of ongoing relationships’ ” (Kymlicka 1990:  270). But, Kymlicka argued, when care theorists want their moral recommendations to apply beyond their immediate circle of family and friends and to encompass, for instance, concern for the poor because of a “shared humanity,” they, like Kantians, are “committed to a principle of universality” and the difference disappears (Kymlicka 1990: 271). Grace Clement, in her book on the ethics of care, asserts that “the difference between the abstractness of the ethic of justice and the concreteness of the ethic of care is a difference in emphasis, not in kind” (Clement 1996: 76). But the difference in emphasis can be important. Other care theorists see the difference as more fundamental. Whether one values the actual person for whom one is caring because this person exemplifies “all persons” or humanity in general, or because of the unique person she is and the particular relationship one has with her, makes a difference (See Held 2006: Ch. 6). We should also consider the significance of starting points. One can hope that, ideally, starting with abstract, impartial principles and considering all the many and complex details of particular cases to which they are applied will lead to recommendations consistent with what care would recommend. But actual moral deliberation is not ideal, and starting with the context of caring for particular others, empathizing with them and responding to their needs, then extending these evaluations may well in actual circumstances yield different judgments than would be arrived at by applying a highly abstract rule to a particular case. Having recognized the values involved in practices of care, one may well extend these values to distant others and to other contexts, sometimes meeting the general norms of a justice approach. But the journeys will be different depending on where one starts. Importantly, the motivations will also be different, between the rational recognition of the implications of a rule or principle, and the empathy and concern of a moral agent moved by the need of a child or the suffering of persons unable to fend for themselves, and then enlarging the evaluations involved. Kymlicka admits that the justice perspective, with its emphasis on objective unfairness rather than subjective hurt, “is only legitimate in certain contexts—​namely, interactions between competent adults” (Kymlicka 1990: 277). It is such persons’ own responsibility to meet many of their needs; they should not expect others to take care of them. Kymlicka acknowledges that “care theorists are right to say that some kinds of relationships must invoke different standards for balancing autonomy and responsibility. For example, we cannot expect children to have the same respect for autonomy and reciprocity as adults” (Kymlicka 1990:  281). For morality between competent

The Ethics of Care    225 adults, however, Kymlicka finds the approach of care inferior. He fears that it would ask us to respond to the feelings of hurt of others who do not deserve our sympathy, perhaps because they have been careless and irresponsible about their economic resources, or because they are pained at giving up privileges to which they were not entitled. Kymlicka’s argument here is in terms of the contrast between the “subjective hurt” he claims to be the basis of judgments of care, and the “objective fairness” that is the focus of justice. This misrepresents the concerns of care, whose practices characteristically aim at meeting genuine needs. They are capable of distinguishing when care really is called for and when it is not. They need not, as Kymlicka fears, reward the profligate and careless or the unfairly advantaged. Good care of children teaches limits on their desires and demands, it does not ask of caregivers a “seemingly ‘unlimited responsibility’ ” (Kymlicka 1990: 282). And his image of the “competent adult” bears a strong resemblance to the “liberal individual” and “economic man” of traditional political theory, with its inattention to the realities of interdependence. In modern economies and societies, even competent adults can fail to find employment, can succumb to illness, and can suffer misfortune. Kymlicka concedes that justice as it has been conceived requires that “the sick, the helpless, and the young are kept safely out of view” (Kymlicka 1990: 283). He concludes that it is “too early to tell” whether we can meet our responsibilities to dependent others without giving up our familiar notions of justice. In the meantime, he admits, justice theorists are on “perilously shaky ground”(Kymlicka 1990: 285). That admission was in 1990. The understanding provided by the care perspective has grown far stronger since then.

10.5  Other Critiques It is frequently supposed that the ethics of care fails to protect us against paternalism, or even that it encourages it. The rugged individual of contract theory is replaced in the imaginations of many by what is conjured up as the domineering mother of the ethics of care. This fear is based on the mistake of failing to appreciate how care must be reflected on and understood from the perspective of recipients, which we all have been, as well as providers. When the ethics of care is appropriately examined, it can be seen that it may, contrary to what is feared, be especially helpful in avoiding paternalism. Serene Khader considers the risks of paternalism in development work in the Global South. She suggests that practices of care are useful in indicating how it can be avoided. She observes that “managing conflict between the self ’s needs and desires and the other’s good is key to avoiding conflict in both development work and caregiving” (Khader 2011: 755). Caregivers risk “confusing their own desires for the others’ desires or projecting their desires onto others,” which may well produce unconscious unjustified paternalism (2011: 755). Among the things that caregivers need to learn is

226   Virginia Held how to avoid doing so, and among the habits that caregivers need to cultivate is the ability to distinguish their own needs and desires from those of the recipients of their care. The ethics of care directs them to do exactly this, and guides them on how. Practices of good care train providers of care to develop sensitivity to how their actions are experienced by recipients, and to understand when recipients feel they are not treated appropriately. Within the standards of good care are respect for the other, or what Sara Ruddick called respect for the “embodied willfulness” of the child or other for whom one is caring (Ruddick 1995: 215). It has often been suggested that the ethics of care is an approach suitable for the private sphere of family and friends but not for the public world beyond. Any number of feminist theorists have argued against this limitation on the ethics of care, showing the clear relevance of care approaches to political life, to social interactions, to economic organization, and to international relations (among early examples are Ruddick 1989; Manning 1992; Held 1993; Tronto 1993). In recent years, Daniel Engster has argued that justice should first of all concern itself with the care that every person needs to survive (Engster 2007). Others have delineated the implications of care in considering what a caring society would look like (Bubeck 1995; Bowden 1997; Sevenhuijsen 1998; Kittay 1999; Noddings 2002; Held 2006; Petterson 2008). Grace Clement showed how social welfare programs, at their best, “exemplify the defining features of the ethic of care” (Clement 1996: 90). Marian Barnes shows how, in particular social services, the approach of care offers guidance on making practices closer to what they should be at their best (Barnes 2012). Joan Tronto asks “How do we know which institutions provide good care?” and “Is there a way to articulate the basis for such judgment?” She argues that evaluating good care by institutions requires explicit attention to the purpose of the care, to how power relations within the institution are handled, and to how well the institution modifies its care for the needs of individual persons. She rejects the application of market models to care institutions such as schools and hospitals, showing why they are inappropriate (Tronto 2010; see also Held 2006: Ch. 7). In her book Caring Democracy  (2013), Tronto examines what democracy guided by the values of care would provide, and shows how contemporary market-​obsessed politics in the United States fails to meet the needs of its citizens. These discussions show that any satisfactory political theory will, of course, have to concern itself with how the care needed by all human beings is to be provided, and they show that the concerns of care clearly have political implications. As to how one can reach the universal norms of morality if one starts with the particular concerns of care, it is not more of a problem for the experience and values and understandings of care to be generalized than it is for traditional moral theories to be able to apply to actual persons embedded in personal and historical circumstances. Neither is easy. Actual persons can fail to be moved by efforts to promote their capacities to care, as they can fail to care whether the principles of justice are respected or not. Those concerned with moral theorizing offer suggestions for those willing to listen.

The Ethics of Care    227

10.6  Care and Justice Care does call for justice, but of a limited kind, in my view, within the caring relations that should form both the wider relations of society and the deeper relations of family and friendship. In the language of Rawls, the ethics of care can be seen as the comprehensive morality that can be accepted along with liberal theories of justice and law. However, a better way to conceptualize the ethics of care, I think, is as the wider morality within which justice should be developed as having priority in a limited domain of relations. Care would recommend a liberal conception of justice, law, and rights, for this more limited domain, though it would reject its expansion to the whole of morality. It would, on the contrary, see the domain of law and its enforcements as necessary but as potentially shrinking as society became more caring in all its other institutions and practices. The ethics of care breaks down the division that became entrenched along with the “liberal individual” at the heart of the dominant moral and political theories since Hobbes and Locke:  the division between “public” and “private.”3 It illuminates the oppression that the public can impose on the private, and it shows the relevance of the personal to the political. As it has matured, the ethics of care can be seen as a political, social, and global moral outlook as well as the one most obviously suitable for the personal contexts of family and friendship. The issues of justice in distribution revolve around questions of law, rights, and the principles that should guide public or governmental policies. Many of the debates surrounding distributive justice concern the different approaches of Kantian morality and utilitarianism. Kantian morality supports respect for rights independently of the contribution this may make to the general welfare. Rights are accorded priority over the achieving of good results. Utilitarianism maintains that a scheme of rights ought to produce what will be most in the interest of the greatest number of people, even if the law accords rights a special priority. Utilitarian moral theory concerns itself especially with the public policies that will serve the general welfare. It is more suited to guide governmental policies and budgets and the enactment of legislation once basic rights have been assured. For doing so, it relies, like Kantian morality, on an abstract, universal moral principle applicable to independent individuals. Kantian morality ultimately appeals to the Categorical Imperative to act only in such a way that one could will the basis of one’s action to be a universal law. Utilitarianism appeals at its most basic level to a similarly universal principle, the principle of utility, advising us to act always to bring about the greatest good of the greatest number of people.

3  As Kymlicka has helpfully outlined, historically in political theorizing the “private” has often meant, not the domestic sphere of the family, but those areas of society other than the “state.” However, I use the word “private” in its most common current usage to mean what he calls the “domestic,” and “civil society” to refer to those other areas of society than the legal/​political.

228   Virginia Held Both these moral and political outlooks can be seen to be theories of justice concerned with rules for determining how the rights and interests of individual persons ought to be acted on and protected. As this volume discusses, debates concerning distributive justice explore such questions as whether justice requires that basic goods be divided according to what Rawls formulated as the difference principle, so that any inequalities of position or reward are to the benefit of the least advantaged, or only according to what meeting the basic needs of all would require. These debates usually concern just­ ice within a society or national state having a government capable of assuring that the requirements of justice are carried out. Similar questions can be raised at the international level. But since this context lacks the world government that could enact and assure what justice would recommend, the answers can be significantly different. Given the dangers that would be presented by the concentration of power that a world government would control, it is doubtful that justice would demand a world government resembling the governments with which we are familiar, the governments of states. But we can still try to clarify what global justice would demand in terms of a fair distribution of the goods of the globe. Thomas Pogge is a foremost advocate of addressing the problem of world poverty by showing how unjust it is for the wealthy of the world to have so much while the poor have so little. He points out the remarkable fact that only half of what the top one-​twentieth have gained in wealth over the past decades of globalization would have been sufficient to bring the world’s poor to an adequate standard of living if only it had been distributed more fairly (Pogge 2001, 2011). Debates concerning distributive justice have divided along the lines of whether we do or do not have a moral obligation to apply at the global level the principles of distributive justice that we can agree to for our fellow citizens. If we do not, as Thomas Nagel has argued, then it is not a violation of principles of distributive justice that some countries have so much more than others (Nagel 2005). If we do, we have an obligation to do more to even the outcomes of global economic activity, but exactly what we are obliged to do is very unclear. And recognition of the requirements of justice has often failed to motivate action to address the problem. Another approach altogether would be to explore what a perspective of care rather than of justice would recommend. Nagel himself has argued for humanitarian assistance, as have many others, acknowledging that it is morally wrong for so many of the world’s people to be so desperately poor. But these theorists have traditionally not connected their arguments to the ethics of care. The approach of care engages the emotions, which, as Hume argued, may be needed for moral action. Care is not mere rational recognition, though to be carried out well, care requires reflection. Care calls for empathy to understand the needs of others, sensitivity in doing so and in responding to them, and especially, respectful concern and activity to actually meet those needs. Care provides motivation, not merely instruction. That literature and the arts can have such an effect on us shows, Hume argued, that we are not indifferent to the miseries of others (Hume 1983 [1751]: section 5).

The Ethics of Care    229 Having myself argued for a basic needs rather than difference principle standard for distributive justice within states (Held 1980: Introduction; 1984: Ch. 10), this argument seems to me even more persuasive at the international level. The latter case is made by Gillian Brock in her book Global Justice (2009). Brock argues that if we were to choose norms for our global interactions on which all could agree, we would choose above all to be “enabled to meet our basic needs and those of our dependents. . . . We would use this as a baseline and endorse social and political arrangements that can ensure and underwrite at least these important goods” (Brock 2009: 52). Brock does not consider, however, the implications for moral theory that her position may suggest. In my view, although I did not recognize this in my own earlier work cited, the basic needs approach indicates the superiority of the ethics of care, which would more strongly support it, to the ethics of justice as a moral outlook for dealing with issues of global poverty and of gross injustice between groups. Whether justice requires that needs be met is a contested issue within theories of justice, both moral and political. It is not contested within theories of care that responding to the urgent needs of interdependent persons must be among the primary aims of morality and social institutions. In her book The Ethics of Need: Agency, Dignity, and Obligation, Sarah Clark Miller argues that instead of thinking almost exclusively about justice, we should understand the arguments for “cosmopolitan care.” She examines the shortcomings of the dominance of justice and rights in cosmopolitan thinking about global moral issues. These shortcomings are their “hyperindividualism, idealization, abstraction, and acontextuality” (Miller 2012: 126). She delineates instead a feminist ethical outlook based on views of persons as relational, interconnected, interdependent, vulnerable, and located in actual contexts. Persons, from this perspective, are “interdependent, vulnerable moral agents with inevitable needs. As embodied and finite, humans share a susceptibility to suffering and needing”(Miller 2012: 124). Hence, she aims to respond to the shortcomings of the emphasis on justice in thinking about global moral problems with an account of feminist global responsibility that focuses on actual contexts of deprivation and oppression. She shows how an emphasis on “human interdependence, vulnerability, need, and care might transform the current . . . discourse of global responsibility.” In her view it would lead to recognition of a global “duty to care.” She concludes that care: . . .  conceived as a practice of taking responsibility for others’ needs in the context of dependency relations, of responding to their needs, of acknowledging and fostering moral atunement to the vicissitudes of their lives, is a moral perspective and practice relevant to . . . ethical, social, and political spheres at the national and international level. (Miller 2012: 127–​8)

Global poverty and hunger are examples of the kind of outrage that is not best addressed through law. It is useful to spell out how respecting human rights should include assuring that persons have the resources they need to live and to act. But when the

230   Virginia Held argument fails to motivate action, as it has for decades, it is questionable whether it is a legal remedy that should be sought. Legal institutions are not well suited to providing the kinds of remedies that would best overcome global poverty and its related problems, though from the point of view of the ethics of care, responding to global poverty is of the utmost urgency. Dealing with the problem that still, in the twenty-​first century, millions and millions of children die each year of easily preventable causes might be more adequately addressed through appeal to the ethics of care, with the motivations as well as reasons it offers, than through appeals to principles of distributive justice. Care requires us to take responsibility.

10.7  Care, Justice, and Violence One could almost say that justice is tied to violence. Justice is tied to law, and law, more than many theorists might like to admit, is tied to enforcement. And enforcement of law, it is often agreed, may require violence. So one hopes to avoid violence, but one prepares, and often enough actually uses violence, in the name of justice. We speak, familiarly enough, of just wars. We cannot possibly speak of caring wars, even though the ethics of care might conclude that violence must reluctantly be resorted to. And it can certainly agree that law often needs to be enforced. The connection between morality and political reality depends on the kind of morality. Morality is where, ideally, law (and much else) is worked out. Moralities of justice and law ask to be accepted because of their rational persuasiveness. Violence should not be needed among rational individuals who universalize the moral judgments they can recognize to be valid. But, alas, resorts to violence are all too frequent, between states, between non-​state groups, between states and non-​state groups, between governments and citizens, between individual persons, between states and groups and individuals. We hope humanity has got beyond the large-​scale violence of war, but millions have recently died in wars in Africa, hundreds of thousands in wars in the Middle East, and preparations for war consume vast quantities of the governmental resources of the United States  and other countries. The smaller-​scale violence of terrorism and mass murder preoccupy rational individuals and disturb their lives. Many terrorists rationally calculate how the limited resources for violence at their disposal can best be employed. And the ranks of the irrational, on all sides, are plentiful. The connection between morality and political reality is different with moralities of care than with moralities of justice. Instead of abstract rational principles applied deductively to the realms of politics and law, the ethics of care builds from the felt experience of caring work and practice. It extends the understanding of the values of care, seen and felt most clearly in the small-​scale practices of everyday care, outward to political life, group conflicts, and global interactions. It understands human beings as vulnerable, dependent, above all interconnected, rather than as the rational, self-​sufficient, autonomous individuals of the liberal traditions on which our reigning models of law and

The Ethics of Care    231 politics, and the moralities that expand them, are built, from Hobbes to Locke to Kant to Mill. With the values and practices of care, violence is never to be used if it can be avoided, although the ethics of care is not a kind of pacifism. Any use of violence is already evidence of failure, failure that calls for new measures and practices. Violence should never be part of the everyday practice of care itself, as it is part of the everyday practice of justice. Police and security forces with their focus on enforcement include the use of violence as part of their normal practices. Together with the courts, police forces are often referred to as “the justice system” of a state. No such inclusion of violence would be part of even the public components of “the care system” of a state, much less of all the caring practices occurring within but not sponsored or controlled by the state. Nor does the effectiveness of care rest on the threat of force or violence, as it does with justice. Care must be built on trust not fear, on sensitivity and responsiveness, rather than rational calculation of advantage, whether of self or other or all. With the growth of care one could realistically hope for the decline of violence. The ethics of care offers a promising and relevant approach to the quest for peace and the reduction of violence in international affairs (see e.g., Ruddick 1989; Robinson 1999; Tronto 2008; Held 2008, 2010). I have argued that the ethics of care would lead to respect for international law in the short term, because of the contributions international law can make to peace and civility and mutual benefit among states and groups (Held 2011). As care would become more influential, however, we could expect that the need for law and the reliance on law and its use of force would decrease, both within states and between them. From the point of view of the ethics of care, there should be a transformation in the different domains within society as they become more caring. The law, for instance, could use more negotiation and reconciliation, and less judgment and punishment. Although its enforcements would remain necessary for certain cases, its use of force could be reduced. Education could be vastly improved, and healthcare no longer driven by market forces where this is the case. Social services could become less imperious and bureaucratic. At the same time, and probably even more important, the relations between the domains should be transformed, with some becoming strengthened and others less influential (see Held 2006). Society could redistribute the resources and attention and prestige it devotes to its various components. In a caring society, childcare and healthcare would not be marginalized—​left for people to deal with on their own as best they could, while the state devoted its efforts to military power and economic dominance. In a society where care was appropriately valued and supported, education, social services, noncommercial cultural expression, and economic activity actually structured to meet people’s needs would be the central concerns of society. Teachers and nurses might earn as much as bankers. And, as adequate support was accorded the institutions and practices of care, there would be diminishing need for law and its enforcements. If the values of care were more effective at a global level, and if global problems were approached as the ethics of care would recommend, there would be less need for the

232   Virginia Held justice of law and the use of force. There would be, in a caring world, far more actual justice in meeting the basic needs of all, and far more cooperation in dealing with such problems facing us all as our shared environment.

References Baier, A. (1994). Moral Prejudices:  Essays on Ethics. Cambridge, MA:  Harvard University Press. Barnes, M. (2012). Care in Everyday Life: An Ethic of Care in Practice. Bristol, UK: Polity Press. Bowden, P. (1997). Caring: Gender-​Sensitive Ethics. London: Routledge. Bubeck, D. (1995). Care, Gender, and Justice. Oxford: Oxford University Press. Card, C. (1995). “Gender and Moral Luck,” in Virginia Held (ed.) Justice and Care: Essential Readings in Feminist Ethics. Boulder, CO: Westview Press, pp. 79–​98. Brock, G. (2009). Global Justice: A Cosmopolitan Account. Oxford: Oxford University Press. Clement, G. (1996). Care, Autonomy, and Justice:  Feminism and the Ethic of Care. Boulder, CO: Westview Press. Engster, D. (2007). The Heart of Justice:  Care Ethics and Political Theory. Oxford:  Oxford University Press. Gilligan, C. (1982). In a Different Voice:  Psychological Theory and Women’s Development. Cambridge, MA: Harvard University Press. Gilligan, C. (2013). Joining the Resistance. Cambridge: Polity Press. Gould, C. C. (2004). Globalizing Democracy and Human Rights. Cambridge: Cambridge University Press. Held, V. (1976). “John Locke on Robert Nozick.” Social Research 43(1) (Spring) 169–​95. Held, V. (ed.) (1980). Property, Profits, and Economic Justice. Belmont, CA: Wadsworth. Held, V. (1984). Rights and Goods: Justifying Social Action. New York: Free Press. Held, V. (1993). Feminist Morality:  Transforming Culture, Society, and Politics. Chicago, IL: University of Chicago Press. Held, V. (ed.) (1995). Justice and Care:  Essential Readings in Feminist Ethics. Boulder, CO: Westview Press. Held, V. (2006). The Ethics of Care:  Personal, Political, and Global. New  York:  Oxford University Press. Held, V. (2008). How Terrorism is Wrong: Morality and Political Violence. New York: Oxford University Press. Held, V. (2010). “Can the Ethics of Care Handle Violence?” Ethics and Social Welfare. 4(2) (July): 115–​29. Held, V. (2011). “Morality, Care, and International Law.” Ethics and Global Politics, 4(3): 173–​94. Open Access journal. Hobbes, T. (1972 [1642]). The Citizen:  Philosophical Rudiments concerning Government and Society, ed. B. Gert. New York: Doubleday. Hume, D. (1983 [1751]) An Enquiry concerning the Principles of Morals, ed. J.B. Schneewind. Indianapolis, IN: Hackett. Jaggar, A. M. (ed.) (2009a). “Global Gender Justice.” Special issue of Philosophical Topics 37(2) (Fall). Jaggar, A. M. (2009b). “The Philosophical Challenges of Global Gender Justice.” Philosophical Topics 37(2): 1–​15.

The Ethics of Care    233 Jaggar, A. M. (2009c). “Transnational Cycles of Gendered Vulnerability.” Philosophical Topics 37(2): 33–​52. Keane, J. (2003). Global Civil Society. Cambridge: Cambridge University Press. Khader, S. J. (2011). “Beyond Inadvertent Ventriloquism: Caring Virtues for Anti-​Paternalist Development Practice.” Hypatia 26 (Fall) 742–​61. Kittay, E. F. (1999). Love’s Labor:  Essays on Women, Equality, and Dependency. New  York: Routledge. Kymlicka, W. (1990). Contemporary Political Philosophy: An Introduction. Oxford: Clarendon Press, 1990. Manning, R. (1992). Speaking from the Heart:  A Feminist Perspective on Ethics. Lanham, MD: Rowman & Littlefield. McLaren, M. A. (2001). “Feminist Ethics: Care as a Virtue,” in Peggy DesAutels and Joanne W. (eds) Feminists Doing Ethics. Lanham. MD: Rowman & Littlefield, pp. 101–​17. Miller, S. C. (2012). The Ethics of Need: Agency, Dignity, and Obligation. New York: Routledge. Nagel, T. (2005). “The Problem of Global Justice.” Philosophy and Public Affairs 33(2): 113–​47. Noddings, N. (1984). Caring: A Feminine Approach to Ethics and Moral Education. Berkeley, CA: University of California Press. Noddings, N. (2002). Starting at Home: Caring and Social Policy. Berkeley, CA: University of California Press. O’Neill, O. (1992). “Justice, Gender, and International Boundaries,” in Robin Attfield and Barry Wilkins (eds) International Justice and the Third World. London: Routledge, pp. 50–​76. Petterson, T. (2008). Comprehending Care:  Problems and Possibilities in the Ethics of Care. Lanham, MD: Rowman & Littlefield. Pogge, T. (ed.) (2001). Global Justice. Oxford: Blackwell. Pogge, T. (2011). “Are We Violating the Human Rights of the World’s Poor?” Yale Human Rights and Development L.J. 14(2): 1–​33. Putnam, R. D. (1994). Making Democracy Work. Princeton, NJ: Princeton University Press. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Richards, D. A. J. (2013). Resisting Injustice and the Feminist Ethics of Care in the Age of Obama. New York: Routledge. Robinson, F. (1999). Globalizing Care:  Ethics, Feminist Theory, and International Relations. Boulder, CO: Westview Press. Robinson, F. (2010). “After Liberalism in World Politics? Towards an International Political Theory of Care.” Ethics and Social Welfare 4(2) (July): 130–​44. Ruddick, S. (1980). “Maternal Thinking.” Feminist Studies 6 (Summer) 342–​67. Ruddick, S. (1989). Maternal Thinking: Toward a Politics of Peace. Boston, MA: Beacon Press. Ruddick, S. (1995). “Injustice in Families:  Assault and Domination,” in Virginia Held (ed.) Justice and Care:  Essential Readings in Feminist Ethics. Boulder, CO:  Westview Press, pp. 203–​23. Saunder-​Staudt, M. (2006). “The Unhappy Marriage of Care Ethics and Virtue Ethics.” Hypatia 21(4) (Fall) 21–​39. Sevenhuijsen, S. (1998). Citizenship and the Ethics of Care: Feminist Considerations on Justice, Morality and Politics. London: Routledge. Sher, G. (1987). “Other Voices, Other Rooms? Women’s Psychology and Moral Theory,” in Eva Feder Kittay and Diana T. Meyers (eds) Women and Moral Theory. Totowa, NJ: Rowman and Littlefield, pp. 178–​89. Slote, M. (2007). The Ethics of Care and Empathy. New York: Routledge.

234   Virginia Held Tronto, J. C. (1993). Moral Boundaries: A Political Argument for an Ethic of Care. New York: Routledge. Tronto, J. C. (2008). “Is Peacekeeping Care Work? A Feminist Reflection on ‘The Responsibility to Protect’,” in Rebecca Whisnant and Peggy DesAutels (eds) Global Feminist Ethics. Lanham, MD: Rowman & Littlefield, pp. 179–​200. Tronto, J. C. (2010). “Creating Caring Institutions: Politics, Plurality, and Purpose.” Ethics and Social Welfare 4(2) (July): 158–​7 1. Tronto, J. C. (2013). Caring Democracy: Markets, Equality, and Justice. New York: New York University Press.

Chapter 11

The Theory an d P ol i t i c s of Rec o gn i t i on Colin Bird

Emmanuel SieyÈs’s “What is the Third Estate?” is not only among the most famous revolutionary pamphlets of all time, but also a founding document of the modern “politics of recognition.” It was written at the outbreak of the French Revolution, as Sieyès and the Third Estate were struggling to overturn the privileges enjoyed by the French aristocracy and the church (the First and Second Estates). What, Sieyès asks, does the Third Estate want? His epoch-​making answer—​“to be something”—​launched a new kind of social movement. In expressing the wish of those disenfranchised by the French ancien régime to “be something,” Sieyès was protesting against the tendency for the French establishment to treat them as if they did not exist, as being of no particular account. Sieyès was not complaining on behalf of any particular group defined by ethnicity, religious affiliation, national identity, economic status, gender, race, or sexual orientation, although he does appeal in a very broad way to notions of social class and a primitive form of nationalism. Since his day, however, social movements claiming to speak for identity groups of all these kinds have become familiar features of the political landscape all over the world. While these movements have been diverse, they all share in Sieyès’s demand that the groups they represent not be ignored, and in that sense “be” something in the eyes of society. This is a demand for recognition. Its longstanding role in the struggle for a fairer, more inclusive, political order guarantees such pleas for recognition a place in the modern rhetoric of social justice. However, philosophers and political theorists have devoted sustained attention to recognitional arguments in politics only since the publication of John Rawls’s A Theory of Justice in 1971 (Wolin 1993; Taylor 1994; Appiah 2005; Kymlicka 1995; Honneth 1996; Fraser 2003; Parekh 2006; Modood 2007; Gutmann 2003; Kukathas 2003). To some extent, Rawls himself paved the way for this ‘recognitional turn’ because, in repudiating a purely utilitarian approach to political philosophy, his approach moved notions of equal status, the terms of inclusion within a society, and questions about the “social bases of self-​respect” to the forefront of academic discussion of distributive justice. On the other

236   Colin Bird hand, Rawls’s approach struck many critics (I believe wrongly (Bird 2007)) as excessively individualist in its orientation. This (mis)judgment fueled an influential ‘communitarian’ reaction against Rawls’s viewpoint (Taylor 1985a; Taylor 1985b; Sandel 1982; Walzer 1983; MacIntyre 1984), much of it inspired by the thought of the German idealist philosopher G. W. F. Hegel, whose social theory accorded the category of recognition great significance. The gathering tendency to acknowledge the “multicultural” character of modern pluralist societies also spurred renewed interest in recognition within political philosophy, for it made a host of questions about how conflicts and disagreements between different cultural groups should be negotiated seem urgent to theorists of justice. None of this means, however, that we have a clear understanding of how the “politics of recognition” and the various ethical expectations it may generate are related to the theoretical debates about justice that have flourished in the wake of Rawls’s work. The purpose of this chapter is to critically explore the many ways in which demands for political recognition might be linked to the project of conceiving and realizing a just society.

11.1  Recognition and Justice: A Conceptual Overview One very basic, perhaps the most basic, way in which complaints about inadequate recognition arise in politics concerns the status that individuals and groups are accorded by their society. Sieyès’s complaints on behalf of the Third Estate exemplify this basic form of status-​recognition. They anticipate the demand, central to modern egalitarian conceptions of citizenship, that legal, political, and social status should never be conditioned on such arbitrary factors as birth, bloodline, sex, race, or creed. Of course, in itself the demand for status-​recognition need not be egalitarian—​regimes of privilege, after all, also sustain themselves by requiring respect for status, though there the relevant forms of status are exclusive and hierarchical, rather than inclusive and horizontal. But in modern times the struggle for recognition has almost always set itself against privilege and so typically assumes an egalitarian cast. However, not all proponents of recognition in politics have thought that securing equal status is sufficient to meet its demands. Increasingly, they have insisted that the recognition of identity (and hence difference) is also required. As Charles Taylor (sympathetically) notes, we owe this interest in “identity-​politics” to nineteenth-​century romanticism, which preached a gospel of individuality, self-​expression, and self-​realization. Clearly, one’s identity can be misunderstood regardless of one’s legal or social status, and so the recognition of identity and difference looks like an independent object of political concern. The relevant post-​Romantic notion of identity is not the philosophical concept of "personal identity,” which has to do with what makes someone the same person over time. The pertinent idea, rather, is the identity of self-​definition—​the set of attributes

The Theory and Politics of Recognition    237 one might cite in attempting to describe and differentiate someone. Identity in this more existential sense is a complex and unruly concept, comprising at least three elements. First, it involves what we might call “census-​identity,” claims about the general categories, classes, and groups to which one belongs (demographic, statistical, gender-​ related, racial, ethnic, national, religious, etc.). Second, it includes many attributes that will be wholly peculiar to individuals (“Derek’s spouse,” “president of Columbia University,” “composer of the Eroica symphony,” etc.). Third, a person’s identity in this sense won’t be an unchanging, stable feature, but rather something that evolves over time and so displays an inescapably narrative character. Exactly how public institutions might “recognize identity and difference” in these senses is a notoriously unclear and often controversial question. Critics see such demands as insatiable or unadjudicable; certainly, the underlying notion of identity is vague. However, it’s easy enough to appreciate what prompts these demands. Agents have a clear interest in their identities not being distorted or misrepresented in the social practices and discourses that envelope them. We can sharpen this by distinguishing between two different, though often overlapping, forms this interest might take. On the one hand, agents seek what I will call authenticity-​recognition. That is, they hope for terms of social acceptance that allow them to live as they really are, rather than requiring them to maintain some inauthentic persona. A  society pervaded by misleading stereotypes and prejudices will fail to secure authenticity-​recognition, for it requires at least some of its members to live behind masks that represent what society expects them to be, rather than to feel free to express themselves without fear. On the other hand, agents have an interest in worth-​recognition: they do not want to be demeaned in the eyes of society for what they are or for what they do. Rather, agents hope to be valued and respected in the eyes of their fellows, their presence welcomed rather than deplored, and their contributions appreciated rather than belittled. These three forms of recognition (status-​recognition, authenticity-​recognition, and worth-​recognition) can’t always be neatly separated, and the same social practices can fail to achieve all three. Sexism is objectionable because it confers unequal status on males and females, propagates inaccurate stereotypes of femininity, and deprecates women along with their contributions. Still, these are not simply three different ways of expressing the same complaint. The conclusion to draw is that, so far as issues of recognition are concerned, sexism is triply problematic, undermining desirable forms of status-​recognition, authenticity-​recognition, and worth-​recognition simultaneously. Some may resist the suggestion that recognition is trifurcated in this way; part of my aim here is to weaken that resistance by laying bare some of the differences between them. How these three types of (mis)recognition might figure in discourse about a just society depends to a large extent on how one construes the concept of justice itself. Although that is a complex matter, for present purposes we can adapt some remarks of Rawls and stipulate that justice comprises: (1) “conditions which we are ready upon due reflection to recognize as reasonable in our conduct towards one another” (Rawls 1999: 514), where

238   Colin Bird (2) those “conditions” refer primarily to the ways in which the framework of public institutions (a)  makes available rights, freedoms, and economic opportunities; (b) determines the allocation of offices and political power; and (c) regulates the distribution of economic wealth flowing from social cooperation. This is a rough-and-ready outline; different conceptions of justice will reach contrasting conclusions about how these elements should be interpreted. But it captures the traditional assumption that justice is what Aristotle termed an “outer virtue” concerned primarily with interpersonal interaction in the context of publicly organized cooperation. So construed, recognitional issues—​which are necessarily interpersonal—​ occupy some of the same territory as the concept of justice. Even though both are concerned with properly ordered relations between persons, however, one could still conceive the two as grounding largely independent claims. Considerations of justice and of appropriate recognition might provide separate criteria by which agents’ conduct toward each other can be judged. Independence in this sense doesn’t entail that criteria of justice and of proper recognition must conflict with each other or issue incompatible recommendations. But it does entail that the two can conflict and that when they do, the resulting conflicts can be resolved only by giving justice or recognition priority over each other. Whether or not Independence holds, recognition might relate to interpersonal conduct in at least three ways. Distribution:  Recognition might be conceived as a kind of good that persons or institutions distribute to each other. If its distribution should be guided directly by principles of distributive justice, complaints about the improper distribution of certain forms of recognition automatically entail charges of injustice. If Independence holds, however, the proper distribution of recognition won’t be a matter of justice: sometimes it can be legitimate to distribute things counter to the requirements of justice, as when (on a Christian view) God provides salvation despite our just deserts, or a bankruptcy court forgives debts to which a creditor has a just title. Treatment: While, as we shall see, it can make sense to treat some forms of recognition as distributable resources, in many instances imperatives of recognition more plausibly refer to direct forms of treatment, unmediated by any sort of distribution. Racists who target members of racial minorities for violence, personal abuse, or ostracism need not be failing to distribute some good properly, but they may nonetheless mistreat their victims. Certainly such mistreatment, and any misrecognition it involves, can be a matter of justice. Justification:  Claims about the importance of certain forms of recognition might also play a deeper role in justifying policies, practices, or actions, whether or not those policies, practices, or actions themselves directly involve forms of recognition. For example, prohibitions on harassment or offensive speech don’t directly involve recognition, either as treatment or as a form of distribution, but might nonetheless be justified for the sake of ensuring that all members of society receive proper recognition.

The Theory and Politics of Recognition    239 In what follows, I will consider these various possibilities in the context of status-​recognition, authenticity-​recognition, and worth-​recognition, respectively. When I discuss Distribution, Treatment, and Justification, I will mostly assume that principles of justice are somehow at stake, but readers should keep in mind that none of these three must be construed as matters of justice. Recognition might be properly distributed, be a form of appropriate treatment, or figure in the justification of social and political practices even if it is a free-​standing desideratum unconnected with justice (i.e., even if Independence holds). Before exploring these possibilities, however, I make two further preliminary points, important because they help to explain why recognitional arguments in political philosophy attract controversy. First, any useful account of recognition must not only identify criteria by which to assess whether required forms of recognition have or have not been shown by some relevantly responsible agency, but also reach conclusions that are at least partly independent of the judgments of those whose recognition is at stake. That is, it cannot be wholly up to self-​described victims of misrecognition to determine that they have been improperly recognized. While the testimony of victims may be crucial in bringing to light forms of misrecognition that would otherwise go unnoticed, complaints from those who feel misrecognized cannot be self-​validating. They may reflect oversensitivity, partiality, attention-​seeking, narcissism, manipulativeness, and/​or insincerity. If one is skeptical that independent criteria allowing us to adjudicate such complaints are available, one is likely to worry that recognitional arguments are too readily politicized to serve as a stable anchor for principled critical reflection about either justice or politics more generally. Second, recognitional arguments in politics characteristically bring an extremely wide range of legal and policy issues into view, and many of these are themselves controversial. Consider the following rough-and-ready list of practices and policies that are often viewed through the lens of recognition: • Formal Equality: The basic entitlements of equal citizenship, including equality before the law; protection of the ‘integrity of the person’; rights to freedom of conscience, speech, and association; and rights to participate on equal terms in democratic governance. • Extensions: Policies that extend existing (a)  legal rights; and (b)  prohibitions to cover members of groups (a)  who have hitherto fallen outside some legally protected status (e.g., extensions of the franchise, or same-​sex marriage); or (b) who are vulnerable to particular forms of abuse that existing criminal law fails adequately to proscribe (e.g., laws prohibiting “hate-​crimes” and “hate-​speech,” or barring specific forms of discrimination/​harassment against tenants, employees, colleagues, or contractors). • Expansions: Policies intended to diversify the range of representation within some public or institutional setting (e.g., reserved seats for certain groups in a legislature, imposition of curricular guidelines in schools and universities guaranteeing that

240   Colin Bird non-​Western cultural inheritances are taught alongside more traditional course material). • Compensations: Policies (e.g., reparations, affirmative action) intended to rectify past, or ongoing, injustices targeted specifically at members of identity-​groups. • Exemptions: Special exceptions from general legal requirements intended to preserve the integrity of vulnerable practices, artifacts, and ways of life (e.g., exemptions from general educational curricula, or permission to use proscribed drugs in religious ceremonies). • Accommodations: Policies intended to help accommodate needs of particular groups (e.g., self-​government rights for national minorities, granting ‘official status’ to minority languages). (For more on Exemptions and Accommodations, see Jones, Chapter 27 in this volume.) As this list indicates, the scope of recognitional concerns in politics is vast, too vast to permit a systematic treatment here. This chapter sets itself the more modest goal of clarifying some of the different routes by which recognitional arguments, whether in combination with, or independently of, claims of justice, might lead one to support or oppose practices and policies of these kinds.

11.2  Status-​R ecognition Political arguments about status-​recognition come in two main varieties. Where individuals already enjoy a certain status (in virtue of possessing certain conventionally accepted moral or legal rights, for example), one can ask whether a particular person or group is accorded appropriate treatment in the light of that standing. When (say) members of racial minorities tend to be punished more severely than members of other groups who commit the same crimes, or are targeted more frequently for police abuse, one will suspect that they are not receiving the treatment appropriate to their status. In such cases, the complaint is that, whatever their de jure status, the targeted groups do not enjoy the same de facto status in relation to others. In other cases, however, the objection is more fundamental: individuals or groups lack even the de jure status that they ought to enjoy. The historical campaigns to end slavery and to accord women the right to vote, along with today’s effort to extend the privilege of marriage to same-​sex couples, exemplify status-​recognition arguments of this second sort. When canvassed in public contexts, both sorts of arguments raise obvious issues of justice. Arguments of the first sort are primarily concerned with just Treatment, where the operative standard of justice is a settled scheme of conventional rights. Here, recognizing persons’ status as (for example) property-​holders, as citizens, and as holders of other publicly recognized rights entails expectations of justice, not only because the relevant entitlements often have overriding public urgency, but also because it becomes a baseline by which unjust discrimination can be recognized. Once two or

The Theory and Politics of Recognition    241 more individuals share the same status, it creates an expectation of similar treatment, and hence the possibility of complaints about partiality, bias, and unfairness. Status-​recognition arguments of the second, more fundamental, kind implicitly make claims about just Distribution as well as Treatment because rights, along with the public funds necessary for their enforcement, are themselves distributable resources. One might question whether the redistribution of wealth necessary to fund the provision of basic rights to protection of person and property and other basic civic entitlements should count as the distribution of recognition. Whether or not one finds that description appropriate, however, the conferral of status in the form of civil rights of one kind or another is itself unquestionably a form of distribution that can be just and unjust. Indeed, the classical notion of “distributive justice,” expressed most clearly by Aristotle, applied in the first instance, not to the distribution of economic goods, but rather to that of forms of civic or official status (e.g., rights to hold office, to vote, to be considered for election to certain magistracies, etc.). These considerations suggest that Independence is implausible in the context of status-​ recognition; the expectation that personal status be properly honored or granted is central to ordinary reflection about distributive justice. Much modern political philosophy has, moreover, insisted that a claim about status-​recognition plays a special foundational role in the Justification of judgments about justice. Indeed, the principle that individuals must be treated “as equals” (a status-​recognition idea) has become virtually axiomatic in philosophical discussion of justice. To recognize persons “as equals” in the relevant sense is to acknowledge limits on the ways in which they may be used, exploited, or sacrificed for the sake of (even a greater number of) others. Although often touted as a “liberal” idea, any association with “liberalism” is ultimately immaterial, because the principle packs an intuitive power that punches well above the weight of any single ideological tradition. Certainly “liberals” (however defined) have no unique claim on it. Note that in this principle the status of ‘an equal’ is implicitly treated, not as an attribute of individual persons considered separately, but rather as a desirable relation of ‘parity’ (Fraser 2003) between them, one that is undermined when the interests of separate persons are traded off in exploitative or sacrificial ways. What exactly that principle entails for the structuring of a just society is, of course, a complex and controversial question, although few deny that it at least requires “formal equality” as outlined above. Most of the debate has arisen over proposals to go beyond “formal equality,” especially by requiring forms of economic redistribution. But however the details fall out, the assumption that critically assessing the justness of different public arrangements requires asking how far they recognize citizens’ moral status as equals is extremely attractive. Few reject that assumption outright, but some critics suspect that status-​recognition of this universal, egalitarian, kind is problematically “difference-​blind” and hence not sufficient to secure full recognition for all members of society. The suspicion arises in the following way. It is easy to think that a commitment to treat all members of society “as equals” presupposes that all persons as such share a generic “human” identity that somehow grounds their equal status. This thought derives some plausibility from the

242   Colin Bird truism that individuals’ standing as moral equals must be somehow connected with their common possession of certain ethically important attributes (autonomy, a sense of a selfhood, the capacity for rational planning, the ability to suffer, etc.). It is easy to construe such claims as postulating a common human identity, and to then worry that they emphasize human “sameness” at the expense of “difference.” The objection is that indiscriminately assimilating everyone to a common human identity provides only a partial, one-​sided, form of recognition that doesn’t leave enough room for sensitivity to agents’ particularities and differences. Properly understood, recognition must be attuned to difference rather than aim to be merely  “difference-​blind,” or so the objection runs. This line of criticism however sets up a false opposition between sensitivity to difference and the recognition of equal personal status. For one thing, as a philosophical matter, we should reject the presumption that egalitarian status-​recognition must be based on claims about a shared identity. To demand that everyone be recognized as “an equal” is simply to assert the equivalence of one person’s claims to certain rights and forms of treatment in relation to others’, but no identity-​claim need be involved. Of course, those who stand in this relation of equality (or “parity”) are all people, and their respective entitlements may have something to do with their possessing such attributes as a sense of selfhood, the capacity for rational planning, or the ability to suffer. But to say, for example, that your suffering matters as much as and no more than mine neither presupposes nor entails that your title to parity in that sense is somehow based on an identity that we share. It is not even to suggest that the propensity to suffer is a feature peculiar to “human identity” (which it obviously isn’t). It is simply to draw attention to a morally relevant consideration about people, and to insist that its influence should not vary from person to person. Not every true proposition about a person implies an identity claim (shared or otherwise)—​any more than facts about my current temperature or subjection to the law of gravity bear on judgments about my identity in the existential sense relevant to debates over recognition. Egalitarian status-​recognition might refer simply to a possible configuration of interpersonal relations whose value is directly a matter of justice in a way that floats free of any identity claims. More generally, we have no automatic reason to assume that those committed to equal status-​recognition in the Rawlsian or Dworkinian sense must thereby expect public institutions to be “blind” to claims made on behalf of particular identity-​groups. To the contrary, many writers committed to that principle accept that group-​differentiated claims must sometimes be accommodated by the state, as Will Kymlicka’s influential work (1995) illustrates. Kymlicka follows Rawls’s lead in suggesting that to fully recognize individuals’ status as equals, a just society must guarantee them adequate access to certain essential social goods and resources (what Rawls calls the “social primary goods”). Rawls did not include “culture” or identity-​related resources on his list of primary goods, but Kymlicka and others argue that, just as individuals motivated to preserve their status as equals will want to guarantee themselves certain formal rights, civil liberties, and economic opportunities (for example), they also have reasons to preserve their access to important cultural resources. In Kymlicka’s own account and several

The Theory and Politics of Recognition    243 others’ (Raz 1994), the value of personal autonomy grounds this need for cultural access, and in turn supports the provision of various group-​differentiated rights. Kymlicka believes that such rights can be reconciled with a commitment to recognize every person as an “equal” by circumscribing them so as to protect individuals against subordination and abuse at the hands of the cultural communities to which they belong. He therefore limits the rights of those communities to impose what Kymlicka calls “internal restrictions” on their own members by, for example, prohibiting cultural communities from denying their members the right to leave the community if they wish to do so. He maintains that such “internal restrictions” needn’t preclude “external protections”—​rights intended to protect the integrity, survival, and self-​determination of cultural communities on which individuals’ autonomy depends. In effect, Kymlicka’s multicultural state is prepared to grant cultural groups certain special accommodations in return for a guarantee that they will themselves recognize individuals’ equal status by refusing to impose “internal restrictions.” To describe this position as “difference-​blind” is misleading. Some doubt the coherence of the Kymlicka strategy, and those like it (Barry 2001; Waldron 1992). Even if they fail, however, it does not follow that those who reject group-​ differentiated rights in the name of the recognition of everyone’s status as an equal are ipso facto insensitive, or somehow “blind,” to “difference.” Some (notably Chandran Kukathas) reject the provision of special accommodations for members of identity-​ groups on the grounds that they conflict with values of free association, toleration, and state neutrality, values that themselves reflect the overriding imperative of equal status-​ recognition at the level of Justification. Kukathas however also rejects Kymlicka’s prohibition on “internal restrictions” because he thinks that it makes the state an agent of cultural assimilation, effectively requiring groups to endorse controversial and often alien “liberal” values. Whatever we think of that conclusion, it is hard to describe it as insensitive to cultural difference and particularity. To the contrary, Kukathas is urging that free societies allow cultural and other identity-​groups a wide latitude to explore nonstandard experiments in living, and in this way to express their cultural identity as they see fit.

11.3  Authenticity-​recognition As already noted, demands for the recognition of identity (and difference) often invoke the importance of authenticity in the context of self-​expression and self-​exploration. Such arguments aim to preserve and promote a certain form of freedom, one that requires agents not to have to choose between conformity to conventional norms and honest self-​expression. To some extent, of course, such conflicts are unavoidable and even sometimes desirable:  self-​concealment is often necessary to avoid offence, and fear of social embarrassment or shame can legitimately deter agents from revealing attitudes, impulses, and characteristics that are better not expressed. But when

244   Colin Bird agents are constrained, under the pressure of sanction, to conceal basic aspects of their personalities in the face of prejudice and unmerited disapproval, their freedom seems illegitimately diminished. To be forced to live a lie is one way to be oppressed. Even though authenticity-​ recognition brings questions about oppression and freedom into view, it doesn’t automatically follow that it is a matter of justice. For one thing, authenticity-​recognition looks more like a certain sort of social outcome than a directly distributable resource like wealth and income, and so any connection with a notion of just Distribution is bound to seem tenuous. This is one reason why both apologists and critics often sharply distinguish the discourse of distributive justice from those aspects of the “politics of recognition” concerned with identity and cultural difference. For a sympathizer like Iris Young, for example, the “recognitional turn” in political theory is a welcome move away from the “Distributive Paradigm” she associated with Rawls and his progeny (Young 1990). Critics like Brian Barry reinforce this contrast when they charge that recognitional concerns (and multiculturalism more generally) objectionably displace more urgent questions about distributive justice (Barry 2001). However, even when construed as somehow implying forms of proper Treatment, authenticity-​recognition might still fall outside the remit of justice. Agents often experience the most intense pressure on self-​disclosure in private settings, beyond the reach of easy or effective public control. This might lead us to endorse Independence in the context of authenticity-​recognition, at least insofar as justice is primarily relevant in public contexts. For example, we might conclude that it is best negotiated in private life, perhaps for the sake of love, decency, solidarity, or fellow-​feeling, rather than as a requirement of justice. I will shortly suggest that in many instances there is a strong case for Independence in the context of authenticity-​recognition. However, to insist that claims about authenticity-​recognition have no role at all in discussions of social and distributive justice would be premature. The possibility that such claims might be cited in Justification for certain public policies, including ones mandating certain forms of Treatment, cannot be discounted. To see how such arguments might run, consider the “stereotypes” that often attach to various census-​identities: the poor are scroungers; women are domestic creatures; African Americans are violent; Jews are greedy, etc. Clearly, such stereotypes fuel prejudices that blind people to individuals’ actual attributes and capacities. They are often objectionable simply because they are derogatory, conveying negative attitudes and judgments (Steele 2010). In that case, they raise concerns about worth-​recognition, but here we are trying to isolate their implications for authenticity-​recognition. How might that become an issue of public concern or a matter of justice? At this point, one might expect proponents of identity-​politics to stress the importance of accurate beliefs about persons, as if there is some set of truths about the authentic identity of each person that would ideally be accepted by those with whom they interact. In fact, however, they rarely argue in this way, and for good reason: surely it is far-​fetched to postulate any sort of entitlement to be seen accurately. Misperception and misunderstanding are the norm in social intercourse, and in many instances can be a salutary rather than problematic experience. Moreover, as we have already noted,

The Theory and Politics of Recognition    245 one’s identity is not a simple datum, frozen in time and space, waiting to be properly perceived. Rather, it is an amorphous and constantly evolving narrative of interpretation and reinterpretation—​in the first instance conducted from one’s own point of view, but ultimately also mediated through dialog with the (also constantly updating) reactions and interpretations of significant others (Taylor 1994; Tully 2000). Identity in the relevant sense is less a static list of personal characteristics susceptible of uniquely “authentic” representation and more a dynamic process of becoming. As recent theorists of recognition repeatedly stress, then, the ethical significance of authenticity-​recognition derives from a set of assumptions about the social conditions under which one’s never fully settled identity can continue to evolve freely and propitiously. On this view, agents will only be able to express their identity authentically if they receive hospitable Treatment, as plants require the right atmospheric conditions, the right soil, the right temperature, etc. to grow correctly. These forms of treatment then define the proper “recognition” for which proponents of authenticity-​recognition call. Of course, the evolution of a human personality is a much more open-​ended project than the biologically programmed growth-​patterns of healthy cacti and petunias. However, even if a person’s identity might develop, under different conditions, in indefinitely many possible directions, we may still be able to diagnose a subset of warped or stunted outcomes, ones that preclude the freedom to express oneself authentically, consigning agents to lives of hypocrisy and alienation. A major preoccupation of theories of authenticity-​recognition is to implicate certain forms of Treatment in such unhealthy conditions, and to recommend (perhaps as a matter of justice) the implementation of various political and legal remedies. Clearly, recognitional arguments along these lines require (on the one hand) a credible psychological account of a “healthy and authentic” sense of one’s identity and (on the other) a plausible explanation of how others’ treatment can support or undermine such identity. Both sets of claims push the project of authenticity-​recognition into controversial territory (McNay 2008). The treatment we receive at the hands of others and the contexts in which we receive it are extremely varied. Claims about how such treatment should or likely will affect the self-​conception of a “normal” individual will often be highly speculative. Fearing that the social and psychological mechanisms involved are too complex and fine-​grained to admit of meaningful public control, many think that theories of recognition are better off to the extent that they avoid reliance on potentially controversial psychological claims of these kinds (Fraser 2003). On the other hand, even if we can’t specify forms of treatment that are strictly necessary or sufficient for healthy identity-​formation, these considerations might still figure loosely in Justification, including justifications that recommend public action as a matter of justice. For example, while they may not completely eliminate racist, homophobic, or sexist attitudes, or guarantee that their targets will be spared a distorted sense of identity, laws prohibiting “hate-​crimes,” or harassment and discrimination against tenants, employees, colleagues, or contractors might still help mitigate an atmosphere hostile to the authentic expression of agents’ identities. At the least, suppressing the public expression of such attitudes must surely reduce their effects on others to some degree.

246   Colin Bird In arguments along these lines, and as our earlier discussion of stereotypes suggested, considerations of authenticity-​recognition and worth-​recognition are easily run together, because the most obviously objectionable feature of (for example) hate speech or various forms of harassment is the contempt and disrespect they involve. However, even when they focus on the disrespectful quality of such conduct, worth-​recognition and authenticity-​recognition arguments handle it in importantly different ways. It is one thing to object to expressions of disrespect because they directly violate expectations of worth-​recognition requiring that agents not be openly demeaned or devalued. It is another to suggest that such expressions are problematic because they tend to infiltrate their victims’ own self-​conception and thereby distort the expression of their identity. This second sort of objection doesn’t invoke a prohibition on demeaning treatment as such (or as uncivil, painful, undeserved, etc.). Rather, it postulates a psychological vulnerability in agents’ capacity for self-​understanding. Viewed from this angle, disrespect is potentially oppressive insofar as it encourages victims to acquire constricted and inauthentic conceptions of themselves and their futures. A related reason to keep worth-​recognition and authenticity-​recognition distinct is that disrespect and other forms of personal devaluation might aid, rather than threaten, free and authentic self-​exploration. Mill’s famous essay On Liberty, while not usually cited as a manifesto for identity politics, nonetheless makes a telling counterpoint here, for he insists that healthy and authentic self-​development often requires that one confront hostile, even offensive, attitudes towards one’s own beliefs and way of life. The value of free discussion in an open society, he insists, lies in the way in which it brings controversy and disagreement out into the open, in turn forcing adherents of different modes of living to confront views and practices that contradict their own presumptions. This is often an unsettling experience, leading agents to question the value of their commitments, in extreme cases shaking core aspects of their identity. But, as Mill plausibly suggests, far from promoting authentic self-​development, a world in which one’s identity and form of life receives only positive affirmation and is never challenged would actually be fatal to it (Mill 1993). Worth-​recognition and authenticity-​recognition are not only different but also sometimes at odds. The tendency to gloss over this contrast is well illustrated in Axel Honneth’s influential theory of recognition (Honneth 1996), which attempts to revive some of Hegel’s early theories about social conditions of identity-​formation by reformulating them in the light of twentieth-​century empirical psychology. Honneth argues that recognition occurs on three levels, in each case corresponding to distinctive social milieux: that of love (the realm of friendship, intimacy, and especially family life), that of respect (the realm of legal rights and equal citizenship), and that of esteem (the realm of civil society, private association, and professional activity). He maintains that these are all necessary if agents are to develop what he describes as “healthy” or “positive self-​relation.” However, because his account emphasizes these strongly evaluative categories (love, respect, and esteem), Honneth’s notion of “positive self-​relation” (Honneth 1996: 163, 173–​4) is ambiguous between two different possible senses. In a first sense, agents have a “positive self-​relation” to the extent that they can see in others’ reactions an

The Theory and Politics of Recognition    247 affirmation of their worth or value. However, “positive self-​relation” might instead refer to a form of healthy self-​acceptance that is undistorted by delusion and in that sense authentic. The distinction between these is subtle, but to see how they can come apart, consider the Christian doctrine of sin. That doctrine offers an account of “healthy self-​relation” that turns on an account of who I am—​a “sinner.” As with contemporary theories of recognition, my identity as a sinner is not, in the Christian account, simply a fixed datum about my nature. It consists, rather, in various dispositions (selfishness, inconsiderateness, aggression, pride, etc.) revealed in my interactions with others over the course of my life, and that sober introspection is supposed to make all too painfully apparent. Achieving an honest apprehension of my moral career as a sinner promotes healthy self-​ relation, on the Christian account, because it exposes my need for salvation. However, it is precisely not to affirm my worth or value: to the contrary, it is to arrive at an unalloyed recognition of my deficiency and to pursue ways of coming to terms with it through various devotional practices. To be sure, it is also part of the Christian conception of salvation that believers are taken to be worth saving. But on the Christian account this element of worth-​ recognition derives, not from reflection on who one is, but rather from a largely separate theological commitment—​the postulation of a creator God who loves His creatures so much that He is prepared to endure penal torment to annul their sins. This is an independent, detachable, component of the Christian view. Modern secular psychological theories (including those cited by Honneth) dispense with it even as they promote something like the honest acceptance of personal deficiency we find in Christianity. An aspiration that many post-​Freudian psychotherapeutic approaches share in common is to help patients to come to peaceable terms with the reality of desires, impulses, and preferences within them, whose value they may never be able to, and even should not, affirm (morally or otherwise). It is perfectly coherent to maintain that these features are authentically part of one’s identity as it works itself out in the struggle of day-​to-​day life, even though grasping them honestly forces one to recognize one’s unworthiness and inadequacy. One might respond that even psychotherapy presumes the patient to be worth curing, and so in that sense necessarily involves a minimal form of worth recognition, at least between patient and therapist. This may often be true, but it isn’t obviously necessary. A therapist might have nothing but contempt for her patient and believe that the pain caused by his neuroses is entirely deserved—​suppose she is a former rape victim who now works in prisons and he is a convicted rapist. Yet she might still think that, in her capacity as a psychiatrist, she ought to set aside such judgments and administer a curative regimen intended to liberate him from psychological suffering. Her situation is not essentially different from that of a doctor who believes her war criminal patient doesn’t deserve the pain relief that she is professionally required to prescribe him. What generates confusion in the therapy case is the fact that the cure involves conscious self-​reflection rather than the application of an external agent like morphine. But it is quite reasonable to think that such self-​reflection will be curative/​emancipatory only to the extent that it encourages

248   Colin Bird patients to confront (rather than deny or be dishonest about) their deficiencies for what they are. If this is right, there will be important differences between defending recognitional policies on the grounds that they promote authenticity-​recognition and doing so in the name of worth-​recognition. Political arguments of these two types are likely to make divergent policy recommendations: think, for example, of the difference between a national history curriculum that seeks to glorify a nation’s past and one committed to an honest acknowledgment of historical atrocities and injustices committed by its leaders. And even when their recommendations converge, they will often arrive at them by importantly different routes. Theorists concerned with the recognition of identity have not, in my view, taken these possibilities seriously enough, and they deserve further investigation. I now return to the postponed question of Independence in the context of authenticity-​recognition: are arguments for certain policies that claim that they promote healthy and authentic identity-​expression likely grounded on justice? With one important exception, I believe that this suggestion is implausible. The exception is group-​ targeted policies—​ affirmative action or reparations providing paradigm cases—​justified as compensation for past and ongoing injustice. Consider one way in which past injustices can continue to adversely affect those in the present generation who belong to groups that have been historically oppressed. Victims of such abuse readily acquire habits and dispositions of fear, servility, diffidence, personal insecurity, reticence, dissimulation, and apathy. They will tend to express such dispositions in the home and in the wider culture of their oppressed community, where children will often unreflectively absorb and imitate them. This pattern can persist across generations. Insofar as such psychological phenomena interfere with the healthy development of individuals in the current generation, long after the original injustices occurred, one could quite reasonably think that they are owed compensatory measures for the sake of authenticity-​recognition. While the entitlement to compensation also rests on other, probably more significant, forms of enduring disadvantage, such identity-​based considerations clearly add something distinctive to the case. (For more on compensatory justice, see Spinner-​Halev 2012.) Beyond this exception, however, Independence is overwhelmingly plausible in the context of authenticity-​recognition. Doubtless societies that institute extensions, accommodations, expansions, and other recognitional measures on the grounds of authenticity-​recognition do something admirable, civilized, and worthwhile, but it seems unlikely that their beneficiaries have a just entitlement to them. Notice that the claim here is neither that justice requires various measures to preserve citizens’ status as “equals” (a “status-​recognition” argument), nor that a just society must resist demeaning and abusive social practices (a “worth-​recognition” argument). The current suggestion is that society has a responsibility founded on justice to ensure that individuals’ identities develop in a healthy and authentic way. This proposal is problematic for at least two reasons.

The Theory and Politics of Recognition    249 First, the enactment of any recognitional measure will impose costs on citizens, both monetary (in the form of taxation) and nonmonetary (e.g., forgone opportunities). Even when particular citizens are not taxed to support such policies, a democratic society still enacts them in their name. On any plausible view, justice constrains the extent to which a government can forcibly impose costs on, and act in the name of, democratic citizens with diverse interests and commitments of their own. To say that citizens are entitled, as a matter of justice, to policies promoting healthy and authentic identity-​formation, then, assumes that states may legitimately impose the costs of those policies on citizens and do so in their name. However, that assumption needs independent demonstration and presupposes a level of agreement among citizens about the requirements of “healthy  and authentic” identity development that seems completely fanciful, at least under pluralist conditions. As our earlier discussion of Christian conceptions of “positive self-​relation” indicates, such notions are highly contested, and often bound up with religious commitments. The absence of any uncontroversial, canonical account of “healthy identity development,” gives a strong reason to locate the ethics of authenticity-​ recognition outside the domain of justice. Second, in practice, it turns out to be very difficult to pursue recognitional policies on authenticity-​based grounds without targeting particular census-​identities for favorable treatment in a way that threatens the impartiality and equality that seem essential to just­ice. Proponents of authenticity-​recognition like Charles Taylor often stress how membership in particular communities contributes crucially to healthy identity-​formation. As a result, their political focus is invariably on policies and legal action intended to help (members of) specified groups. Taylor’s own case for special accommodations for the Francophone community in Quebec is exemplary in this regard (Taylor 1994). He sometimes defends such action as required by “equal recognition,” but that invocation seems rather tendentious here. Targeted accommodations justified on these grounds seem to say that the achievement of authenticity for members of some groups is a more urgent public concern than it is for those unaffiliated with those groups. Far from promoting equality, one could quite reasonably see such accommodations as instituting unjust privileges (Bird 2004). On the other hand, the source of this problem isn’t so much the interest in authenticity-​ recognition itself, as the strong stress that Taylor and those he has influenced place on communal and group-​identity as a precondition for authentic identity-​expression (Parekh 2006). However, as many critics have pointed out, collective identities are themselves often normalizing and exclusionary, extracting considerable psychological taxation as the price of conformity (Butler 2006; Brown 1995; Connolly 2002; Markell 2003). Assimilating to a particular identity-​group may actually come at the cost of authenticity and healthy identity-​expression, not be a condition for it. Worries along these lines turn concern for authenticity-​recognition as we have construed it here against some of the political causes with which it is historically associated. For here authenticity-​ recognition shows its true colors as a child of nineteenth-​century romanticism, and its characteristic commitment to the expression of individuality. The ultimate protagonist in authenticity-​recognition arguments may then be the romantic hero, striving

250   Colin Bird to be true to themselves in the face of a hostile social environment. This is an ironic result, in that proponents of identity-​politics often cast themselves in the tradition of great collective resistance movements like the Marxist campaign on behalf of the proletariat, mobilizing various communal identities to combat the abstract individualism of liberal democracy. Against this, one can quite reasonably conclude—​with Mill and his followers—​that a concern for authenticity-​recognition points toward, rather than away from, an individualist orientation.

11.4  Worth-​R ecognition It is not always easy to distinguish between claims about a person’s “status” and their “worth.” To illustrate, consider the invocation of the ambiguous concept of “dignity” in ethical contexts. That concept hovers uneasily between evocations of social and official status (think of the term “dignitary” in the context of international diplomacy, for example) and significations of worth and worthiness (built into the etymology of the word itself, and exposed in the German translation “Würde”). The same ambiguity crops up elsewhere. Judgments about relative importance, inferiority, and superiority are very often cited as reasons for assigning people differential status. Conversely, once status-​orders are in place (whether vertical and hierarchical, or horizontal and egalitarian), and especially once they become common knowledge (Chwe 2003), people inevitably evaluate themselves in their light. That is one reason why low status can often cause individuals to develop what Adler famously baptized as an “inferiority complex.” Punishment is another social practice in which the two overlap, for punishment is at once a public demotion (a status-​modification) and a disgrace, vividly communicating society’s active disapproval of offenders. Still, worth and status aren’t interchangeable categories. The contrast between them is not that between categorical and scalar dimensions of assessment:  something can be judged categorically unworthy or “worthless” just as status can be all-​or-​nothing. Conversely, people can rank more or less highly within some status hierarchy and be judged more or less worthy in relation to some standard of value. The contrast, rather, lies in this: status-​claims characteristically classify something by applying a formal rule of some kind. Their usual function is to put things in their proper places, by locating them (inside/​outside, near/​distant, above/​below, etc.) in relation to a given classification. Of course, classifying people into various categories sometimes implies (various sorts of) evaluations, but it doesn’t have to do so, and clearly valuing need not be a form of classification (to report one’s love for a person is to say something about how one values them, but it is not, thank goodness, to categorize them). Indeed, status-​judgments and value-​judgments often move in opposite directions. For example, to be the champion is to enjoy a certain status (supremacy) in relation to the tournament and its rules, but champions can be worthy or unworthy victors. For that reason, I deny the widely held assumption that having “status as a person” and “worth as a person” are equivalent

The Theory and Politics of Recognition    251 (although I concede that some philosophical theories, like that of Kant, equate the two). The fact that we are both persons certainly implies that we have the same status, but it does not imply that we are equally worthy, nor that personality itself instantiates any sort of worth. Personality might be construed purely as a status, in which case it refers simply to membership in a formal classification. Proponents of recognition implicitly draw this distinction when they argue that principles of toleration are insufficient to fully realize recognitional ideals. Regimes of toleration revolve around various forms of status-​recognition, because they subsist by upholding formal rights to freedom of worship, association, and discussion. Such rights classify religious communities and other groups as having a protected status, but they are not necessarily granted, nor are they upheld, because anyone values the identities and forms of life receiving those protections. To the contrary, toleration necessarily presupposes disapproval of, or at least indifference toward, the tolerated. Because it doesn’t require the affirmation of the value of identity, culture, religion, or diversity etc., it may seem to offer recognition of a problematically thin and grudging kind. To get beyond mere toleration and status-​recognition, then, theorists of recognition like Honneth turn to the concepts of love, esteem, and respect, which were mentioned earlier. Each refers to a distinctive mode in which one might be affirmed in the eyes of others, a different dimension along which worth-​recognition might be plotted. Do these forms of worth-​recognition implicate principles of justice, or are they independent of them? Independence seems most plausible in the context of love. While love certainly involves recognizing a beloved in very particular ways, assessing our loves against expectations of fairness and justice seems close to a category-​mistake, as the proverb “All’s fair in love and war” suggests. The dissolutions of friendships, marriages, and love affairs are often bitter, painful, shabby, and cruel, but the forms of (mis)treatment involved needn’t be unjust. And rarely do we think of love as subject to just and unjust Distribution. Love, moreover, seems more a private than a public matter. To be sure, feminists have powerfully warned against assuming that the “private” sphere is a zone in which norms of justice cannot apply. They have done so, however, not because they want to ensure that love is justly shown, but rather because they see (say) family life as a venue harboring unjust power-​relations between men and women. More generally, the language of duty, obligation, reciprocation, and rights around which norms of just­ ice revolve seems incongruous in the context of loving recognition (but see Liao 2015). Relationships become friendless and loveless to the extent that the parties reduce their association to a list of entitlements and responsibilities, though that is not to say that successful friendships can dispense entirely with expectations of just treatment within them (Gaus 1990: 289–​92; Waldron 1993). Matters seem rather different in the case of esteem, which would seem to admit of Distribution. Indeed, Geoffrey Brennan and Philip Pettit have recently developed an elaborate model of the “economy of esteem” in which the circulation of esteem is analyzed in the way economists understand market exchange (Brennan and Pettit 2005). In ordinary language, we commonly speak of achievements receiving due

252   Colin Bird recognition (or undue neglect), and so it seems natural to think of esteem as being justly and unjustly distributed. Such recognition need not be merely private, as the cases of prizes and public honors illustrate. Interestingly, however, Treatment makes little sense in those cases. Someone who receives the Nobel Prize is not receiving “treatment” at the hands of the prize committee. One possible reason for this is that we see prizes as “earned,” such that the committee is merely ratifying an independent truth about what their recipients are due, rather than introducing anything new. Another possibility is that prize-​giving is simply a special case of judging, special because unlike focal cases of judging (which often occur in the privacy of personal rumination and don’t involve any sort of treatment), it involves a specific public act. However, although esteem seems distributable, and moreover in just and unjust ways, one can doubt—​as do Brennan and Pettit—​that it counts as a kind of recognition in the first place. Esteem, they argue, presupposes, but is not itself, a form of recognition. As it stands, this conclusion is too sweeping given the ordinary use of the term “recognition” but, nomenclature aside, an important point undergirds it. One cannot have esteem or disesteem for someone or something without at least giving it or them enough attention to form a judgment as to its/​their value. In this sense, even disesteem requires a degree of recognition. Overlooking, ignoring, or (worse) deliberately treating someone as unworthy of any attention whatsoever is, in contrast, neither to esteem nor disesteem them. It involves a far more radical form of nonrecognition, in which one fails or refuses to reckon a person as worthy of any engagement at all. Ostracism is perhaps the most extreme case of such nonrecognition, though milder forms are legion. Ostracism’s psychological potency can be attested by anyone who has experienced its most common contemporary form—​the “silent treatment.” Its distressing quality is not a matter of one’s formal status going unrecognized. It is profoundly disturbing because one’s total exclusion from another’s consideration conveys utter contempt. It gives the impression that one is not worth attending to or reckoning with, and hence that one is effectively worthless in the eyes of another. This sort of worth-​recognition involves the bestowal and withdrawal of respect. These considerations lead me to resist Honneth’s tendency to cordon respect within the realm of legal, civic, and other formal types of recognition. It is not that respect is irrelevant in this context, but that our interest in respect (and abhorrence of contempt) transcends the issue of formal status and reflects a deeper concern about worth-​recognition. Respect engages this concern because, as Kant perceived, respect commands the will, on the basis of some attention-​grabbing encounter with the respect-​object but in ways that are unwelcome, even painful, to the respecter (here respect and love contrast strongly). Gaining respect is a matter of extorting a kind of recognition from others that comes at some cost to the respecter, a cost measured by the heed they are prepared to “pay” them. Judging others’ willingness to incur such attentional costs and to signal it in their conduct is thus one way to gauge our worth in the eyes of others. This is a vital insight about how the respect of others affirms the social worth of its objects. The ability to reliably divert others’ attention from their interests and onto one’s own claims, especially when those others have incentives to resist them, reflects one’s de facto importance in

The Theory and Politics of Recognition    253 others’ eyes. Conversely, the inability to command the attention of others, to be excluded from their consideration, can shatter any confidence that one counts for much among one’s fellows. It would be very surprising if respect in this sense had nothing to do with justice. As Kantians insist, respectful worth-​recognition seems obviously relevant to reflection about justice at the level of Justification, grounding an ideal of the Rechtstaat. But the justificatory link between a formal conception of moral worth and a juridical ideal of civic equality need not be the only, or even the most interesting, way in which respect and distributive justice are connected. For consider, in closing, the widening income inequality that afflicts Western—​ especially Anglophone—​liberal democracies. Over the past 40  years these societies have devoted less and less attention to the plight of the disadvantaged, choosing to celebrate instead the badges of affluence, conspicuous success in the marketplace, and the value of autonomous self-​help. To the extent that such a trend exists, one can reasonably conclude that disadvantaged members of these societies increasingly lack the power to command attention to their predicament. Rather than showing a willingness to incur the costs of committing themselves to serious remedial measures, their fellow citizens are conspicuously indifferent to the problem. In some cases, this is just a matter of complacency. In others, it is something more sinister, as when commentators indulge ignorant and self-​flattering stereotypes of “welfare Queens” and “scroungers” to divert attention away from the real conditions faced by disadvantaged members of our societies. The contempt and disrespect that this expresses is not an abstruse matter of justification, but can be regarded on its face as an unjust, or at least objectionable, form of Treatment. And, insofar as we think of respectful attention as a distributable resource, one could also see it as a form of unjust Distribution. In either case, what is at stake is social worth and its proper recognition. One could quite reasonably conclude that the economically disadvantaged (and other, similar situated, groups) are the Third Estates of our societies. If so, their wish to be something in the eyes of their fellows is neither a demand for authenticity-​recognition nor for status-​recognition. Rather, they seek a kind of power: the power to extort respect from an unwilling and self-​absorbed society, and thereby to know that their social worth is affirmed rather than denied.

References Appiah, A. (2005). The Ethics of Identity. Princeton, NJ: Princeton University Press. Barry, B. (2001). Culture and Equality: An Egalitarian Critique of Multiculturalism. Cambridge, MA: Harvard University Press. Bird, C. (2004). “Status, Identity and Respect.” Political Theory 32(2) (April): 207–​32. Bird, C. (2007). “John Rawls, Lectures on the History of Political Philosophy: Lectures on the History of Political Philosophy.” Ethics 117(4) (July): 784–​90. doi:10.1086/​518675. Brennan, G. and Pettit, P. (2005). The Economy of Esteem: An Essay on Civil and Political Society. Oxford; New York: Oxford University Press.

254   Colin Bird Brown, W. (1995). States of Injury:  Power and Freedom in Late Modernity. Princeton, NJ: Princeton University Press. Butler, J. (2006). Gender Trouble: Feminism and the Subversion of Identity. Routledge Classics. New York: Routledge. Chwe, M. S.  ​Y. 2003. Rational Ritual:  Culture, Co-​ordination, and Common Knowledge. Princeton, NJ: Princeton University Press. Connolly, W. E. (2002). Identity, Difference:  Democratic Negotiations of Political Paradox. Expanded edn. Minneapolis, MN: University of Minnesota Press. Fraser, N. (2003). Redistribution or Recognition? A Political-​Philosophical Exchange. London; New York: Verso. Gaus, G. F. (1990). Value and Justification: The Foundations of Liberal Theory. Cambridge Studies in Philosophy. Cambridge, UK and New York: Cambridge University Press. Gutmann, A. (2003). Identity in Democracy. Princeton, NJ: Princeton University Press. Honneth, A. (1996). The Struggle for Recognition:  The Moral Grammar of Social Conflicts. 1st MIT Press edn. Studies in Contemporary German Social Thought. Cambridge, MA: MIT Press. Jones, P. (2018). “Cultural and Religious Minorities,” in Serena Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 576–96. Kukathas, C. (2003). The Liberal Archipelago:  A Theory of Diversity and Freedom. Oxford Political Theory. Oxford: Oxford University Press. Kymlicka, W. (1995). Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford; New York: Clarendon Press; Oxford University Press. Liao, S. M. (2015). The Right to be Loved. Oxford: Oxford University Press. MacIntyre, A. (1984). After Virtue. 2nd edn. South Bend, IN: Notre Dame University Press. Markell, P. (2003). Bound by Recognition. Princeton, NJ: Princeton University Press. McNay, L. (2008). Against Recognition. Cambridge: Polity Press. Mill, J. S. (1993). Utilitarianism; On Liberty; Considerations on Representative Government; Remarks on Bentham’s Philosophy. New edn. The Everyman Library. London; Rutland, Vt; Dent, MN: Tuttle. Modood, T. (2007). Multiculturalism. Themes for the Twenty-First Century. Cambridge, UK and Malden, MA: Polity Press. Parekh, B. C. (2006). Rethinking Multiculturalism:  Cultural Diversity and Political Theory. Second edn. Basingstoke, UK; New York: Palgrave Macmillan. Rawls, J. (1971). A Theory of Justice. Cambridge, MA:  Belknap Press of Harvard University Press. Rawls, J. (1999). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press. Raz, J. (1994). “Multiculturalism: A Liberal Perspective,” in Joseph Raz (ed.) Ethics in the Public Domain. Oxford: Clarendon Press, pp. 155–​76. Sandel, M. (1982). Liberalism and the Limits of Justice. Cambridge: Cambridge University Press. Spinner-​Halev, J. (2012). Enduring Injustice. Cambridge, UK; New York: Cambridge University Press. Steele, C. M. (2010). Whistling Vivaldi: And other Clues to How Stereotypes Affect Us. New York and London: W. W. Norton & Company. Taylor, C. (1985a). “Atomism”, in Philosophical Papers vol. II. Cambridge: Cambridge University Press, pp. 187–​210. Taylor, C.  (1985b). “What’s Wrong with Negative Liberty”, Philosophical Papers vol. II. Cambridge: Cambridge University Press, pp. 211–​29.

The Theory and Politics of Recognition    255 Taylor, C. (1994). Multiculturalism:  Examining the Politics of Recognition. Princeton, NJ: Princeton University Press. Tully, J. (2000). “Struggles over Recognition and Distribution.” Constellations 7(4): 469–​82. Waldron, J. (1992). “Minority Cultures and the Cosmopolitan Alternative.” University of Michigan Journal of Legal Reform 25(751): 751–​94. Waldron, J. (1993). “When Justice Replaces Affection: The Need for Rights,” in Liberal Rights. Cambridge: Cambridge University Press, pp. 370–​91. Walzer, M. (1982). Spheres of Justice. New York: Basic Books. Wolin, S. (1993). “Democracy, Difference, and Re-​Cognition.” Political Theory 21(3): 464–​83. Young, I. M. (1990). Justice and the Politics of Difference. Princeton, NJ:  Princeton University Press.

Pa rt  I I I

F OU N DAT IONA L DI SP U T E S

Chapter 12

Distribu tive J u st i c e and Hum an Nat u re Paula Casal

12.1 Introduction Political philosophers have invoked different conceptions of human nature, from the darkest to the most sanguine, to support different conclusions about, for example, the role of government and education, and the limits of legitimate coercion and private property. The debate between these rival conceptions continues to this day, as evidenced, for example, in the contrast between John Gray’s The Silence of Animals (2013) and Steven Pinker’s The Better Angels of our Nature (2011). While the former, appealing to history, literature, and anecdotes, denies progress and the transformative power of reason, the latter, appealing to evolutionary science and statistical analysis, argues there has been a decline in violence across time and defends the power of empathy and reason. Being part of political philosophy, distributive justice is not isolated from these debates about what human beings are like. But how is it affected by them? Few would deny that understanding human nature is instrumentally relevant to designing institutions that protect distributive justice. But is the kind of animal we are also non-​instrumentally relevant to distributive justice? Can evolutionary scientists uncover facts that would lead us to revise our view of what justice requires? The defence of the non-​instrumental relevance of human nature is often associated with conservatism, as traditionally it has been conservatives from Aristotle, Hobbes, Spencer, Oakeshott, and Schmitt to John Gray himself who have invoked nature to support anti-​egalitarian conclusions and argued that the demands of distributive just­ ice are far more modest than progressive reformers claim. Distributive justice, they have often insisted, cannot require radical reforms, because justice cannot require implementing proposals that, given human nature, are bound to fail, perhaps with disastrous consequences.

260   Paula Casal A different appeal to nature, made to deny that distributive justice favours significant social reforms, claims that even if various egalitarian reforms are feasible, justice does not require them, because justice favours redressing only inequalities that are the product of unjust social institutions or practices, such as racial discrimination, and does not oppose inequalities with a natural origin. This second strategy replaces the empirical claim that the change progressives propose is impossible with the normative claim that it is not required. Although both may have similar implications, whilst the first stresses the continuity between the natural and the social and may even understand the social order as a prolongation of the natural order, the second strategy stresses the discontinuity between the natural and the social in order to justify different institutional responses to inequalities that may otherwise be identical but have different origins.1 Like the first, this second strategy also implies that scientific findings can alter our understanding of what distributive justice requires. So, for example, it could become justified to reduce compensations or social support for some medical conditions if scientists discover that they are caused mainly by genetic factors. There is also a third view grounded in evolution that may also weaken the impulse to reform society. On this view, even if reform is feasible and necessary to eliminate socially created inequalities, we are not really required to engage in any egalitarian reform. This view argues that any egalitarian convictions and sentiments we possess are attitudes that were evolutionarily advantageous for our ancestors to possess regardless of whether they were justified or unjustified (see, e.g., Street 2006, 2011; Joyce 2006). Philosophers who endorse evolutionary debunking explanations of morality may still support certain reforms they believe will have consequences, such as reductions in waste or unnecessary suffering, which they find preferable or desirable. But they would not defend them by appeal to justice or equality as they see those values as mere useful fictions or tricks of nature for achieving certain adaptive results. There is not sufficient space here to address the latter two, more specific views. Instead, this chapter focuses on the widespread assumption, present across the political spectrum, that evolutionary data favours conservative doctrines, offering ammunition against progressive social reform. It also discusses various ways in which human nature may have more than merely instrumental relevance to distributive justice. It proceeds thus. After section 12.2 sketches some aspects of the kind of animal we are, to start with a less vague notion of ‘human nature’, section 12.3 explores various responses to the claim that evolutionary science has conservative implications. sections 12.4 and 12.5 then, turn to various ways in which human nature can be non-​instrumentally relevant to distributive justice. Section 12.4 focuses on the rights or interests certain kinds of creatures may have, and section 12.5 on the principles of distribution that may apply 1 

This view is most clearly associated with Thomas Nagel (1997), who advocates this position, employing a distinction between natural and social inequalities or disadvantages derived from John Rawls. The distinction is also present in other philosophers of Rawlsian origin, such as Norman Daniels (2007) and Thomas Pogge (1989: sections 15–​16; 1995; 2000), but does not have the same implications in all authors.

Distributive Justice and Human Nature    261 to them, such as sufficiency, priority, and equality. Section 12.6, finally, concludes with a further response to the conservative appeal to nature.

12.2  Liberté, Egalité, Fraternité:  A Few Words on Humans and Other Mammalian Persons Like other great apes, elephants, and some cetaceans, humans belong to a group of unusually intelligent self-​conscious mammals, capable of mirror self-​recognition, which I shall abbreviate as mammalian persons. All these persons possess exceptionally large brains, are well equipped with mirror neurons and spindle neurons, and display the capacities for imitation, cooperation, reciprocity, communication, and cultural transmission (Cavalieri and Singer 1993; White 2007; De Waal 2009; Goodall 2010; Casal 2011; see also Nussbaum 1992: 216–​20).

12.2.1  L  iberté Mammalian persons can imagine themselves in places other than their present location, have a sense of themselves as thinking beings that exist over time, make plans for the future, and can display long-​term persistence in realizing those plans. As they are also endowed with long-​term emotional memory,2 these plans often involve other individuals chosen on the basis of past behaviour. Some argue that these uncommon features give rise to an interest in liberty that is non-​reducible to the interest in suffering-​ avoidance that they share with other sentient creatures (Cochrane 2012). Fish, including goldfish, have much longer memories than commonly assumed, and can experience a decline in welfare if their tank is too small or dirty, or if trapped by the waves in a small pool where they cannot exercise adequately. But they do not resent having been prevented from going somewhere with their friends and relatives, as they typically lack the capacity to form such attachments and to imagine themselves somewhere other than where they in fact are. A person trapped in a cage, however, may feel indignation towards a deliberate interference with her plans, and experience not only a decline in welfare but also the deprivation of liberty that results from confinement to a life other than the one she had planned, and continues to want.

2  This type of memory is different, for example, from the spatial memory required to find one’s nest, migration route, or where a nut was buried, and involves the involuntary re-​emergence of specific emotions associated with somebody or something—​like the evocative muffin Marcel Proust described—​ and is associated with reactions like gratitude and resentment.

262   Paula Casal

12.2.2  F  raternité Unlike most animals, mammalian persons are not r-​strategists but extreme K-​strategists, that is, they invest enormously in very few offspring rather than very little on an enormous number of descendants, as most animals, being r-​strategists, do. The attention required by each offspring is such that pregnancies are typically singleton and followed by several years of lactation and infertility. Among orangutans, for example, births are interspersed by thirty-eight-​week pregnancies followed by six to eight years of lactation, thereby avoiding sibling competition, and allowing mothers to devote a decade to educating each offspring. In contrast with species where survival depends on the ability to kill, outcompete, or even eat one’s siblings, among great apes, like humans, though certain rivalries also exist, fraternity is a survival-​enhancing sentiment, particularly welcome for younger siblings and overworked mothers. The great educational investment needed by members of these species requires enormous empathy and maternal self-​denial, but pays off because mammalian persons are very long-lived. Elephants, for instance, gestate for twenty-two months and often remain under their mother’s supervision until their mid-​teens, but can then survive eighty years or more. Major educational investments tend to make more sense in more longevous species and also tend to involve additional traits. For example, the magnitude of mammalian persons’ maternal investment explains females’ interest in deciding whose offspring they will devote themselves to, and thus their determination to maintain the choice of a mate, and the distress they experience when raped or forced to cohabit with a male they do not want. The magnitude of this investment also explains mothers’ determination to keep their offspring within sight, and the distress they experience when their offspring are taken or killed. Mothers persistently seek and then attempt to resuscitate their offspring, sometimes carrying a corpse for weeks, with the group slowing down its march to accommodate the mother’s erratic behaviour, despite the collective disadvantages of corpse-​carrying and other death rituals. Corpse-carrying has not only been observed in apes, but also in some whales and dolphins. Gorillas and elephants even assist with burials and may accompany mothers who revisit graves. When it is the needed mother that dies, older siblings can act as their surrogates, which is another reason why fraternity enhances survival for mammalian persons.

12.2.3  Egalité While males can in principle sire thousands of offspring, females can have very few offspring, and having invested in long pregnancies and lactation, they tend not to abandon them, even when they are abandoned themselves. This permits fathers to abandon their offspring to pursue other females, who will raise more offspring for them. This imbalance in procreation investments creates a tendency towards inequality. It makes male reproductive success depend on the ability to kill or defeat the greatest number of sexual

Distributive Justice and Human Nature    263 rivals, rather than on empathy and altruism towards one’s offspring, as the offspring’s mother’s empathy and altruism normally suffices to avoid the offspring’s death. In virtually all animals this polygynous pattern (involving one male mating with several females) results in males acquiring certain secondary traits. Compared to females, males become (i) larger and better armed or ornamented; (ii) more aggressive; (iii) more drawn to competitive interaction and aggressive play; (iv) more likely to engage in escalating violence, leading to injury or death; (v) more prone to high-​risk behaviour, particularly when pursuing females; (vi) more eager to mate; (vii) less discriminating about mates; (viii) more likely to die prematurely in accidents, combat, or from disease; (xi) less long-​lived through physiological malfunction, such as testosterone-​induced heart attacks; and (x) conceived and born in larger numbers, which roughly balances their dying prematurely in larger numbers from violence, disease, malfunction, and imprudence (Thornhill and Palmer 2000: 37). In animal species like pipe​fish or jacanas where the males do the parenting work, it is large, territorially aggressive females that compete for male labour, as it is not maleness as such but rather inequality in childcare that gives rise to this trait set (Trivers 1972). Besides being the product of inequality in childcare, these secondary differences can lead to further differences or tertiary inequalities, such as when the larger, more aggressive sex monopolizes the most nutritious food source, or when the larger, more aggressive individuals achieve greater reproductive success and multiply their resource control through extended families. Throughout human history, political and religious leaders have gained power through armies of descendants raiding other groups, thus increasing their resource control and impregnating—​genetic testing suggests—​astonishing numbers of women (Beitzig 1986). For example, 8 per cent of North Asians, and 0.5 per cent of humans appear to descend from Genghis Khan and his close male relatives (Zerjal et al. 2003). Polygyny is thus connected not only to inequality between the sexes and within each sex, but also to both intra-​group and intergroup inequality and violence. These conflicts can in turn be prolonged and exacerbated by various forms of in-​group bias, such as nepotism or racism. Ethnographic data on primitive human societies shows both that they were clearly stratified and also that polygamy was common. Marxists were in fact quite wrong to assume that securing sexual access to women through resource control was a recent bourgeois invention and that humans were equal until technological development enabled them to produce a surplus that could feed an idle class (Cancian 1966). Over time, however, humanity appears to have become less dimorphic, polygynous, and brutal (Pinker 2011; Hrdy 2011; Casal 2013). The human species now clearly displays a growing tendency to become and remain pair-​bonded, like the small apes (Henrich et al. 2012). Some have even argued that, in fact, humans combine aspects of the three reproductive strategies of the other apes, including the promiscuous bonobo (Stewart and Thomas 2013). Both the parenting cooperation of the monogamous species and the matriarchal social organization of the bonobo are associated with reduced inequality. Until recently, moreover, humans lived in very extended families, with fraternal bonds linking many in one cohort, which could contribute to the existence of egalitarian relations within that circle. Humanity’s success is also due to its high level of cooperation (Bowles

264   Paula Casal and Gintis 2011), a trait associated with egalitarian instincts in other highly cooperative species like the capuchin monkey (De Waal 2009: 171, 176, 187, 193). Finally, we were foragers during most of our evolution,3 a mode of subsistence associated with deeply egalitarian, luck-​sharing practices (Henrich et al. 2005; Hauser 2006: 90ff). Given all these traits, humans could be expected to possess strong egalitarian propensities. Inequality among humans, however, is enormous, and growing, even as monogamy also grows (Milanovic 2011; Piketty 2014). Like other primates, humans display respect for property through practices like begging (Brosnan 2011), considering unpicked fruits and empty caves to be unowned, and fruits in somebody’s hand and occupied caves as taken (Maynard-​Smith and Parker 1976). They consider losing something as worse than not gaining it and, much like other species that depend on constructions, like the birds that attack those who steal twigs from other birds’ nests, they may defend, not only their own producer entitlements, but also those of others (Stake 2004: 1763). None of this, however, necessarily leads to inequality. A tendency to respect first occupancy, for example, can protect weak occupants and work against anyone occupying more than one spot, and the same applies to a tendency to defend the nests or other constructions of absent or weaker individuals. Humans may well have natural propensities with contrary effects (see, e.g., Boehm 1999). However, the views that are encountered more frequently are (i)  those which attribute all behavioural patterns to specific cultures; and (ii) those which see the persistence of inequality throughout history as proof of homo sapiens’ non-​egalitarian nature. I believe both of these common views to be mistaken, but turn to the latter first.

12.3  Responses to the Conservative Conservatives often argue that inequality is a fact of life, or at least of human life, and that any attempt to eliminate it is futile. Evidence of inequality, either in past societies or in other species, confirms for them that inequality is not just widespread but a fixture of society:  ‘The poor will always be with us!’ In response to the claim that we should at least try to reduce inequality to whatever extent it can be reduced, they argue that attempts to reduce it will have undesirable effects or even make matters worse (Hirschman 1991). For example, charity creates dependency, food aid causes increased population growth among the poor, and a reduction in incentive payments leads to drops in productivity, detrimental to society at large. It is thus better, they argue, to let ‘nature’ take its course. In response to evidence of successful reductions in inequality and poverty, they then argue that, even if reducing inequality were feasible to some

3 

Agriculture appears to have emerged around 11,000 bc, half a million years after homo sapiens, and 100,000 years after anatomically modern humans and Neanderthals.

Distributive Justice and Human Nature    265 extent and not self-​defeating, it would not be worthwhile, as it would jeopardize something else of great value, such as privacy or freedom, or would require intolerable levels of indoctrination. Whatever social disadvantages or inequalities reformers attempt to address, human nature will make the reform futile, counterproductive, or cause something else of great value to be lost, because whatever institutions are in place, our true nature will always assert itself. This type of discourse diverts our attention away from genuine normative discussions about whether a distribution is fair or unfair, towards evidence from biology or social science that allegedly shows that the best feasible social order is some kind of prolongation of the natural order, with its hierarchy, minimal government, traditional families, and sexual division of labour. Such discourse, vivid in structural functionalists like Herbert Spencer, Kingsley Davis, and Wilbert Moore, often insists on the functional necessity of stratification, or portrays everything as fitting together into an organic whole in such a way that changes in one element will cause severe problems elsewhere. The most common response to the conservative is to challenge (i) the natural origin of the facts they invoke; and (ii) their unalterable character. Philosophers often challenge also (iii) the normative implications conservatives attribute to the facts. For example, they argue that we cannot assume that, to use Pope’s verse, ‘Whatever is, is right’, for the fact that something is natural or universal lacks normative implications.

12.3.1 Challenging the Natural Origin of the Alleged Facts The most common response to the conservative’s ‘appeal to nature’, particularly among political philosophers, is to blame all unattractive human features on human culture instead, and insist on educational malleability. J. S. Mill and G. A. Cohen, for example, argued that employing economic incentives to make talented individuals work hard at socially useful occupations would be unnecessary under different social and cultural circumstances (Mill 1848: 211ff; Cohen 2009). Others seem to assume, more generally, that expressing a priori scepticism towards any evolutionary findings is an essential aspect of egalitarian or feminist philosophy. As Louise Anthony (2000: 8) writes, ‘[M]‌ost feminists today are suspicious of any appeal to the notion of “human nature” in normative or political theory.’ Even universal patterns of behaviour we share with most mammals are assumed to originate in each society’s specific culture.4 The conservative, undiscriminating appeal to the natural origin of all problems or traits is thus replaced by an equally undiscriminating appeal to the social origin of everything. Such a reaction is unnecessary because unjust social practices can be rejected as unfair even if they are not deemed the sole cause of all problems. One may, for example, criticize gender 4  For example, Rob Sparrow argues that if women in all cultures have a lesser propensity to violence than males, it is because of the expectations each culture has of males. So, for example, ‘US troops went to war in Iraq . . . to benefit others’, as such altruism is expected of North American men (Sparrow 2013: 734).

266   Paula Casal discrimination, without also deeming discrimination the only cause of all gender differences that, combined with social factors, give rise to objectionable inequalities. After all, a social practice that worsens a major, pre-​existing problem could be even worse than one that creates a new one. Reformers can also emphasize that conservatives have not shown that our admittedly limited malleability is incompatible with some specific changes. It is also wrong to assume that insisting on the power of education is necessarily progressive or liberating. ‘Give me a child and I’ll shape him into anything’, said Skinner, with Jesuits and Leninists making similar remarks. The idea that anybody can be moulded into anything given sufficient training can be very oppressive, as have been assumptions about the social construction of any behavioural differences between men and women or gay and straight individuals. Such beliefs have supported practices like that of turning boys without normal male organs into girls or re-​educating individuals’ sexual orientation (see Kipnis and Diamond 1998). An indiscriminate aversion to any evolutionary findings is, moreover, self-​defeating because the assumption that evolutionary findings can only undermine proposals for social reform is in fact far more worrying than anything conservatives could ever say: everybody expects your political opponents to undermine you. It is far worse if science proves you wrong. Instead of rejecting scientific findings, social reformers can invoke findings about societies that achieved what conservatives deem incompatible with human nature. For example, if Cubans and Scandinavians are human, whatever levels of domestic equality, environmental protection, and cosmopolitan solidarity these humans display is not something humans are biologically incapable of reaching. Is scepticism towards human evolutionary science so widespread on the Left merely because conservatives appeal to it? Not necessarily. The aversion might also be due to egalitarians sharing the conservative belief that whatever has a natural origin cannot be changed. But this belief is also questionable.

12.3.2 Challenging the Inalterability of Natural Facts We assign an evolutionary origin to our bodies and health, and yet we are extremely effective at modifying them. We attribute our mood and personality to culture and experience, and yet we repeatedly fail to manage even minor alterations. Thus, some authors argue that if we understood human behaviour in more natural terms, we could also become more effective at modifying it. For example, if people could test and advertise their possession of the so-called ‘monogamy gene’, they could contribute to the spread of such genes. Criminal behaviour has also been linked to genes, brain lesions, and physiognomy, and so there may be scientific methods of reducing proclivity to crime (Rainer 2013). Or consider, for example, the ultimatum game, where individuals are allowed to keep 100 units of a benefit provided that they find somebody willing to share it with. There appears to be a genetic basis to judgements of fairness, as twins tend to make or reject the

Distributive Justice and Human Nature    267 same offers (Wallace et al. 2007). In addition, there is evidence that, besides triggering birth and breast milk production, oxytocin causes individuals to make more generous sharing offers, whilst testosterone is associated with meaner options and greater willingness to punish those who do not offer what one wants (Zak 2007, 2009). Oxytocin also boosts trust and in-​group empathy, whilst serotonin can increase cooperation and reduce aggression, and ritalin seems to reduce violent hostility. In fact, an increasing number of drugs, routinely taken for a variety of conditions, are found to have an impact on individuals’ moral conduct, for example, by making people either calmer or more irritable. Given these findings, some believe that in a not-​so-​distant future it will be technologically possible to enhance, not only the welfare, but also the moral character of humanity. Reflecting on such a possibility, some philosophers have begun to advocate investing in research into how to make humans more empathetic and altruistic through biomedical means such as drugs and genetic manipulations (Persson and Savulescu 2012). This new reply to the conservative view is that we cannot resist social change by appealing to human nature, because human nature can itself be changed. And it could be argued that by locating the origins of our social failings in human nature, it is conservatives that have inadvertently paved the way for this new conception of social reform.

12.3.3 Challenging the Normative Implications of the Alleged Facts Those who reject moral bio-​enhancement may still endorse the general form of the preceding response, which is that all the conservative view can offer is some reasons why attempting change in certain ways may not work. Conservatives cannot prove that all means, including those not yet discovered, are doomed to fail. And so long as creating a fairer society remains merely difficult, rather than impossible, it still makes sense to discuss just shares. This is the basis of a third response to the conservatives that many would find appealing, and perhaps more appealing than the preceding responses of either the social-​constructivist or the advocates of moral enhancement. Suppose, for example, that scientists discover that blue-​eyed students are faster at calculus but slower at problem-​solving than dark-​eyed students. Whether such natural differences result in economic or political inequalities depends on social rules. In themselves they need not create any differences, even in the students’ marks in mathematics, their exams include questions involving a variety of tasks. Since scientific findings do not normally include the additional finding that nothing whatsoever could ever be done about the newly found facts, the newly discovered facts may lack non-​instrumental consequences for policy. Political implications or claims about just outcomes cannot be read off mechanically from evolutionary findings, unless we also accept a conservative conception of justice which privileges the status quo, accepts only minimal intervention, and does not condemn a society where individuals’

268   Paula Casal prospects in life can be predicted from facts about their birth. Thus, whether a particular finding has consequences for the allocation of just shares may not depend only on the finding, but also on the theory of distributive justice we adopt. Consider, for example, the issue of the greater risk of accidental suffocation for newborns that sleep with their fathers, or step-​fathers, rather than with their mothers.5 However large this increase in risk may be, we need not infer that mothers must be the primary carers. We may instead require that additional tasks be performed only by fathers in compensation or that babies sleep in cots. When we then find that co-​sleeping with babies causes a drop in testosterone in fathers, which in turn makes them become better parents (Gettler et al. 2012), we may then encourage men to co-​sleep with their babies placed in a protective in-​bed cot. Evolutionary findings, thus, rather than guiding our goals, may provide valuable information about how to achieve them safely, so that, for example, greater gender equality is not accompanied by increased infant mortality. Does this mean that the Left is therefore stuck with the claim that evolutionary findings can only be of instrumental importance? Not necessarily. One can, of course, describe the significance of any finding as merely instrumentally useful for attaining the ultimate goal of protecting certain important interests, or of reducing suffering or unfairness in the world. However, without denying this very general claim, one may argue that the kind of creature we are may alter what constitutes suffering, our critical interests, or unfairness, and also alter more specific distributive norms regulating the achievement of these overarching goals in different contexts. Thus, whether we choose to describe the relevance of science as merely instrumental may also depend on our philosophical standpoint. For utilitarians, for example, everything may be of mere instrumental importance relative to utility maximization. Other philosophers with more Aristotelian inclinations, like John Rawls or Martha Nussbaum, for example, will not see the relevance of human nature as merely instrumental, because instead of positing a prior good, like utility, that can be independently defined, they think that the sort of creature we are determines what is good for us, and not only what makes us flourish but also what constitutes flourishing for us. Section 12.4 elaborates on this point.

12.4  How Human Nature may Matter: Interests Reacting to the conservative emphasis on human nature, some have denied not only the relevance but even, as Marx famously claimed, the existence of human nature. As

5  UK statistics led the NHS to issue leaflets in 2003 discouraging father–​baby co-​sleeping. It seemed prudent because, even leaving hormonal effects on sleep patterns aside, men are larger, heavier, more likely to be heavily smoking or drunk shortly after the birth, and lack the compensating benefits of breast ​ milk.

Distributive Justice and Human Nature    269 Norman Geras (1993) has argued, however, not even Marx would have denied that human nature imposes certain needs that not only play an explanatory role in the study of our past, but also a normative role in explaining what a just society should not deny us. The sort of creature we are explains, for example, why we have certain fundamental interests and why all normal members of our species have them. Consider, for example, the interests in non-​interference, pain  ​avoidance, or continued existence to which I referred earlier.

12.4.1 The Right to Life, Liberty, and the  Pursuit of Happiness 12.4.1.1 Liberty Like other mammalian persons, we possess a long-​term emotional memory that enables us to experience gratitude and resentment and recall with whom it is safe to cooperate without risk of being exploited or betrayed. Such sentiments and capacities have played an important role in our developing an interest in justice. But they are also connected to an interest in others not interfering with our plans to engage in activities with certain other individuals, whose past performance with us we recall. Thus, while children and animals may lack an interest in non-​interference, or even have a deep interest in our interfering with their unsuccessful or ill-​conceived efforts (Horta 2013), humans and other mammalian persons have a profound interest in non-​interference not only with their life plans, but also with more specific endeavours. And if liberty matters independently of its possible contribution to some other goal like aggregate utility, then some of our moral arguments and claims are not independent of the kind of creature we are, and the relevance of human nature is not merely instrumental.

12.4.1.2 Happiness The fact that children or animals may lack an interest in being free from other individuals’ interference is consistent with their having an equal interest in pain avoidance. In fact, children and non-​human animals can sometimes experience greater suffering than adult humans because of their inability to make sense of the pain they experience or their fear of something harmless like mere darkness. Certain capacities, however, can also increase suffering. For example, while a fish could reach the end of the pond repeatedly and see it each time as if for the first time, creatures with brains designed to be constantly active, transmitting and absorbing complex information, can suffer deeply from boredom and monotony. The informational desert of four prison walls, for instance, whilst innocuous for some animals, can be devastating for cultural animals evolved to process information constantly. Moreover, being intensely social and emotional creatures, solitary confinement can be worse than pain for such animals. Physical pain itself can be intensified by the anticipation of a torturer’s future actions and their likely long-​term consequences, and by fearing comparable fates for

270   Paula Casal our loved ones. The long-​term emotional memory characteristic of mammalian persons also prolongs the resulting suffering, transforming us forever into individuals who were tortured once. Despite all this, we may, however, prefer torture to some alternative outcome. For having second-​order preferences, we may endure hunger, for instance, to later have enough food for our loved ones. The kind of animal we are, thus, determines how we suffer, as well as the not always predictable ways in which we may find happiness, perhaps by choosing to suffer for certain goals.

12.4.1.3 Life Consider now our interest in continued existence. Like other mammalian persons, humans have long lives during which they acquire and transmit their culture and discover part of their world along with others with whom they form intense bonds, weaving interesting biographies. They thus lose something very valuable when their lives are taken from them. If this were the only consideration (Nagel 1979: 1ff), however, we would have to conclude that the death of a child would be less bad than a very early miscarriage, as nobody has a greater life-​expectancy than a very early embryo. One may argue that the death of a child is (or appears to be) worse, because it is more traumatic for those who have bonded with the child over the years. Given our intensely social, interdependent, and emotional nature, our death is likely to affect others for whom our death may be even worse than for ourselves. These deep bonds could make the death of children not only worse for their family, but also worse for the children, because they too can have an interest in their family faring well, and because children who will leave a devoted family behind have, in that respect, more to lose when losing their life. But this cannot be the whole story either, for an early miscarriage is never as bad as the death of an orphan, who lacks parents but perhaps hopes to find some, and has other hopes, aspirations, and plans. We are such purposeful, future-​oriented creatures, that once we become capable of thinking beyond the moment, we develop our most deep and intimate relationship with our future selves. This connection eventually becomes so intense that we start to live mentally in our future, directing virtually all present actions towards it. Thus, death will not only deprive the orphan of potential experiences but will suddenly sever this deep bond with the future which gives meaning to our lives. The badness of death is harder to explain in animals that have no psychological continuity and so do not appear connected to their earlier or future selves (McMahan 2002). This is not only because individuals who forget what happens to them very quickly cannot engage in the sort of activities that will make their lives more valuable to them. If the importance of psychological continuity were explained merely by the contribution it makes to our having a more valuable existence, we would not have made any progress in explaining why the death of a child is worse than an early miscarriage, for embryos too will eventually have a life with the added value of all the activities that psychological continuity permits. Psychological continuity must thus matter at least in part because killing us involves severing the deep and intimate bond we have with our future self.

Distributive Justice and Human Nature    271

12.4.1.4 Rights Given the importance of psychological continuity, it is hard to explain why it matters if the fish we see in the pond is the same we saw last year or its identical descendant, when the fish itself does not know this and cannot care about whether this is the case. When we learn from the pond carer that one fish died and five offspring survived, or that the old one had to be put down so that the other five could survive in the pond, we do not storm out appalled. By contrast, it would be shocking to hear about an old tourist being drowned so that five young ones could enjoy the pool. Some argue that this is because the sort of creature we are determines also our moral status. Agnieszka Jaworska (2007), for example, agrees with Judith Jarvis Thompson (Thomson 1990: 292) that we can kill a chicken to save five, but denies we can kill a mammalian person like a gorilla to save five, because gorillas are creatures that ‘care’ and thus possess full moral status. Whatever our view on the permissibility of such trade-​offs, it is because of the kind of animal we are that we believe that both the gorilla and the tourist have a very weighty interest in continued existence, non-​interference, and suffering avoidance that any ethical theory on the fair distribution of benefits and burdens must accommodate. In fact, a theory of justice should also accommodate other fundamental interests of normal members of our species. For example, if we had thousands of limbs that we could easily replace by re-​growing them, we would have less stringent prohibitions on the non-​ consensual use of our limbs than we currently have (Lippert-​Rassmusen 2008). If instead of being nomads and explorers we were the sort of animal that never leaves its birthplace and cannot even survive elsewhere, we would not have several articles of the Universal Declaration of Human Rights referring to the right to live anywhere within one’s state, to abandon it, to live elsewhere, and to return to it. It is also appropriate that, being extreme K-​strategists for whom it is extremely important to maintain mate choice, we should insist that ‘marriage should be entered into only with . . . free and full consent’ (United Nations 1948: art. 16). It is possible to imagine other rational creatures for whom some of what matters to us, such as bodily integrity, free movement, mate choice, or occupational freedom, were not as important as they are for us. As Rawls (1999: 372–​80) notes, discussing what he calls the ‘Aristotelian Principle’, the fact that we prefer performing complex tasks over simple tasks that require no training is a contingent fact with deep ethical implications. The same could be said about other contingent facts about humans, such as our fascination with unique works of art that very few can own or produce. Scientific explanations of how and why doing certain things to members of certain species displaying some contingent traits causes extremely deep, lasting, or even irreparable harms can support arguments in favour of protecting those individuals from those harms. Such scientific explanations could then cause us to increase the stringency of certain prohibitions and alter our judgements about the distribution of certain burdens and benefits, and our ethical theory more generally. For Rawls, for example, the Aristotelian Principle bears on discussions of the human

272   Paula Casal good, self-​realization, self-​respect, rational life-​plans, and the principles that protect them, thus playing a central role in developing a plausible theory of justice for humans, though perhaps not for extraterrestrials (Rawls 1999: 364ff, 372ff 380ff, 398, 413, 500ff, 510).

12.5  How Human Nature may Matter: Principles Some will grant that human nature is relevant to defining basic needs and capabilities, human rights, or conditions for flourishing or self-​realization, but consider it irrelevant to determining the principles of justice that evaluate the distribution of those benefits. Let us examine this option by considering the main principles in turn.

12.5.1 Principles of Sufficiency, Priority, and Equality The principle of equality is a comparative principle which states that it is bad if some are worse off than others through no choice or fault of their own. By contrast, the principles of sufficiency and priority are non-​ comparative principles of distributive justice. Sufficiency states that what matters is that individuals have enough, and the principle of priority holds that the moral value of receiving a benefit or the moral disvalue of bearing a burden diminishes as its recipient becomes better off. Roger Crisp (2003: 749–​50) dismisses equality as an illusion derived from our evolutionary past, proposing a debunking explanation of egalitarian sentiments as side-​ effects of envy, generalized through sympathy. He, instead, advocates sufficiency, which he thinks not only escapes evolutionary debunking explanations but is a principle which can be applied without any reference to the sort of creature we are. On his view, ‘eighty years of high-​quality life on this planet is enough, and plausibly more than enough, for any being’ (Crisp 2003: 762). It is odd, however, that a particular number of Earth orbits that suspiciously coincides with Crisp’s own life-​expectancy should be the right principle for any creature from any galaxy. Even within our planet we know there are species that live over two centuries and others that live only hours, and species that survive by hibernating or slowing down their metabolism, and others that have short, hectic lives. It thus seems very odd to say that all these very different creatures would have a sufficiently good life when they reach their eightieth birthday, which is something only a handful of species can do. Most of the members of the small club of the longevous, moreover, such as some urchins, lobsters, or reptiles, may never have lives of ‘high quality’ because they lack most of what we think gives value to life. According to Crisp, benefits to those who have less than enough have lexical priority over benefits to those who have

Distributive Justice and Human Nature    273 enough, and the principle does not hold just for humans. Thus, we must benefit prawns over lobsters, snakes over tortoises, quails over macaws, and mice over eighty-​year-​old humans, elephants, or whales (as according to Crisp, even bowhead whales, the most longevous mammals, have had enough by the end of their eighth decade). It seems odd, moreover, that longevity should be taken into account in species lacking psychological continuity. Lobsters may be no more connected to their childhood than a prawn to another prawn. It thus seems more plausible to understand that whether or not we ‘have enough’ also depends on the kind of creature we are, just as whether we have enough to survive or enough to flourish depends on the sort of creature we are. This flexibility could actually be seen as an advantage of sufficiency over other principles, for it provides a way of avoiding the implausible implications of an inter-​specific application of the principle (Vallentyne 2005). Once we accept that what counts as ‘enough’ (to survive, to flourish, and so on) cannot be the same for all species, we become liberated from the implausible commitment to dropping the dolphin to aid the sardine. Let us now turn to the principle of priority (Parfit 1995). The principle is sometimes understood as a variant of utilitarianism (on other variants, see Smart and Williams 1973). Since utilitarianism is a consequentialist, welfarist, aggregationist doctrine, the priority view, understood as ‘tilted utilitarianism’, is also assumed to be consequentialist, welfarist, and almost aggregationist (see Temkin 1993: xii, 8, 31ff, 69, 245ff, 282, 303; Scanlon 1998: 235; Arneson 2000). Like utilitarianism, priority takes into account not only ‘how numerous’ are the potential beneficiaries and ‘how much can they benefit’, but also (pace utilitarianism) how badly off they are. If prioritarianism is thus understood, it would be very difficult for it to avoid giving priority to the sardines (Vallentyne 2005). For how much is ‘enough’ can vary from one species to another, but ‘the worst off ’ sentient creature could be a fish. To be sure, the principle’s implementation would still require species-​specific information. Imagine we want to apply it to armadillos. We would need to know what dimensions of armadillo existence should be evaluated, when clear misfortunes in other species such as blindness or solitude, for example, do not matter to them much. We need to have some sense of what, if anything, besides a sense of smell and sharp claws, makes armadillo life go well. ‘Tilted utilitarianism’ however, would still require giving priority to the worst-​off armadillo and to armadillos over humans, if armadillos are worse off. This implication and other objections to the priority view (e.g. Otsuka and Voorhoeve 2009) can be avoided if we understand it in different terms, for example, not as a consequentialist ‘telic prioritarian’ principle, but as ‘deontic prioritarianism’ (Nagel 1979: 106ff; Parfit 1995; Williams 2012). In addition, we may also stop seeing prioritarianism as a welfarist doctrine and focus on capabilities, resources, or primary goods instead. Thus reformed, the priority view would have more resources than utilitarianism (tilted or otherwise) to avoid the conclusion that we must never help the dolphins before the sardines. For example, it can be seen as part of a contractualist doctrine, like Nagel’s, or simply as a principle to settle the competing claims between individual members of the same group, like the group of persons or moral agents. If so, we may face decisions

274   Paula Casal involving dolphins and other mammalian persons, but not sardines, and can thus avoid the problematic implications of an inter-​specific application of the principle. Let us finally consider equality. An inter-​specific application of some understanding of the principle will also, counter-​intuitively, require us to devote our resources to make all individuals from species with lower welfare or capabilities as well off as us (Vallentyne 2005). Equality, however, seems to make sense only within certain species. It would be odd, for example, to claim that all armadillos are equal. This may be due to speciesist bias (Persson 1993: 192) or to other reasons. For example, the standard interpretation of the principle of equality applies across complete lives, which assumes a psychological continuity armadillos lack. Armadillos are not persons, do not live in a society that can be equal or unequal, and would not contest an unequal distribution or be affected by it. Despite being born always in groups of four identical quadruplets, they go on to live solitary lives, and cannot experience anything like what Rawls (1999: 468) called ‘excusable envy’ or probably any sort of envy at all. There are non-​envy-related reasons to contest unequal distributions, but contesting unequal distributions seems to be something that generally, solitary non-​cooperative creatures like armadillos would not do. By contrast, the more dependent on cooperation, luck-​sharing, and reciprocity a species or tribe is, the more it is likely to reject individuals or states of affairs contrary to the sharing spirit that is essential for its survival. For example, since hunting depends on luck and since meat is so nutritious that even a very small amount of meat can save somebody’s life, it is adaptive for hunters to develop a strong ethos of luck-​sharing. The same applies to other precious findings like honey. So when the foraging Ache of Paraguay were told they could keep 100 units of something if they found somebody to share it with, they offered potential sharers over 40 per cent, during an international, interdisciplinary study of fairness that employed the ultimatum game (Henrich et al. 2005). This response fitted with the general cooperative tendencies of the Ache, where hunters invariably share their catch with the rest of the camp (Hauser 2006: 91). This cultural explanation of their response is compatible with the idea that responses to the ultimatum game have a genetic basis (Wallace et al. 2007), since it would make sense for the Ache to have inherited tendencies appropriate to their environmental adaptation. The lowest offers of all the people sampled in this international study were made by the slash-​and-​burn horticulturalist Machiguenga of Peru, who offered as little as 15 per cent. This also fitted with their lifestyle, which involves little cooperation, trade, or sharing beyond the family unit. Within all groups, moreover, individuals also displayed different dispositions to share when they were told different stories about other individuals’ contribution to the good up for distribution. The general conclusion of the study was that humans in any part of the world would not make offers above 50 per cent and would reject offers of less than 15 per cent, and just like the birds that defend other birds’ nests, they believe that those who construct something have stronger claims over it than others. Like different tribes, different species which depend on sharing, reciprocity, and cooperation to different degrees will display stronger or weaker propensities to share, reciprocate, or cooperate. Capuchin monkeys, for example, may need cooperation to survive even more than the larger chimpanzees, and so may display a stronger disposition

Distributive Justice and Human Nature    275 to suffer in order to punish non-​cooperators in ultimatum games (Jensen et al. 2007; De Waal 2009: 187; Proctor et al. 2013). It would be very odd if a preference for egalitarian distributions and a disposition to punish non-​cooperators emerged among armadillos, as it would be surprising for an egalitarian sentiment to have emerged if inequality were innocuous for the species that appears to have such sentiments. Inequality does appear to be harmful for species that reject it. Wilkinson and Pickett (2009), for example, have identified over a dozen important problems in any society that worsen as inequality grows. Inequality correlates, for example, with worse mental and physical health, higher crime rates, more drug addiction, greater pollution and waste, and lower life-​expectancy. It is possible, of course, that the sorts of creature that possess egalitarian sentiments are also more vulnerable to the stress of grossly unequal relations. If so, one may argue that the harmfulness of inequality could not, or not fully, explain the emergence of such sentiments, since the harmfulness of inequality plausibly depends, partly, on the existence of egalitarian sentiments. After all, it is plausible to assume that inequality must be perceived in some way, at least subconsciously, in order to inflict harm. It is also plausible, however, that certain egalitarian practices, such as food sharing, and certain egalitarian sentiments, such as discomfort in situations where only some have food, have co-​evolved. The harmful effects of departing from equality could be the predictably detrimental effects of departing from an adaptive practice and the sentiments that sustain it. According to Wilkinson and Pickett, if some could not eat, the group’s health, including the health of those who eat, would be worse to an extent that goes beyond what could be attributed to the mere food abstinence of one individual.6 Consider now Wilkinson and Pickett’s case of the rise in teenage pregnancy as inequality grows. In K-​species like ours with long life-​expectancy and very few offspring, females do not start procreating at the earliest opportunity but delay maternity while their prospects improve. The prospects of career women typically improve as they age, encouraging them to delay maternity to the limit of what biology permits, and even beyond, through in-​vitro fertilization. Their very late maternity contrasts with the very early pregnancy of poor women whose only asset is their youth and their not-​yet-​ damaged health. Young women that live in pockets of poverty, crime, and drug-​dealing in developed societies see their prospects rapidly deteriorating, with friends in their age group arrested, involved in gang feuds, or damaged by drugs. Life in such environments can cause fertility and teenage pregnancies to increase, giving rise to a perverse cycle of poverty and premature pregnancies. One may argue that the issue here is poverty, rather than inequality, but Wilkinson and Pickett could reply that various related factors such as crime are in turn linked to inequality. Also, some adversities, like lacking a father or decent clothes are particularly hard when others have both supportive parents and decent clothes. In addition, teenage mothers are competing in both the love and the labour markets with others who have assets other than youth. So the relevant comparison is 6  Note that accepting that practices and sentiments have evolved in some way does not in itself imply a debunking explanation. There could still be correct norms for individuals who have evolved certain needs and capacities.

276   Paula Casal not just between current and future selves, but between a young woman and women of various ages with assets other than youth. Motherhood, moreover, is a path to a desired adult status others can achieve through employment and other means. Finally, the authors’ statistics link the problem to inequality rather than poverty (Wilkinson and Pickett 2009: 116, 122–​3). In any case, and regardless of how this particular dispute is resolved, the example illustrates once more how evolutionary information can be relevant to our distributive goals by informing us of how harmful some distributions can be to a certain type of creature and how vicious circles may emerge. Thus, evolutionary data may not only help us identify the best policies to achieve our distributive goals but may also help us see how urgent or beneficial reaching these goals could be. In light of this conclusion, should egalitarians continue to respond to conservatives by dismissing evolutionary findings and denying the importance of biology in general? Consider an alternative strategy some authors employ.

12.6  Addressing Conservatism Conservatives, it seems, were right after all to argue that human nature is relevant to identifying, not only the means, but also the goals (at least the intermediate, rather than the ultimate goals) of distributive justice. Conservatives also advocate some values for which one can find additional support in evolutionary science. For example, there are differences between men and women with evolutionary roots, such as women’s lower propensity to crime (US Department of Justice 2000; Wilkinson and Pickett 2009: 132). Also, functional families are extremely important for all their members and society at large, and are associated with greater psychological stability and lower crime rates. In addition, the bond between mother and child is important to the development of normal individuals, and although both in humans and other primates it is a vehicle for the reproduction of class inequality, we should not attempt to sever it (Bowlby 1951). It is erroneous, however, to derive from these sound conservative judgements that we should respect not only the parent–​child bond but also any inequality that results from this and other less valuable associations. This is a non sequitur. Leaving inequality intact may in fact be detrimental to the realization of family values or other values conservatives typically advocate, such as stability, security, law and order, health, longevity, or drug control. Conservatives may think otherwise because they may be right about the importance of human nature but wrong about human nature itself. For a very unequal society of chronically unemployed youths, absent parents, and areas of poverty, crime and drug ​use is not conducive to the protection of the very values they cherish. If we want to reduce teenage pregnancies, births out of wedlock, chronically unemployed individuals turning to crime, drug ​use, violence, social instability, illness and premature death, it is good to have public education and health care, and a more equal distribution of wealth and work, both between men and women and within each sex. Conservatives may be right about the importance of protecting certain family values but wrong about

Distributive Justice and Human Nature    277 how to secure their protection (see Brighouse and Swift 2014). For example, they could be wrong about what works nowadays or wrong about our past. In fact, some of the most central conceptions about human evolution conservatives employ are currently being challenged by scientists. Frans de Waal (2006, 2009), for example, has collected a large number of conservative statements advocating a mistaken view of human nature he labels ‘veneer theory’. Veneer theory portrays humans as selfish, callous, greedy individuals, precariously restrained by a thin veneer of civilization but aching to come out and realize their nasty goals. In fact, as de Waal and others have shown, the animal inside is already an empathetic creature, capable of sharing, cooperation, reciprocity, and self-​sacrifice, with such forms of conduct becoming increasingly well documented among many mammals. In addition, comforting the bereaved, adopting orphans, helping the elderly, or caring for other species is not even uncommon among female apes (De Waal 2006, 2009; Goodall 2010). Biology, thus, offers no evidence that ‘if you scratch an “altruist” you see an “egoist” bleed’ (Ghiselin 1974: 247; De Waal 2006: 10) and contains in fact overwhelming contrary evidence. Empathy, sharing, cooperation, and reciprocity are already part of our nature. Consider now the myth of ‘Man the hunter’, a widespread view of our origins which depicts useless females passively awaiting the return of heroic male hunters on whom they entirely depended. Seen in this light, female economic dependence on men appears as an immutable aspect of the human condition, rather than as a product of childcare inequality, discrimination, and labour market segregation. It is thus unsurprising that appealing to ‘Man the hunter’ should be a recurrent conservative theme. Scientists like Jared Diamond (1993), however, have argued that ‘Man the hunter’ was, and still is in hunter-​gathering societies, largely an embellished male fantasy. The humans that settled near the North Pole in the last stage of human expansion depended on hunting. The survival of most humans for most of our history, however, depended, like that of other tool-​using apes, on a humble search for fruits, nuts, roots, shoots, insects, molluscs, and so on, with hunting large animals being most likely the exceptional feat, commemorated in cave walls, and repeated in fireside tales (Diamond 1993: 32ff). Among other mammalian persons, females do not need males, or need them only to cope with other males, and are entirely self-​sufficient at feeding themselves and their offspring. In fact, it is only or mainly females we see termite fishing or sweeping, nut-​cracking with hammers and anvils, and, in the case of dolphins, employing sponges for protection whilst fishing and air bubbles as toys. In fact, some argue that it was only because of the great alimentary efficiency of female humans that males could afford to go on hunting expeditions without the risk of starving if they had no luck: they could always fall back on the female harvest, particularly after its caloric content was increased by female efforts to make it more tender for children by mashing and cooking it (Wrangham 2009: 130ff). Females benefited from males going away hunting even if they did not bring back much because then they could eat in peace with their offspring, instead of having to share their niche with larger, hungrier, and more aggressive individuals. The resulting exploitation of different niches is widely believed to have been key to human success. It also fits with

278   Paula Casal the evidence of early women focusing on agriculture whilst men continued foraging (Ogilvie and Hilton 2011). Questioning alleged implications is easier for philosophers than checking alleged facts, and discourses that emphasize the irrelevance of human nature tend to discourage the more arduous search for accurate data. It is possible, and preferable, however, to perform both tasks. Stephen Macedo’s response to the conservative opposition to gay marriage illustrates nicely the effectiveness of this dual strategy. Macedo begins by citing—​with some approval—​a passage from Germain Grisez (1993: 662, in Macedo 1996: 27) arguing that whilst sexual liberation was supposed to increase general welfare, it did the opposite, for: . . . the pain of sexual frustration is slight in comparison to the misery of abandoned women and unwanted children, of people lonely for want of true marital intimacy, of those dying wretchedly from sexually transmitted disease. Moreover, unchastity’s destructive effects on so many families impact on the wider society, whose stability depends on families. . . . Boys and girls coming to maturity without a solid foundation in a stable family are ill-​prepared to assume adult social responsibilities.

Macedo’s critical response is not that conservatives like Grisez or John Finnis weight these different values incorrectly. Nor does he deny that traditional marriages are stability enhancing. He just notes that promiscuous, sensation-​seeking behaviour is part of male rather than gay nature. Lesbian couples are remarkably stable and faithful, and sex is less frequent or important in their relationship than it is in heterosexual couples (Macedo 1996: 33). This is not the case with male homosexuals (Posner 1992: 306ff), because they are male, not because they are gay. If heterosexual males did not have wives and children and the institution of marriage, perhaps they would be as promiscuous as gay men. Thus, if marriage is for the faithful, lesbians have stronger claims to it than heterosexuals; and if marriage is meant to encourage faithfulness and we want gay men to settle down, the worst we can do is to deny them the legal right to get married and the human right to found a family. Thus, revising the alleged facts on gay nature, Macedo turns the conservative arguments on their head. This sort of response to the conservative view seems preferable to responses that rely on insisting that everything is a matter of education or that dismiss any evolutionary evidence on the grounds that behaviours which were previously believed to be natural or hard to change later turned out not to be so. Our having been mistaken in the past does not license permanent disbelief. After all, we have historically held false beliefs on virtually everything, and do not now reject all scientific findings as a result. Furthermore, when conservative views have found more support this has been in religion, superstition, and tradition rather than in science. Had society been better informed about evolutionary science earlier, it would have laughed at the suggestion that homosexuality was not natural, that native Americans were not human, that women lacked the intellectual or moral capacities to be rulers or priests, and it would have denied any credence to

Distributive Justice and Human Nature    279 the numerous dogmas of the Inquisition: no scientifically informed society would have believed that torture did not hurt innocent suspects or that women could fly. Such a rational, minimally well-​informed society would not have viewed homosexuals or women as inferior and would have seen whites and albinos for what they are: humans with little pigmentation but without special powers. Political philosophers (and ideally, society as a whole) ought to become better informed about the kind of creature we are, instead of assuming scientific findings must be irrelevant or have, as conservatives believe, only conservative implications. First, even when genetic tendencies exist, it is also scientists that note the importance of environmental triggers and epigenetic factors. Second, genes can be weeded out or be made to become more common, which is just one of the possible means we can now explore to alter human nature. Third, as Louise Anthony (2000: 10) stresses, we should neither agree nor appear to agree that feminist (or more generally, egalitarian) goals depend on ‘there being nothing about us that is in any sense “true by nature” ’. As noted earlier, findings have no direct implications because no finding includes the additional finding that nothing whatsoever can be done about the newly found fact. Finally, there is no reason to assume that evolutionary findings can only legitimize current injustices rather than dispel prejudices. It could well be that, as Lenin said, the truth is not conservative but revolutionary.

Acknowledgements For helpful comments I thank John Horden, Serena Olsaretti, and especially Andrew Williams.

References Anthony, L. (2000). ‘Natures and Norms’. Ethics 111: 8–​36. Arneson, R. (2000). ‘Luck Egalitarianism and Prioritarianism’. Ethics 110: 339–​49. Beitzig, L. (1986). Despotism and Differential Reproduction. New York: Aldine de Gruyter. Boehm, C. (1999). Hierarchy in the Forrest. The Evolution of Egalitarian Behaviour. Cambridge, MA: Harvard University Press. Bowlby, J. (1951). Maternal Care and Mental Health. Geneva: World Health Organization. Bowles, S. and Gintis, H. (2011). A Cooperative Species. Princeton, NJ;  Oxford:  Princeton University Press. Brighouse, H. and Swift, A. (2014). Family Values: The Ethics of Parent–Child Relationships. Princeton, NJ: Princeton University Press. Brosnan, S. (2011). ‘Property in Non Human Primates’. New Directions for Child and Adolescent Development 132: 9–​22. Cancian, F. (1966). ‘Maximization as Norm, Strategy and Theory: A  Comment on Programmatic Statements in Economic Anthropology’. American Anthropologist 68: 465–​70. Casal, P. (2011). ‘Love not War. On the Chemistry of Good and Evil’, in A. Gosseries and Y.  Vanderborght (eds) Arguing About Justice:  Essays for Philippe Van Parijs. Louvain-​la-​ Neuve: Louvain University Press, pp. 145–​57.

280   Paula Casal Casal, P. (2013). ‘Sexual Dimorphism and Moral Enhancement’. Journal of Medical Ethics 39(12): 722–​8. Cavalieri, P. and Singer, P. (eds) (1993). The Great Ape Project:  Equality beyond Humanity. New York: St Martin’s Press. Cochrane, A. (2012). Animal Rights without Liberation. New York: Columbia University Press. Cohen, G. A. (2009). Rescuing Justice and Equality. Oxford: Oxford University Press. Crisp, R. (2003). ‘Equality, Priority and Compassion’. Ethics 113: 745–​63. Daniels, N. (2007). Just Health. Cambridge: Cambridge University Press. De Waal, F. (2006). Primates and Philosophers: How Morality Evolved. Princeton, NJ: Princeton University Press. De Waal, F. (2009). The Age of Empathy. New York: Harmony. Diamond, J. (1993). The Rise and Fall of The Chimpanzee. London: Vintage. Geras, N. (1993). Marx on Human Nature: Refutation of a Legend. London: Verso. Gettler, L. T., Agustin S. S., McDade T. W., and Kuzawa C. W. (2012). ‘Does Cosleeping Contribute to Lower Testosterone Levels in Fathers? Evidence from the Philippines’. PLoS ONE 7(9): e41559. Ghiselin, M. (1974). The Economy of Nature and the Evolution of Sex. Berkeley, CA: University of California Press. Goodall, J. (2010). Through a Window:  My Thirty Years with the Chimpanzees of Gombe. New York: Mariner. Gray, J. (2013). The Silence of Animals. London: Penguin. Grisez, G. (1993). The Way of the Lord Jesus. Vol. II. Chicago, IL: Franciscan Herald Press. Hauser, M. (2006). Moral Minds. New York: Harper Collins. Henrich, J. et  al. (2005). ‘“Economic Man” in Cross-​ Cultural Perspective:  Behavioral Experiments in 15 Small-​Scale Societies’. Behavioural and Brain Sciences 28: 795–​855. Henrich, J., Boyd, R., and Richerson, P. (2012). ‘The Puzzle of Monogamous Marriage’. Philosophical Transactions of the Royal Society B: Biological Science 367: 657–​69. Hirschman, A. (1991). The Rhetoric of Reaction, Perversity, Futility, Jeopardy. Cambridge, MA: The Belknap Press of Harvard University Press. Horta, O. (2013). ‘Zoopolis, Intervention, and the State of Nature’. Law, Ethics and Philosophy 1: 113–​25. Hrdy, S. (2011). Mothers and Others. Cambridge, MA: The Belknap Press of Harvard University Press. Jaworska, A. (2007). ‘Caring and Full Moral Standing’. Ethics 117: 460–​97. Jensen, K., Call, J., and Tomasello, M. (2007). ‘Chimpanzees are Rational Maximizers in an Ultimatum Game’. Science 318: 107–​9. Joyce, R. (2006). The Evolution of Morality. Cambridge, MA: MIT Press. Kipnis, K. and Diamond, M. (1998). ‘Pediatric Ethics: The Surgical Assignment of Sex’. Journal of Clinical Ethics 9(4): 398–​410. Lippert-​ Rasmussen, K. (2008). ‘Against Self-​ Ownership:  There are No Fact-​ Insensitive Ownership Rights over One’s Body’. Philosophy and Public Affairs 36(1): 86–​118. Macedo, S. (1996). ‘Against the Old Sexual Morality of the Natural Law’, in R. George (ed.) Natural Law, Liberalism and Morality. Oxford: Oxford University Press, pp. 27–​48. MacMahan, J. (2002). The Ethics of Killing. Oxford: Oxford University Press. Maynard-​Smith, J. and Parker, G. A. (1976). ‘The Logic of Asymmetric Contests’. Animal Behaviour 24: 159–​75.

Distributive Justice and Human Nature    281 Milanovic, B. (2011). The Haves and the Have-​Nots: A Brief and Idiosyncratic History of Global Inequality. New York: Basic Books. Mill, J. S. (1848). Principles of Political Economy. London: Longmans [1909], Book II. Nagel, T. (1979). Mortal Questions. Cambridge: Cambridge University Press. Nagel, T. (1997). ‘Justice and Nature’. Oxford Journal of Legal Studies 17(2): 303–​21. Nussbaum, M. (1992). ‘Human Functioning and Social Justice:  In Defence of Aristotelian Essentialism’. Political Theory 20: 202–​46. Ogilvie, M. and Hilton, C. E. (2011).‘Cross-​Sectional Geometry in the Humeri of Foragers and Farmers from the Pre-​Hispanic American Southwest: Exploring Patterns in the Sexual Division of Labor’. American Journal of Physical Anthropology, 144(1): 11–​21. Otsuka, M. and Voorhoeve, A. (2009). ‘Why it Matters that Some are Worse Off than Others:  An Argument against the Priority View’. Philosophy and Public Affairs 37(2): 171–​99. Parfit, D. (1995). Equality or Priority? Lawrence, Kansas: University of Kansas Press. Persson, I. (1993). ‘A Basis for Interspecies Equality’, in P. Cavalieri and P. Singer (eds) The Great Ape Project. New York: St Martin’s Press, pp. 183–​93. Persson, I. and Savulescu, J. (2012). Unfit for the Future. Oxford: Oxford University Press. Piketty, T. (2014). Capital in the Twenty-​First Century. Cambridge, MA: The Belknap Press of Harvard University Press. Pinker, S. (2011). The Better Angels of our Nature. London: Penguin. Pogge, T. (1989). Realising Rawls. Ithaca, NY: Cornell University Press. Pogge, T. (1995). ‘Three Problems with Contractarian-​Consequentialist Ways of Assessing Social Institutions’. Social Philosophy and Policy 12(2): 241–​66 and in E. F. Paul et al. (eds) The Just Society. Cambridge: Cambridge University Press, pp. 247–​50. Pogge, T. (2000). ‘Justice for People with Disabilities—​The Semi-​Consequentialist Approach’, in L. Francis and A. Silvers (eds) Americans with Disabilities: Exploring Implications of the Law for Individuals and Institutions. New York: Routledge, pp. 34–​53. Posner, R. (1992). Sex and Reason. Cambridge, MA: Harvard University Press. Proctor, D. et al. (2013). ‘Chimpanzees Play the Ultimatum Game’. PNAS 110, 2070–​5. Rainer, A. (2013). The Anatomy of Violence. New York: Pantheon Books. Rawls, J. (1999). A Theory of Justice. Cambridge, MA: Harvard University Press. Scanlon, T. (1998) What We Owe to Each Other. Cambridge, MA: Harvard University Press. Smart, J. J. and Williams, B. (1973). Utilitarianism: For and Against. Cambridge: Cambridge University Press. Sparrow, R. (2013). ‘Sexism and Moral Enhancement’. Journal of Medical Ethics 39(12): 732–​5. Stake, J. A. (2004). ‘The Property “Instinct”’. Philosophical Transactions of the Royal Society B: Biological Sciences 359(1451): 1763–​74. Stewart, S. and Thomas, A. G. (2013). ‘The Ape that Thought it was a Peacock:  Does Evolutionary Psychology Exaggerate Human Sex Differences?’ Psychological Inquiry 24(3): 137–​68. Street, S. (2006). ‘A Darwinian Dilemma for Realist Theories of Value’. Philosophical Studies 127(1): 109–​66. Street, S. (2011). ‘Evolution and the Normativity of Epistemic Reasons’. Canadian Journal of Philosophy 35: 213–​48; supp. vol. on Belief and Agency, ed. D. Hunter. Temkin, L. (1993). Inequality. Oxford: Oxford University Press. Thompson, J. J. (1990). The Realm of Rights. Cambridge, MA: Harvard University Press.

282   Paula Casal Thornhill, R. and Palmer, C. T. (2000). A Natural History of Rape. Cambridge, MA and London: MIT Press. Trivers, R. L. (1972). ‘Parental Investment and Sexual Selection’, in B. Campbell (ed.) Sexual Selection and the Descent of Man 1871–1971. Chicago, IL: Aldine, pp. 136–​79. US Department of Justice (2000). ‘Women Offenders, Bureau of Justice Statistics Special Report 2000’. https://www.bjs.gov/content/pub/pdf/wo.pdf ​(accessed 1 March 2018). United Nations (1948). Universal Declaration of Human Rights. http://​www.ohchr.org/​EN/​ UDHR/​Documents/​UDHR_​Translations/​eng.pdf/​ (accessed 16 November 2017). Vallentyne, P. (2005). ‘Of Mice and Men: Equality and Animals’. The Journal of Ethics 9: 403–​33. Wallace, B. et  al. (2007). ‘Heritability of Ultimatum Game Respondent Behaviour’. PNAS 104(40): 15631–​4. White, T. (2007). In Defence of Dolphins: The New Moral Frontier. Oxford: Wiley-​Blackwell. Wilkinson, R. and Pickett, K. (2009). The Spirit Level: Why Equality is Better for Everyone. London: Allen Lane. Williams, A. (2012). ‘The Priority View Bites the Dust?’ Utilitas 24: 315–​31. Wrangham, R. (2009). Catching Fire: How Cooking Made us Human. New York: Basic Books. Zak, P. J., Stanton, A. A., and Ahmadi, S. (2007). ‘Oxytocin Increases Generosity in Humans’. PLoS ONE 2(11): e1128. Zak, P. J. et al. (2009).‘Testosterone Administration Decreases Generosity in the Ultimatum Game’. PLoS ONE 4(12): e8330. Zerjal, T. et  al. (2003). ‘The Genetic Legacy of the Mongols’. American Journal of Human Genetics 72: 717–​21.

Chapter 13

P olitical a nd Distribu tive J u st i c e Chad Van Schoelandt and Gerald Gaus

A devotion to distributive justice seems to inevitably involve the political, that is, social institutions creating and coercively enforcing authoritative rules. To most philosophers, unmet claims based on distributive justice imply a political injustice—​some have a complaint of justice against their political system. But how are claims arising from distributive justice linked to claims about the justice of the political? This chapter is divided into two main parts. In section 13.1 we consider a variety of ways that distributive justice has been coupled with justice in the political sphere. Starting with instrumental accounts that view the justice of the political largely in terms of securing independently defined requirements of distributive justice, we move to accounts that more intimately couple the distributive and political, such as those depicting distributive justice as arising out of certain types of political arrangements. Having analyzed a variety of ways of coupling political and distributive justice, section 13.2 argues that the tie between political and distributive justice should be greatly weakened. A justified polity with legitimate authority cannot be required to conform to any specific theory of distributive justice.

13.1  Coupling Distributive and Political Justice 13.1.1 Distributive Justice as an External End of the Political 13.1.1.1 The Exogenous Instrumental View It is widely supposed that political institutions ought to secure distributive justice; but there are different ways that they might do so. On the simplest view, the requirements

284    Chad Van Schoelandt and Gerald Gaus of distributive justice are identified entirely independently of facts about political institutions, which are viewed as means for securing the end of distributive justice. So we can identify: The Exogenous Instrumental View: The requirements of distributive justice, DJ, can be known apart from facts about political institutions, including their existence. Political institutions and norms are modes of regulation that are justified as contextually appropriate means to achieving DJ. Understood thus, distributive justice is an exogenous or external good (in relation to political institutions), while political institutions are just only if they advance distributive justice. On this Exogenous Instrumental View, distributive justice is similar to a good such as health. It may be supposed that a just state promotes the health of its citizens, but we can identify the good of “health” independently of any facts about political institutions. To be sure, a state’s policies might be causes of good or ill health, but a conception of “health” is not endogenous to political institutions, as if health was a good different in democracies and dictatorships. Politics and political institutions do not figure into identifying the nature of the good. G. A. Cohen provides a quintessential instance of this Instrumental View. Distributive justice, according to Cohen (2000: 132), “consists in a certain egalitarian profile of reward.”1 The state of affairs itself, the actual possession of benefits and burdens, is just or unjust. It is irrelevant to distributive justice, on this understanding, whether that profile of advantages was specifically produced by a political policy or by natural chance. It is also irrelevant on this account whether those holding the advantages are co-​members of a single state or lack political relations with each other. It is important to realize that the Instrumental View is compatible with the following statement: Distributive Justice is not Definitive: An institution or policy can be politically justified even if it sacrifices satisfaction of the requirements of distributive justice for other political values. Some may hold that distributive justice (or more generally, justice) is the first and overriding virtue of political institutions (see section 13.1.3), but Cohen does not. He upholds Distributive Justice is not Definitive, as he insists that distributive justice is merely one value among many that a political institution must accommodate; consequently “unjust inequality” may “be part of a package of policy that is, all things considered, superior to any other (because values other than justice weigh in its favor)

1  Cohen (2008: 7) holds that “an unequal distribution whose inequalities cannot be vindicated by some choice or fault or desert on the part of (some of) the relevant affected agents is unfair, and therefore, pro tanto, unjust, and that nothing can remove that particular injustice.”

Political and Distributive Justice    285 or even  . . .  more just than any other (because considerations of nondistributive justice weigh in its favor)” (Cohen 2008: 7, emphasis added). Even if a policy that involved forced labor and gross violation of freedom of occupation would bring about the correct egalitarian distribution, Cohen maintains that it is important for the state to respect these basic freedoms.2 It thus seems clear that on Cohen’s account, even if one aspect of political justice involves advancing distributive justice, that is not the only aspect; the justice of political institutions involves more than the distributions of the benefits those institutions bring about or fail to bring about. That would not entail that the society is not unjust, but that it had moral reasons to not do everything in its power to pursue distributive justice. Though the political institutions are justified in maintaining the inequalities, the society “cannot be through-​and-​through just” (Cohen 2008:  7). Distributive justice also might require more than any political institution can accomplish. Just as a state may be unable to prevent some illnesses no matter how much of the state’s resources are devoted to health, the state may face insoluble problems in promoting distributive justice. Just as a state policy can be merely one factor among many in determining health, state actions might be simply one factor affecting the justice of the final distribution of advantages.3 Thus Cohen also upholds: The Non-​Uniqueness of the Political: Political means are not the only ways to secure distributive justice; they are not generally sufficient and might (eventually) not be necessary. Beyond any political requirements, Cohen (2000: Chs 8–​9) argues that distributive just­ ice requires individuals to guide their private decisions by an “egalitarian ethos.”4 Such an ethos would lead individuals to accept their highest productivity career prospects while forsaking any income beyond what distributive justice would allot them. Indeed, ultimately distributive justice should not require the political at all: In the ideal socialist society, equal respect and concern are not projected out of society and restricted to the ambit of an alien superstructural power, the state. If the right principles are, as Marx thought, the ones that are right for everyday, material life, and if they are practiced in everyday life, as the socialist ideal utopianly envisages they will be, then the state can wither away. (Cohen, 2008: 1)

2 

Cohen (2008: 186) writes that “old-​style Stalinistically inclined egalitarians might have responded by setting their faces against freedom of choice of occupation . . . But my own inclinations are more liberal, so that way out is not for me.” 3  Christiano (2008: 53) similarly holds that political institutions are not logically necessary, but are “one way in which individuals attempt to establish justice among themselves.” 4  Cf. Neufeld and Van Schoelandt’s (2014) discussion of the relation between political and ethos justice.

286    Chad Van Schoelandt and Gerald Gaus

13.1.1.2 Pure State-​Driven Judgments The Exogenous Instrumental View initially appeared to couple distributive and political justice in a simple and compelling way. Distributive justice is understood as a good, independent of the political, and so the task of the political seems crystal clear: to promote distributive justice. When distributive injustice is discovered, we can proclaim: “Something should be done, the political task is to secure distributive justice!” However, we have seen that even on this Instrumental View matters are more complex: the political may have other goals that override distributive justice (Distributive Justice is not Definitive), and there may be requirements of distributive justice that cannot, or should not, be achieved through the political (The Non-​Uniqueness of the Political). Notice that both of these additional claims attenuate the link between distributive justice and political justice. This attenuation reduces the clarity of the political task. This attenuation becomes much more severe when we add a third claim characteristic of Cohen’s analysis: Distributive Justice as Pure State Judgments:  A judgment of distributive justice is the judgment of whether the distribution of some good (G) in a state of affairs (S) corresponds to the requirements of distributive justice (DJ). Because, on this view, judgments of justice are purely judgments about states of affairs concerning G, there is no necessary connection between judgments of justice and what is to be done (Cohen 2000: 132). But recall that according to the Instrumental View, political institutions and norms are modes of regulation that are justified as contextually appropriate means to achieving distributive justice. Now that we have added our three additional claims, the connection between the political institutions and achieving distributive justice becomes rather too attenuated to say anything definite about the appropriate political institutions. In principle, an account that understands judgments of distributive justice as purely state-​oriented judgments might have nothing to say about what to do politically, or might not yield actual imperatives, policies, or interpersonal claims, should it be the case that there are no realistic options that produce just conditions without excessive costs to other politically relevant values. Provocatively, David Estlund (2008: 264) defends such “hopeless visions of justice”: Consider a theory that held individuals and institutions to standards that it is within their ability to meet, but which there is no reason to believe they will ever meet. . . . It would be morally utopian if the standards were impossible to meet, but, again, by hypothesis, they are not. Many possible things will never happen. The imagined theory simply constructs a vision of how things should and could be, even while acknowledging that they won’t be. . . . So far, there is no discernable defect in the theory, I believe. For all we have said, the standards to which it holds people and institutions may be sound and true. The fact that people will not live up to them even

Political and Distributive Justice    287 though they could is a defect of people, not of the theory. For lack of a better term, let us call this kind of theory a version of hopeless realism.5

In a similar vein, Cohen (2008: 20) insists that the infeasibility of a vision of justice does not “defeat the claim of a principle.”6 Such theories might be described as conceptions of “cosmic justice.” Charles Larmore (2013:  292n) contrasts views of cosmic justice with his own concern for “justice insofar as it has some import for social life.”7 Aaron James (2013: 104) similarly contrasts such cosmic justice with a political philosophy that we expect to be “normative for us.” Elizabeth Anderson (2012: §1 1) also criticizes such views, insisting that justice is fundamentally relational and tied to interpersonal accountability. We might also understand justice as having functions, such as adjudication of competing claims and facilitation of cooperation among people pursuing diverse, often conflicting, ends.8 Accounts of Distributive Justice as Pure State Judgments are unsuited for fulfilling that function, given their conceptual independence from our actual moral and physical circumstances. All these complaints pick up on the gulf between judgments of distributive justice and judgments about what a just state must do, require, or prohibit.

13.1.2 Distributive Justice as Weakly Endogenous to the Political: An Alternative Instrumentalist View 13.1.2.1 An Endogenous Instrumentalist View Recall that we commenced with an apparently attractive view of the relation between the political and distributive justice: a critical aim of the political was to pursue distributive justice. The problem with the exogenous account we have explored is that, seeing distributive justice as a good entirely external to the political that describes states of affairs rather than actions or policies, it becomes possible that distributive justice might be a hopeless aspiration which is beyond the ken of the political. There is, however, an alternative conceptualization of the instrumentalist idea that assures an instrumental link between the political and distributive justice. Arash Abizadeh (2007: 324) identifies a view along the lines of:

5 

For an extended analysis, see Estlund (2011). Christiano (2008: 36) likewise holds that distributive justice “does not obey the ‘ought implies can’ principle entirely[,]‌” and may thus present an infeasible ideal. 6  At times Cohen seems to go beyond even Estlund in the defense of the hopeless; while Estlund is a firm defender of “ought implies can” (and so the vision cannot be impossible to meet), Cohen (2008: 20, 250–​4) is ambivalent about the principle. 7  For criticism of this project, see Cohen 2000: esp. 302ff. Many will find “cosmic justice” to be a misnomer unless there is some agent who culpably ordered the cosmos in these unjust ways, be that entity a god or perhaps the cosmos itself as a pantheistic agent. 8  Rawls (1951: 1–​2) and (1999: § 23).

288    Chad Van Schoelandt and Gerald Gaus The Political as a Condition for Distributive Justice: A requirement (DJ) of distributive justice is justified only when (a) there is a current political institution that can secure DJ or (b) it is feasible to create an institution that can secure DJ. The Political as a Condition for Distributive Justice takes seriously Michael Blake’s point that “states of affairs that are open to human control are, morally speaking, distinct from those that are not” (2001: 273). The general upshot is that a principle of distributive justice can only condemn a state of affairs that is “in some sense amenable to control by human agency” (2001: 273). Though being subject to human control seems a weak condition, it is an important qualification imposed by the Political as a Condition for Distributive Justice on any proposed principles of distributive justice. Let us consider two ways in the literature that the common sort of proposals for distributive justice seem to instrumentally require institutions and may be shown inapplicable when the necessary institutions cannot legitimately be established.

13.1.2.2 Institutions to Coordinate Distribution Efforts Supposing a minimally complex principle of distributive justice and a large-​scale society, it is implausible that a just distribution can be secured by the decentralized efforts of uncoordinated individuals. “Given the scale of a modern economy,” Samuel Scheffler (2010: 115) writes, “the establishment and preservation of background justice requires far too much information and is far too complex a task to be accomplished by any set of rules that might plausibly guide individual conduct.” Even if there was a shared substantive understanding of what distributive justice required, the uncoordinated efforts of individuals are unlikely to achieve it. Any proposed distributive standard will be informationally demanding if it involves relative shares, more so if those are shares across time (e.g., requiring equal lifetime income, as opposed to such equality at each moment), and further demanding yet if what the principle demands is at all choice sensitive. These informational demands prevent individuals from reliably even knowing what changes would be moves toward or away from distributive just­ ice. The lack of coordination in action amplifies the problem, for the actions of diverse people changing the distribution may cancel each other out, or even create a less just distribution. Considering the need to gather information and coordinate activity, it seems that most distributive principles will require institutions, and most likely political institutions empowered to command the relevant actions. We note that clause (b) of the Political as a Condition for Distributive Justice (section 13.1.2.1) allows that demands of distributive justice may arise even for agents when they do not currently have the instrumentally necessary institutions. In such a case, it can be added that political justice demands that, if within some range of feasibility, they bring about the required institution. Even so, there are a number of reasons we may find that the relevant political institutions are infeasible. Most obviously, they may not be feasible given resource scarcity or other basic physical limitations, or because they would be unstable or morally dangerous. For example, Rawls (2001:  36), following Kant, argues that a world state

Political and Distributive Justice    289 would be either likely to break up from the disputes between different nations within it, or would become despotic.9 If Kant and Rawls are right on that point, this gives us some reason to believe that cosmopolitan principles of distributive justice requiring a world state are not justified. We should also note that many proposed principles of distributive justice must be rejected because some information gathering and coordination is infeasible even with political institutions. Rawls (1999: 81), for instance, argues that the “currency” of justice cannot be utility, because interpersonal utility comparisons are infeasible. He proposes instead the metric of “social primary goods” such as income, in part because these resources are measurable in a way likely to produce inter-​subjective agreement. We can effectively rule out any principles that use criteria we cannot reliably assess. If no institutions for coordinating distributions are relevantly feasible, then no principles of distributive justice at all may be justified. For instance, Hayek (1945: 528–​9) argues that we rely on decentralized decision-​making within a market order not only for economic growth, but simply to maintain our economic position through changing conditions (cf. Mises [1920] 2009: 105–​6). Drawing on decentralized decisions, he argues, results in distributions that fail to match any substantive distributive principles, and attempts to make the distribution satisfy those principles are destructive of the market order. We can thus see Hayek as arguing that we have sufficient reason to reject institutions that would attempt to coordinate the realization of almost any distributive principles. Such principles, having no acceptable means for their realization, would all be invalidated. Distributive justice is shown to be, as Hayek put it, a mirage.

13.1.2.3 Assurance Among the conditions most difficult to satisfy is that a proposed principle of justice must be such that, should we seek to realize it, we can solve the assurance problem.10 According to Rawls (1951: 236), “reasonable” people are motivated to do their part in a fair system of cooperation “when we believe that others, or sufficient many of them, will do theirs.”11 Though other aspects of morality may be unconditionally binding, for Rawls and many others, distributive justice assumes only the disposition to conditionally comply with fair cooperative schemes: it does not demand leaving oneself open to exploitation though unilateral compliance. Principles of justice, then, organize cooperative relations among those whose cooperation is conditional on the expectation of the cooperation of others. As Paul Weithman (2013: 47) writes, if “citizens are to act from

9 

Many anarchists argue that the threat of despotism applies also to all states. For instance, Mikhail Bakunin ([1873] 1972: 328) writes: “Every state power, every government, by its very nature places itself outside and over the people and inevitably subordinates them to an organization and to aims which are foreign to and opposed to the real needs and aspirations of the people.” We can thus see the outlines of an anarchistic case against distributive justice. 10  For example, Rawls (2005: 49–​50). Cf. John Thrasher and Kevin Vallier (2013). 11  For empirical evidence about the importance of empirical expectations, see Cristina Bicchieri and Alex Chavez (2010).

290    Chad Van Schoelandt and Gerald Gaus their sense of justice consistently, each must have some assurance that others will consistently act justly as well.” Assurance problems do not arise simply because people are self-​interested; if one cannot advance distributive justice alone, then complying with norms of distributive justice simply wastes resources that could have been devoted to mitigating poverty, promoting virtue, or saving souls (James 2013:  108). Solving the assurance problem for any proposed principle is a part of showing that the principles could be stable for the right reasons—​stable through the ongoing rational and reasonable acceptance of, and compliance with, the principles, rather than mere stability through force or manipulation. As James (2013: 104) argues, “proposed principles must credibly address the available human means for the public resolution of problems of assurance.” This line of reasoning leads to: Distributive Justice as a Distinctively Public Duty: Securing distributive justice is a political, or social project, that generates duties for an institutionally organized collectivity, and only derivatively to individuals as members of such a collectivity. Thomas Nagel (1981: 200) endorses this claim, maintaining that individuals might not be morally obligated to do privately what it would be appropriate for the state to force them to do, in part because of the need for assurance. Interestingly, Cohen seems attracted to a version of this claim. He argues that, assuming an egalitarian conception of distributive justice, “no one is obligated to sacrifice so much that she drops to a level worse than what she would be at in an egalitarian society” (Cohen, 2000: 176). Without assurance that others will maintain the egalitarian distribution, however, anyone who privately brings herself down to the egalitarian level risks being driven below that level by “the buffetings of uncompensated bad luck” (2000: 176). So, he concludes, “what goes for the public domain need not go for the private. One might say that for assurance reasons, equality is, necessarily, a social project” (2000: 176).12

13.1.3 Distributive Justice as Strongly Endogenous to the Political: The Justice of Distributing Systems 13.1.3.1 From Skepticism to the Justice of a Distributor A strand of skepticism about the scope of distributive justice common in the classical liberal tradition advances a view nearly the opposite of Distributive Justice as a Distinctively Public Duty (section 13.1.2.3): to such classical liberals, all rules of justice apply primarily to individual acts, and only derivatively to distributional matters. Robert

12 

Though Cohen does not think this need mean a specifically political or institutional project, given his support for an egalitarian ethos.

Political and Distributive Justice    291 Nozick’s famous slogan, “From each as they choose, to each as they are chosen,” expresses this individual decision-​ based understanding of “distributive justice” (1974: 160). According to Nozick, there is no correct pattern of distribution of goods except the actual pattern that happens to emerge from the decentralized decisions of people exercising their individual rights and respecting those of others. One could, of course, point to a current distribution as unjust and indicate what distribution would be just, but such claims would be based on the history of transactions, both rights-​ respecting and rights-​violating, and not merely the distributive pattern. Judgments of the distribution, then, are fundamentally judgments of the just or unjust actions of individuals. Hayek essentially agrees, insisting that the justice of a distribution is generally derivative of the justice of the individual decisions that led to it, where the justice of the action is determined by compliance with social rules and does not reference the distributive outcomes.13 A fundamental pillar of Hayek’s (1976: 117) argument is that the specific distribution emerging from a market order has “not been brought about deliberately.” As he argues, the distribution that emerges is unplanned and not the result of anyone’s intentional choice. The closest we can get to questions of distributive justice for the unplanned market order is to ask whether the system of rules in which individuals make their choices is a good system in that it provides each with good ex ante prospects, though there is no guarantee of how they will eventually do.

13.1.3.2 Distributive Justice as Norms for Political Distributors The classical liberal critique of distributive justice leaves open the possibility that distributive requirements do indeed apply when there is a distributor of resources, such as a state. Hayek (1976: 32–​3) is clear that the rules of justice, as well as the concept of distributive justice, apply to governments and their actions. Consider, for instance, a state official allocating resources controlled by the state, such as police or education services. We may scrutinize the distribution the official chooses, and rightly criticize her if she distributes the resources disproportionately without some appropriate justification. She may not, for instance, favor her family and friends with the public resources, though she may appropriately do so with her private finances. Stanley Benn advances an account of distributive justice that takes as central this idea of the state as distributor. He bases his analysis on: The Principle of Equal Consideration of Interests:  “To justify discrimination in the attention accorded to the interests of different persons, one must be able to point to a difference between them relevant to the discrimination being made. This might be formulated negatively as the principle that there cannot be first-​class and second-​class persons.” (1988: 117)

13 

See Hayek (1976: Ch. 10), particularly pp. 117 and 122. Cf. Lister (2013).

292    Chad Van Schoelandt and Gerald Gaus This principle regulates decisions and actions, and is not in the first instance a criterion for evaluating states of affairs in terms of the degree to which the interests of different persons are satisfied. (Thus it is very nearly the opposite of Distributive Justice as Pure State Judgments, section  13.1.1.2) According to Benn (1988:  118), the Principle of Equal Consideration of Interests does not apply to every decision, but only when “the decision maker had an obligation in this case to deal fairly—​to set aside his preferences for the sake of universally acceptable reasons.” So, for example, he tells us that Caroline need not justify the “discrimination” she expresses in choosing to marry one person rather than another on the basis of personal preference. The fact that the interests of other potential spouses are affected is not sufficient to generate claims against Caroline. For Benn, liberty is the default and the demands of impartial treatment are a special case. On Benn’s analysis, far-​reaching requirements of distributive fairness arise for modern states only because discharging the tasks they have undertaken, as a matter of history, brings into play the Principle of Equal Consideration of Interests. Where everyone now, as a moral person, had a right to vote, and economic management and income distribution became a regular and recognized feature of State policy, everyone’s need became a subject of claims on State consideration, and this as a matter of just distribution. (1988: 250)

Instead of seeing just distributions as an independent good that the state can or should promote, Benn argues that these matters of justice only arise for political institutions that have taken up the tasks of managing and allocating resources. If a state is to take up such tasks, extending beyond the “night-​watchman state,” it must do so in a way that gives equal consideration to the interests of each citizen.14 Benn’s equal consideration principle creates only a limited scope for principles of distributive justice. Mere distributions in the world are of no special concern. As Hayek and Nozick stressed, the distributions that emerge unintentionally from the decentralized decisions of individuals exercising their rights, including the right to show partiality, are not appropriate objects for judgments of distributive justice. Distributive justice in the guise of the Principle of Equal Consideration of Interests applies only to those distributive activities for which impartiality is required. Indeed, Benn’s principle is a norm governing organizational actions, including those of the state, and only indirectly about securing distributions. Distributive justice thus arises as a problem within politics, not as an external problem for politics to solve.

14  According to Benn, “Human welfare rights arise, not directly from a right that one’s needs, even one’s basic needs, be satisfied, but rather from a right to fair treatment, to the equal consideration of one’s interests along with those of every other beneficiary in the social distribution of goods. If there are to be institutions allocating resources which can take account of needs in the distributive process, then no one’s needs for the less basic goods should be satisfied before everyone’s need for the most basic” (1988: 250, emphasis added).

Political and Distributive Justice    293

13.1.3.3 The Existence of Political Institutions and the Value of Distributive Justice A special claim of distributive justice arising from the activity of states can be grounded in other ways. As Nagel (2005: 120) puts it On the political conception, sovereign states are not merely instruments for realizing the preinstitutional value of justice among human beings. Instead, their existence is precisely what gives the value of justice its application, by putting the fellow citizens of a sovereign state into a relation that they do not have with the rest of humanity; an institutional relation which must then be evaluated by the special standards of fairness and equality that fill out the content of justice.

The dispute within this family of views regards what, more precisely, is the institutional relation that triggers those special demands of justice. For instance, Dworkin (2002: 6) claims that the special standards arise because a political institution that demands from citizens “allegiance and obedience to its laws, must take up an impartial, objective attitude toward them all.” Such a view may emphasize that not all institutions demand allegiance and obedience, and a state does not typically demand allegiance and obedience from everyone in the world. Instead, a state has this special relation to its citizens, and because of this it must give them a certain sort of impartial treatment. In a related way, Blake (2001: 258) holds that a “concern with relative economic shares” is appropriate for “individuals who share liability to the coercive network of state governance.”15 On Blake’s account, each state systematically coerces people within its own borders, including coercively enforcing a system of property rights. It is because of this particular activity, and the way coercion prima facie conflicts with the status of people as free and equal, that the state must meet a special justificatory standard in which relative shares matter. Nagel, in contrast, argues that the relative shares between co-​citizens matter specifically because the state—​most obviously, the democratic state—​claims to rule in their name. It is, for Nagel, the fact that the laws are claimed to arise from the citizens that gives rise to egalitarian requirements, but only as applied among citizens. Since the state does not claim to rule in the name of those outside its borders, relative distributions of shares with and among those people do not require justification. These views exhibit important disagreements about what, precisely, triggers the distributive demands, but they concur on the core claim that paradigmatic political institutions, because of the particular way they relate to their own citizens or denizens, are under special demands of justice that do not arise for most private action. They thus endorse Distributive Justice as a Distinctively Public Duty.16 The overall point

15 

Cf. Scheffler (2006: 124–​5). Blake (2001: 258ff.) and Nagel (2005: 118) each endorse principles of assistance for alleviating absolute, as opposed to relative, deprivation, but they distinguish such principles from the distributive principles for co-​members of a society. 16 

294    Chad Van Schoelandt and Gerald Gaus for philosophers such as Benn, Blake, Dworkin, and Nagel is that the demands of distributive justice only arise for certain existing distributors or political agents engaged in activities that make distributive justice matter. On this view, demands of distributive justice are not simply about the way diverse people happen to have advantages, or even how the fruits of cooperation are distributed through markets, international trade, or other types of social institutions.17

13.1.3.4 The Distribution of Political Rights A special case of distributive justice being strongly endogenous to the political may be the distribution of political power itself. Political institutions, particularly in coercively imposing a unified legal system on all, potentially embody public subordination or domination independent of other distributive issues. For instance, Thomas Christiano holds that “mere mutual impacts among people is not sufficient to ground democratic institutions” (2008: 83), but that co-​members of a unified legal system must have equal voice in shaping that system. Imposing a legal system on people who do not have equal political rights, Christiano argues, publicly expresses disregard for the fundamental interests of the marginalized. Phillip Pettit (2012: Ch. 3) likewise argues that a political state, even one otherwise promoting justice, practices public domination by leaving some under the power of political elites unless each member equally shares in the control of the state. For these theorists, the distributions that affect the justice of the political institutions are perhaps maximally endogenous to the political in that they regard goods, like voting rights, that are not only distributed by the political institutions, but furthermore are themselves part of those institutions or at least matter for the sake of controlling those institutions.

13.2  Loosening Political Justice’s Coupling to Theories of Distributive Justice We have surveyed a variety of ways that theories of distributive justice have been coupled to judgments of the justice of states and their policies, from the attenuated to the conceptually constitutive. Even the more attenuated accounts, however, suppose that there is typically an important tie between the moral requirements of a theory of distributive justice and judgments of the justice of a state. In the remainder of this chapter we argue that this is an error. Under contemporary 17  On such an account, immigration restrictions plausibly bring people outside of a state’s borders into the scope of distributive justice for that state. Cf. Abizadeh (2007: 348ff.), Gaus (2011: 478–​9), and Huemer (2010).

Political and Distributive Justice    295 conditions of moral disagreement, one’s conception of a politically just state should be largely uncoupled from one’s conviction about the best account of distributive justice.

13.2.1 The Political Authority of Moral Conviction—​and its Problem 13.2.1.1 The Political Authority of Moral Conviction Suppose you have come to a principle of distributive justice, DJ, which determines whether the distribution of some critical good in your society is just. Suppose according to this principle it is not. This is a judgment you believe is supported by sound and impartial reasoning, and is intended to provide (as James says) normative guidance and to be (as Larmore says) relevant to social life. Given this, your moral judgment leads to a moral imperative or requirement that your society’s distribution be brought into compliance with DJ; call this “Conform to DJ!” Because, further, on your view distributive and political justice are coupled, you claim that insofar as your society now fails to conform to DJ, a politically just state is one that acts on the imperative “Conform to DJ!,” say by enacting the relevant legislation. If your state does not obey that imperative it is characterized by significant injustice. It has failed in its duty to promote distributive justice. To be sure, as Cohen stresses (section 13.1.1.1), there may be other morally relevant considerations in determining whether your state is, overall, morally justified (or even, overall just), but it is a serious injustice of your state that it fails to obey the imperative (remember, we are examining the distributive justice of a critical good). This is a prevalent attitude. Recall the back cover of Rawls’s 2001 Justice as Fairness: “Rawls is well aware that since the publication of A Theory of Justice in 1971 American society has moved further away from the idea of justice as fairness. . . . This book demonstrates that moral clarity can be achieved even when a collective commitment to justice is uncertain” (2001; emphasis added). It is extraordinarily difficult not to interpret this condemnation as based on the supposed failure of American society to better conform to Rawls’s “second principle” of justice, dealing with distributive justice. Regarding the protection of the “first principle” basic liberties, it is hard to see any obvious retrograde movement between 1971 and the turn of the century (quite the contrary: think of the great advances made during this period in civil rights, especially in the south, and gender equality). Underlying this line of thinking is something like the following idea: The Political Authority of Moral Conviction: If Alf conscientiously and sincerely believes that the principle of distributive justice DJ is morally justified and that DJ deems his society’s distribution unjust, then he is committed to accepting that, pro tanto, his current state should, as a matter of justice, employ its political authority (or legitimate political power) to move the distribution into alignment with the requirements of DJ. If it fails to do so, his state exemplifies serious injustice.

296    Chad Van Schoelandt and Gerald Gaus If Alf continues to hold a well-​reasoned belief that DJ is violated, then (at least pro tanto) he must also hold that justice requires that the state use its authority or legitimate power to do something about it. Note that Alf has moved from a judgment of justice to a claim that political authority should advance his judgment about justice. To be sure, Alf may seek to attenuate this transition from personal moral judgment to claims on political authority. He might be uncertain whether his judgment about DJ is correct, and so draw back from calling on the state to act on it. Of course, for the same reason, he might make other moral judgments and refuse to issue any imperatives or demands based on them. Yet if Alf requires certainty in his judgments before he issues imperatives, the practical nature of distributive justice, so stressed by Anderson, James and others, would be undermined. At some point, if reasoning about distributive justice is to give us practical moral guidance, there must be agents who have reasonable confidence that DJ calls for some action, and be prepared to call on others to heed this. If, as we think is the case with most political philosophers, something along the lines of the Political Authority Of Moral Conviction is accepted, this ultimately leads to calls on the state to make all citizens act in conformity with DJ.

13.2.1.2 Whose Judgment? Suppose that Alf has thought things through, identified what is the best principle of distributive justice, and demands policy α; Betty has thought things through, identified what is the best principle of distributive justice, and she demands the state use its authority or power not to enact α, but instead the incompatible β. And there is also Charlie, Doris and everyone down to Zedra, each with their judgments and demands. Well-​intentioned, thoughtful people come to divergent conclusions about the fundamental requirements of distributive justice. As Rawls (2005: xlvi–​xlviii, 5ff., 54ff.) argued, the exercise of reason under free conditions leads to diversity in views of philosophy, religion, the nature of the good life, and even of justice. The past forty years of philosophical reflection on distributive justice—​as this very Handbook attests—​has resulted in a rather astounding array of competing theories, calling for different and often opposite movements in state policy. Whose judgment is to guide the state? The answer seems easy: the correct one. Let us grant that. Yet this is of no help at all, for our disagreement is about whose judgment is the correct one, so our problem remains. But perhaps this is too quick. To many philosophers, this really is no problem at all for Alf. Having reasonable confidence that he is correct, he must insist that the just state conform to his judgment, for he has identified the correct policy, α. So, he asks, “What’s the problem?” And since Betty, Zedra, and all the rest will reason in precisely the same way, they won’t see any problem either. And so no one will see any problem, so we might be tempted to conclude that there is no problem. Hobbes provided perhaps the earliest and most incisive analysis as to why there truly is a problem. In this dispute Alf (and the same holds for each of the others) thinks himself “wiser than all others,” and clamours and demands “right reason for judge, yet seek no more but that things should be determined by no other men’s reason” but his own. But this, says, Hobbes ([1668] 1994: Ch. 5, ¶3), is “intolerable in the society of men,”

Political and Distributive Justice    297 for our collective problem is that we seek to live together under a common framework (e.g., the state), but our use of our reasoning to decide what “right” or correct reasoning requires of the state lands us in interminable disagreement. For someone to seek to resolve this fundamental social and political problem by saying that others should do as he says, since his reason is right reason, betrays his lack of right reason by the very claim he lays to it; he fails to grasp what the social and political problem is, and would have us take his errors, prejudices, and appetites as the dictates of right reason—​even when he is in dispute with others about what right reason is. As Rawls (2005: 61) observed, Of course, those who do insist on their beliefs also insist that their beliefs alone are true: they impose their beliefs because, they say, their beliefs are true and not because they are their beliefs. But this is a claim that all equally could make; it is also a claim that cannot be made good by anyone to citizens generally.

The problem is that political justice concerns a social and political life under common rules and policies, and this requires significant decoupling of personal convictions about morality from the rules and policies of our common social and political existence. Kant ([1797] 1999: 116 [§43]) understood the state of nature as one in which each claims “the right to do what seems just and good to him, entirely independently of the opinion of others” (emphasis added). In such an apolitical condition, each can simply insist on following their own judgments, and can even insist that all others do as they say. But unless they gather great force on their side, others will not obey, for their moral judgment leads them in other directions. Kant ([1797] 1999:  116 (§43)) believed that such a condition of pure individual moral guidance would be one of endemic conflict in which people “can never be certain they are secure against violence from one another.” For Kant the state is constructed to resolve this problem of discordant moral judgment; but if each marches under the banner of the Political Authority of Moral Conviction, the state becomes simply a site of our conflict, not its resolution.

13.2.2 The Normative Autonomy of Political Justice The social contract tradition always recognized that moral judgment is an exercise of private reason and, consequently, widespread and intractable dispute about the claims of morality are inevitable. Hobbes’s ([1668] 1994: Ch. 18, ¶1) preferred solution is as unappealing as it is elegant: to nominate the personal reason of one of us as the mark of right reasoning, and to “authorize all the actions and judgments, of that man, or assembly of men, in the same manner, as if they were his own.” The idea of accepting and authorizing the judgment of the sovereign occurs repeatedly in Leviathan. It is important that for Hobbes ([1668] 1994: Ch. 7, ¶2), while the will concerns deliberation about action, judgment is “the last opinion in the search of the truth.” Hobbes identifies public reason with the judgment of the sovereign, and so the sovereign provides a public determination of the truth, for example, of a claim that a miracle has occurred ([1668]

298    Chad Van Schoelandt and Gerald Gaus 1994: Ch. 37, ¶13). Even “when it comes to confession of . . . faith, the private reason must submit to the public” ([1668] 1994: Ch. 37, ¶13, emphasis added).18 For Hobbes, because any dispute arising from differences in private judgment can lead to conflict, we must always be prepared—​on any issue—​to take the judgment of the sovereign as the mark of right reason, and so the correct judgment. This requires that each be prepared to abandon acting on their personal judgment whenever the sovereign deems it necessary: “we are not every one,” says Hobbes ([1668] 1994: Ch. 37, ¶13), “to make our own private reason, or conscience, but the public reason, that is, the reason of God’s supreme lieutenant, judge.” This sort of radical abandonment of reliance on personal conscience in favor of the sovereign’s is not quite so anachronistic as one might think. More radical theories of democracy, which are attracted to the maxim Vox populi, vox dei (“the voice of the people is the voice of god”), advocate a similar deference to public judgment, albeit of the majority. Not much is changed if we add that the democratic process has significant epistemic tendencies to be correct; as long as the epistemic properties of democracy are not so overwhelming that only a rationally flawed person would refuse to change her judgment in the face of its pronouncements, a person would still be committed to taking its output as, for public purposes, the correct judgment even though her best personal reflections lead her to disagree.19 Hobbes’s view is so radical in the social contract tradition, not because he thinks that a just political life requires abandoning private judgment, but because he insists that one is committed to abandoning private judgment on any issue if the sovereign decides public judgment is required.20 It is almost always overlooked that Locke agrees with Hobbes that personal moral judgments about civil matters must be excluded in a just political order. He concurs that diversity of private judgment about not only religion, but the dictates on natural law, is a fundamental social fact that must be reconciled with the demands of social order. Peace and justice, Locke ([1689] 1960: §87) concludes, can only be secured by “all private judgment of every particular Member being excluded, [so that] the community comes to be Umpire[d]‌by settled standing Rules, indifferent, and the same to all Parties” (emphasis added). It is the task of government to serve as Umpire—​the voice of public reason; it yields a common judgment that provides a definitive, public, understanding of the demands of morality. It should be stressed that a political justice that requires abandoning private judgment about public morality in no way precludes distributive justice from being part of political justice. If that is what the “umpire” (the sovereign, the majority) decides, then that 18  Although the sovereign cannot directly command us to believe a proposition since belief is not

under the control of the will, Hobbes appears to hold (i) the sovereign has authority to declare public truths from which we are obligated not to dissent; (ii) we can be obligated to publicly affirm these truths; and (iii) the sovereign has authority to shape the environment in which opinions are formed ([1668] 1994: Ch. 32, ¶5, Ch. 37, ¶13). 19  Christiano (2008: 55, Ch. 3) and Pettit (2012: Ch. 2) both argue that citizens must respect democratic processes as a means of publicly recognizing their equality despite substantive moral disagreement. 20  Cf. Gaus (2015).

Political and Distributive Justice    299 is what public morality requires. If the political process determines that the principle DJ is part of political justice, then so be it. This view does, however, reject at its very root the Political Authority of Moral Conviction: no private citizen can claim that political justice requires DJ because according to her own moral conscience DJ is correct. That is precisely the claim that the social contract is intended to eliminate. More than that, though, on Hobbes’s view, should one’s private conscience determine that DJ is correct and DJ deems distribution x to be unjust, while political justice based on public reason declares x just, as a citizen seeking to be politically just one must uphold the public just­ ice of x. Thus we are led to: The Normative Autonomy of Political Justice: Alf, who believes that on the best view of morality, some distribution x is unjust, can reasonably endorse x as politically just, or just from the political point of view. Unless the Normative Autonomy of Political Justice obtains, political justice cannot be decoupled from personal judgments of justice. The social contract/​public reason solution to the problem of disagreement requires that a reasonable person must be able to admit the normativity of the requirements of public justice, while holding contrary views based on their personal deliberations and their preferred “theory of distributive justice.” Yet many will ask: how can anyone coherently conform to such a principle?

13.2.3 Three Routes to the Normative Autonomy of Political Justice 13.2.3.1 Hobbesian Self-​Effacing Authorization As has already been mentioned (section 13.2.2), at the center of Hobbes’s account was a contract in which one authorizes another’s judgment as if it were one’s own. This authorization is self-​effacing in the sense that one’s own personal judgment about morality becomes irrelevant to political justice. It is important that Hobbes insisted that the sovereign could violate the laws of nature, and a citizen could be correct in her judgment that the sovereign did so ([1668] 1994: Ch. 31, ¶7, Ch., 26, ¶24, Ch. 28, ¶¶22–​23, Ch. 30, ¶15). But because of the social contract, the judgment of the subject has no public standing, and so is no complaint against the sovereign; more than that, the subject has authorized the judgment of the sovereign on this matter, and so (political) justice demands conformity to (indeed public endorsement of) his decision. Reasonable people could only be driven into such a thoroughly self-​effacing authorization if, as Hobbes appears to insist, it is absolutely necessary for any stable social existence. In the face of the eruption of diverse moral and political judgments and political conflict in the seventeenth century, the only solution Hobbes saw as viable was for each to renounce all implications for political, public, justice of one’s personal moral and religious judgments. Pace Hobbes, early modern citizens were not prepared to purchase

300    Chad Van Schoelandt and Gerald Gaus social stability at this price, and, most emphatically, neither are twenty-​first-century liberals.

13.2.3.2 Rawlsian Political Constructivism The power and originality of Rawls’s (2005: xlii, 12) political constructivism in his political liberalism project was to seek a distinctively public, political, understanding of just­ice that was free-​standing in relation to the diverse moral views and comprehensive theories of justice.21 As Rawls said in his first version of his Dewey Lectures22 ([1980] 1999: 306–​7), the aim is to: search for reasonable grounds for reaching agreement rooted in our conception of ourselves and in our relation to society. . . . The task is to articulate a public conception of justice that all can live with. . . . What justifies a conception of justice is not its being true to an antecedent order to and given to us, but its congruence with our deepest understanding of ourselves and our aspirations, and our realization that, given our history and the traditions embedded in our public life, it is the most reasonable doctrine for us. We can find no better basic charter for our social world. [emphasis added]

The aim is to construct a public conception of justice out of the common elements of the public culture. Unlike Hobbesian public reason, this political conception would be one that we all (or at least the overwhelming majority) could endorse, as it is the view that can be derived from what we share. Thus Rawls does not require us to be nearly so self-​ effacing as does Hobbes; the public conception is one that, from the public point of view, we all share. Rawls originally thought that from these distinctively public, shared materials a public conception of distributive justice could be constructed. Such a political conception has precious little in common with a view such as Cohen’s (section 13.1.1.1): it is not seeking to develop the morally correct theory of distributive justice, but one that can serve as the basis of public life and normative order in a diverse society. As is well known, as Rawls’s thinking about this distinctively public, political conception developed, he became increasing skeptical that there was a uniquely justifiable public criterion of distributive justice. Given the deep disagreements about distributive justice, his original aim could only be seen as extraordinarily optimistic. Instead of establishing the unique reasonability of his two principles of justice as the criterion of public justice, Rawls ultimately concluded that the public construction identified a family of “liberal conceptions of justice.” Although he believed all members of the family issued requirements concerning the provision of resources, the family as such was not committed to any 21 

As the reader will have noted, we have included Rawls’s thinking under more than one category. This attests not only to the development in his theory over his career, but to the radically different ways in which his theorizing from the same period is interpreted by the philosophical community. We do not seek to resolve these controversies here. 22  Political Liberalism was presented as a revised and expanded version of these lectures.

Political and Distributive Justice    301 criterion of distribution, though he still believed that his “difference principle” was part of one such reasonable public construction (Rawls, 2005: xlvi–​xlviii, 5ff.). It is, though, hard not to be skeptical that in our highly diverse societies the nearly-​universally shared elements are thick enough to build the rather robust political constructions (fair equality of opportunity, the difference principle, and restrictions on advertising!) that Rawls (2005: 365) envisaged. It is important to stress that not only did the free-​standing argument seek to show that endorsement of political justice did not require citizens to be self-​effacing, since they concur on the elements of the construction, but, crucially that, once they had considered their entire set of moral and religious commitments, they would continue to endorse the conception of political justice. At least, he hoped that such an “overlapping consensus,” providing further support for the free-​standing political conception, would occur (2005: Lecture IV). Notice that if such an overlapping consensus—​in which each continues to endorse the political conception given their full set of moral and religious commitments—​occurs, Rawls could establish the Normative Autonomy of Political Morality without requiring citizens to be self-​effacing at all: they all endorse the political conception while drawing on their full set of moral judgments. This would be an especially happy result; Rawls could avoid both the public moral anarchism of the Political Authority of Moral Conviction (section 13.2.1.1) and any hint of Hobbesian self-​ effacement to public reason. However, as some of the current generation of Rawlsians have recognized, if political justice really needs to be endorsed by the full moral and religious commitments of citizens, the prospects of a robust egalitarian distributive liberalism being justified as part of political justice look pretty slim. In the eyes of Jonathan Quong (2011: 167–​9), Rawls’s claim that overlapping consensus is required for full justification holds political justice “captive” to those with inegalitarian moral commitments.23 Quong thus eliminates overlapping consensus as a test for the acceptability of the outcome of Rawls’s free-​standing argument from the original position. Quong, then, seems to acknowledge that a liberal political justice with commitment to a robust criterion of distributive justice requires many citizens to be significantly morally self-​effacing; classical liberal citizens, for example, must set aside their wider moral and philosophical commitments to endorse public justice (cf. Gaus 2012: 7–​15).

13.2.3.3 Convergent Normativity Quong seems to present Rawlsian liberals, upholding the Normative Autonomy of Political Justice (section 13.2.2), with something of a dilemma. On the one hand, the Rawlsian can stay true to their robust egalitarian theory of political justice (including, say, fair equality of opportunity and the difference principle), but only at the cost of eliminating the idea that political justice must be supported by an overlapping consensus. Thus citizens are required to be significantly self-​effacing with respect to their wider moral views: they must set aside their personal moral convictions (including their preferred

23 

Cf. Rawls (2005: 389).

302    Chad Van Schoelandt and Gerald Gaus theories of distributive justice) in order to endorse the normativity of political justice. Quong, as we have seen, takes this more Hobbesian-​inclined path. The alternative is to accept Quong’s conclusion that the Rawlsian cannot have their robust egalitarianism and overlapping consensus too, and to pursue overlapping consensus. On this view, the aim is to try to minimize the extent to which a political conception of justice requires citizens to set aside their moral and religious convictions when they participate in a common, moral, political life. The aim of what we might call “convergent normativity theorists” (Gaus 2011; Vallier 2014) is to show that a very wide range of moral and religious perspectives can converge on endorsing a set of political institutions and basic rules as normatively acceptable ways to live together. Given our moral disagreements, we will not agree on what rules and institutions are best. When one is building a house alone, one can follow only one’s judgment about the optimal house (consistent with feasibility qua budget constraints); but when building a common dwelling, each cannot go about building their optimal house, nor insist that all work together to build their optimal house. What we seek is a common structure—​there are perhaps many that, as Rawls said, “all can live with”—​that all can see as sufficiently answering to their important moral concerns, and which does not violate their moral integrity by insisting that they be self-​effacing when deciding on what they can live with. That was always the aim of the social contract tradition: to devise rules and institutions that all could agree sufficiently promote their concerns to make the institution worth the willing endorsement of each. The institution, while not best in the eyes of all, is something each could live with. This means, though, that an adequate theory of political justice will not present inspiring ideals. It will not be aspirational, or paint pictures of a society with perfect distributive justice—​perfect, that is, on one vision. Rather, it will seek to investigate the types of arrangements and structures that allow those in a diverse society to live together in ways that all (or, at least, as many as is compatible with a moral order) endorse as normatively acceptable, a normative structure that allows each to seek to flourish as they understands that protean ideal. We all are free to think through the moral world for ourselves, and our collective political justice should not require us to ignore these reflections. But neither can an ordered and cooperative social life exist when each elevates her own “theory of justice” into the authoritative blueprint for all.

13.3 Conclusion More than half a century ago Bertrand de Jouvenel (1957: 139) described “social” or “distributive” justice as the “obsession of our time.” It has certainly been the obsession of political philosophy. For the most part this obsession has manifested itself in rather intricate and controversial “moral theories of distributive justice” that seek to shape political justice. We have analyzed the many different ways these theories have been

Political and Distributive Justice    303 coupled with the political, but in our view they all fail to appreciate the fundamental requirement of a moralized political life in a diverse society—​that political justice must be significantly autonomous of controversial moral “theories” as well as religious convictions. The moral theorist of distributive justice has far too often taken on the role of the high priest, sending down from their ivory tower instructions for political authority. Although it is too often—​we fear, usually—​overlooked, the genius of Rawls was his realization that such moralized political theory is inappropriate as a basis of political justice in our diverse society. His proposal was striking: we must follow the social contract tradition in recognizing the relative normative autonomy of the political, yet we can construct a distinctively political standard of distributive justice that all can endorse. As we have argued, we believe this project failed, and those who appreciate his genius must now choose what to pursue: their devotion to a theory of distributive just­ ice, or the ideal of a conception of political justice that all good-​willed, free, and equal citizens can endorse without self-​effacement.

References Abizadeh, A. (2007). “Cooperation, Pervasive Impact, and Coercion: On the Scope (not Site) of Distributive Justice.” Philosophy and Public Affairs 35(4): 318–​58. Anderson, E. (2012). “The Fundamental Disagreement between Luck Egalitarians and Relational Egalitarians.” Canadian Journal of Philosophy, Supplementary, 36: 1–​23. Bakunin, M. [1873] (1972). “Statism and Anarchy,” in Bakunin on Anarchism, ed. and trans. S. Dolgoff. New York: Alfred A. Knopf, pp. 323–50. Benn, S. I. (1988). A Theory of Freedom. Cambridge: Cambridge University Press. Bicchieri, C. and Chavez, A. (2010). “Behaving as Expected: Public Information and Fairness Norms.” Journal of Behavioral Decision Making, 23(2): 161–​78. Blake, M. (2001). “Distributive Justice, State Coercion, and Autonomy.” Philosophy and Public Affairs 30(3): 257–​96. Christiano, T. (2008). The Constitution of Equality:  Democratic Authority and its Limits. Oxford: Oxford University Press. Cohen, G. A. (2000). If You’re an Egalitarian, How Come You’re So Rich? Cambridge, MA: Harvard University Press. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Dworkin, R. (2002). Sovereign Virtue: The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Estlund, D. (2008). Democratic Authority:  A Philosophical Framework. Princeton, NJ: Princeton University Press. Estlund, D. (2011). “Human Nature and the Limits (if Any) of Political Philosophy.” Philosophy & Public Affairs, 39(3): 207–​35. Gaus, G. (2011). The Order of Public Reason: A Theory of Freedom and Morality in a Diverse and Bounded World. Cambridge: Cambridge University Press. Gaus, G. (2012). “Sectarianism without Perfection? Quong’s Political Liberalism.” Philosophy and Public Issues, 2(2): 7–​15.

304    Chad Van Schoelandt and Gerald Gaus Gaus, G. (2015). “Public Reason Liberalism,” in S. Wall and C. Kukathas (eds) The Cambridge Companion to Liberalism. Cambridge: Cambridge University Press, pp. 112–​40. Hayek, F. A. (1945). “The Use of Knowledge in Society.” The American Economic Review 35(4): 519–​30. Hayek, F. A. (1976). The Mirage of Social Justice. Chicago, IL: University of Chicago Press. Hobbes, T. [1668] (1994). Leviathan, with Selected Variants from the Latin Edition of 1668, ed. Edwin Curley. Indianapolis, IN: Hackett. Huemer, M. (2010). “Is There a  Right to Immigrate?” Social Theory and Practice 36(3): 429–​61. James, A. (2013). Fairness in Practice: A Social Contract for a Global Economy. New York: Oxford University Press. Jouvenel, B. de (1957). Sovereignty. Cambridge: Cambridge University Press. Kant, I. [1797] (1999). The Metaphysical Elements of Justice, 2nd edn. ed. and trans. John Ladd. Indianapolis, IN: Hackett. Larmore, C. (2013). “What is Political Philosophy?” Journal of Moral Philosophy, 10(3): 276–​306. Lister, A. (2013). “The ‘Mirage’ of Social Justice:  Hayek against (and for) Rawls.” Critical Review, 25(3–​4): 409–​44. Locke, J. [1689] (1960). Second Treatise of Government in Two Treatises of Government, ed. Peter Laslett. Cambridge: Cambridge University Press. Mises, L. von [1920] (2009). “Economic Calculation in the Socialist Commonwealth,” in F.  A.  Hayek (ed.) Collectivist Economic Planning. Auburn, AL:  The Ludwig von Mises Institute, pp. 87–​130. Nagel, T. (1981). “Libertarianism without Foundations,” in P. Jeffrey (ed.) Reading Nozick: Essays on Anarchy, State, and Utopia. Totowa, NJ: Rowman & Littlefield, pp. 191–​205. Nagel, T. (2005). “The Problem of Global Justice.” Philosophy and Public Affairs, 33(2): 113–​47. Neufeld, B. and Van Schoelandt, C. (2014). “Political Liberalism, Ethos Justice, and Gender Equality.” Law and Philosophy, 33(1): 75–​104. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. Pettit, P. (2012). On the People’s Terms:  A Republican Theory and Model of Democracy. Cambridge: Cambridge University Press. Quong, J. (2011). Liberalism without Perfection. Oxford: Oxford University Press. Rawls, J. (1951) [1999]. “Outline of a Decision Procedure for Ethics,” in S. Freeman (ed.) Collected Papers. Cambridge, MA: Harvard University Press, pp. 1–​19. Rawls, J. (1980) [1999]. “Kantian Constructivism in Moral Theory,” in S. Freeman (ed.) Collected Papers. Cambridge, MA: Harvard University Press, pp. 303–​58. Rawls, J. (1999). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press. Rawls, J. (2001). The Law of Peoples. Cambridge, MA: Harvard University Press. Rawls, J. (2005). Political Liberalism, expanded edn. New York: Columbia University Press. Scheffler, S. (2006). “Is the Basic Structure Basic?,” in C. Sypnowich (ed.) The Egalitarian Conscience:  Essays in Honour of G.A. Cohen. New  York:  Oxford University Press, pp. 102–​29. Scheffler, S. (2010). “The Division of Moral Labor,” in his Equality and Tradition: Questions of Value in Moral and Political Theory. New York: Oxford University Press, pp. 107–​29.

Political and Distributive Justice    305 Thrasher, J. and Vallier, K. (2013). “The Fragility of Consensus: Public Reason, Diversity and Stability.” European Journal of Philosophy. doi: 10.1111/​ejop.12020). Vallier, K. (2014). Liberal Politics and Public Faith: Beyond Separation. New York: Routledge. Weithman, P. (2013). Why Political Liberalism? On John Rawls’s Political Turn. New  York: Oxford University Press.

Chapter 14

C onsequ en t ia l i sm, De on tol o gy, C ontractua l i sm, and Equa l i t y Jonathan Quong

Theories of distributive justice tell us what belongs to whom. They tell us whether the current distribution of income, wealth, liberties, and other advantages in our world is just, and if it is not, why not.1 Claims of justice are also often presented as having lexical or strongly weighted priority over other considerations. As John Rawls famously said: Each person possesses an inviolability founded on justice that even the welfare of society as a whole cannot override. For this reason justice denies that the loss of freedom for some is made right by a greater good shared by others. It does not allow that the sacrifices imposed on a few are outweighed by the larger sum of advantages enjoyed by many . . . the rights secured by justice are not subject to political bargaining or to the calculus of social interests . . . an injustice is tolerable only when it is necessary to avoid an even greater injustice. (Rawls 1999: 3–​4)

Even many of those who disagree sharply with Rawls about the principles of distributive justice nevertheless share his view that individuals’ claims of justice take priority over other considerations (Nozick 1974: 155–​60). In this respect, distributive justice might seem clearly incompatible with a consequentialist moral framework. 1 

Some philosophers reject the term distributive justice because they believe it implies an unduly narrow conception, one where the exclusive focus is on the distribution of goods or advantages, rather than the establishment of appropriate social relations. I will continue to use the term distributive justice partly for the sake of convenience, but also because I do not believe the term needs to be so narrowly construed.

Consequentialism and Deontology    307 But many theories of distributive justice—​for example, those that include egalitarian or prioritarian principles—​seem to have a distinctive consequentialist character. These theories—​as Robert Nozick warned—​call for certain distributive patterns to be achieved. Consider the following canonical statement of luck egalitarianism: “an unequal distribution whose inequality cannot be vindicated by some choice or fault or desert on the part of (some of) the relevant affected agents is unfair, and therefore, pro tanto, unjust” (Cohen 2008: 7). On this view, states of affairs can be judged just or unjust, and what justice requires, one might plausibly infer, is to bring about states of affairs that are more, rather than less, distributively just. Just distributive patterns are simply good consequences that can be incorporated into a consequentialist moral framework. Debates about distributive justice might therefore seem to represent yet another front in the never-​ending war between consequentialism and deontology. Consider, for example, the debate between two prominent conceptions of egalitarian justice: democratic or relational egalitarianism and luck egalitarianism. Elizabeth Anderson explains that one of the central divisions between the two camps is that luck egalitarianism is consequentialist, whereas democratic or relational egalitarianism is not (Anderson 2010). In order to know what kind of egalitarian you are (insofar as you are an egalitarian), you must first choose your side in the wider war in normative ethics: are you a consequentialist or do your sympathies lie with the loose alliance of deontological moral theories? One of the aims in this chapter is to resist this way of mapping of the philosophical terrain. As others have argued in recent years, there is no single and useful way of distinguishing consequentialism from deontology. Indeed, as Rawls pointed out long ago, any sensible deontological theory will also be concerned with the consequences of adopting proposed moral principles (Rawls 1999: 26). Instead, it is more helpful to consider how different theories of distributive justice answer the following three questions: 1. What are the relevant objects of justice and injustice? 2. What makes something just or unjust? 3. What is it for something to be just or unjust? Addressing these questions does more to illuminate recent debates about egalitarian justice. In particular, I suggest that luck egalitarianism and democratic egalitarianism, far from being rival accounts of the grounds of distributive justice, can be understood as complementary answers to different questions. I focus on the debate between luck egalitarianism and democratic egalitarianism because it has shaped a great deal of the recent literature on distributive justice, and the positions that have emerged sometimes create the misleading impression that there is a deep and unbridgeable divide between two views of equality. But some of the lessons from this chapter apply more generally. Whenever two views about distributive justice are presented as incompatible because one is consequentialist and the other is deontological, or because one is consequentialist and the other is contractualist, we should be immediately suspicious, since these are not mutually exclusive categories.

308   Jonathan Quong The chapter is structured as follows. Section 14.1 explains why the distinction between consequentialism and deontology is unhelpful, and identifies more helpful distinctions. Section 14.2 briefly describes the divide between democratic and luck egalitarians. Sections 14.3–​14.5 examine, respectively, the three questions just listed, and show how focusing on these questions sheds new light on democratic and luck egalitarian conceptions of distributive justice, and makes a reconciliation of these views possible. Section 14.6 defends this proposed reconciliation from an important potential objection, and Section 14.7 concludes with a summary.

14.1  Consequentialism vs Deontology? The distinction between consequentialism and deontology is invoked all the time in normative ethics but, as with many other sweeping distinctions in philosophy, there is no single and widely accepted account of the distinction. Below I identify four different distinctions that are sometimes conflated with the alleged contrast between consequentialism and deontology.

14.1.1  Consequentialism vs Nonconsequentialism Consequentialist theories define moral rightness exclusively in terms of what produces the best consequences. Nothing apart from the consequences of an action or rule can be relevant in assessing whether the action or rule is morally right. Nonconsequentialism is then simply the denial of consequentialism. Nonconsequentialists affirm the weak thesis that things apart from consequences can sometimes be relevant in determining the rightness of an act or rule, though they need not deny that consequences are also relevant in determining rightness (Kamm 2013:  261). Nonconsequentialism does not affirm the extreme thesis that rightness must be defined without any reference consequences.

14.1.2  Teleology vs Deontology Teleological theories begin with an independent definition of goodness, and then define rightness as maximizing goodness (Rawls 1999: 21–​2). Deontological theories, by contrast, reject at least one of the two features of the teleological framework. In deontological theories, either the right is independent of the good in the sense that moral rightness is not defined as maximizing the good, or else goodness is not defined in a manner that is independent of moral rightness (Rawls 1999: 26). Utilitarianism is the paradigmatic example of a teleological theory. The good is independently identified as happiness, and then moral rightness is defined as maximizing

Consequentialism and Deontology    309 happiness. Perhaps because utilitarianism is a species of consequentialism (as defined above), philosophers often present deontology as the polar opposite of consequentialism, but this is a mistake. As Rawls says, It should be noted that deontological theories are defined as non-​teleological ones, not as views that characterize the rightness of institutions and acts independently from their consequences. All ethical doctrines worth our attention take consequences into account in judging rightness. One which did not would simply be irrational, crazy. (Rawls 1999: 26)

Indeed, not only can deontological theories take consequences into account in defining what is right, deontological theories can be consequentialist. I  might endorse a theory in which the rightness of actions or rules is entirely determined by the consequences produced, but which is non-​teleological in the sense that the right is not defined as maximizing the good. A  theory, for example, which defined goodness as happiness, and defined rightness exclusively in terms of producing states of affairs where happiness is equally distributed would be both consequentialist and deontological since the aim is not to maximize goodness, but rather to distribute goodness in a manner that is deemed just or fair. Although it is counterintuitive, it might also be possible for nonconsequentialist theories to be teleological. Teleological theories define the right as maximizing the good. While this fits most naturally within a consequentialist framework, it’s possible to hold a view whereby goodness is not a consequence that can be brought about via acts or rules, but rather inheres as a property of certain acts or rules, and thus aiming to maximize goodness will not entail consequentialism. Suppose, for example, goodness inheres entirely in acts performed with beneficent intentions, and further suppose it is not possible to cause others to act with such intentions.2 Maximizing goodness would then require always acting with beneficent intentions, but the rightness of the acts would not be defined in terms of the good consequences they produce, rather the rightness of the acts would inhere in a feature of the acts themselves. If acts can be separated from their consequences in this way,3 then teleological nonconsequentialism is a possible (albeit unlikely) moral position. The distinction between teleology and deontology thus cuts across the distinction between consequentialism and nonconsequentialism. The regular way consequentialism is contrasted with deontology is thus deeply unfortunate and likely to mislead rather than illuminate.

2 

This latter clause ensures that one cannot maximize goodness by causing others to act with beneficent intentions. Goodness is thus not something that can be a downstream consequence of one’s acts. 3  This claim is controversial, and denied by those who argue that all moral theories can be “consequentialized.” See, for example, Dreier (1993) and Steiner (1994: 138–​41).

310   Jonathan Quong

14.1.3  Agent Neutral vs Agent Centered A distinction that is often conflated with the divide between consequentialism and its rivals is the distinction between agent neutral theories and those that are in some way agent centered. As James Dreier explains, A theory is agent neutral if it gives to everyone the same advice or aims. According to an agent neutral theory, your (theory-​given) aims are better fulfilled exactly when mine are. By contrast, an agent centered theory gives to us at least some prescriptions or advice or aims which include indexicals. Sometimes the things an agent centered theory advises me to do will conflict with the things the same theory advises you to do. (Dreier 1993: 22)

Again, perhaps because utilitarianism is often presented as the paradigmatic consequentialist theory, it is sometimes assumed that consequentialist theories are all agent neutral, but as Dreier argues, this assumption is mistaken. A view can be consequentialist in the sense just defined—​the rightness of actions is entirely determined by the consequences of performing the act—​but also agent centered, for example, hedonistic egoism (Dreier 1993: 22–​3). Even theories that are not thoroughly egoistic, but simply direct agents to attach disproportionate weight to their own lives or interests relative to those of others can remain resolutely consequentialist. It is also possible for a theory to be both nonconsequentialist and agent neutral. Suppose, for example, you endorse a divine command theory whereby what makes acts morally wrong is the fact that God prohibits them. This account is nonconsequentialist, as I have defined it, and yet it could also be agent neutral if God issues only agent neutral commands. Alternatively, you might endorse a view in which what makes acts morally wrong is the fact that they are prohibited by a particular contractualist or interpersonal test,4 but where the reasons contracting parties can appeal to are all agent neutral (e.g., perhaps there is only one reason-​giving value: utility), such that the resulting moral directives are strictly agent neutral. These examples illustrate that the consequentialism vs nonconsequentialism distinction is also independent from the agent neutral vs agent centered distinction.

14.1.4  Forward-​Looking vs Backward-​Looking There is a further relevant distinction that is implicit in many discussions of consequentialism and deontology, perhaps particularly when evaluating the moral

4 

Sections 14.5 and 14.6 will offer some reasons to doubt this is, in fact, the best way to understand contractualism.

Consequentialism and Deontology    311 desirability of social institutions. The distinction I have in mind appears most clearly in the debate about the justification of punishment. Punishing wrongdoers typically imposes serious harms on the wrongdoers, and also comes at a substantial social cost. A natural question to ask about this practice is this: Does it do any good? That is, if we punish wrongdoers now, will this have any future benefits? Those who believe the practice of punishment cannot be permissible unless it produces future benefits are typically described as consequentialists and are contrasted with retributivists, who deny that the practice of punishment needs to do any good in the future in order to be permissible. This way of presenting the terrain, however, risks conflating separate moral distinctions. Let’s say that you endorse a forward-​looking justification of harm imposition if you endorse the following proposition: any act (or rule or practice) that causes some harm cannot be permissible unless it also produces at least some expected benefits in the future. Alternatively, those who endorse a backward-​looking justification of harm imposition deny this. They assert that acts (or rules or practices) that cause some harm and produce no expected future benefits can nevertheless sometimes be permissible as appropriate or fitting responses to what has happened in the past. Although it’s extremely tempting to collapse this distinction into the distinction between consequentialism and nonconsequentialism, the distinctions are independent. First, consequentialism evaluates the permissibility of options in terms of whether an option produces the best consequences relative to the alternatives, whereas the forward-​ looking view is weaker: it merely requires that, in order to be permissible, harm-​causing acts produce some future benefit, but it does not require that I choose the option which produces the best consequences. Second, unlike the forward-​looking view, consequentialism does not have temporal directionality as a built-​in feature. Consequentialism requires opting for the best consequences or state of affairs, but what happens in the future might be only one component of the consequences or state of affairs produced by one’s action. Suppose, for example, I believed that the best state of affairs is one where each person gets all the happiness or suffering he or she deserves. A consequentialism informed by this assumption will be one where the value our acts create depends heavily on what has happened in the past. Consequentialists could thus be retributivists, at least as I understand these terms.5 And those who advocate a forward-​looking conception of punishment need not be consequentialists. Although I have illustrated the distinction between backward and forward-​looking moral views by looking at the justification of punishment, this same distinction plays a role in debates about distributive justice. In this section I have argued that the frequently drawn distinction between consequentialism and deontology is an unhelpful mistake. There are, rather, at least four different distinctions philosophers sometimes have in mind when they discuss the difference between consequentialism and rival moral views. We will make more

5 

This point is made in Tadros (2011: 35).

312   Jonathan Quong progress in understanding rival moral theories, and also rival conceptions of distributive justice, if we keep these different distinctions in view.

14.2  Luck and Democratic Egalitarianisms In the remainder of this chapter, I examine the debate between luck egalitarians and democratic or relational egalitarians (henceforth democratic egalitarians). This debate is sometimes presented as one manifestation of the wider divide in normative ethics between consequentialism and deontology. But since, as I have just argued, this distinction is non-​existent and misleading, this cannot be the best way to understand what divides luck and democratic egalitarians. Luck egalitarianism, as I shall understand it here, is committed, at a minimum, to the following proposition:  luck-​generated inequalities are unjust, whereas choice-​ generated inequalities are presumptively just. Luck egalitarians famously differ regarding the appropriate currency of egalitarian justice—​resources, welfare, and access to advantage have all been proposed as the appropriate equalisandum. For ease of exposition, I will use the neutral terms “advantages” and “disadvantages” as placeholders for the equalisandum. To simplify somewhat, luck egalitarians, whatever their other disagreements, are united in their belief that a key feature of distributive justice is ensuring that the distribution of advantages and disadvantages amongst persons should be equal, unless inequalities can be vindicated by appeal to genuine choices made by the affected individuals against a fair background of opportunity. Democratic egalitarianism, as I shall understand it, affirms the following thesis: just­ ice requires that all citizens stand in a relationship of equality. Although democratic egalitarianism and luck egalitarianism are almost always presented as rival conceptions, it is not immediately apparent—​based on the descriptions provided thus far—​why they are rival views. It is far from obvious why standing in a relationship of equality is inconsistent with the aim of eliminating luck-​generated inequalities. So why are these views commonly perceived as competing conceptions of egalitarian justice? Democratic egalitarians argue there are at least four main differences between the two views (Anderson 1999; Scheffler 2003, 2005).6 First, luck egalitarianism focuses on the distribution of some equalisandum—​it tells us that egalitarian just­ ice is achieved when the equalisandum has been distributed in accordance with the

6 

This is not to deny that there are other points of regular dispute between the two conceptions. Also, it is a noteworthy feature of the debate that democratic egalitarians tend to emphasize the stark differences between the two theories, whereas luck egalitarians have often responded by emphasizing the ways in which the views overlap or converge. For an example of this latter response, see Stemplowska 2011. A notable exception to this latter trend, however, is Tan (2012).

Consequentialism and Deontology    313 appropriate egalitarian principle—​whereas democratic egalitarianism focuses on the relationships in which persons stand with regard to one another. Some democratic egalitarians argue that an egalitarian distribution of advantages is neither necessary nor sufficient to ensure persons (or citizens) stand in relationships of equality. Second, democratic egalitarians present equality as a political ideal, one that is somehow grounded in our political relationship as citizens, and whose scope does not extend beyond the major political and social institutions of society. Luck egalitarians, by contrast, tend to present equality as a moral value or principle that applies to the relationship between all persons, and whose grounding and scope is not distinctively political, though it has clear political implications.7 Third, democratic egalitarians typically reject the central role that luck egalitarians assign to personal responsibility in the allocation of advantages and disadvantages. Democratic egalitarians argue that luck egalitarianism’s extreme emphasis on individual responsibility has various unacceptable implications; for example, it implies that egalitarian justice countenances leaving the imprudent to suffer serious harm or even death when the imprudent are responsible for their plight. Democratic egalitarians typically present their view as one that requires all citizens be guaranteed the resources needed to function as equal citizens. Fourth, and relatedly, democratic egalitarians deny that justice should be primarily concerned with eliminating the influence of luck on patterns of distribution. The influence of luck only threatens egalitarian justice when it prevents persons or citizens from participating as equals in political life, but there are many areas of life where the influence of luck on the distribution of advantages and disadvantages should be of no concern to a conception of justice (e.g., whether one is lucky in love, lucky enough to enjoy one’s job, or fortunate to have good friends and a cheerful disposition). Why do democratic and luck egalitarians appear to differ so sharply regarding the content and shape of distributive justice? One explanation is to see the debate as a manifestation of a wider disagreement in normative ethics. Elizabeth Anderson, for example, argues that the main source of disagreement between democratic (she uses the term “relational”) egalitarians and luck egalitarians is that the latter adopt a third-​personal account of political justification, whereas the former deploy a form of second-​personal justification (Anderson 2010:  2–​3). On the former model, the identity of the person making an argument, or the identity of the person to whom the argument is addressed, are irrelevant, whereas on the latter model these features play a crucial role in evaluating justifications. And this difference, she claims, explains another crucial difference between the two views, namely, that luck egalitarianism is consequentialist: “it takes the value of a state of affairs as foundational and gives all agents the common aim of realizing it.” Democratic equality rejects this view in favor of one where justice is a virtue of agents, one where agents are disposed to treat one another in accordance with principles that “express, embody, and sustain relations of social equality” (Anderson 2010: 2).

7 

Tan is an interesting exception to this general luck egalitarian view. See Tan (2012: Chs 2–​3).

314   Jonathan Quong In what follows, I offer an alternative map of the debate between democratic and luck egalitarians. The map is organized around three questions: 1. What are the relevant objects of justice and injustice? 2. What makes something just or unjust? 3. What is it for something to be just or unjust? Egalitarians sometimes present luck egalitarianism and democratic egalitarianism as rival answers to the second question (Anderson 2010; Arneson 2011: 42; Cohen 2008: 7, 279; Scheffler 2003:  21–​3). But when the two positions are presented in this way, it makes each view less plausible than it needs to be. Instead, I suggest that democratic equality is most plausibly construed as an answer to the third question, and luck egalitarianism as one part of the answer to the second question (with their answers to the first question being derivative). Thus, far from being rival conceptions of distributive justice ensconced in fundamentally different models of normative ethics, luck egalitarianism and democratic egalitarianism might be more plausibly understood as offering different answers to different questions, but answers that can be integrated into a broader and coherent conception of egalitarian justice.

14.3  Objects of Justice Something is an object of justice or injustice when it is appropriate to describe the object as just or unjust. Narrowing our focus to the topic of distributive justice, potential objects of justice include at least the following: patterns of distribution, actions, rules, institutions, persons, and dispositions. Most theories of distributive justice will include all of the items on this list as potential objects of distributive justice, but different theories will divide on the following question: which are the fundamental objects of distributive justice, and which are merely derivative? Something is a fundamental object of distributive justice when claims about justice apply directly to the object in question in virtue of features inherent to the object. Something is merely a derivative object of just­ ice when claims about justice apply to it only because of the way the object is connected to a fundamental object. For example, if patterns of distribution are the only fundamental objects of distributive justice, then actions, rules, and dispositions are only appropriate objects of justice because of the way actions, rules, and dispositions can help to bring about particular distributive patterns. Similarly, if the basic structure of society (in Rawls’s sense) is the only fundamental object of distributive justice, then individuals’ actions would be objects of distributive justice only in the derivative sense—​they would be of interest to a theory of justice only insofar as they bear some causal or other connection to the basic structure. With this distinction in hand, let’s consider the debate between luck egalitarians and democratic egalitarians. Anderson might be taken as suggesting that for luck

Consequentialism and Deontology    315 egalitarians, states of affairs are fundamental, whereas for democratic egalitarians it is agents’ actions and dispositions that are fundamental. But this may be too simple. There are certainly some conceptions of luck egalitarianism where states of affairs appear to be the sole fundamental objects (Arneson 2011: 41–​2; Cohen 2008: 7, 268).8 But some luck egalitarians explicitly reject this formulation. Kok-​Chor Tan, for example, develops a form of luck egalitarianism where the major public institutions and rules governing society—​and not states of affairs—​are the fundamental objects of distributive justice (Tan 2012). More broadly, there is nothing about luck egalitarianism that entails viewing states of affairs as the sole fundamental objects.9 Conversely, it seems clear that democratic egalitarianism need not take individual actions and dispositions as the fundamental objects of justice. A democratic egalitarian might, instead, believe that what is of fundamental importance is a state of affairs where individuals stand in relations of equality that can be defined, at least in part, independently of actions and dispositions, with the latter having to that extent only derivative significance. The more general point is this. Democratic egalitarians may focus on relationships, whereas luck egalitarians may focus on patterns of distribution, but this doesn’t tell us what the fundamental objects of distributive justice are, since whether something is a fundamental or merely derivative object of justice depends on answering prior questions about the grounds and nature of distributive justice. To answer the question about the fundamental objects of justice we must decide what makes something just or unjust.

14.4  Grounds of Justice Suppose you are deciding whether to ϕ, and I tell you that ϕ-​ing would be unjust. You might ask, “What makes ϕ-​ing unjust?” In asking this question, you are asking about the normative grounds of justice and injustice. You want to know what fact about ϕ-​ing is the injustice-​making fact; what is the feature of ϕ-​ing in virtue of which it is unjust? There are, as with our previous question, fundamental and derivative answers to this question too. Suppose, in answer to your question, I reply “ϕ-​ing is unjust because it would involve breaking a promise to Albert, and this will cause Albert pain.” On this view, promise-breaking is only a derivative ground of injustice. The more fundamental ground is causing pain, though on further inspection it may turn out that “causing pain” is also only derivative of some more fundamental ground, perhaps “failure to maximize pleasure and minimize pain.” In any such normative investigation about the grounds 8 

Note that neither Arneson nor Cohen is explicit about whether states of affairs are the sole fundamental objects, but both can be plausibly interpreted in this way. 9  In this respect, the characterization of luck egalitarianism I offered in a previous paper was unduly narrow. In that paper I identified one version of luck egalitarianism—​albeit a prominent version—​but the view need not be construed in this way. See Quong (2011: 76–​7).

316   Jonathan Quong of justice, there will come a point at which no more fundamental feature explains the justice-​making or injustice-​making nature of what is under consideration. When we reach this point, we have found one of the fundamental grounds of our conception of justice. Luck egalitarianism and democratic egalitarianism are sometimes presented as offering rival answers to this question about the fundamental grounds of justice. Luck egalitarianism, in its most extreme form, has been understood as offering an exhaustive account of the fundamental grounds of distributive justice.10 On this view, once the equalisandum is in place, all claims of distributive justice are ultimately grounded by appeal to the distinction between luck-​generated versus choice-​generated inequalities. Every true statement of the form, “ϕ is distributively unjust”, will ultimately be grounded by an appeal to the way ϕ contributes to, or constitutes, an arbitrary inequality in the relevant equalisandum. Insofar as democratic egalitarianism holds that there can be types of distributive injustice that are not grounded in a claim about the injustice of arbitrary inequality, democratic egalitarianism and luck egalitarianism are rival positions. Democratic egalitarians typically do hold that there can be instances of distributive injustice that are not grounded in facts about arbitrary inequality. Timothy Hinton, for example, says, We ought to reject the claim that what makes inequality morally objectionable is its origins in brute luck. Take the system of racial domination that existed under the apartheid regime. That system required for its functioning an array of social and political institutions that enabled white people to dominate and exploit black people. What made the system evil, surely, was the way that black people were forced to live: their continual subjection, humiliation, and deprivation; their lack of access to education, decent jobs, and to the means of legal and political redress. The evil did not consist in the fact that the color of one’s skin is largely a matter of brute luck. (Hinton 2001: 79)11

Democratic egalitarianism is thus sometimes presented as offering a different account of the fundamental grounds of justice. The aim is to understand what principles of justice persons would agree to under fair conditions, or what principles of justice no person could reasonably reject.12 When our actions and institutions are governed by 10  G. A. Cohen claims “equality constitutes distributive justice,” and by that I think he most plausibly means that his preferred luck egalitarian conception of equality constitutes the sole principle of distributive justice (see Cohen 2008: 279). Arneson also affirms a view whereby luck egalitarianism is “all of distributive justice,” though his version of luck egalitarianism includes a prioritarian component (see Arneson 2011: 42). Also see Seligman (2007: 268) for an interpretation of luck egalitarianism as exhaustive of distributive justice. 11  For similar remarks see Anderson (1999: 288–​89); and Scheffler (2005: 12). 12  For explicitly contractualist presentations of democratic or relational egalitarianism, see, for example, Anderson (1999: 322); Anderson (2010: 3); or Tan (2012: 97). Scheffler also grounds his discussion of democratic egalitarianism via a discussion of Rawls’s theory, which is, of course, presented in contractualist terms (see Scheffler 2003).

Consequentialism and Deontology    317 such principles, we instantiate the appropriate relationship of equality amongst persons or citizens. For any particular judgment about justice, the ultimate explanation will thus stop with something like the following: “ϕ-​ing is unjust because it is inconsistent with living with others in relations of social or political equality.” Democratic and luck egalitarianism are, according to this picture, rival views about the ultimate grounds of distributive justice. I believe, however, that there are good reasons to reject this characterization since it renders both views less plausible than they might otherwise be. Let’s begin with luck egalitarianism. If luck egalitarianism is presented as offering an exclusive answer to the question about the grounds of distributive justice, it yields an account that is too narrow. The account is too narrow since it denies that anything else—​apart from how chance and choice bear on advantage—​has fundamental significance for distributive justice, but this is implausible. For example, an individual’s absolute level of advantage surely sometimes has non-​derivative significance when determining the distribution of goods. Even if Albert is fully responsible for being at a lower level of absolute advantage than others, I have at least some reason to give his claims greater moral weight the worse off he is, when deciding how to distribute some unexpected new benefit. Even if one believes that his responsibility for his lower level of absolute advantage overrides or outweighs the salience of the fact that he is at a lower absolute level of advantage, the fact he is at the lower level seems to have non-​derivative importance for the question of how the windfall ought to be distributed. Similarly, it is implausible to deny that some notion of sufficiency is relevant to questions of distributive justice. As many critics (and also some defenders) of luck egalitarianism have insisted, it appears at least pro tanto distributively unjust to deny someone life-​saving medical treatment on the grounds that she is responsible for her life-​threatening injury or illness when the treatment can be provided without denying anyone else anything of comparable moral importance.13 But democratic egalitarianism fares no better when presented as an account of the fundamental grounds of distributive justice. Consider a case where the current generation faces a policy question about the rate of savings for future generations. If the current generation chooses policy X which involves high rates of consumption now, then a future generation 150 years from now will lead very difficult lives with very few resources, whereas if the current generation chooses policy Y, which involves far less current consumption, the future generation will have roughly the same standard of living as the current generation. On one interpretation of democratic egalitarianism, because no members of the current generation will overlap with any members of the future generation, they cannot stand in any sort of relationships, egalitarian or otherwise, and thus democratic egalitarianism cannot deem this choice to be a matter of distributive justice. This, however, seems an unduly narrow account of the scope of distributive just­ ice. Proponents of democratic egalitarianism might argue that the current generation does, somehow, stand in a relationship with the future generation, and this relationship

13 

For discussion of these points, see Arneson (2011).

318   Jonathan Quong is the ultimate basis for why it is unjust to choose policy X over policy Y. This, however, seems implausible. If it is unjust for the current generation to choose X over Y, what ultimately makes this choice unjust is not the bad relationship it creates between the two generations. Rather, it seems clear that the basic injustice-​making grounds have to do with how badly off the members of the future generation will be. We can, if we like, say that in choosing X over Y, the members of the current generation create an objectionably inegalitarian relationship with the members of the future generation, but this seems parasitic on the prior fact that choosing X will make the future generation so badly off. This latter fact is what seems explanatorily more basic.14 In sum, although luck egalitarianism and democratic egalitarianism are often presented as rival accounts of the ultimate grounds of distributive justice, neither position looks particularly plausible when presented in this way.15 In what follows I propose an alternative framework.

14.5  What Justice Is Recall the three questions presented earlier: 1. What are the relevant objects of justice and injustice? 2. What makes something just or unjust? 3. What is it for something to be just or unjust? Thus far I have suggested that the answer to the first question depends on our answers to the second question, and I’ve also argued that luck egalitarianism and democratic egalitarianism do not offer very plausible answers to the second question, at least when presented as exclusive or comprehensive answers. Instead, I think luck egalitarianism is best understood as one of a plurality of answers to the second question, and democratic egalitarianism is best understood as an answer to the third question (a more precise explanation of the third question is provided later). As I  have already noted, democratic egalitarianism is sometimes presented in contractualist terms. As Anderson puts it, relational or democratic egalitarians endorse a “second person or interpersonal conception of justification. This follows from their 14  Of course, egalitarians (of any type) might explain duties of justice owed to distant future generations by appeal to the duties that arise amongst overlapping generations and thereby link the current generation to the distant one without requiring the existence of a direct duty of justice between the current generation and the distant one. Although this may be the best way to explain some duties owed to future generations, it cannot explain all the duties owed to future generations, since we can easily imagine cases where the current generation faces a choice that has no impact on any of the intervening generations, but a great impact on some distant generation. 15  Which also explains why it is possible to construct different versions of both views that provide different answers to the first question: what are the fundamental objects of justice?

Consequentialism and Deontology    319 contractualism. Contractualism is the view that the principles of justice are whatever principles free, equal, and reasonable people would adopt to regulate the claims they make on one another” (Anderson 2010: 3).16 Contractualism can, however, be understood in two different ways. On one view, contractualism provides the account of what makes something morally wrong or unjust; for example, ϕ-​ing is wrong because it is prohibited by a principle that no one can reasonably reject. When presented in this way, contractualism is an answer to the second question above. But contractualism can, alternatively, be understood, not as an account of what makes something morally wrong or unjust, but rather as an account of what it is for something to be morally wrong or unjust. This is how T.  M. Scanlon presents his version of contractualism. Contractualism is offered as an account of what moral wrongness is—​an account of a normative kind, where the property that all acts of this kind share is the property of being prohibited by a principle that no one could reasonably reject (Scanlon 1998: 10–​13). This property, Scanlon suggests, “seems to be connected in the right way with our reasons for thinking those actions to be wrong; and . . . it provides a plausible interpretation of reasons for avoiding such actions and criticizing those who perform them” (Scanlon 1998:  13). Wrongness is thus not a mysterious non-​analyzable property, and the contractualist analysis of wrongness competes with others (e.g., utilitarianism or divine command theory) as an explanation of what wrongness is and why we might think we have distinctive reasons to avoid wrongdoing and react to wrongdoers in particular ways. But failing the contractualist test is not what ultimately makes some act wrong. What makes some particular act wrong will be some more specific feature of the act, for example that it causes gratuitous suffering, or that it distributes benefits in a manner that is insensitive to individual responsibility, and so on. As Scanlon emphasizes, in response to an objection from J. J. Thomson, contractualism: . . . is intended as an account of what it is for an act to be wrong. What makes an act wrong are the properties that would make any principle that allow it one that it would be reasonable to reject. (Scanlon 1998: 391)

I propose a similar reconceptualization of the democratic egalitarian project. Democratic equality, on this view, provides an account of what social or distributive just­ice is. All individual instances of injustice share a common feature: to say that X is unjust is to say that a principle (designed to regulate the distribution of advantages and disadvantages between persons) prohibiting the act is one that could not be reasonably rejected by suitably situated contractors.17 This account of what distributive injustice is 16 

Also see Anderson (1999: 322); or Tan (2012: 97). I use this Scanlonian formulation of contractualism for expositional purposes in what follows, but the proposed reconciliation of democratic and luck egalitarianism does not depend on endorsing this particular version of contractualism. 17 

320   Jonathan Quong has a good deal of intuitive plausibility. To borrow Scanlon’s language, it seems to be connected in the right way with our reasons for thinking that particular acts or rules are unjust. And it provides a plausible interpretation of the special reasons we have to avoid social injustice and to criticize unjust actions and institutions:  injustice undermines a relationship of equality between persons or citizens, a relationship that depends on being able to live together on terms that no one can reasonably reject. But contractualism (or some similar test of interpersonal justification) does not, on this proposal, provide the fundamental explanation of what makes some act or rule unjust. Institutions that deny citizens the minimal resources needed to survive are unjust because of the great harm this unnecessarily causes to individuals. Institutions that discriminate on the basis of race in hiring employees are unjust because they do not give individuals an equal chance at a valuable social good for spurious and socially demeaning reasons. Rules that allow wealthy citizens to evade paying tax on their income and wealth are unjust because they countenance an unacceptable form of free-​ riding and they conflict with basic notions of reciprocity. These, I  suggest, are the fundamental grounds of injustice in each example. It is also true that each rule or institution would not be permitted by a principle that no one could reasonably reject, but this fact is best seen as a descriptive feature that supervenes on the more basic injustice-​ making grounds of each individual instance. Some proponents of democratic egalitarianism will resist this reconceptualization of the view. They insist that what makes various social arrangements unjust is the fact that they instantiate inegalitarian relationships or create social or political hierarchies amongst citizens who ought to regard one another as equal.18 In what sense, then, am I offering a recognizable version of the democratic egalitarian project? I am not offering an interpretation of the project as its current proponents understand it. I am offering a revisionary account, one that emphasizes some features of what democratic egalitarians have said, while rejecting or ignoring other features of the view. Like any revisionary view, it is bound to conflict with what some of its proponents believe. Democratic egalitarianism needs at least some revision, since it currently makes claims that are not consistent. Suppose we take seriously the idea that democratic egalitarianism is contractualist, in the sense described by Anderson, where the principles of justice are whatever principles free, equal, and reasonable people would adopt to regulate the claims they make on one another. If this is the right way to understand democratic egalitarianism, then democratic egalitarians cannot also claim to provide an account of the fundamental grounds of injustice. The reason is simple: contractualism is not a plausible account of the fundamental grounds of injustice. What makes a hereditary aristocratic social and economic system unjust is not the fact that such a system would be rejected by the relevant contractualist procedure. Rather, what makes it unjust is that it’s unfair for people to have hugely different opportunities in life for no good reason. It may also be unjust because it instantiates inegalitarian social relations

18 

Anderson (1999); and Hinton (2001) both emphasize this idea.

Consequentialism and Deontology    321 amongst people who ought to stand as civic equals. But if these are some of the injustice-​making facts, then failing the contractualist test is not the fundamental grounds of injustice. I’ve suggested democratic egalitarians should jettison the claim to be providing an account of the fundamental grounds of injustice and instead focus on the way their contractualism might provide an account of what social injustice is. But a proponent of democratic equality could, of course, draw the opposite inference. She could jettison the contractualist component, and present the view solely as an answer to the question: what makes something unjust? On this version of the view, democratic equality might amount to the following claim: DE1  The fundamental grounds of all social injustice are inegalitarian relations or wrongful hierarchies of status between citizens. But DE1 seems obviously false. The injustices of religious persecution, or dire poverty, are not primarily explained by appeal to inegalitarian relationships or wrongful hierarchies. Perhaps democratic egalitarians that aim to answer only the second of our questions thus ought to endorse a more modest claim: DE2  Whenever inequality is unjust, what makes it unjust are inegalitarian relations or wrongful hierarchies of status between citizens. On this view, democratic equality only purports to explain what makes inequality unjust, without claiming to explain all forms of social injustice. Although DE2 is more plausible, it’s still false. Suppose Albert steals some of Betty’s wealth via an electronic transfer in such a way that he never knows whose money he has stolen and Betty never discovers who has stolen her money. Albert spends this money on a year-​long trip around the world that Betty will now never be able to afford. It’s clear that there is an unjust inequality between Albert and Betty, but what makes this inequality unjust is not best explained by appeal to inegalitarian relations or wrongful hierarchies of status. It’s also easy to imagine societies where citizens regard one another as civic and moral equals—​where there are no hierarchies of status—​and yet unjust inequalities remain, for example certain employment and educational opportunities are allocated arbitrarily at birth. In light of such examples, proponents of DE2 can either implausibly insist that these are not instances of inegalitarian injustice, or else retreat further and endorse: DE3  What makes some inequalities unjust are inegalitarian relations or wrongful hierarchies of status between citizens. I think DE3 is probably true, but if democratic egalitarians retreat this far, democratic egalitarianism is no longer a complete account of egalitarian justice (let alone social just­ ice), nor is it necessarily in conflict with more modest forms of luck egalitarianism. In sum, those who believe democratic egalitarianism is better conceptualized as an answer

322   Jonathan Quong to the second question—​what makes some things unjust—​will end up with a view that is at least as revisionary as my account of democratic egalitarianism. Luck egalitarianism, on the other hand, does seem most plausible as an account of what makes certain actions, rules, or institutions unjust. But the appeal to the chance/​ choice distinction must be seen as a single consideration amongst a plurality of others that sometimes grounds judgments about justice and injustice.19 A  society where a person’s opportunities in life are largely determined by lottery—​for example, the lottery of who one’s parents happen to be—​is unjust precisely because it is unfair to structure society in a way that ensures some people will be much worse off than others for arbitrary reasons. Similarly, if one person becomes poor due to a responsible decision to spend her resources on gambling and parties, it is unjust to treat this person’s claims to social assistance as having the same weight as someone who has become poor through no choice of her own. It is the arbitrariness of the inequality in the first case, and the fact that one person’s poverty can be traced to their responsible choices in the second case, which grounds the relevant judgments about justice. But a focus on how chance and choice bear on advantage does not ground all of our judgments about distributive just­ ice. It does not ground the judgment that it is unjust to fail to provide life-​saving medical treatment to those who need it, regardless of their degree of responsibility. It does not ground the judgment that it can be just to create a system of inequality-​generating economic incentives when those incentives work to the benefit of the least-​advantaged members of society. Once luck egalitarianism and democratic egalitarianism have been interpreted in this way, they are no longer rival views about the nature of egalitarian justice.20 Instead, they can be seen as complementary answers to distinct questions about distributive just­ ice. Luck egalitarian explanations of what makes some particular action or institution unjust can comfortably co-​exist with a democratic egalitarian account of what it is for something to be unjust. Whenever luck egalitarianism does provide the grounds for a judgment about the injustice of some action or rule, it will also be the case that this action or rule is one that would be prohibited by a principle that no one could reasonably reject. The advantages of reconceptualizing both democratic and luck egalitarianism in this way are, hopefully, clear. I have already emphasized that neither view seems particularly plausible when presented as the sole grounds for judgments about distributive justice. The proposal allows each view to play a role for which it is better suited. Luck egalitarians need not implausibly claim that all judgments about distributive justice are reducible to judgments about whether an inequality is the result of brute luck or responsible choice. And democratic egalitarians—​on this proposed view—​need not insist that what ultimately makes unfettered inheritance, or the failure to provide for future generations, unjust is the fact that doing so undermines relations of social or political equality. The

19 

20 

Consider, for example, how Rawls understands the principle of redress (see Rawls 1999: 86–​7). For a different way of reconciling the two views, see Schemmel (2012).

Consequentialism and Deontology    323 proposal also has the advantage of clarifying how the considerations that make some rule or institution unjust are connected to the distinctive reasons I have to care about justice and to react to injustice in particular ways. A  complaint sometimes pressed against luck egalitarianism, for example, is that there don’t seem to be any distinctive reasons to care about arbitrary inequality as such. The mere fact that there is a brute-​ luck generated inequality between persons does not, at least for some, seem appropriately connected to any of the distinctive reactions we have to paradigmatic instances of social injustice. But insofar as we have reasons to care that the terms on which we interact with others are interpersonally justifiable, and if the fact that some inequality has its source in brute luck is sometimes the crucial fact that makes it the case that some act, rule, or institution is not interpersonally justifiable,21 then we have an explanation of how the diverse grounds of injustice connect to a set of related, but distinctive reasons to care about living on just terms with others. The proposal described in this section is clearly only a sketch, and many details need to be filled in. In section 14.6 I address one important objection to the proposed reconciliation of democratic egalitarianism and luck egalitarianism.

14.6  An Objection Skeptics might worry that the proposal in section 14.5 depends on an untenable distinction between what makes something unjust, and what it is for something to be unjust. In particular, how can contractualism be an account of what injustice is without also being the ultimate explanation of what makes something unjust? Consider the existence of arbitrary inequality between persons. On the view proposed above, this is sometimes the fundamental ground for a judgment that some institution X, is unjust. But it is not the only grounds of injustice: there are others. Moreover, the skeptic might plausibly insist, there will be cases where an institution may create or sustain arbitrary inequalities between persons and yet not be unjust. Arbitrary inequality is thus sometimes, but not always, the grounds for correctly assessing some act or institution as unjust. But this implies that there is some deeper, or more comprehensive, explanation of when arbitrary inequality does, or does not, constitute the basis for an attribution of injustice. The contractualist ideal seems to provide this deeper explanation: arbitrary inequalities are only unjust when they would be prohibited by a principle that no one could reasonably reject. If so, our skeptic insists, then it is really the contractualist test that ultimately explains what makes X unjust, since the fact that X causes arbitrary inequality only makes X unjust when this fact is one that makes it true that X is prohibited by the contractualist test. On this view, the distinction between what injustice is and 21  It is beyond the scope of the current chapter to consider whether the test of interpersonal justification that identifies instances of social injustice applies only to some objects (e.g., acts, rules, or institutions) and not others.

324   Jonathan Quong what makes something unjust cannot be sustained, at least not when contractualism is presented as the account of what injustice is. This objection, however, depends on a misunderstanding of contractualism. Contractualism, as I understand it, offers a particular way of modeling our fundamental moral or political values (e.g., the idea of persons as free and equal), and is also designed and modified to cohere with our most strongly held considered convictions (e.g., torturing people for pleasure is unjust). Given that the contractualist standard is something that is designed in light of our moral convictions—​to reflect those convictions and see how we might organize our existing judgments into a single framework—​it is implausible to see the contractualist test itself as providing the fundamental reasons grounding judgments of justice and injustice. Even if contractualism is part of the wider method of reflective equilibrium—​one where particular considered convictions might be revised if they conflict with the more abstract principles that contractualism recommends—​this does not make the contractualist test basic in the normative order. Consider the fact that there are many different versions of contractualism—​the information available to contractors, the contractors’ aims and motives, and the question the contractors address are just some of the dimensions along which different models of contractualism can vary. To evaluate competing conceptions of contractualism, we must appeal to our existing normative convictions—​the convictions contractualism is meant to model and help us better understand. But if we use normative convictions to decide which version of contractualism is best, or at least which one is better, then it is implausible to view the contractualist device as providing the fundamental grounds of our judgments about justice and injustice. Here is an analogy. Suppose we have different conflicting maps that purport to accurately map the physical terrain. We choose the map that seems to do the best job of making sense of the physical terrain, given our current and independently formed beliefs. Even if the map we have chosen turns out to be accurate in every respect—​and even if we subsequently adjust some of our pre-​existing convictions about distances or locations in light of the map’s explanatory power—​it is not the map itself that makes it true that the mountain to the west is five hundred miles away. Contractualism is, in this sense, like a map. We use our existing beliefs about justice and injustice to construct a map or theory that provides a more complete picture of what justice and injustice are. But even if we come to believe our preferred version of contractualism perfectly represents the truth about justice and injustice, this does not mean the contractualist test is what makes any particular act or institution just or unjust.22

22  This view of contractualism also makes sense of the fact that ordinary persons (i.e., non-​ philosophers) can understand what makes various actions or institutions unjust, without requiring those people to understand a complex form of contractualism. To insist that the ultimate ground of injustice is the fact that the act or rule in question would be prohibited by a principle that suitably situated contractors could not reasonably reject is to assume that the vast majority of people do not understand the basic grounds of ordinary instances of injustice.

Consequentialism and Deontology    325 There is, however, an important disanalogy between physical maps and contractualism. We don’t usually have distinctive reasons to care about the information we receive from maps of the physical world, but contractualism is different. The contractualist map does not simply identify which acts, rules, and institutions are just and unjust—​it also identifies the property that is common to all instances of injustice and all instances of justice. Just acts, rules, and institutions are those that can be reasonably justified to all the members of the relevant constituency, whereas unjust acts, rules, and institutions are characterized by their failure to be interpersonally justifiable. It’s very plausible to suppose that we have distinctive reasons not to create or sustain relations with others where those relations cannot be reasonably justified to all the members of the relationship, and so contractualism identifies the distinctive reasons we have to avoid and combat injustice.

14.7 Conclusion In this chapter I have made two main suggestions. First, I argued that existing debates about egalitarian justice are not instantiations of a broader dispute in normative ethics between consequentialism and its alleged competitors. Second, although luck egalitarianism and democratic egalitarianism are often presented as rival conceptions of justice—​in particular, as offering competing accounts of the grounds of distributive justice—​I have suggested this may be a mistake, since it makes each view less plausible than it otherwise might be. Instead I proposed an alternative view where luck egalitarianism and democratic egalitarianism can be understood as complementary answers to different questions. Luck egalitarianism identifies one of the fundamental grounds of justice and injustice, whereas democratic egalitarianism is better conceptualized as offering an account of what it is for something to be just or unjust. What it is for X to be unjust, on this view, is for X to be prohibited by an appropriate test of interpersonal justification. Consequentialism, deontology, and contractualism are often presented as rival accounts of normative ethics.23 We are told that one must choose between these competing views of the ethical landscape. If this is correct, then the topic of egalitarian justice might require the same choice: in order to understand what equality requires with regard to questions of distributive justice, one first must decide whether the topic is best approached from a consequentialist, deontological, or contractualist perspective. But if the two main claims in this chapter are correct, then this is a false choice. It might be that these different normative perspectives are all at least partially correct, or that they each correctly capture one part of the truth about egalitarian justice.

23 

Though Derek Parfit rejects this view in Parfit (2011).

326   Jonathan Quong

Acknowledgments I am very grateful to Serena Olsaretti, Zofia Stemplowska, and Rebecca Stone for comments on earlier drafts of this chapter. Some of the ideas in this chapter were initially conceived with Zofia Stemplowska, but she is not responsible for the way I have developed and presented them.

References Anderson, E. (1999). “What is the Point of Equality?” Ethics 109(2): 287–​337. Anderson, E. (2010). “The Fundamental Disagreement between Luck Egalitarians and Relational Egalitarians.” Canadian Journal of Philosophy 40(sup1): 1–​23. Arneson, R. (2011). “Luck Egalitarianism—​A Primer,” in C. Knight and Z. Stemplowska (eds) Responsibility and Distributive Justice. Oxford: Oxford University Press, pp. 24–​50. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Dreier, J. (1993). “Structures of Normative Theories.” The Monist 76(1): 22–​40. Hinton, T. (2001). “Must Egalitarians Choose between Fairness and Respect?” Philosophy & Public Affairs 30(1): 72–​87. Kamm, F. M. (2013). “Nonconsequentialism,” in H. Lafollette and I. Persson (eds) The Blackwell Guide to Ethical Theory. Oxford: Wiley-​Blackwell, pp. 261–​86. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. Parfit, D. (2011). On What Matters. Vol. 1. Oxford: Oxford University Press. Quong, J. (2011). “Left-​Libertarianism:  Rawlsian not Luck Egalitarian.” Journal of Political Philosophy 19(1): 64–​89. Rawls, J. (1999). A Theory of Justice, revised ed. Oxford: Oxford University Press. Scanlon, T. M. (1998). What We Owe to Each Other. Cambridge, MA: Belknap Press. Scheffler, S. (2003). “What is Egalitarianism?” Philosophy & Public Affairs 31(1): 5–​39. Scheffler, S. (2005). “Choice, Circumstances, and the Value of Equality.” Politics, Philosophy & Economics 4(1): 5–​28. Schemmel, C. (2012). “Luck Egalitarianism as Democratic Reciprocity? A Response to Tan.” Journal of Philosophy 109(7): 435–​48. Seligman, M. (2007). “Luck, Leverage, and Equality:  A Bargaining Problem for Luck Egalitarians.” Philosophy & Public Affairs 35(3): 266–​92. Steiner, H. (1994). An Essay on Rights. Oxford: Blackwell. Stemplowska, Z. (2011). “Responsibility and Respect: Reconciling Two Egalitarian Visions,” in C. Knight and Z. Stemplowska (eds) Responsibility and Distributive Justice. Oxford: Oxford University Press, pp. 115–​35. Tadros, V. (2011). The Ends of Harm: The Moral Foundations of Criminal Law. Oxford: Oxford University Press. Tan, K. (2012). Justice, Institutions, and Luck: The Site, Ground, and Scope of Equality. Oxford: Oxford University Press.

Chapter 15

Ideal Th e ory David Schmidtz

Morality describes the conduct of perfect men, and cannot include in its premises circumstances that arise from imperfection. That rule which attains to universal sway when all men are what they ought to be, must be the right rule, must it not? (Herbert Spencer, Social Statics 1871)

15.1  Ideal Theory: What It Was Here is the idea that launched a debate and defined ideal theory. The intuitive idea is to split the theory of justice into two parts. The first or ideal part assumes strict compliance and works out the principles that characterize a well-​ordered society under favorable circumstances. It develops the conception of a perfectly just basic structure and the corresponding duties and obligations of persons under the fixed constraints of human life. My main concern is with this part of the theory. Nonideal theory, the second part, is worked out after an ideal conception of justice has been chosen; only then do the parties ask which principles to adopt under less happy conditions. (Rawls 1999a: 216)

Rawls goes on to say, With the presumption of strict compliance, we arrive at a certain ideal conception. When we ask whether and under what circumstances unjust arrangements are to be tolerated, we are faced with a different sort of question. We must ascertain how the ideal conception of justice applies, if indeed it applies at all, to cases where rather than having to make adjustments to natural limitations, we are confronted with injustice. The discussion of these problems belongs to the partial compliance part of nonideal theory. (1999a: 309)

328   David Schmidtz As Rawls conceded, “Obviously the problems of partial compliance theory are the pressing and urgent matters. These are the things that we are faced with in everyday life” (1999a: 8). Despite this, Rawls adds, “I consider primarily what I call strict compliance as opposed to partial compliance theory” (1999a: 8). Why? “The reason for beginning with ideal theory is that it provides, I believe, the only basis for the systematic grasp of these more pressing problems. . . . At least, I shall assume that a deeper understanding can be gained in no other way” (1999a: 8). As Rawls knew, ideal and nonideal theory are not rivals. Each plays a role in political theory done well. Yet, as Rawls also knew, where we start is critical. Assuming perfect compliance leads down one road. Realistic assumptions lead down another. We assume perfect compliance not because more realistic assumptions go nowhere so much as because they do not go where Rawls wants theorizing about justice to go. As Rawls puts it, “We want to define the original position so that we get the desired solution” (1999a: 122). We can see why he would say that. However, precisely because Rawls is right that it matters where we start, we cannot afford to be uncritical about where he starts. Rawls also says, “until the ideal is identified, at least in outline—​and that is all we should expect—​nonideal theory lacks an objective” (1999b:  90). Really? Alexander Graham Bell had a sense of where long-​distance communication was in his day and where it could go. Thomas Edison had a sense of the possibility of recorded sound. They did not lack for objectives, and their objectives were rooted in their sense of what was possible, not in any sense of what was ideal. They did not need a theory about whether ideal sound transmission would be faster than the speed of light. Progress in solving problems was relevant and meaningful; perfection was neither. Where there are facts, subject to change in ways that matter, and where there is something we can do, we have a problem. Where there are problems, there can be objectives. Where there are objectives, there can be theorizing with a point. Problems set parameters for what to count as a solution, and ultimately for what would be our best—​our ideal—​response. Likewise, some of our best legal reasoning comes from judges aiming, not to approximate perfection, but to solve a problem—​to resolve disputes brought before their bench by real litigants. It is not law’s purpose to be an ideal answer to all possible questions. Systems evolve, new conflicts emerge, and judges will one day need to say more than has yet been said about what would make a newly emerging kind of dispute easier to resolve or avoid in the future. Real solutions do not track perfection but instead play a perpetual game of catch-​up with evolving problems. That is how progress occurs in fields where progress occurs—​not necessarily, of course, but as a matter of fact.1 1  In an ideal world, would there be progress, or would progress be inconceivable? It depends on whether an ideal society is an ideal outcome or an ideal process. Let’s say a utopian ideal is a “peak” outcome from which further progress is inconceivable, whereas a realistic ideal is a process that makes progress likely. A utopian, seeing problems, imagines what it would be like not to have them. A realist, seeing problems, imagines what it would be like to solve them.

Ideal Theory   329 Realistic idealism is theory about ideals, but does not start with ideals. Realistic idealists know that it matters where we start, so they do not start with a desired conclusion. Instead of letting an answer dictate what they count as a question, theorists who succeed at being realistic let real, located, practical questions dictate what they count as an answer. Suppose we call something ideal. Then we pause to wonder whether there is any reason—​any reason whatever—​to call it ideal. Suppose Thomas Edison says, “in an ideal world, our incomes would be equal, and there would be no carnivorous plants.” Under what circumstances would that answer a real question? Here is a suggestion: theorizing goes astray, not when it bears on ideals, but when it fails to bear on problems. By failing to bear on problems here and now, it fails to bear on what we have reason to regard as an ideal response to here and now, and instead bears only on what would be ideal if we depart far enough from the human condition as it really is. Realistic idealism, that is, theorizing about ideal responses identifies x as worthy of aspiration here and now, starting from a sober assessment of real problems here and now. Disciplined idealism—​ realistic idealism—​manifestly is worth doing well.

15.2  What It Could Be Suppose we define a perfectly just world as one that could not be more just. What else would a perfectly just world be? Utopian idealism, imagining a world so perfect that nothing remains for justice to demand, is one alternative. Realistic idealism, by contrast, consists of imagining a perfect response to circumstances that demand a response. Note: the best available response is not always what we would call ideal. We call the best available response ideal only when we accept some fairly strong version of the thought that we could not have done better. Suppose I  am imagining that cooking lasagna would be ideal for our dinner tonight. Then you mention that you are allergic to tomatoes. I should reply, “That fact calls for a change of plan, because given that fact, serving lasagna tonight would be far from ideal.”2 Suppose instead that I reply by saying, “So far, there is no discernible defect in my plan to make lasagna for our dinner tonight. 2 

I gesture here at a contrast between my view and David Estlund’s. Estlund says, “People could be good, they just aren’t. Their failures are avoidable and blameworthy, but they are also entirely to be expected as a matter of fact. So far, there is no discernible defect in the theory, I believe. For all we have said, the standards to which it holds people might be sound and true. The fact that people will not live up to them even though they could is a defect of the people, not of the theory” (Estlund 2008: 264). Estlund’s insight is that it takes two to make a motivational failure. The bare fact of people responding badly to a standard does not entail that the standard is faulty. Of course we may predict that students will fail my exam without blaming my exam. Still, responsible reflection on a predictably bad outcome does

330   David Schmidtz It is you that is defective, not my lasagna.” Why does this sound immature? Perhaps because it disregards the fact that there are people for whom my lasagna would not be ideal, and you are one of them. The problem is not that lasagna is infeasible but that it is undesirable under the circumstances. If I find that I literally cannot execute my plan (perhaps tomatoes are unavailable), then I am finding that serving lasagna is not feasible. By contrast, if I find out what a bad idea it would be to serve you lasagna, then I am finding that serving lasagna is undesirable. Being infeasible does not entail that I should stop calling my plan ideal. But being undesirable does. Realistic idealism does not give up on ideals; it gives up on starting with ideals (see Galston 2010 and Elkin 2006.) To those who start with ideals, it can seem okay to have an intuition that everyone ideally would have the same income, and to wish such intuitions were untestable—​fact insensitive—​in a world that has indeed tested those intuitions (to the extent that worlds of uncontrolled variables ever test the hypotheses of social science) and found them wanting. But realists who seek ideal responses cannot look at it that way. To them, there is a general rule: for any problem P, there may be an ideal solution, but for S to be an ideal solution to problem P, S must first be a solution to problem P. From the perspective of a realistic idealist, it is not ideal to set aside the problem, and it is not ideal to set aside whether S solves the problem. Because utopian idealists aim to describe an ideal world rather than an ideal response to the world, it follows that what they call ideal does not entail feasible. So, far, there is nothing to quarrel about. But note how much more peculiar it is if what they call ideal does not even entail desirable. And yet, whether S is desirable is as fact-​ sensitive as whether S is feasible. Some theorists concede that we cannot, on humanitarian grounds, put (their intuition about) justice into practice.3 But if a vision is of a system characteristically at odds with humanitarianism, then the problem with the envisioned system is not that we lack any way to get it, but that we lack any reason to want it. If a system’s logic makes its inhumanity predictable when premised on human nature as it is, the problem with instantiating that logic is not that we can’t do it, but that we shouldn’t want to. Edward Hall, similarly, concludes that: . . . the insistence that political philosophy must start with the acceptance of various facts, such that in politics people do not display the sort of unity of purpose that they not simply make that point and then stop. That students predictably misread double negations is not a defect in my exam, but littering my exam with double negations is. Once I know this about students, the implication is clear: My exam is bad and I need to fix it. It might be ideal for ideal students, but it is not ideal for mine. Or, suppose it is entirely to be expected that insects will evolve resistance to what seemed like an ideal insecticide. Does that count against insects or against the insecticide? Should we infer that the insects are defective? To be sure, we have a choice, but the only honest answer is that we never had an ideal insecticide. Further, whether it was ideal was never a question to answer by envisioning counterfactual conditions under which the insecticide would have worked. 3  Cohen embraces this possibility (2008: 271–​2 or 302–​4). To Cohen, saying humanitarian values sometimes trump egalitarian values is (oddly) not the same as saying the normativity of egalitarian values is fact-​sensitive.

Ideal Theory   331 do in Cohen’s camping trip, is not a “feasibility” concern at all, but rather the more fundamental requirement that we actually address the practice with which we claim to be concerned. (2013: 180)4

But again, this is not a broadside against ideal theory. My target here is a specific kind of ideal theory: a utopian idealism that is not even trying to discover an ideal response to a recognizably human problem. Much of what we currently call ideal theory is an exercise in imagining how we would reinvent the world if only we could start with a clean slate and did not need to start from here. In truth, we live in a Neurath’s boat and cannot do better than to replace one plank at a time. What actually needs doing had better be guided by ideals of a kind, but if we think we need to be guided by a vision of the ideal boat, we are mistaken. Real solutions to real problems characteristically exhibit twists of historical contingency that armchair philosophy cannot anticipate. The convention of driving on the right solves a problem in some societies. It makes no difference whether driving on the right is a necessary condition for solving a problem; it makes no difference whether a convention of driving on the left would have been at least as good. When a convention of driving on the right emerges and solves an important problem, members of that society become warranted in expecting each other to drive on the right.

15.3  Toward a Philosophy that Starts at the Start The assumption that bargainers choose for a closed society, Rawls admits, “is a considerable abstraction, justified only because it enables us to focus on certain main questions free from distracting details” (1993: 12). A theory articulates an understanding, which entails deciding what to ignore. Although we face the human condition in its full complexity, theorizing begins with describing, and, as Peter Godfrey-​Smith observes of biological theories, “ignoring some features in a description of a system is inevitable to some extent in any description” (2009: 48). Yet, it is easy to slide from ignoring to ignoring “with prejudice”: setting details aside not because they don’t affect the answer but precisely because they do. Can we ignore the fact that humans live in increasingly open societies? Yes, provided that this detail makes no difference to the question we aim to answer. In practice, to begin with a problem is to begin more problematically with a description of a problem. That is, descriptions are idealizations. They all set aside details, 4 

See also Ronzoni (2012).

332   David Schmidtz notwithstanding the fact that what gets set aside as a detail is not guaranteed to turn out to be merely a detail. Yet, however controversial our description inevitably must be, there remains such a thing as the human condition. Simon Hope (2010: 135) warns against working on idealized representations with cherry-​picked features rather than on the situated problems of real human beings. But as Hope also notes, the bare fact of articulating a question is not enough to convict us of begging the question. There is no uniquely accurate account of the human condition, yet there are more and less accurate accounts.5 When Rawls agrees with Hume that we live in a world of real but limited generosity, we see the truth in what they say. First, they are not telling us everything of any conceivable relevance—​of course they are simplifying—​but neither are they making it up. Second, Hume and Rawls agree, we live in a world of scarcity. Third, the scarcity we face is manageable if we cooperate. Fourth, we are intelligent; cooperation is possible. None of us can know everything, decide everything, or do everything. Yet, we anticipate and respond to each other, communicate, give our word, by our deeds make our word count for something, and in time learn to trust one another. These are enduring features of the human condition. They motivate us, and enable us, to be a kingdom of ends. Rawls says society is a cooperative venture for mutual advantage. Hume saw that, as Rawls would concede, the emergence of a cooperative venture is an achievement. Thus, Hume asked, which basic structures foster society as a cooperative venture? Hume saw this as an empirical question and saw social science as a key to real answers. He sought “to introduce the Experimental Method of Reasoning into Moral Subjects.”6 Rawls launched his greatest work with the thought that justice is the first virtue of social institutions (1999a: 3), suggesting that a theorist’s main task is to articulate principles of justice. Rawls’s sentence resonates. It is lyrical, poetic, compelling. But it is not right. To Bernard Williams, the first institutional question concerns not justice so much as “the securing of order, protection, safety, trust, and the conditions of cooperation. It is ‘first’ because solving it is the condition of solving, indeed posing, any others” (Williams 2005: 3). Historically, we make progress when we acknowledge that the first virtue of social institutions is that they enable us to be neighbors. Being neighbors is not everything, but it is a start. In practice, we first secure a framework of mutual expectation that keeps the peace well enough to enable society to be, in the most rudimentary and non-​theory-​laden sense, a cooperative venture for mutual advantage. Then we start talking about justice.7 In passing, the further implication is that when we get to a point of being able to talk about justice we realize that what Rawls calls justice is not even the first question of justice. That is, Rawls’s opening question is about 5 

See Larmore (2013) and Part 1 of Schmidtz (2006). This is the subtitle of Hume’s Treatise of Human Nature. 7  Williams had no interest in underwriting the status quo. To say we need people to form mutual expectations and rely on them enough to form a community is not to valorize any particular order, including the status quo. 6 

Ideal Theory   333 distributing the fruits of cooperation, but before we begin to distribute what separate persons bring to the table, fairness mandates figuring out how to respect the fact that they brought it.8 Of course, we cannot settle disputes by deciding that our vision has a right to be colonial and that we can condescendingly dismiss rival visions as unreasonable. Real resolution starts by aiming for real resolution. To be in the grip of a vision—​any vision—​is problematic. What we need is not to envision but to listen. That is, we need politics. Of course, neither Hume nor Rawls represented their maps of the human condition as complete. Maps at best supplement rather than replace good judgment. Maps represent a problem. They do not constitute it. Moreover, maps can mislead. Thus, a challenge for any theorist is to distinguish what can safely be set aside as a distracting detail from what cannot. Rawls’s thought experiment, like most thought experiments meant to elicit intuitions about distributive justice, has us facing the above circumstances as adults, meeting to divide a pie from nowhere, devoid of historical baggage. In reality, we arrive at different times to a world knee-​deep in baggage.9 We each arrive on the scene to find a pie already produced, divided, claimed, and in use by someone else. Philosophers for whom the pie’s history is nothing but a distracting detail are primed to see equal shares as presumptively fair shares. But except when we actually do arrive simultaneously, there is no fact of the matter regarding what to count as an equal share. Philosophers are trained to set this detail aside, but it changes everything. Suppose twenty-​year old Smith is hired today at half the wage of his forty-​year old colleague Jones but twice the inflation-​adjusted wage that Jones was paid when Jones was hired twenty years ago. Does one of them have a valid complaint? If so, which one? Must we avoid progress, lest people born today have a higher quality of life than (and thus an arbitrary advantage with respect to) people born a generation ago? Further, must equal shares look like equal shares at any particular moment? Or is it okay for shares to even out over a lifetime, where highly skilled fifty-​year-​olds have incomes dwarfing what unskilled teenagers are making at that moment?10 Another sometimes pivotal feature of the human condition is that we not only arrive at different times: we arrive helpless, incapable of the reciprocity that grounds relations among free and equal adults. We arrive to a community that does not yet need us, and whether it some day will be better off with us than without us is not a given. That we make it to the table at all implies that we had help.11 Just as there tends to be no fact 8 

Rawls saw that question coming, but it is hard to imagine a worse answer than to say, “Rational contractors regard the producers they represent as bringing nothing to the table but morally arbitrary accidents of the genetic and social lottery to which they have no claim. Said producers, having been represented so capably, will strictly comply with whatever we condescend to distribute to them.” 9  See Part 4 of Schmidtz (2006). 10  For a survey of choices that need to be made in order to bring ideals of equality to bear on institutional practice, see Brighouse and Swift (2014). See also McKerlie (1989). 11  See Robeyns (2008: 358). See also Part 5 of Schmidtz (2006).

334   David Schmidtz about what counts as an equal share, there is not always a settled truth about what counts as returning a favor, or to whom a favor is or could be owed. Do we have debts to the next generation because of what the previous one gave us?12 Answers come not from philosophical analysis so much as from ongoing contingent mutual adjustment of expectations within communities. (I distil this lesson from observed histories of people muddling through, avoiding and resolving conflict. It is not something I would have seen simply by surveying my intuitions.) What people around us call justice is not something we lightly set aside as a distracting detail—​not even if we fail to see the point of seeing things their way. Presupposed by all this—​the most primordial political fact of all—​is the fact that I am not alone. I live among beings who decide for themselves, not pawns that I move and sacrifice at will. I may feel that people cannot reasonably reject my deepest convictions about justice. But they can, and they know it. This fact makes politics what it is, and just­ ice what it is.13

15.4  Setting Aside the  Human Condition As noted, we must set aside distracting details and focus on the problem—​on the human condition if we are theorizing about justice—​even though this invites accusations of begging someone’s version of the question. But one thing we must not set aside is the problem. Tucson’s city government once sought to manage traffic flows by designating inner lanes of major roads as one-​way lanes toward the city center during the morning rush. During the evening rush, the same lanes reversed and became one-​way lanes from the city center. At off-​peak times, inner lanes reverted to being left-​turn lanes. In a world of ideal drivers, it might have solved the problem. In Tucson, with its steady influx of elderly drivers unfamiliar with local conventions, where one indecisive driver is enough to create a dangerous mess, the system was a recipe for traffic jams, accidents, and road rage. One way of describing the mistake is to say traffic managers set aside the problem. Alternatively, they solved an idealized problem rather than a real one. A proper aim of serious traffic management is to minimize collisions, not to be a system that would minimize collisions between ideal drivers. Solving idealized problems is not a way of being a serious idealist. That a traffic management scheme would work for ideal drivers says nothing in its favor even as an ideal.

12  13 

On the imperfect duty of reciprocity, see Part 3 of Schmidtz (2006). See Sleat (2014), and Jubb (2012).

Ideal Theory   335 We can go badly astray if we strive for what Rawls called a “systematic grasp of more pressing problems” by assuming away those very problems. And yet, egregious though this sounds, it is not always a mistake. A physicist can demonstrate friction’s importance by modeling frictionless systems. Ronald Coase demonstrated transaction cost’s significance by modeling an economy that lacks the friction of transaction cost. This was no mistake; it would, however, be a mistake to think solving idealized problems yields approximations of solutions to real problems. It would be almost comical to see Coase’s exercise—​showing what is efficient in the absence of transaction cost—​as showing what is approximately efficient in a world like ours. Nothing of the kind follows. It likewise is almost comical to see Rawls’s exercise—​showing what is fair in the absence of compliance problems—​as showing what is approximately fair in a world like ours. Here too, nothing of the kind follows. In both cases, what is set aside is anything but a detail. It also would be a mistake to think that solutions to idealized problems can set the bar regarding what to count as having high standards. In our world, driving apt for a world of perfect drivers would be driving, not to a higher standard, but to a grossly inferior one. A genuine ideal portrays driving well in a world of other drivers, which is to say, being a defensive driver. Whether drivers ideally would not need to drive defensively has nothing to do with what counts as being an ideal or even a minimally competent driver in the real world. The same could be said of the idea that people ideally would not need to anticipate that the people around them are capable of strategic behavior. Such a conjecture has no bearing on how to respond ideally or even competently to a world like ours.14 An ideal traffic manager’s idealism rises to the level of seeking ideal responses to real problems. If a proposed solution would not help, the serious conversation is over. There is no further question about whether we can imagine it helping under ideal circumstances.

15.5 Separateness Treating unconditional giving as an ideal is a way of setting aside rather than honestly addressing the problem of politics. Here is the argument. When Rawls says society is a cooperative venture for mutual advantage, this opening line tells readers that the theorizing to come is an exercise in realistic utopianism. Utopian because a real society is more, and sometimes less, than a mutually advantageous cooperative venture; Rawls is describing a society of reciprocators as it could be and should be, not how it is. Realistic too, because what Rawls is portraying is not an ideal society so much as our own society’s ideal response to the human condition. What 14  Andrew Mason observes that we are not always sensitive enough to whether theorizing has “identified a principle of justice for a different kind of being rather than a more ultimate principle of justice” (2012: 539).

336   David Schmidtz Rawls was picturing was not a dream world, but rather how our own society truly does look when we respond to the human condition with the best strategy we have: namely, with reciprocity. To briefly review the game-​theoretic background, The Prisoner’s Dilemma models a key insight into the human condition. In the example from which the name derives, you and your partner Jane face criminal charges, and up to ten years in jail. You need to decide whether to betray Jane by testifying against her. The prosecutor makes you an offer: If Jane keeps silent, you get a ten-​year sentence reduction if you testify, or a nine-​year reduction if you also keep silent. Alternatively, if Jane testifies against you, you get a one-​year reduction for testifying, or zero reduction if you keep silent. Jane has received the same offer. The essence of a Prisoner’s Dilemma is that, collectively, all players are better off cooperating, while individually, each player is better off defecting. For each of you, keeping silent optimizes your collective sentence reduction, while testifying optimizes your individual sentence reduction. Testifying is a dominant strategy: each of you individually is better off betraying the other (one year better off in this example), no matter what the other does. In the Prisoner’s Dilemma, “withholding” is a dominant strategy:  if your partner cooperates, you are better off withholding your contribution; if your partner “withholds,” you still do better to withhold. Now let the game be repeated. In an “iterated” Prisoner’s Dilemma, you still choose whether to cooperate or withhold, but repeated play allows for conditional strategies. In particular, you can play “tit-​for-​tat” (moving as your partner moved in the previous round), and in that way reciprocate in the simplest way possible, and thereby make it pay for your partner to cooperate. Although withholding pays more in a one-​shot game, reciprocated cooperation pays more than reciprocated withholding in the long run.15 This bit of game theory can help us to distinguish real from spurious ideals. Arguably, there is an ideal strategy in a repeated Prisoner’s Dilemma: namely, reciprocity. Why? Because (1) it matters whether your partner cooperates; yet (2) you do not choose whether your partner cooperates. However, (3) you can make it pay for your partner to cooperate. This logic makes reciprocity an ideal strategy in a repeated Prisoner’s Dilemma. In this context, reciprocity is like defensive driving. It is “driving” to a high standard: it acknowledges fellow players as players. There is no better strategy for making cooperation pay.16 By contrast, a spurious ideal of unconditional giving makes free-riding pay. It is a recipe for conflict. 15 

Axelrod 1984. I allow that the strategic case is not the only case. There are nonstrategic situations—​ early child rearing, say—​where children taking full advantage of a free-ride is the relationship’s point. Thus, it is a mistake to assume ideal compliance and partial compliance are mutually exclusive by definition. When strict compliance is the ideal (the compliance most worthy of aspiration) it is not settled by stipulation. It is interesting that Rawls, for no apparent reason, comes close to denying this when he stipulates that, “Principles are to be chosen in view of the consequences of everyone’s complying with them” (1999a: 114). 16  Reciprocity interpreted along Axelrod lines involves responding to cooperation with cooperation and to defection with defection in an iterated Prisoner’s Dilemma, thereby making it pay to be a cooperator.

Ideal Theory   337 One possible view here is that an ideal of unconditional giving “may play an inspirational role”17 but “it is not feasible. It is therefore naïve, and ineffective, to hold existing societies to account on the basis of such demanding moral standards.”18 Such a view treats unconditional giving as too much to ask of all-​too-​human beings. It reaches too high. My view, by contrast, is that treating unconditional giving as ideal asks too little, not too much. My claim is not that realistic idealism must ask less than utopian ideal theory asks: on the contrary, it must ask more. Realistic idealism is more demanding insofar as it demands that we take responsibility for motivating the people around us, that we do not take their motivation for granted, and that we do not call them defective just because we don’t want the responsibility of doing what it takes to motivate them. Consider the perspective of a chess player. I imagine how ideal it would be to move my pawn to K4, but if I fail to anticipate my partner’s response, then to any chess player my so-​called imagination is the paradigm of failed imagination. It takes imagination to be a realist. Players who look far enough ahead to anticipate what can go wrong are the ones whose imagination is admirable. Imagining what would be ideal in a solipsistic world is no substitute for imagining what is ideal in a world of strategically responsive agents. Solving real problems may require vastly more creativity and knowledge of (for example) social science. Still, that is the challenge to which a serious ideal theorist rises. To a realist, the challenge is to be responsive to the whole environment, not merely to a convenient fragment of it. David Estlund takes as given that idealistic theories posit standards that can be unrealistically high. Surely he is right about that, but he does not consider the corollary that they also can be unrealistically low. He says, A hopeless theory can be dangerous, of course. The soundness of the standards might lead some to take actions in their pursuit, and this might be bad. Actions in pursuit of what will never be achieved can be wasteful or even disastrous. A theory that counsels action in pursuit of high standards that are not sufficiently likely to be achieved, where the costs of failing are very high, often deserves to be chastised as utopian. On the other hand, some people might be led by unrealistically high standards to improve themselves or their institutions, even though not all the way (full achievement is hopeless after all, by hypothesis). (2014: 7–​8)

My claim: what makes utopian standards every bit as dangerous as Estlund says is that they tend to be so low, not that they tend to be so high. Revolutionaries treat utopian standards as licenses to ignore what humanity has learned from experience. Thus, to conceive of justice as demanding unconditional giving is to conceive of justice, not as asking too much, but as openly inviting the free-riding that unconditional giving openly 17  Stemplowska and Swift (2012: 387) say this, making a point about unrealizable ideals in general rather than unconditional giving in particular. 18  Valentini (2012: 659) says this, likewise making a point about unrealizable ideals in general.

338   David Schmidtz invites. To come to the table as a reciprocator is to demand that partners cooperate; to instead come as an unconditional cooperator, demanding nothing, invites free-riding.19 That an action guide would be ideal if competently implemented by all people together has nothing to do with whether it is ideal in a world where the antecedent does not hold.20 Idealization does not relieve us of responsibility for actually defending the substance of what we call justice. G. A. Cohen offers what seems to present itself as a heroic defense of the thesis that not all principles of justice are fact-​sensitive. But no one doubts that there are fact-​insensitive fragments of justice. Here is an obvious one: punishing the innocent per se is not required by justice. Nevertheless, the fact remains that overall conceptions of justice are not actually defended until they are defended with reference to facts. G. A. Cohen supposes that principles are more fundamental than rules of regulation, whereas in reality, the trail of justification leads both ways. We can defend a proposition by arguing that it rests on something deeper and foundational. But to defend a foundation qua foundation is to judge it by what people build on it. When we look at what gets built on a foundation, we enter a world of fact-​based empirical contingency. Looking at what gets built on a foundation does not prove anything, if by proof we mean being able to deduce that the principle necessarily leads to that result. Still, at some point, as a matter of science and common sense, we look at what keeps happening and we conclude that this is a mistake we should not keep repeating. We cannot make a particular conception just, merely by calling it just. What Cohen is calling justice is an intuition: ungrounded, untested, barely articulated, and massively fact-​influenced (by his personal history). Presumably Cohen’s conception of justice is internally consistent, as are infinitely many conceptions that experience counsels us to reject.21 What sorts out conceptions is information—​facts—​about whether they bear any relation to anything that any situated human being has reason to treat as worthy of aspiration. To demand an ideal response is, first of all, to demand a response. The problem with unconditional giving is not that it asks too much, but that it is unresponsive to the human condition. Humans are social. They are separate. And for better or worse, they respond. If an ideal response is cooperative, then an ideal way of conducting oneself among people who decide for themselves is a way most apt to induce that cooperative response. Unconditional cooperation is not an ideal but a strategy for fostering an actual ideal:  namely, cooperation. But whether unconditional cooperation truly fosters the ideal of cooperation under realistic conditions is a matter of observation. Theorizing can help us predict what we will observe, but theorizing does not settle the matter.

19 

As Rawls warned, “If a conception of justice is unlikely to generate its own support, or lacks stability, this fact must not be overlooked” (1999a: 125). Section 15.6 returns to this point. 20  I employ the phrase “all people together” with a nod to David Estlund’s view that a theory can be “hopeless” if too few will comply to make it work as a framework for cooperation, yet still “prescribe action in a certain way” insofar as it “counsels all people, together, to behave differently” (Estlund 2008: 266). 21  I thank Mario Juarez for the thought.

Ideal Theory   339 We might say reciprocity is realistic while unconditional cooperation is utopian as a strategy for realizing the ideal of cooperation. Rousseau opens Social Contract by announcing that, “we study men as they are and laws as they might be.”22 The part about laws signals that the theory will be idealistic; the part about men signals that the idealism is meant to be realistic, not utopian. Crucially, an idealism that aspires to identify institutions ideal for men as they are, taking into account how men predictably take shape within those institutions, remains for all its idealism a work of genuine political economy. It yields testable predictions about how institutions work, and thus a real basis for deciding whether an institution is worth wanting. A work of political apology might, by contrast, defend institutions that observably turn people as they are into monsters, where defense consists of imagining such institutions turning people into angels instead. A political apology imagines utopian conditions under which realistically indefensible institutions would be defensible. Those institutions are an ideal response to nothing in human experience, but utopians find them inspiring all the same. When Rawls says society is a cooperative venture for mutual advantage, this is not a description, but a vision of what a society of reciprocators would be like. A real society is more, and also sometimes less, than a mutually advantageous cooperative venture. So, Rawls is telling us about “laws as they might be,” not society as it is. Yet, intuitively, there is nothing silly about Rawls’s idea—​despite a certain lack of realism—​that society is a mutually advantageous cooperative venture. Mutually advantageous cooperation is not mere wishful thinking. It is a useful approximation, an accurate (even if simplified) description of society at its best. It is an ideal. But what makes the Rawlsian ideal worthy is not its portrayal of an ideal society, but its portrayal of a society’s ideal response to the actual human condition. He is describing what our society looks like when we respond to the actual human condition with the best strategy we have for responding to that condition. In our world, reciprocity is exactly right as a response to a defining problem of political theory, namely specifying terms of engagement that make people willing and able to trust each other enough to launch and sustain society as a cooperative venture. Among separate agents, reciprocity nurtures a solidarity worth wanting, a norm we want to see spreading through a population of players if we care about them.

15.6 Compliance The concept of compliance has a confusing place in the literature on ideal theory, as every consumer of the literature knows. Ideal theory is not to be simply equated with the assumption of full compliance, but the latter is the literature’s lightning rod, attracting

22 

I thank Robert van’t Hoff for the thought.

340   David Schmidtz much of the critical attention. We sense something wrong with assuming full compliance. We also intuitively see that if we were to achieve perfect justice, one feature of that achievement would be that everyone would be in compliance with its requirements. Between those intuitive understandings, there is room to observe that if compliance is ideal, then taking compliance for granted is not. Full compliance is a hard-​won achievement. It is achieved only in certain settings. That means that if there is a realistic ideal, it will be one of those settings. That is, a realistic ideal is a setting in which full compliance is realistically achievable. Against Rawls, Cohen objects: . . . if we assume, following Rawls, that individuals are motivated to comply with justice, then the need to trade off equality and well-​being disappears. It only arises in the first place because talented people demand incentive payments to become more productive. But people who are motivated to realize justice fully would not demand incentive payments but rather increase productivity without them. (as Cohen 2008 is paraphrased by Hamlin and Stemplowska 2012: 57)

Cohen has a point, but what he misses here is that Rawlsian contractors accept a tougher assignment: they are contracting for a social world populated by agents other than themselves. Bargainers cannot stipulate that human psychology is something other than what it is. For bargainers to know human psychology, as Rawls assumes, is for bargainers to know that psychologies of citizens at large are exactly what they are (see also James 2012: Ch. 4). Human psychology being what it is, whether people respond compliantly to what we ask depends on what we ask. Suppose we aim to sell coffee pots. We ask what a profit-​ maximizing price would be. Are customers disposed to pay what we ask? Obviously, that depends on what we ask. But suppose we imagine that setting aside this pivotal fact lets us focus on the “main” question of what we should charge for our product. Thus liberated from “distracting details,” we have an open road to the desired conclusion that customers ideally will give us everything they have. What can go wrong? Crucially, whether customers will give everything they have for a coffee pot determines, not whether asking for everything they have is feasible, but whether asking for everything they have is desirable. In other words, whether a bizarrely high price is the ideal price depends on whether customers are bizarrely compliant in the first place. It will not do to argue that in ideal theory agents are bizarrely compliant and therefore asking a bizarrely high price is ideal. That does not follow (and is not true). All that follows is that if the posited agents were sufficiently unlike real people in a particular way, then asking a bizarrely price would be ideal. Theorists who doubt that ideal prices (or any other institutional arrangements) are sensitive to motivational features of agents (e.g., Estlund 2011: 227) do not know what makes ideal prices ideal. An ideal price does not set aside the fact that the point of setting prices is to sell a product. To be an ideal price is to be an ideal response to customers as the separate agents they are.

Ideal Theory   341 If strict compliance is part of an ideally just society’s essence, that precludes rather than mandates taking strict compliance as given. If indeed “an important feature of a conception of justice is that it should generate its own support” (Rawls 1999a: 119)—​then a serious investigator does not stipulate that a conception has that pivotal feature (when in the real world it patently does not). A serious investigator checks.23 Every conception has its own characteristic compliance problem. We cannot stipulate that a conception has a compliance problem other than the one it actually will have in a particular setting. When we choose a package of principles to build into a basic structure of society, we choose whichever compliance problem goes with that package. We choose it! To set aside our chosen compliance problem as a detail best ignored is to set aside the nature of what we are choosing as a detail best ignored. Accordingly, it is a mistake to set aside compliance problems when theorizing about what would be an ideal response to the challenge of forming a community: the challenge of specifying terms of engagement that leave people knowing (a) what to expect from each other; (b) what to regard (and expect others to regard) as a person’s due; and (c) how to sustain a cooperative venture. To set aside that we live amongst agents—​beings who decide for themselves whether to comply—​is to set aside the defining problem of political theory.

15.7  Conflict and Conclusion The topic of ideal and nonideal theory is seductive. The urge to write on this topic is like the urge to gamble: we can’t win, so there is no natural satisfaction and no way to get it out of our system. Still, it is time to take stock. I accept that ideal theory can be done well. I defend ideal theory in two ways. First, ideal theory simplifies. All maps and all theories idealize in the sense of leaving out details. Simplifying is risky:  whether a detail is crucial or merely a distraction will depend on the purposes of particular users. So, are we illuminating real problems or making them invisible? Setting aside distracting details is not a mistake, but setting aside the problem is (O’Neill 1987). Moreover, knowing the difference is an achievement. There is no formula. It takes insight. Authors choose which problem to try to solve. Readers decide whether a chosen problem is important, and when authors try too hard to contrive something resembling a solution they end up changing the problem to something not worth solving. Relatedly, any decent theory simplifies in order to be useful, thereby leaving itself open to counterexamples. Theories are tools, and one thing we cannot build into useful 23 

Rawls’s views about “strains of commitment” likewise suggest that he would be comfortable with the idea that compliance is an achievement that cannot be taken for granted, and thus that his conception of the nature of defensible ideals and his concerns about stability were evolving in the direction of the view defended here.

342   David Schmidtz tools is a guarantee that they could never be misused by evil geniuses. Still, although much can go wrong when tools are useful enough that it becomes possible to misuse them, simplifying per se is not a mistake. Second, ideal theory articulates objects of aspiration. This too is legitimate in principle. Rawls was right that ideal and nonideal theories are complements rather than rivals. Articulating objects of aspiration is not a mistake. But there is a difference between aspiration and worthy aspiration. Calling something worthy does not make it so. We need theorizing about what makes one aspiration worthier than another. Calling worthiness fact-​insensitive does not help. If instead we start with a problem, then our starting point has the potential to discipline our reflection on what to count as a solution. From there, our ideals can emerge from our theorizing as warranted conclusions about the real world rather than as unwarranted premises. Taking compliance as given—​imagining that people do not decide for themselves whether to comply with what we call justice—​is a way of setting aside rather than working on the defining problem of politics: namely, that people choose for themselves, yet still manage under favorable conditions to launch cooperative ventures for mutual advantage. Crucially, favorable conditions are those that induce compliance with norms of mutual advantage. But seriously studying what induces compliance does not start by taking compliance as given. If anything needs to be set aside and treated as a mere distraction from work worth doing, it is visions of how well a system would work but for the recalcitrant reality of human beings. My further conclusions are more speculative. The question with which moral philosophy begins is a question of how to live. To recover a measure of relevance to the question of how to live here and now, theorizing about justice would need to consider which principles have a history of being demonstrably the organizing principles of actual thriving communities at their best. Ideals emerging from such research are not precipitates of armchair theory but are a distilling of social scientific insight from observable histories of successful community building (Pennington 2011). What emerges from ongoing testing in the crucible of life experience will be path dependent in detail if not in general outline, partly because the challenges to which a way of life is a response are themselves contingent and path dependent (Mason 2004). Finally, habitable ideals—​principles real people can live with—​are those by which courts and citizens avoid and resolve conflict. We are tempted to presume justice is more fundamental than conflict resolution, but philosophy done well is neither as autonomous as that nor as naïve. Hampshire goes so far as to say, “fairness in procedures for resolving conflicts is the fundamental kind of fairness” (2001: 4). Darwall says modern philosophy’s break with scholasticism began with a secular insight that a harmony of interests is not guaranteed (1993: 416). Moral theory done well tracks truth about the human condition: what it actually takes for people to live in peace. Thus, the relation between justice and conflict resolution is not a one-​way relation of more to less fundamental: it is a dance of mutual specification, anchored to evolving facts about how people manage traffic, settle disputes, feel at peace with settlements made, develop a mutually intelligible and mutually liberating

Ideal Theory   343 sense of what it means to mind their own business, and work toward making sure their neighbors are better off with them than without. Social structures that make it easier to resolve and avoid conflict go a long way toward fostering society as a cooperative venture for mutual advantage. To the extent that a society is such a venture, it is responding well to the human condition. People are learning to trust each other far enough, and adjust their expectations far enough, to constitute themselves as a kingdom of ends.

Acknowledgments Work on this chapter was supported by a grant from the John Templeton Foundation. The opinions expressed here are mine and do not necessarily reflect views of the Templeton Foundation. I’m also grateful to the Property and Environment Research Center in Bozeman for welcoming me as Julian Simon Fellow in the summers of 2012 and 2013, and to the Earhart Foundation for support in the fall of 2013. I learned much about this topic from Arià Paco Abenoza, Jonny Anomaly, Sameer Bajaj, Jacob Barrett, Matt Bedke, Michael Blake, Jason Brennan, Bruce Brower, Allen Buchanan, Ronna Burger, Tom Christiano, Ben Cilwick, Andrew J. Cohen, Alison Denham, Michael Duval, David Estlund, Jerry Gaus, Avital Hazony, Chris Howard, Mario Juarez, Robert Jubb, Brian Kogelmann, Christian Koons, Loren Lomasky, Eric Mack, Andrei Marmor, Michael McKenna, Elijah Millgram, Alex Motchoulski, Mike Munger, Carmen Pavel, Guido Pincione, Sarah Raskoff, Jeremy Reid, Jonathan Riley, Greg Robson, Dan Russell, Alex Schaefer, Lucy Schwarz, Simone Sepe, David Shoemaker, Matt Sleat, Stephen Stich, John Thrasher, Hannah Tierney, John Tomasi, Bas van der Vossen, Chad van Schoelandt, Steve Wall, and Fabian Wendt. Above all, I thank Serena Olsaretti, whose extensive and incisive commentary, anticipating reader responses, was repeatedly instrumental in helping me decide what to say and how to say it.

References and Further Reading Axelrod, R. (1984). The Evolution of Cooperation. New York: Basic Books. Brennan, J. (2014). Why Not Capitalism? New York: Routledge Press. Brighouse, H. and Swift, A. (2014). “Putting Educational Equality In Its Place,” in K. Meyer (ed.) Education, Justice, and the Human Good. London: Routledge, pp. 444–66. Carens, J. (1981). Equality, Moral Incentives, and the Market. Chicago, IL:  University of Chicago Press. Cohen, G. A. (2003). “Facts and Principles.” Philosophy and Public Affairs 31: 211–​45. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Darwall, S. (1993). “Motive and Obligation in Hume’s Ethics.” Noûs 27: 415–​48. Elkin, S. L. (2006). Reconstructing the Commercial Republic:  Constitutional Design after Madison. Chicago, IL: University of Chicago Press. Estlund, D. (2008). Democratic Authority. Princeton, NJ: Princeton University Press. Estlund, D. (2011). “Human Nature and the Limits (if any) of Political Philosophy.” Philosophy & Public Affairs 39: 207–​37. Estlund, D. (2014). “Utopophobia.” Philosophy & Public Affairs 42: 113–​34. Farrelly, C. (2007). “Justice in Ideal Theory: A Refutation.” Political Studies 55: 844–​64.

344   David Schmidtz Forcehimes, A. T. and Talisse, R. B. (2013). “Clarifying Cohen: A Response to Jubb and Hall.” Res Publica 19: 371–​9. Galston, W. (2010). “Realism in Political Theory.” European Journal of Political Theory 9: 385–​411. Gaus, G. (2011). The Order of Public Reason. New York: Cambridge University Press. Gaus, G. (forthcoming). The Tyranny of the Ideal. Princeton, NJ: Princeton University Press. Godfrey-​Smith, P. (2009). “Abstractions, Idealizations, and Evolutionary Biology,” in A. Barberousse, M. Morange, and T. Pradeu (eds) Mapping the Future of Biology. Dordrecht: Springer, pp. 47–​56. Hall, E. (2013). “Political Realism and Fact-​Sensitivity.” Res Publica 19: 173–​81. Hamlin, A. and Stemplowska, Z. (2012). “Theory, Ideal Theory, and the Theory of Ideals.” Political Studies 10: 48–​62. Hampshire, S. (2001). Justice Is Conflict. Princeton, NJ: Princeton University Press. Hart, H. L. A. (1961). The Concept of Law. Oxford: Oxford University Press. Hope, S. (2010). “The Circumstances of Justice.” Hume Studies 36: 125–​48. James, A. (2012). Fairness In Practice. Oxford: Oxford University Press. Jubb, R. (2012). “Tragedies of Non-​ Ideal Theory.” European Journal of Political Theory 11: 229–​46. Larmore, C. (2013). “What is Political Philosophy?” Journal of Moral Philosophy 10: 276–​306. Mason, A. (2004). “Just Constraints.” British Journal of Political Science 34: 251–​68. Mason, A. (2010). “Rawlsian Theory and the Circumstances of Politics.” Political Theory 38: 658–​83. Mason, A. (2012). “What is the Point of Justice?” Utilitas 24: 525–​47. McKerlie, D. (1989). “Equality and Time.” Ethics 99: 475–​91. Miller, D. (2013). Justice for Earthlings. Cambridge: Cambridge University Press. Mills, C. W. (2005). “Ideal Theory as Ideology.” Hypatia 20: 165–​84. Murphy, L. B. (2000). Moral Demands in Nonideal Theory. New York: Oxford University Press. O’Neill, O. (1987). “Abstraction, Idealization, and Ideology in Ethics,” in J. D. Evans (ed.) Moral Philosophy & Contemporary Problems. Cambridge: Cambridge University Press, pp. 55–​69. Pennington, M. (2011). Robust Political Economy. Cheltenham: Edward Elgar. Philips, M. (1985). “Reflections on the Transition from Ideal to Non-​Ideal Theory.” Noûs 19: 551–​70. Rawls, J. (1993). Political Liberalism. New York: Columbia University Press. Rawls, J. (1999a). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press. Rawls, J. (1999b). Law of Peoples. Cambridge: Harvard University Press. Rawls, J. (1999c). Collected Papers, ed. S. Freeman. Cambridge, MA: Harvard University Press. Robeyns, I. (2008). “Ideal Theory in Theory and Practice.” Social Theory & Practice 34: 341–​62. Ronzoni, M. (2012). “Life is not a Camping Trip—​On the Desirability of Cohenite Socialism.” Philosophy, Politics, and Economics 11: 171–​85. Rousseau, J. J. (1762/​1913). The Social Contract, G. Cole (trans.). London: J. M. Dent. Russell, D. C. (2013). Happiness for Humans. New York: Oxford University Press. Schmidtz, D. (2000). “Natural Enemies.” Environmental Ethics 22: 397–​406. Schmidtz, D. (2006). Elements of Justice. New York: Cambridge University Press. Schmidtz, D. (2011). “Ideal Theory: What it is and What it Needs to Be.” Ethics 121: 772–​96. Schmidtz, D. (2016). “After Solipsism.” Oxford Studies in Normative Ethics 6: 145–​65. Sen, A. (2009). The Idea of Justice. London: Allen Lane. Simmons, A. J. (2010). “Ideal and Nonideal Theory.” Philosophy & Public Affairs 38: 5–​36.

Ideal Theory   345 Sleat, M. (2014). “Realism, Liberalism, and Non-​Ideal Theory.” Political Studies. 64: 27–​41. Spencer, H. (1871). Social Statics. New York: Appleton. Stemplowska, Z. (2008). “What’s Ideal about Ideal Theory?” Social Theory & Practice 34: 319–​40. Stemplowska, Z. and Swift, A. (2012). “Ideal and Nonideal Theory,” in D. Estlund (ed.) Oxford Handbook of Political Philosophy. Oxford: Oxford University Press, pp. 373–​89. Swift, A. (2008). “The Value of Philosophy in Nonideal Circumstances.” Social Theory & Practice 34: 363–​87. Valentini, L. (2012). “Ideal vs. Nonideal Theory:  A Conceptual Map.” Philosophy Compass, 7: 654–​64. Vanderschraaf, P. (2006). “The Circumstances of Justice.” Philosophy, Politics, & Economics 5: 321–​51. Williams, B. (2005). In the Beginning was the Deed. Princeton, NJ: Princeton University Press.

Chapter 16

C onstruc t i v i sm, Intu itioni sm, a nd Ecum e ni sm Aaron James

In political philosophy, constructivism and intuitionism are often seen as opposed methods for justifying moral principles for political institutions. John Rawls, for example, sharply contrasts his broadly Kantian constructivism with rational intuitionism in the English tradition of Clarke, Price, Sidgwick, Moore, and Ross, suggesting that the two approaches are not only distinct but inconsistent (Rawls 1999b: 343).1 And on this point even Rawlsian constructivism’s most thorough critic, G.  A. Cohen, fully agrees:  constructivist and intuitionist methods can’t both account for distributive justice’s fundamental nature. For Cohen, only an intuitionistic style of justification does, in which case it, rather than constructivism, should be the fundamental method of political philosophy (Cohen 2008: 3, 300–​1). There is another possible position—​what may be called ecumenism—​according to which “intuitionism” and “constructivism” are different but unopposed. On this view, there are different kinds of question we might ask about “distributive justice,” and each style of reasoning can yield fundamental principles, as long as each is responding to a different kind of question, as appropriate (e.g., Williams 2008, esp. §V). While this may seem to simply push back the crucial dispute, to an argument about which moral question of distributive justice we should be asking in the first place, the ecumenist suggests, instead, that there is plenty of room in political philosophy’s garden: we can rightly take up different questions, with different equally valid theoretical objectives, and use either the intuitionistic or the constructivist forms of argument, as appropriate, in hopes of cultivating a thousand blooming flowers.

1 

Rawls also mentions the dominance of rational institutionalism since Plato and Aristotle, until it was questioned by Hobbes and Hume, and in Leibniz and Wolff.

Constructivism, Intuitionism, and Ecumenism    347 In this discussion, I elaborate this neglected position. I pay special interest to Rawls’s constructivism and his evolving dialogue with the classical intuitionists, especially in view of his methodological affinity and acknowledged debt to the intuitionist Henry Sidgwick, who used and articulated what Rawls came to call the method of “reflective equilibrium,” though with much greater emphasis on “self-​evidence.”2 (I will return to Cohen’s well-​known views, but they won’t be my focus.) As will become clear, Rawls’s main difference from Sidgwick lies in the way he frames the question of right or just­ ice in the first instance. This leads Rawls to a quite different conception of the task of justification, and to a more “constructive” form of argument. My general suggestion is that the nature of this difference—​in how the moral question of justice is understood—​ brings out both the possibility and the attractions of the ecumenist conception of political philosophy.

16.1 Preliminaries As I will understand it, ecumenism is a doctrine about the political philosophy of distributive justice, and therefore neutral about meta-​ethical disputes in which forms of “intuitionism” and “constructivism” may well be genuinely opposed. Is ethical truth “independent of us” and known by quasi-​perceptual rational insight (e.g., as according to intuitionists such as Plato or Clarke), or instead simply a function of idealized human practical reasoning (e.g., as according to constructivists such as Kant or Rawls)?3 Ecumenism, in the present sense, does not purport to say. It concerns only the methodology of substantive justification; that is to say, how we, as theorists, are to properly resolve moral questions of distributive justice, and what forms of reasoning are necessary or sufficient if we are to justify proposed distributional principles as against alternatives. Meta-​ethical questions about the status of such principles and such reasoning are viewed as a separate issue, at least in the first instance.4 Beyond meta-​ethical neutrality, I will assume that “intuitionism” and “constructivism” share several further assumptions. A principle of distributive justice is a general answer to some specified question of distributive justice. A principle is fundamental if it is derived from no other principle of the same normative kind. And in contrast with skepticism about what we can know or justifiably believe, both approaches assume that some such principles can indeed be justified: an appropriate form of reasoning

2 

Rawls (1971: 51 n26), says as much, with reference to Schneewind (1963). I argue that the meta-​ethical contrast is genuine in James (2007 and 2012). 4  On separating substantive/​methodological and meta-​ethical questions, see Rawls (1999a). T. M. Scanlon (2012), mixes what he regards as constructivism about right and wrong, in a sense related to substantive justification, with minimal realism about the meta-​ethics of practical reasons. My own meta-​ ethical constructivism about practical reasons in James (2007, 2012) is consistent with “intuitionistic” styles of practical reasoning. 3 

348   Aaron James and argumentation, when correctly carried out, is indeed sufficient to justify a principle as fundamental, at least with respect to the distributive justice question being pursued. Barring further elaboration, then, intuitionistic and constructivist approaches differ mainly in their methods; that is to say, in what forms of argument or reasoning they regard as necessary if sufficient grounds for accepting a principle are to have been provided. While forms of reasoning may vary on either approach, the two approaches might be contrasted in their bare essentials as follows. According to (minimal) intuitionism: a question of justice is to be answered in a general search for reflective equilibrium (i.e., a state in which our intuitively made judgments, of any level of generality, are adjusted so as to cohere, in a process of reflection). The intuitionist thus says little or nothing about how reflective equilibrium is to be attained, beyond any requirements that apply in the method of reflective equilibrium in general (e.g., a requirement to resolve incoherence, at least eventually). We can, for example, simply test general proposed principles of justice in light of examples and counterexamples, in whatever ways we find clarifying and convincing, with little or no self-​conscious attention to procedure. In contrast, according to (minimal) constructivism:  beyond the general demands of the method of reflective equilibrium, a question of justice is to be answered according to specified conditions of adequacy. While versions of constructivism vary according to the conditions of adequacy imposed, the general hope is to “construct” principles that conform to the stated design parameters—​to show that they should be accepted, by showing that they can be justified in the required way. As just contrasted, the two approaches are inconsistent when addresssed to the same question. For the constructivist, the question will have to be answered in a particular way. For the intuitionist, no particular form of reasoning is necessary, beyond what adds to our confidence in the search for coherence. (This is true even if the intuitionist also happens to use the constructivist’s favored reasoning, without assuming that it is necessary for an adequate justification.) And yet, for all we have said so far, the two methods are fully compatible when addressed to moral questions of different kinds. For this it need only be the case that some questions do, while others do not, generate special justificatory demands. For the ecumenist, that is the general situation: the flatly stated question, “What does distributive justice require?” is open to different interpretations, which in turn dictate different expectations for justification, different reasoning styles, and what may be different but equally fundamental principles.

Constructivism, Intuitionism, and Ecumenism    349

16.2  A Question of Fair Division In order to see the initial attraction of ecumenism, it will be helpful to compare two questions of fairness, which seem to call for different kinds of reasoning—​in one case intuitionistic, in the other constructivist. The first question is a simple question of fair division. Suppose there is some good to be distributed among certain people—​say, some tidy sum of money. We might then ask, what distribution of the funds would be fair if we know nothing else about the situation? Knowing nothing else, we immediately seem to have a natural and intuitively correct answer: each person should have an equal share. This answer may be made still plainer to intuition if we add certain stipulations. Suppose that all involved are moral equals; that all have the same interest in a greater rather than a lesser share; and that none have any special claim, such as a claim of need or special entitlement. If we then ask what division would be fair, it seems perfectly obvious, and indeed difficult to deny, that the answer is equality of shares. Our answer might well change if we relax the stipulated conditions. As long as they are assumed to hold, inequality of shares seems manifestly unfair. On at least this question of distributive justice, then, we might find ourselves in agreement with the dogmatic intuitionist Samuel Clarke, who tells us that reasonable and equitable expectations are perfectly “evident,” “plain,” and “manifest” (Raphael 1991: 202). By this he means not simply that he finds them evident, as a mere biographical report, but that they are undeniable by anyone who might reasonably consider the matter. As Clarke elaborates, “The unalterable rule of right and equity . . . necessarily and unavoidably determine the judgment . . . and force the assent of all men that use any consideration.” The situation, for Clarke, is precisely akin to our assent to an evident mathematical truth “to which the mind cannot but give its assent, necessarily and unavoidably, as soon as they are distinctly proposed to it” (Raphael 1991: 206). One must of course pay careful attention to the proposed moral proposition, and a “skillful eye” may be required to clarify where there is vagueness or confusion (Raphael 1991: 197–​8). But once the demands of right and equity are plainly articulated, Clarke tells us, only “the extremest stupidity of mind, corruption of manners, or perverseness of sprit, can possibly make any man entertain the least doubt concerning them” (Raphael 1991: 194). If Clarke is certainly not one for understatement,5 his main point does not depend on any very quick dismissal of those of the “extremest stupidity” who disagree. As the more temperate twentieth-​century intuitionist W. D. Ross would explain, even a “self-​evident” proposition needn’t be conclusively proven or shown, in the sense of finally settling an investigation (see also Stratton-​L ake 2002: 18–​23). The fairness 5  Though this was perhaps the spirit of the times; other intuitionists, such as Ralph Cudworth (Raphael 1991: 105–​19), have a similar penchant for bombast.

350   Aaron James of equality of distribution under the stipulated conditions can be merely what Ross calls a “prima facie duty,” which is nevertheless self-​evident, though: not in the sense that when we have reached sufficient mental maturity and have given sufficient attention to the proposition it is evident without any need of proof, or of evidence beyond itself. It is self-​evident just as a mathematical axiom, or the validity of a form of inference, is evident. . . . In both cases we are dealing with propositions that cannot be proved, but that just as certainly need no proof (Ross 1988: 29)

Ross’s remarks suggest that the self-​evidence of a particular judgment is consistent with “evidence beyond itself.” Justification can also (but perhaps needn’t always) proceed holistically, perhaps much as Sidgwick, (1981/​1907) had by then argued, in a way that set the stage for Rawls’s method of reflective equilibrium. In Rawls’s presentation, justification is to proceed by considering proposed principles (at any level of generality) in view of our considered judgments, where each is open to revision in light of the other (Rawls 1971: 46–​53). In Sidgwick’s accounting, among the main versions of the Intuitional Method, as he calls it, it is only an “ultra-​intuitional” version that “recognizes simple immediate intuitions alone and discards as superfluous all modes of reasoning to moral conclusions.” This is unsatisfactory, Sidgwick tells us, because any particular intuition may have an “element of error” (Sidgwick 1981/1907: 211). The general model of reasoning is therefore “inductive,” in the sense that Aristotle attributed to Socrates: the definition of Justice would be sought by comparing different actions commonly judged to be just, and framing a general proposition that would harmonize with all these particular judgments. (1981/1907: 99)

Sidgwick adds that the particular judgments are not simply our own, but those of “common consent,” due to disagreement. Not only may we ourselves “have inconsistent intuitions at different times,” we often “become aware that the moral perceptions of different minds, to all appearance equally competent to judge, frequently conflict: one condemns what the other approves.” In these ways, our “doubts are aroused,” and “we are led to endeavour to set these doubts at rest by appealing to general rules, more firmly established on a basis of common consent” (Sidgwick 1981/1907: 100). But we needn’t always consult the opinions of others; if we are inclined to answer our initial fair division question in the simple and direct way suggested, we might well find the equality solution convincing by a process of clarification and understanding alone. In this case, those who disagree may be rightly seen as confused (perhaps for having failed to distinguish between cases in which the stipulated circumstances do and do not hold). As Sidgwick suggests, it is only when our judgments face a considerable possibility of error, perhaps in view of real and sustained disagreements, that we have reason to at least consult “common consent.”

Constructivism, Intuitionism, and Ecumenism    351

16.3  Distributive Justice in Practice So much seems plausible as far as it goes. But now suppose we ask a very different question of distributional fairness in society, albeit in an idealized way. Suppose a certain group of well-​motivated people must decide together how valued goods or resources are to be shared among them. Their question, let us assume, is not a simple question of what distributive outcome is ideally preferable, even, perhaps, in a world more favorable to distribution than in the human condition. The question is rather how those involved themselves ought to adjust the expectations, rules, and institutions that decide who gets what and how much within their ongoing social practice, given the distributive arrangements available to them, under what may be less than ideal circumstances (e.g., the human condition). We may assume that each is resolute in his or her commitment to justice to fulfill whatever distributive justice should demand. Justice is assumed to decisively settle normative questions about their shared distributional practice, almost whatever the cost to other worthy values. And because all are well motivated, we can assume that none will leave the matter to past privilege or current bargaining advantage. All will look to reason, for possibilities of reasoned argument, in hopes of being guided together toward a principled resolution that is reasonably acceptable to everyone. Now notice that, when we pose a question of distributive justice in this way, equality of distribution may not seem self-​evidently required. Our earlier judgment in favor of equality of shares had stipulated conditions, which now have no obvious or immediate application. The value of equality, as justified under the previously stipulated conditions, may have no relevance at all under the present circumstances, unless we have further reason to think the two types of situation are relevantly similar. And while the relevance and value of equal distribution might now receive a fresh defense, the present question of justice does not seem resolvable in any comparably clear-​cut way. Equality of shares is one possible answer, but other reactions seem initially credible, depending on how the social context is specified. Indeed, initially, the most natural reaction might be a simple blank stare. We might further specify a type of social context (e.g., the major institutions of domestic society), but even then there may be no very simple or clear-​cut way to balance or prioritize all the relevant values. For Ross, even a mere prima facie duty isn’t a “proper duty” that arises from a particular situation’s “whole nature” (Ross 1988:  20). In the present case, if the goods or resources to be distributed are not manna from heaven but the fruit of social cooperation, then the “whole nature” of the situation may include a range of relevant conflicting claims, including claims of need, of special entitlement, and of the size of the distributable pie given the way production relations work over time. One could, of course, argue that certain normal situations and their various relevant values are nevertheless relevantly similar to our initial fair division situation, but the argument would indeed have to be made.6 Once the question is how reason might 6 

As, for example, in James (2012) for the case of the global economy.

352   Aaron James offer a principled resolution that all can agree upon in practice, the work of justification has yet to begin. Thus we might say that any answer to the essentially social question of distribution, at least once we appreciate its point, is subject to certain adequacy constraints. The task is still to discern what is true or correct under some relevant range of conditions, and here clarification may be helpful and indispensible. But now an answer requires reasoning of an appropriate sort, along with some conception of what kind of reasoning that would be, and an argument about where, specifically, that sort of reasoning leads. For any number of reasons, this may or may not permit any direct appeal to an otherwise intuitively attractive distributive ideal. For example, if indeed any correct principles must supply a presumptively conclusive reason for action, and if reasons for action are always reasons for some agents, then the relevant reasoning arguably must be fit to justify principles that are addressed to specified agents, in light of such things as the social options available to them and their potentially limited governance powers, given the circumstances within their (perhaps collective) control. In that case, certain attractive distributional states of affairs may well be “unavailable” to the group of agents being addressed, either because they are insufficiently feasible or sufficiently esoteric so as to be beyond people’s grasp. Alternatively, an otherwise attractive ideal might not be reasonably acceptable to all, or simply not close enough to general acceptability to be eligible for serious consideration. If any such reasoning must consider what distributional arrangements are reasonably acceptable from potentially divergent perspectives, then it must also take into account how different perspectives balance or prioritize values in different ways. Depending on what potential divergence must be accounted for, this may leave one without a straightforward argument for one’s own favored way of weighing up the values. To the extent that such constraints leave us without plain answers, we have seen Sidgwick’s advice: consult “common consent” (1981/1907: 100). Perhaps those involved have already found ways of getting on together, on the basis of certain shared views. In that case, certain assumptions could be presumed as both available to all and at least prima facie reasonable from the different perspectives of each. Even so, this may only narrow the field of credible arguments, and we may not be permitted to affirm our most favored propositions about just distribution among the field of contenders. Sidgwick himself was surely too optimistic about the commonsense resonance of his own utilitarianism, and not simply because of utilitarianism’s by-​now-​familiar problems. For any proposed propositions about complex matters, it will be a significant further question whether they really could or would command the assent of anyone who carefully considers them, on pain of being unreasonable (e.g., due to corrupting motivation, a failure of competence, or some non-​strictly-​question-​begging error of judgment, etc.) If one’s own most favored conclusion is to be finally justified in those terms, it can’t be presumed evident to all or most. At the very least, a substantial case for it must be presented. This, in essence, is Rawls’s “constructivist” conception of the problem of distributive justice. The question of principle is posed, in the first instance, as a question about what principles could be justified for service in a certain role in the regulation of public

Constructivism, Intuitionism, and Ecumenism    353 affairs. They are to regulate public affairs by and through the addressed parties’ own powers of reason, including their powers of “public reason,” but in a way that is subject to human reason’s “inevitable limitations” (Rawls 1999b: 347). When human reason readily supplies plain answers to questions about the permissibility of murder or deception in a state of nature, or about fair division under our initial stipulations, there may be no need for any argument, “constructivist” or otherwise, that displays possibilities of agreement: clarification and plausible elaboration may be sufficient.7 When we turn to the central problem of societal justice, however, the possibilities of convergence cannot be presumed ahead of time. Without further argument to the contrary, reason could well be fundamentally disunified or fragmentary in its recommendations; there may well be no single correct principles of societal distribution that would be seen as manifestly correct to every reasonable person who considers the matter, even, perhaps, in view of careful arguments. In this context, justification has the task of making the correctness of principles evident, or at least much more evident, to all involved. The task calls for a method that might show the possibility of a basis for agreement where it would not otherwise be plain to all. In that sense, an adequate answer to the essentially social question of just distribution must be justified where justification might fail, simply because human reason itself fails to provide the necessary materials. It is from this cautious starting point that Rawls undertakes the breathtakingly ambitious task of defending robustly egalitarian principles, along with “priority rules” that place them strictly before other values. Ideally, we would “construct” an answer in a way that is akin to a quasi-​mathematical demonstration, as by “moral geometry” (1971: 121). The social requirements of justice would be made manifest to all who competently consider the argument, in much the way that a theorem can be seen to follow from given axioms to everyone of sufficient mathematical competence who carefully considers the demonstration. These are Sidgwickian aspirations. And yet, as we will see, they address a challenge for reason deeper than any Sidgwick appreciated. Rawls poses the question of social justice as a question of mutual recognition in the first instance, and for that reason not even a formal proof will suffice for a justification: the argument must also be socially founded, in a sense I return to momentarily.

16.4  Deep Ecumenism All of this suggests that the intuitionist and constructivist styles of argument are not directly opposed. To the extent we think each style is natural and attractive in response to different fairness questions, we aren’t forced to choose between them: each can be appropriate depending on what question of fairness we are asking. It is a further question 7  Note that this still doesn’t imply meta-​ethical realism; one can still favor meta-​ethical constructivist interpretation, as Scanlon (1998) does for judgments of right and wrong, or as I do for practical reason judgments in James (2007).

354   Aaron James (which I cannot pursue here) what reasoning is in fact required, but we can at least see, in an intuitive way, why our two questions might be treated differently. In our initial question of fair division, the constructivist limits on argumentation are surely a gratuitous detour away from the simpler task of setting out a plainly correct answer. But when we pose a quite different question about bases of reasonable agreement for distribution in ongoing society, where reasonable people tend to disagree, the intuitive reasoning that settles a simple question of fair division has no similarly immediate or obvious application. And yet, for all that says, each of the different forms of argumentation might be perfectly appropriate for the moral question at hand. Let us call “shallow ecumenism” the view that different questions of distributive just­ ice can call for different styles of argument—​intuitionistic in some cases, constructivist in others—​where no further assumptions are made about the relative status of the different questions we might ask. No assumption is made, for instance, about which questions are relatively basic or otherwise the ones that political philosophy should address. Accordingly, shallow ecumenism is relatively uncontroversial. An intuitionist could happily grant that a constructivist style of reasoning is appropriate for certain questions, but reject or marginalize those questions themselves. On one version, they are relevant for questions about rules for social regulation, but not for questions of distributive justice per se.8 Likewise, a constructivist could admit an intuitive argument for our initial fair division question, but regard it as practically irrelevant. Perhaps its stipulations are so restrictive that the question is little more than an intellectual curiosity, and so of little or no relevance to political philosophy, which should only be “practical” in its concerns.9 What we might call “deep ecumenism” weighs into this further dispute about what questions are basic or worthy of political philosophy’s attention. On this view, we can rightly ask different questions of distributive justice, which can reflect fundamentally different ways of conceiving of the problem and task of moral justification; each (or at least many) such question(s) may be basic in its own way and fully worthy of study and argument. (A still more radical ecumenism would hold that all or many of such different conceptions are equally worthy. This is an ultra-​relaxed version of the “deep” ecumenism just stated.) Deep ecumenism stands in a particularly strong dialectical position, since it can be maintained by an easy strategy of partial concession. Suppose the intuitionist or the constructivist wishes to maintain that his or her favored question should be political philosophy’s exclusive or primary concern. It should then be said what crucial feature of the favored question makes this so. But whatever the answer, the deep ecumenist rejoinder can simply be, “Fine. Let’s say that certain questions do have the crucial feature, and must be addressed accordingly. But let’s also admit that other questions, which don’t have the crucial feature, are also basic and worthy of inquiry in their own way. Why not

8 

9 

Cohen (2008: 3). Joshua Cohen (2001) may be read this way.

Constructivism, Intuitionism, and Ecumenism    355 let a thousand flowers bloom?” The non-​ecumenist might seek to raise the stakes for possession of the crucial feature, perhaps by urging that the crucial feature is needed for us to properly speak of “distributive justice,” or “fundamental distributive justice,” or “political philosophy worthy of its name.” But the ecumenist’s rejoinder can again be concessive: “OK, sure. Then let’s allow that these terms have different senses, or a general sense that admits of different sub-​concepts, some of which assume the crucial feature, while others don’t. After all, what’s in a name? A rose by any other name smells just as sweet.”10 Though Rawls never officially endorses deep ecumenism, his own evolving engagement with intuitionism has something of this concessive character. He gradually shifted from what can appear as an outright rejection of rational intuitionism in “Kantian Constructivism in Moral Theory” (1999b) to the more modest position in Political Liberalism (1996: 95) that, as we might put it, it is merely inappropriate for the question of social justice as he understands it. That is, intuitionism is not necessarily wrong, but merely inappropriate for the social question of Rawls’s central concern, in which case intuitionistic argument, too, might be a fully appropriate way of resolving important but different questions of political philosophy that lay beyond Rawls’s scope of concern.11 Here much depends on the nature of Rawls’s “political turn” (see Weithman 2011). As Rawls (1996) came to explain, intuitionism is insufficiently “political,” and so inappropriate for reasoning about justice in the basic structure of society (at least as regards “constitutional essentials” and “matters of basic justice”), in just the way that any general philosophical position or “comprehensive doctrine” is unfit as a basis for general agreement where perspectives reasonably tend to diverge. If Rawls is taken to mean that political philosophy should only address “political” questions in his later, quite restrictive sense, then he is not a deep ecumenist. Yet the later Rawls is better seen as taking the special demands of the “political” to apply only to a proper subset of political philosophy’s rightful concerns. Otherwise, Rawls’s own “pre-​political” works, including A Theory of Justice, come out as a half-​ baked constructivist project, which might be better recast in intuitionist terms,12 instead of as mature and developed answers to the questions he in fact then posed. I submit, however, that Rawls’s “political turn” is merely a shift of emphasis within a largely continuous (and well-​baked) constructivist project. Rawls came to see A Theory of Justice as “unrealistic” about the extent to which reasonable people disagree. But even before this shift, his chief objection to intuitionism was always that it was inappropriate (rather 10  Compare Cohen (2008: 304), who also raises this question, though without endorsement of what I am calling “deep ecumenism.” The present suggestion is that this is where this line of thinking leads. 11  Rawls’s famous “intuitive” argument for the difference principle, as against the system of natural liberty (1971: 72), might be seen as an intuitionistic kind of argument, which Rawls used to supplement his argument from the original position. He might have done so on the assumption that it would not be sufficient to justify the difference principle on its own, and that it merely had a suggestive and corroborative role. 12  Cohen (2008) may be read as offering this suggestion about Rawls’s appeal to the difference principle, beyond his critiques of ideas of Rawls’s treatment of realism, publicity, and incentives.

356   Aaron James than simply false) for his essentially social question of just societal distribution. This was for a general “foundational” (rather than merely “political”) reason: it is inappropriate simply because it fails to appreciate the “inevitable limitations” of human reason. Rawls simply became clearer about this as he gradually came to see how different his constructivism was from Kant’s (of which more follows). If this is right, then Rawls’s position is at least consistent with deep ecumenism. When human reasoning does readily answer certain moral questions—​whether about the permissibly of murder or deception in a state of nature, or about fair division under our initial stipulations—​the essentials of intuitionism can be granted as fine and appropriate for those questions. But insofar as other equally or more important questions are not so readily answered, we require a method that self-​consciously displays the possibility of an otherwise non-​obvious basis of agreement. Rawls insists upon the centrality of such questions, at least as regards major institutions. But giving them pride of place does not require excluding questions more amenable to intuitionistic resolution. The questions are separate questions, but both can be seen as an important concern of political philosophy.13 In the remainder of my discussion, I further suggest that Rawls could at least have easily adopted the deep ecumenist position. I do this by more closely examining his evolving discussions of classical intuitionism, in view of how he revises his characterization of the role of self-​evidence and of autonomy. Both revisions bring out Rawls’s long-​standing view of the limits of reason and of what is needed to overcome them. The demands of the “political,” as the later Rawls came to see them, merely reflect a sharper version of this longer-​standing concern about what the limits of reason are.

16.5  Moral Geometry In “Kantian Constructivism in Moral Theory,” when Rawls first introduced the comparison to rational intuitionism, he emphasized that it regards irreducibly moral principles as self-​evident once they are correctly stated (1999b: 343). He took the point

13   Accordingly, Cohen (2008) needn’t be seen as substantively wrong about the issue of justice he’s after. His mistake can lie in seeing his position as an alternative answer to Rawls’s question, when it is in fact simply a different, equally fundamental answer to a different question of justice. Cohen does argue for his unaccommodating position. Principles justified under factual assumptions, including specified social conditions, he says, cannot be “fundamental” on the grounds that basic principles must be “fact-​ insensitive.” But this argument fails. Any principles justified for a stipulated type of social circumstances are “fact-​insensitive” in the following trivial sense, which Cohen never rules out: for any hypothesized social reality F, and any principles that apply P, it will be true that “If F, then P,” whatever the actual world facts. In that case, different social realities can yield different fact-​insensitive conditional propositions (of the form “If F, then P”), whether actual world facts or hypothesized facts are being considered. Such principles count as both “fact-​sensitive,” because sensitive to the social circumstances they are justified for, and “fact-​insensitive,” in the sense that such conditional propositions are also true.

Constructivism, Intuitionism, and Ecumenism    357 to apply in political philosophy. For intuitionism, it is the “recognition of self-​evident truths about good reasons” that founds “the agreement in judgment which is so essential for an effective public conception of justice” (1999b: 344). Later, in “Themes in Kant’s Moral Philosophy,” however, he would say that self-​evidence is “not essential.” The agreement needed for justice to effectively regulate public affairs could be founded merely upon a “shared recognition of truths” (1999a: 511). Rawls may have so de-​emphasized self-​evidence in order to better account for moderate intuitionists such as Sidgwick and Ross. Again, while Clarke emphasized and perhaps required self-​evidence on a proposition-​by-​proposition basis, Sidgwick rejects this “ultra-​intuitionist” version of the Intuitive Method, in a way that recognizes the possibility of error and advises finding a “basis in common consent.” In the passages indicated earlier, Sidgwick goes on to tell us that there are two very different ways to go about this; one “Dogmatic,” and one “Philosophical,” depending on how far we seek to rise above and offer a “rational synthesis” of common opinion. Neither approach will simply “follow rules generally received, not intuitively apprehended”; we will still “see for [our]selves the truth and bindingness of all or most of these current rules,” “only as supporting the individual’s intuition, not as a substitute for it or as superseding it.” Yet the Dogmatic version remains too confining; it assumes only “that we can discern certain general rules with really clear and finally valid intuition,” which are “implicit in the moral reasoning of ordinary men, who apprehend them adequately for most practical purposes.” The philosopher’s task is then merely “to state them with proper precision,” which “requires a special habit of contemplating clearly and steadily abstract moral notions” (1981/1907: 101). This is too confining, or too un-​philosophical, Sidgwick tells us, because even carefully crafted rules may then amount to “an accidental aggregate of precepts,” wherein “we may yet require some deeper explanation why it is so.” It is because we should seek a deeper explanation that we are led to seek a “philosophical basis which [even carefully crafted rules] does not itself offer,” a basis in which we “get one or more principles more absolutely and undeniably true and evident, from which the current rules might be deduced” (p. 102). Accordingly, Sidgwick regarded his own favored principle—​utilitarianism—​as self-​ evident, not simply because it is supposed to be evident taken by itself, but because it could be derived from further premises, seen as themselves still more clearly true. But, as suggested earlier, Sidgwick relied here on what Rawls would call the method of reflective equilibrium; he allowed that seemingly self-​evident judgments could be mistaken, and that we might resolve what to believe in light of systematic comparisons and our reflective engagement with common opinion. Rawls himself never rejects the possibility of self-​evidence in theory; he merely gives the idea little or no special role, emphasizing instead the possibilities of revising one’s judgments or one’s principles in light of the other. And yet he was nevertheless greatly influenced by Sidgwick’s ambition to consider the possibilities systematically, with aspirations of a kind of “moral geometry.” While Sidgwick thought he had found “principles more absolutely and undeniably true and evident,” from which common opinion might be deduced, Rawls at least hoped that his original-​position argument might illuminate the question of justice in a way

358   Aaron James that could demonstrate the correctness of his favored egalitarian principles, in something like the way a geometer provides a proof. The crucial difference between Sidgwick and Rawls lies elsewhere. For Rawls, Sidgwick nevertheless failed to give the right kind of argument: he sought mere proof, where something more than proof was needed. The task of moral geometry, as Rawls understood it, is not simply one of adducing logical relations between propositions, even from supposedly self-​evident axioms that happen to have moral content. For, Rawls explained, “we may accept the axioms as much because of the theorems they lead to as the other way around” (Rawls 1996: 242). What is needed, in addition, is a kind of a social acceptability, given the views basically reasonable people simply happen to have, that presents a more difficult justificatory challenge, which Sidgwick failed to appreciate, or at least rise to. That is, for Rawls, to successfully give a “justification” of principles in political philosophy is to offer a basis upon which shared arrangements can be justified by one person to another, on publically recognizable grounds (Rawls 1999b:  305;  see also Freeman 2007). Principles that provide the “basis” for justification cannot then simply be evidently true propositions, unless they are also appropriate for a social role as a public basis of mutual recognition. If a proposed proposition about distribution is unsuited for the role, it is simply not a principle of distributive justice, in the relevant sense, no matter how self-​evident it may appear considered independently. Further, being suitable for the social role in question is a demanding condition: it requires being grounded in common opinion, among the people who need the basis of mutual recognition in their ongoing shared relationship, given their limited and culturally shaped powers of reason. As Rawls explains: Since justification is addressed to others, it proceeds from what is, or can be, held in common; and so we begin from shared fundamental ideas implicit in the public political culture in the hope of developing from these a political conception that can gain free and reasoned agreement in judgment. (Rawls 1996: 100, emphases added)

Though this passage appears in Political Liberalism, it is not a simple function of Rawls’s political turn. Rawls put the same idea in A Theory of Justice this way: justification is “designed to reconcile by reason” and “proceeds from what all parties to the discussion hold in common” (Rawls 1971: 580). He elaborates in a way that disqualifies “mere proof ”: Ideally, to justify a conception of justice to someone is to give him a proof of its principles from premises that we both accept, these principles in turn having consequences that match our considered judgments. Thus mere proof is not justification. A  proof simply displays logical relations between propositions. But proofs become justification once the starting points are mutual recognized, or the conclusions so comprehensive and compelling as to persuade us of the soundness

Constructivism, Intuitionism, and Ecumenism    359 of the conception expressed by their principles. It is perfectly proper, then, that the argument for the principles should proceed from some consensus. This is the nature of justification. (Rawls 1971: 581)

We might say, then, that Sidgwick did not quite appreciate what “justification-​to” requires; he did not pose and address the question of right and justice as a question about mutual recognition among people with potentially divergent perspectives. But might Sidgwick simply reply that his own philosophical argument shows utilitarianism to be self-​evident, and thereby does offer a basis for mutual justification from things people hold in common, at least provided a bit of reflection? In theory, perhaps. Rawls tells us that Sidgwick indeed “takes for granted that [the different methods of ethics] aim at reaching true judgments that hold for all rational minds,” and that they are “found in the moral consciousness of humankind and worked into familiar historical systems” (Rawls 1999b: 341). But Sidgwick was surely too sanguine about the commonsense credentials of utilitarianism, if only because it is surely open to reasonable disagreement (especially given that it demands what many will regard as an extraordinary level of sacrifice). And, Rawls would add, Sidgwick failed to appreciate the depth of human reason’s “inevitable limitations” and the difficulty of overcoming them. The point is not that we should lower our expectations for the possibilities of justification; it is that the possibilities of justification are socially enabled in the first instance: “[t]‌he constructionist view accepts for the start that a moral conception can establish but a loose framework for deliberation” in which “our powers of reflection and judgment . . . are not fixed once and for all, but are developed by a shared public culture and hence shaped by that culture” (Rawls 1999b: 347). In theory, reason may be so fragmented that even those of good will are left empty-​handed, even upon extensive reflection and reasoned discussion. Our best hope for unification is therefore reason in practice, as shaped by a “shared public culture,” which partly consists in shared moral precepts, or at least shared ideas that might serve as a basis for fashioning points of deeper agreement, once people seriously consider the matter. But how are people’s assumptions within common opinion supposed to bear on a question of moral principle? Aren’t people’s moral beliefs and the correctness of certain moral principles quite different things? The answer, I take it, is that Rawls joins practice and principle though constructive interpretation, as explicated in the following three main argumentative stages (the order of which is irrelevant within a broader reflective equilibrium methodology) (James 2005: 18–​28). The first stage is individuation:  we single out an object of social interpretation and moral evaluation, at first in relatively uncontroversial terms, with reference to various interpretive “data points” or “source materials” that any further conception of the practice should take into account and explain (or “explain away”). Thus Rawls identifies “modern constitutional democracies” as his domestic target, and, in his late work, “international law and practice” on the global scene. The second stage is framing characterization: we work up a general characterization of the practice, in light of its

360   Aaron James distinctive structure and (presumed legitimate) purposes. It is here that Rawls invokes general conceptions of society and persons, offering a moralized interpretive characterization of what those self-​conceptions are (e.g., a “fair scheme of cooperation,” “equal citizens”). Moralized characterization is appropriate at the stage, so long as it comports with familiar interpretive constraints of consistency, coherence, explanatory power, simplicity, and so on. The third and final stage is substantive argument: we engage in substantive moral reasoning about what the various relevant reasons in play support, as framed and guided by the specified framing conception. It is at this stage that Rawls offers the original-​position argument for his principles of distribution. This involves substantive judgments about what is both morally relevant and morally irrelevant, for purposes of justification, but all ultimately in the service of elaborating what the more basic conceptions of person and society ultimately support. In “Kantian Constructivism in Moral Theory,” Rawls worked with highly abstract conceptions of person and society, following A Theory of Justice’s relative optimism about the extent of disagreement in society. As he later came to regard A Theory of Justice as “unrealistic” about how much agreement could be assumed—​especially in democratic societies marked by freedom and divergence of thought—​he became increasingly concerned to plausibly identify, rather than merely assume, a richer conception of democratic practice and its cultural basis, which fellow members of democratic society could be plausibly said to share (Rawls 1996: xvii). Accordingly, Rawls shifted from a “protestant” to a more “catholic” social interpretive methodology. A “catholic” interpreter assigns special authority to how participants in a social practice themselves interpret what they are doing together, and so will be especially sensitive to both shared understandings and disagreement in participant interpretations (see Postema:  1987). A  “protestant” interpreter gives no special authority to participant interpretation. Each interpreter is beholden only to his or her own interpretation of the practice; how participants interpret their practice, and whether they agree or disagree, represents only so much initial data or source material from which the interpreter must construct his or her own best characterization of the practice overall (Dworkin 1986). So while the early and middle “protestant” Rawls simply works out his own proposed best understanding of domestic societal practice, without great concern for people in the practice who might disagree, the later, more “catholic” Rawls accepts a special burden to take disagreement into account and addresses it on its own terms. In particular, he increasingly turns to a constructive interpretation of overall public practice as a source of argumentative pressure against potentially resistant elements of people’s moral or philosophical outlooks. To overcome or at least narrow their disagreements, the argument turns even more self-​consciously to the self-​conceptions that people can be said to share. If the limits of reason are inevitable, they may nevertheless be contingent. It could have been the case, in a more ideal world, that all or most principles are revealed to everyone through mere intellectual inquiry, as though by a discovery in mathematics. Provided sufficient good will, justice could then readily advance through cooperation that flows easily from each person’s ready grasp of the evidently reasonable and good course for collective action. But, alas, tragic or unfortunate as it may be, in the human

Constructivism, Intuitionism, and Ecumenism    361 condition as it is, rather than as it could have been, we face a sharper predicament. The possibilities of reasoned demonstration require the unifying effect of social practice. And so the task of justification is in part one of reasoning from practice, that is, a task of characterizing a basis for agreement between reasoning people who are prone to disagreement, where this is done by “constructing” principles that aren’t yet obvious to them, from the “materials” of agreement already on hand, along with further moral judgments that, upon further reflection, all might find acceptable, partly in view of what they already accept. So it would not be enough to show that certain moral “theorems” follow from certain moral “axioms,” simply because one could readily reject the conclusion by giving up one or more premises in which one has no particular stake. Rawls seeks to foreclose this possibility, or at least dramatically raise the stakes: he hopes to conclusively settle the standard of public mutual justification, in favor of certain principles, and to do this by resting his demonstration upon weak and widely shared premises, which no one could reasonably reject—​where no one could reasonably reject them in part because they are firmly rooted in commonsense.

16.6  Must Sidgwick and Rawls Disagree? Here Sidgwick could reply that he is simply not attending to the special demands of ‘justification-​to’, and that such “social correctness” is unnecessary or a separate issue. It is quite enough, he might claim, that a justification displays the evident correctness of an ethical proposition, much as in mathematics. In that case, however, Rawls can offer the following deep ecumenist concession: “Fine by me. As long as you agree that ‘justification-​to’ is of central importance for public affairs, there’s nothing necessarily wrong with asking other sorts of ethical question, which admit of ‘justification’ in a different sense and by a different method of argument. The principle of utility might even be said to apply in certain cases, as a conclusion about what people ought to do, as long as it only applies after the principles of social justice have been satisfied.” Though concessive, this is a forceful reply. It cannot be said that Rawls’s special idea of “justification-​to” is of no special importance for society, in which case it rarely or never displaces utilitarian demands. Perhaps the intuitionist might agree. In any case concerns of justifiability to others can indeed be developed as a distinctive moral domain in intuitively resonant terms—​what T. M. Scanlon calls “what we owe to each other”—​which apply both within society and beyond, even in a “state of nature” (Scanlon 1998). The Sidgwickian might therefore offer a sharper rejoinder. He or she might admit that moral justification in society is a matter of justification to all involved, and perhaps self-​evidently so. Yet, it may be said, Rawls is mistaken to suppose that this implies accountability to political culture. Instead, principles count as fully justified, from within a theorist’s own reflective equilibrium, without any special accountability to what

362   Aaron James may be the divergent opinions of others. What people could reasonably accept or reject would depend only on the evident or plausible nature of the proposed principles and the various reasons offered in support of them, without essential reference to an actual endorsement or endorsability in the society to which the principles in question are addressed.14 Here Rawls would say that, while this may be fine for certain questions of justifiability to others, the argument is no longer engaged with the societal question of distribution as he understands it. As I have suggested, that question is essentially about what bases for “agreement in judgment” could effectively govern societal affairs, in a way that allows people to be “reconcile[d]‌by reason” through what parties to a discussion “hold in common,” even as they often sharply disagree. Or as we might put the idea, what is needed is a basis for people to hold each other accountable in practice, through reasoned discussion, where reasoned discussion is founded upon certain sorts of agreements, for instance about the kind of relationship the discussants take themselves to be maintaining and what it roughly requires, in practice. In this, we might say, civic friendship is much like friendship. Friends, as well as lovers and life-​partners, work out their problems through reasoned discussion about who is accountable and for what, according to their shared understandings about the nature of their relationship, subject to ongoing reasoned re-​negotiation. A  theorist’s own substantive reflective equilibrium may or may not be fit for these particular social roles, in friendship, intimate partnership, or larger society. If it is appropriate, this must be made evident to the others involved if it is to provide a basis for concord and reconciliation in the relationship in actual practice. While this abstract point might be developed in various ways, the key point for present purposes is that Rawls’s requirements of social engagement needn’t disqualify or minimize (perhaps Scanlon-​style) questions of “justification-​to” that aren’t socially engaged. Rawls can still be a deep ecumenist: he can grant that many other such questions are worthy of inquiry, provided only that they are seen as a different kind of moral or justice question. The theorist could happily work out his or her own favored view of what we owe to each other in distributive affairs, inviting others to consider whether they might agree. While the theorist should perhaps explain why his or her question is of interest, the success of such an explanation might be assessed for each question taken on its own terms. There would seem to be no reason to assume, ahead of time, that only the class of socially engaged questions will turn out to seem worthwhile. On the other hand, Rawls does suggest that he would not be so accommodating. In “Kantian Constructivism in Moral Theory,” he calls the task of answering his socially engaged question of mutual recognition “the real task” of political philosophy, which might seem tantamount to insisting that it is the only task of political philosophy (Rawls 1999b: 306). Yet this remark is easily regarded as passing rhetoric. Certainly little

14 

Scanlon (1998) may be read this way.

Constructivism, Intuitionism, and Ecumenism    363 depends on it. Rawls surely means to urge the central importance of his question. He’d object if his task were neglected. But otherwise it can be seen as the really hard task, rather than the only or the only worthy task of political philosophy.

16.7 Autonomy I believe this is true despite the fact that, in “Kantian Constructivism in Moral Theory,” Rawls did flirt with Kant’s argument that intuitionism is not simply inappropriate but flatly mistaken. For Kant, we have principles for “autonomous” governance only when the argument that founds them rests upon self-​conceptions. People must be able to see themselves, on the basis of the argument, as bound to action through their own awareness of their freedom as persons. Similarly, Rawls says that principles of distributive justice must provide a basis for “autonomous” social self-​governance, in a way that itself precludes governance by independent moral truth, known through intuition, which makes no essential reference to self-​conceptions (Rawls 1999b: 345). Although Rawls did not explain why intuitionists could not also appeal to self-​conceptions (though he noted that they traditionally don’t), the idea, it seems, was that a basis for agreement must be of a certain sort. In theory, people could fully agree upon the moral facts simply because reason makes them blindingly obvious; few with the most basic powers of reflection could deny them, even among those who try in vain to look the other way. And yet, the idea goes, even such robust agreement would not provide a basis for an “autonomously” rather than “heteronomously” governed society, unless some self-​conceptions are also introduced. Here it is important that Rawls does not entirely follow Kant in “Kantian Constructivism in Moral Theory.” Rawls presumably found Kant’s own basis for construction, a conception of the person, too sparse. He therefore also includes a society’s public self-​conception, as found within its political culture, among the possible grounds for principles for autonomous public regulation. As Rawls puts it, “On the Kantian view that I shall present, conditions for justifying a conception of justice hold only when a basis is established for political reasoning and understanding within a public culture” (Rawls 1999b: 305). But despite this quite un-​Kantian grounding, Rawls did seem to hold, in Kantian style, that intuitive agreement, based merely on the shared recognition of moral truths, is insufficient to establish principles in their most important social role in autonomous public regulation. Principles must be also justified for, or in the light of, “a basis . . . for political reasoning and understanding within a public culture” (by constructive interpretation, as elaborated earlier). On the other hand, while Kant certainly rejected intuitionism wholesale, Rawls’s own argument from autonomy does not necessarily preclude the deep ecumenist view that it is merely inappropriate for the question of distributive justice as he understands it. Although Rawls didn’t address the matter, he could still admit other worthy questions

364   Aaron James of distributive justice that simply aren’t about the possibility of autonomous governance, and so might admit of intuitive justifications, without appeal to self-​conceptions. And even if Rawls did follow Kant’s wholesale rejection of intuitionism for a time, this would not become his settled position. Just as he later de-​emphasized the contrast with intuitionism over self-​evidence, he de-​emphasized the Kantian argument from autonomy, recasting it in a “doctrinal” rather than “constitutive” form (Rawls 1996: 99). Intuitive style justifications are not then precluded when and simply because they aren’t grounded in self-​conceptions; the problem with intuitionism, Rawls would explain, lies elsewhere. It is worth noting here that the Kantian version of the point was arguably always something of a stretch. “Autonomy,” in the central and most important sense, as Rawls came to explain, was supposed to mean that principles are not “externally imposed,” as they would be, for instance, by a simple morality of divine commands (Rawls 1996: 98). But in that case, why would each person’s mere intuitive appreciation of certain moral truths be insufficient as a basis for autonomous governance? If people conformed to certain principles of distribution, sensing that this was right and required of them, exercising their powers of reason, why wouldn’t this be good enough? It would surely be wrong to say they are simply acting upon authority, whether “externally imposed” or otherwise, especially if the people in question could adduce various reasons why they took the distribution in question to be just. Why would treating the exercise of reason as an act of “rational intuition” undercut such good motivation, if that amounts mainly to a way of appreciating the good reasons for action there are? Even if Kant would staunchly disagree, one might even regard this as “autonomy” par excellence. There is, moreover, a better explanation of Rawls’s original resistance to intuitionism in matters of public governance: it simply reflects his long-​standing skepticism about our powers of reason, what he calls their “inevitable limitations” (Rawls 1999b: 347), which his political turn merely depends on and elaborates. This at least means that Rawls does not presume that practical reason is sufficiently unified to command the required shared appreciation of truths; if certain conclusions are to be commanded by reason, this must be shown, and they cannot be shown in any straightforward way that would readily convince any good-​willed person who considered the matter carefully. And while that could lead to skepticism, Rawls is finally hopeful about giving a philosophical argument of the right sort. If the propositional representation that Sidgwick aspired to won’t likely command the assent of all who put reason to use, Rawls sees better chances in the argument from autonomy in its socially constructive form, precisely because it invokes and interprets the way those addressed conceive of themselves and their society. When principles are grounded in the way people conceive of themselves, rather than merely articulating what may be abstract propositional truths, people can see argument for themselves in a particularly accessible and convincing way: they see the argument for themselves by seeing themselves in the argument. For that reason, Rawls assumes, the socially rooted argument from autonomy offers the best hope of justifying principles in a way that might not only command their assent but also sustain their allegiance over time.

Constructivism, Intuitionism, and Ecumenism    365 The argument from autonomy thus turns out to be necessary, but only contingently. In theory, reason could well have been more unified and efficacious than it is, issuing unavoidable commands rather than whispering to those already listening. In that case, pace Kant, a shared recognition of reasons would suffice for autonomous governance; people could uphold them because they would readily see for themselves that the principles apply. And yet, Rawls assumes, the needed shared appreciation happens not to be in the cards—​at least not in the conditions of modern social life, and especially not under free institutions that encourage diversity of opinion and divergence of thought. The source of the trouble ultimately lies in the human condition itself: in the limited nature of our reasoning powers.

References Cohen, J. (2001). “Taking People as they Are.” Philosophy and Public Affairs 30: 363–​86. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Dworkin, R. (1986). Law’s Empire. Cambridge, MA: Harvard University Press. Freeman, S. (2007). “The Burdens of Public Justification: Constructivism, Contract­ualism, and Publicity.” Politics, Philosophy and Economics 6: 5–​43. James, A. (2005). “Constructing Justice for Existing Practice:  Rawls and the Status Quo.” Philosophy and Public Affairs 33 (3): 1–​36. James, A. (2007). “Constructivism about Practical Reasons.” Philosophy and Pheno­menological Research 74(2): 302–​25. James, A. (2012). “Constructing Protagorean Objectivity,” in J. Lenman and Y. Shemmer (eds) Constructivism in Practical Philosophy. Oxford: Oxford University Press, pp. 60–​80. Postema, G. (1987). “‘Protestant’ Interpretation and Social Practices.” Law and Philosophy 6: 283–​319. Raphael, D. D. (1991). British Moralists, 1650-​1800. Volume 1: Hobbes–​Gay Indianapolis, IN/​ Cambridge, MA: Hackett Publishing Co. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (1996). Political Liberalism. New York: Columbia University Press. Rawls, J. (1999a). “The Independence of Moral Theory,” in S. Freeman (ed.) Collected Papers. Cambridge, MA: Harvard University Press: pp. 286–​302. Rawls, J. (1999b). “Kantian Constructivism in Moral Theory,” in S. Freeman (ed.) Collected Papers. Cambridge, MA: Harvard University Press, pp. 303–​58. Ross, W. D. (1988). The Right and the Good. Indianapolis, IN/​Cambridge, MA:  Hackett Publishing Co. Sidgwick, H. (1981/​1907). The Methods of Ethics, 7th edn. Foreword by J. Rawls. Indianapolis, IN/​Cambridge, MA: Hackett Publishing Co. Scanlon, T. M. (1998). What We Owe to Each Other. Cambridge, MA: Harvard University Press. Scanlon, T. M. (2012). “The Appeal and Limits of Constructivism,” in J. Lenman and Y. Shemmer (eds) Constructivism in Practical Philosophy. Oxford: Oxford University Press, pp. 226–​42. Schneewind, J. B. (1963). “First Principles and Common Sense Morality in Sidgwick’s Ethics.” Archiv fur Geschichte der Philosophie, Bd. 45: 137–​156.

366   Aaron James Stratton-​Lake, P. (ed.) (2002). Ethical Intuitionism: Re-​Evaluations. Oxford: Oxford University Press. Weithman, P. (2011). Why Political Liberalism? On John Rawls’s Political Turn. New  York: Oxford University Press. Williams, A. (2008). “Justice, Incentives, Constructivism.” Ratio 21(4): 476–​93.

Chapter 17

C on cep tual Ana lysi s a nd Distribu tive J u st i c e Matthew H. Kramer

The enterprise of conceptual analysis is often taken by philosophers to consist in the specification of necessary and sufficient conditions for the applicability of a given concept to anything. However, because most concepts cover marginal instances as well as straightforward instances of the phenomena which they comprehend, the enterprise of analysis just described can sometimes lead to unhelpfully thin results. Accordingly, it is often supplemented or even supplanted by approaches that sacrifice some breadth for the sake of depth. For example, instead of seeking to distill necessary and sufficient conditions for the applicability of a concept C to absolutely any instance of the phenomenon which C denotes, many philosophers seek to distill necessary and sufficient conditions for the applicability of C to any paradigmatic or standard instance of that phenomenon. Still, in discussions of distributive justice during the past several decades, a more common approach resides in distinguishing between concepts and conceptions. Whereas a concept is thinly abstract in the way just suggested, a conception is an elaboration of the concept that fleshes out its elements by developing them in specific directions. Associated with the concept of distributive justice, then, are multiple conceptions advanced from diverse perspectives. Wariness is advisable, however, for the concept/​conception dichotomy is misleadingly tidy. An invocation of that dichotomy conveys the impression that only two levels of theoretical elaboration are involved. In fact, we can best understand and locate the principal contemporary accounts of justice among several different layers of theoretical development. This chapter’s first three main sections, then, will explore the concept/​conception distinction in its multilayered complexity in the context of present-​day debates over distributive justice. Of course, my discussion of that distinction will not aim to chart the sundry positions that have been taken in those debates by a host of philosophers. Instead, we shall look at a couple of prominent recent accounts of justice in order to gain

368   Matthew H. Kramer some sense of how they unfold over several levels of increasing specificity. Thereafter, the chapter will take up the vexed question whether expositions of the concept of justice can ever be austerely analytical or formal rather than morally value-​laden. To address that question satisfactorily, the chapter will distinguish between value-​independence and value-​neutrality. It will maintain that, although some possible accounts of justice are at least partly value-​neutral, no accounts are ever value-​independent.

17.1  Concepts versus Conceptions: Rawls Fittingly enough, the book that has been the fountainhead of disputation about just­ ice among contemporary philosophers—​John Rawls’s A Theory of Justice—​is also, within political philosophy, the locus classicus for the distinction between concepts and conceptions. To be sure, that distinction did not originate with Rawls. He himself attributed it to H. L. A. Hart (Rawls 1971: 5n1), and it existed in various guises long before Hart. Still, the following remarks by Rawls have been particularly influential in shaping the methodological orientation of philosophers who write about distributive justice: Men [and women] disagree about which principles should define the basic terms of their association. Yet we may still say, despite this disagreement, that they each have a conception of justice. That is, they understand the need for, and they are prepared to affirm, a characteristic set of principles for assigning basic rights and duties and for determining what they take to be the proper distribution of the benefits and burdens of social cooperation. Thus it seems natural to think of the concept of justice as distinct from the various conceptions of justice and as being specified by the role which these different sets of principles, these different conceptions have in common. Those who hold different conceptions of justice can, then, still agree that institutions are just when no arbitrary distinctions are made between persons in the assigning of basic rights and duties and when the rules determine a proper balance between competing claims to the advantages of social life. Men [and women] can agree to this description of just institutions since the notions of an arbitrary distinction and of a proper balance, which are included in the concept of justice, are left open for each to interpret according to the principles of justice that he accepts. These principles single out which similarities and differences among persons are relevant in determining rights and duties and they specify which division of advantages is appropriate. Clearly this distinction between the concept and the various conceptions of justice settles no important questions. It simply helps us to identify the role of the principles of social justice. (1971: 5–​6, footnote omitted)

As this passage suggests, the concept of distributive justice is thin by Rawls’s reckoning. It can best be understood as denoting the appropriate apportionment of the benefits and

Conceptual Analysis and Distributive Justice    369 burdens of social interaction among the members of any community. Such a formulation leaves open what counts as “appropriate,” what count as “benefits” and “burdens” and “social interaction,” and what counts as a “community.” Hence, it covers illiberal conceptions of justice as well as all liberal conceptions. However, more than one level intervenes between that thin concept and Rawls’s conception of justice as his two lexically ordered principles that are applicable to the major legal–​governmental institutions of any society.1 Rawls’s conception is one of many liberal conceptions, each of which differs from any illiberal conceptions. For Rawls, the hallmark of the heterogeneous liberal conceptions of justice is convergence on certain tenets about the nature of persons and of the societies in which they live. (Because this chapter is exploring methodological matters rather than any more concretely substantive matters, I will not consider here the extent to which the focus by Rawls on persons—​with a consequent neglect of animals and mentally infirm human beings—​is a shortcoming in his theorizing.) More specifically, all the liberal conceptions of justice concur in deeming persons to be free and equal and in affirming that any morally legitimate society is a fair arrangement for social cooperation. They further share the premise that individuals who are not subject to coercive or manipulative thought-​control will reasonably diverge from one another in the course of arriving at a multiplicity of views about what is valuable in life. The notion of fairness invoked at this point is not yet a distinctively Rawlsian notion; it is a much more capacious idea that can accommodate many theories of justice that are strongly opposed to Rawls’s theory. For example, many varieties of utilitarianism will be classifiable as liberal conceptions in the broad sense that is captured by the tenets just broached. Similarly, Robert Nozick’s libertarianism and quite a few other varieties of libertarianism will qualify as liberal theories. By including in his theory a Lockean proviso as a constraint on the legitimacy of private appropriation, Nozick aligned himself—​ albeit somewhat tenuously—​with those who take legitimate societies to consist in fair arrangements for social cooperation. A  host of other free-​market thinkers, such as Friedrich Hayek and Milton Friedman, are likewise classifiable as liberals by Rawls’s reckoning. Notwithstanding that many libertarian and utilitarian accounts of justice could be recognized by Rawls as broadly liberal, he did not regard all such doctrines as fully reasonable. He of course did not believe that his own conception of justice is the lone fully reasonable conception, but he did maintain that liberal theories are not fully reasonable unless they partake of certain features of his own theory.2 Specifically, they have to 1  My thinking about matters of conceptual analysis in relation to Rawls’s theorizing, especially in relation to the theorizing in his later work, has been helped by my perusal of the sixth chapter of Quong 2011. However, Quong heavily emphasizes the constructivist aspect of Rawls’s methodology, whereas I play down that aspect. 2  On Rawls’s account, two or more theories can be fully reasonable—​in that they each accept all the tenets recounted here—​even though one of the theories is more reasonable than the other(s). Rawls explicitly took reasonableness to be a scalar property: “[A]‌conception of justice is reasonable in proportion to the strength of the arguments that can be given for adopting it in the original position” (1971: 352).

370   Matthew H. Kramer accept that every citizen is endowed with basic rights and liberties, and that those basic rights and liberties enjoy special priority over other considerations of political morality. They must likewise accept that every citizen should possess adequate general means for the exercise of the basic rights and liberties. At this level of theoretical elaboration, then, nearly all varieties of utilitarianism and some further varieties of libertarianism are excluded as not fully reasonable. In reaching the level of theoretical elaboration just mentioned, Rawls proceeded with a constructivist methodology. That is, he posited a situation of rational agents who all endorse the fundamental tenets of liberalism recounted above, and he pondered the conclusions at which they would arrive under ideal conditions of deliberation—​the “Original Position”—​from which all morally arbitrary or improper factors have been excluded by a Veil of Ignorance. However, the constructivist methodology is just one heuristically valuable way of supplying a framework for processes of moral reasoning that are subject to certain normative constraints (the constraints that produce the idealization of the agents’ deliberations in a constructivist approach). What is crucial for Rawls’s project is the satisfaction of those constraints, rather than the specific method of satisfying them. Having moved from the level of liberal theories of justice to the level of fully reasonable liberal theories of justice, Rawls then expounded his own conception of just­ ice as the two complex principles that have been so extensively discussed ever since the publication of A Theory of Justice. (Of course, the levels or stages in the elaboration of Rawls’s conception of justice are not as tidily demarcated in his writings as this chapter might seem to be suggesting. His unpacking of the concept of justice proceeded through some of those levels or stages simultaneously; nonetheless, they can and should be distinguished here analytically.) Rawls’s conception of justice is in rivalry with other contemporary conceptions that also qualify as fully reasonable liberal accounts of justice: prioritarianism, sufficientarianism, competing versions of egalitarianism, some varieties of libertarianism, and so forth. Unfolding through the different levels that have been recounted here, his elucidation of the concept of justice acquired enough definiteness to place it in competition with these other theories. This chapter obviously does not aim to appraise the success of Rawls’s conception as a substantive theory of justice. Rather, the point here has been to consider how his theorizing about justice can be fruitfully understood as a mode of conceptual analysis. At each level of his ruminations, we find necessary and sufficient conditions for the inclusion of various theories at that very level. Those conditions are of course themselves matters of political morality, rather than austerely analytical matters. Nonetheless, to some degree, they methodologically assimilate Rawls’s endeavors to those of other philosophers who work in quite different areas of the discipline.3 3  In the philosophy of law, for example, legal positivists in the aftermath of Hart have often sought to specify necessary and sufficient conditions for the status of any theory of law as an instance of jurisprudential positivism in opposition to natural-​law doctrines. Some of those philosophers have

Conceptual Analysis and Distributive Justice    371

17.2  Rawls on Reflective Equilibrium I have not yet mentioned Rawls’s most famous contribution to the methodology of philosophy: his notion of reflective equilibrium. Within the structure of conceptual analysis expounded above, reflective equilibrium is most notably operative in the move from liberal theories of justice to fully reasonable liberal theories of justice. Rawls prodded his readers to contemplate their own firmest convictions about justice. Although those convictions can pertain to highly abstract matters as well as to more concrete matters, Rawls especially had in mind our convictions about specific policies and particular cases. His examples in his opening discussion of reflective equilibrium indicate as much: There are questions which we feel sure must be answered in a certain way. For example, we are confident that religious intolerance and racial discrimination are unjust. We think that we have examined these things with care and have reached what we believe is an impartial judgment not likely to be distorted by an excessive attention to our own interests. These convictions are provisional fixed points which we presume any conception of justice must fit. (1971: 19–​20)

With those provisional fixed points in view, Rawls’s readers are then to ponder his specification of the conditions that constitute the Original Position. Those conditions generate certain principles of justice, of course. What Rawls’s readers need to ask themselves, on the way to attaining reflective equilibrium, is whether the principles of justice derivable from the Original Position are congruent with their considered convictions about justice. If a reader notes any discrepancies between those principles and her convictions, she then needs to determine how harmony between them can best be effected. She might decide that she needs to modify some of her considered convictions in order to bring them into line with principles of justice which she finds compelling, or she might instead decide that Rawls’s principles are themselves in need of modification. If she does reach the latter conclusion, she will also have to specify the appropriate alterations in the conditions that make up the Original Position. As a reader seeks to reconcile the principles of justice with her own considered

further sought to specify conditions necessary and sufficient for the status of any theory of law as an instance of Inclusive Legal Positivism in opposition to Exclusive Legal Positivism, or for the status of any such theory as an instance of descriptive legal positivism in opposition to prescriptive legal positivism. These sundry specifications of the tenets of theoretical positions have been in the service of analyses of the concept of law.

372   Matthew H. Kramer convictions, she might well undertake adjustments in both directions through several iterations of the process. As Rawls declared: We can either modify the account of the [Original Position] or we can revise our existing judgments, for even the judgments we take provisionally as fixed points are liable to revision. By going back and forth, sometimes altering the conditions of the contractual circumstances, at others withdrawing our judgments and conforming them to principle, I assume that eventually we shall find a description of the [Original Position] that both expresses reasonable conditions and yields principles which match our considered judgments duly pruned and adjusted. This state of affairs I refer to as reflective equilibrium. (1971: 20)

Some of the controversial features of Rawls’s Original Position, such as the ascription of maximin reasoning to the parties therein, are best understood as products of the attainment of reflective equilibrium. They are features which serve to reconcile Rawls’s principles of justice—​the principles of justice that emerge from the Original Position—​ with his considered convictions. In his later writings, Rawls distinguished explicitly between narrow and wide varieties of reflective equilibrium.4 If we seek narrow reflective equilibrium, we are aiming simply to ensure that our considered convictions and our principles of just­ ice fit together coherently. If we instead seek wide reflective equilibrium, we not only aim to achieve congruity between our convictions and our principles; in addition, we subject the convictions and principles to challenges by assessing them in the light of alternative perspectives. That is, a theory of justice developed in wide reflective equilibrium is a doctrine whose tenets and implications have been tested against the claims of the major competing theories of justice. When such a theory has been fashioned and refashioned in a process of attaining wide reflective equilibrium, its substantive soundness as well as its internal consistency has been put to the proof. Though a quest for reflective equilibrium is not the sole element in Rawls’s understanding of the way in which the concept of justice should be analyzed, it is the central element. Rawls’s highlighting of it reveals that, in his view, the endeavor of conceptual analysis—​that is, the elucidation of the boundaries and implications of a concept—​is in part a collaborative enterprise. To achieve wide reflective equilibrium in one’s conception of justice, one must engage not only in reflections on one’s own ideas but additionally in dialogues with some rival outlooks.

4  The distinction between the wide and the narrow is first drawn explicitly in those terms in §1 of Rawls 1974. However, the distinction itself (without the “wide”/​”narrow” terminology) is present in Rawls (1971: 49–​50).

Conceptual Analysis and Distributive Justice    373

17.3  Concepts Versus Conceptions: Dworkin Ronald Dworkin invoked the distinction between concepts and conceptions in various parts of his work. With regard to issues of distributive justice, however, he doubted the availability of any formulable concept. As he wrote: Political philosophers can . . . try to capture the plateau from which arguments about justice largely proceed, and try to describe this in some abstract proposition taken to define the “concept” of justice for their community, so that arguments over ­justice can be understood as arguments about the best conception of that concept. Our own philosophers of justice rarely attempt this, for it is difficult to find a statement of the concept at once sufficiently abstract to be uncontroversial among us and sufficiently concrete to be useful. Our controversies about justice are too rich, and too many different kinds of theories are now in the field. . . . Perhaps no useful statement of the concept of justice is available. If so, this casts no doubt on the sense of disputes about justice, but testifies only to the imagination of people trying to be just. (1986: 74–​5)

Dworkin maintained that debates over justice are meaningful and coherent not because they are oriented toward some unifying formulation of a concept, but because the participants in those debates share a pre-​theoretical understanding of the range of matters covered by the concept of justice, and because the participants likewise concur on an array of paradigmatic specimens of justice or injustice. In the final book that he published during his lifetime, Justice for Hedgehogs, Dworkin further developed the account of interpretive concepts which he had first presented at length in the 1980s. As he wrote: “We share these [interpretive] concepts [such as the concept of justice] not because we agree in their application once all other pertinent facts are agreed upon, but rather by manifesting an understanding that their correct application is fixed by the best interpretation of the practices in which they figure” (2011: 158). Dworkin took those practices to be held together not by the sharing of formulable criteria but by widespread agreement on paradigmatic cases. Philosophers who dispute with one another about the concept of justice are offering rival sets of principles—​rival conceptions of the concept—​under which the paradigmatic cases of justice can be subsumed and thereby vindicated. Those principles extend beyond the paradigmatic cases to other matters on which there is no widespread agreement. Because the disputants adhere to divergent sets of principles, they wrangle intractably over non-​paradigmatic cases even though their argumentation is anchored in their shared recognition of the paradigms as such. Now, although Dworkin’s approach to pinning down the extension of the concept of justice is plainly different from Rawls’s, the two are not perforce incompatible. Dworkin,

374   Matthew H. Kramer however, took Rawls to task for seeking to specify a concept of justice that sets forth criteria or decisive tests for the identification of items within the concept’s extension. Everyone who shares the concept of justice is in agreement on the applications of those criteria or tests. Such is the view which Dworkin ascribed to Rawls: In [A Theory of Justice] John Rawls says that people who disagree about justice nevertheless “agree that institutions are just when no arbitrary distinctions are made between persons in the assigning of basic rights and duties and when the rules determine a proper balance between competing claims to the advantages of social life.” It is far from plain, however, that people do agree on criteria even at that very abstract level. It is a popular view in some parts of the world, for instance, that political institutions are unjust when they fail to respect God by providing authority and preference to his priests. That opinion objects not when arbitrary distinctions are made but when necessary ones are not made, and the complaint contains no claim about the proper distribution of advantages created by social life. (Dworkin 2011: 166–​7, footnote omitted, quoting Rawls 1971: 5)

In his criticism of Rawls, Dworkin returned to his contention that the concept of justice does not lend itself to any helpful encapsulation: It is unclear that we can find any form of words, however abstract, that describes a consensus among those we take to share the concept of justice. But even if we could, that consensus would not describe a decision procedure for identifying justice or injustice. On the contrary, it would simply point to further apparent disagreements, whose nature as genuine disagreements would then have to be explained. If we accepted Rawls’s suggestion, for instance, we would have to identify criteria that people who disagree about justice all accept for determining which distinctions are “arbitrary” and what is a “proper” balance of advantages. There are no such criteria. (Dworkin 2011: 167)

Though Rawls’s very early work is open to Dworkin’s reading, his later work does not warrant any such gloss. My overview of the structure of Rawls’s theorizing—​a structure within which Rawls explicated the concept of justice by striving for reflective equilibrium in dialogue with rival theories—​ does not rely at all on the notion of decision procedures whose outcomes or applications are recognized by every competent participant in discussions of justice. Rawls did not conflate determinacy and demonstrability in the manner suggested by Dworkin. To be sure, each of the layers in Rawls’s reflections is marked by necessary and sufficient conditions. However, as has been indicated, those conditions at each level pertain to the inclusion of theories at that level. They are moral principles or values on which certain theories converge through moral reasoning from more abstract points of convergence. Theorists who arrive at alternative moral principles or values have not misapplied some decision procedure that was immanent in the more abstract points of

Conceptual Analysis and Distributive Justice    375 convergence.5 Rather, they have engaged in alternative paths of moral reasoning. They have attached different weights or different understandings to the values and principles on which they converged with other theorists at the more abstract level(s). In other words, the processes of conceptual explication in Rawls’s work are broadly similar to those undertaken by Dworkin. In neither case does the elucidation of the concept of justice proceed through criteria that are presumed to be applied alike by all competent participants in debates over justice. As is conceded in the last of the quotations from Dworkin above, the rejection of criterialism (in Dworkin’s sense) is separable from the question whether the concept of justice lends itself to being pithily formulated. Dworkin and Rawls did disagree over that latter question, but neither of them was thereby committed to the implausible model of conceptual explication which Dworkin imputed to Rawls. As for the matter of coming up with an encapsulation of the concept of justice, Dworkin abandoned the quest rather precipitately. In the penultimate quotation above, for example, his objections to Rawls’s formulation depend on construing “when” as “when and only when” in the putative principle that “political institutions are unjust when they fail to respect God by providing authority and preference to his priests.” If “when” is so construed, that putative principle affirms a view which—​as far as I am aware—​has never been espoused by anybody. Hence, Dworkin’s invocation of that view does not tell against the success of Rawls’s concept-​distilling efforts. Moreover, when Dworkin specified a few of the paradigms which he took to be the unifying field of the concept of justice, he left himself vulnerable to some caviling doubts akin to those which he raised against formulations of that concept. He maintained that one of the paradigms is the unjustness of a situation in which a government takes steps “to convict and punish someone known to be innocent of any crime” (2011: 161). If a convicted person’s innocence is known only to legal-​governmental officials—​and is unlikely ever to be known by anyone else—​then there are many possible circumstances in relation to which a hard-​core proponent of utilitarianism will be prepared to hold that the conviction of the person for some crime is not unjust. Even if a convicted person’s innocence is known widely to members of the general public, there can be situations in which those members of the public will regard the conviction of the person for some crime as just. For example, by the end of the trial in To Kill a Mockingbird, everyone in the courtroom knows that Tom Robinson has not perpetrated the crime of which he has been accused. Nonetheless, the jurors convict him, and they undoubtedly perceive their heinous verdict as just. In their eyes, the verdict is a rightful affirmation of

5  A defender of Dworkin might here point out that, in the structure of Rawls’s theorizing, there is a transition from liberal accounts of justice to fully reasonable liberal accounts of justice. However, as has been remarked, that transition leaves ample room for theories of justice in rivalry with Rawls’s own theory. Moreover, as has also been remarked, reasonableness is a scalar property; hence, the fact that certain theories fall short of being fully reasonable does not entail that they are straightforwardly unreasonable and therefore incompetent.

376   Matthew H. Kramer the status of a black man who has been accused by a white woman. Dworkin committed himself to saying that the jurors do not have any grasp of the general concept of justice, whereas Rawls could more plausibly maintain that their grasp of that general concept is combined with their hideously illiberal understanding of what counts as an arbitrary distinction.

17.4  Value-​Independence versus Value-​Neutrality Most philosophers realize that the challenge of explicating the concept of justice is a morally laden endeavor rather than an austerely analytical or formal endeavor. However, there are a few dissenters. A good point of departure for one’s reflections on this matter is an important distinction between value-​independence and value-​neutrality.6 If a project of theoretical elaboration is value-​independent in the sense specified here, it is not grounded on any ethical values. That is, the justifications for its theses do not consist in invocations of ethical values. Some values do of course underlie any such project, but they are theoretical-​explanatory or logical/​mathematical or aesthetic rather than ethical. Value-​neutrality, in the sense specified here, is different. If certain theses are value-​ neutral, then we can gauge the fulfillment or non-​fulfillment of those theses in any number of circumstances without having to undertake any ethical judgments. In other words, the states of affairs recounted by those theses can be ascertained (as present or absent) without any ethical assessments. The distinction between value-​ independence and value-​ neutrality can perhaps best be illustrated with an example from a closely related area of political philosophy. Ian Carter in his theory of political and social freedom has addressed questions concerning whether individuals are free or unfree to do various things, and he has likewise addressed questions concerning how free each person is overall (Carter 1999). Under Carter’s theory, the answers to those questions can in principle be discovered—​through formidably complex investigations—​without any recourse to ethical judgments. We can answer those questions by ascertaining what people are able to do or unable to do, and by carrying out complicated aggregations of the magnitudes of their abilities and inabilities. Hence, by his lights, overall freedom and particular freedoms and unfreedoms are value-​ neutral properties. However, his justifications for specifying the nature of freedom as he does are ethical through and through. Carter maintains that the value-​neutral property which he singles out as overall freedom is of great moral significance in a number of respects which he carefully delineates. He contends that that property is a desideratum 6 

My thinking about this distinction has been sharpened by some exchanges with Ian Carter, to whom I am grateful.

Conceptual Analysis and Distributive Justice    377 with reference to which we can make good sense of the theories of justice that prescribe how freedom should be distributed. Similarly, when Carter specifies the nature of particular freedoms and unfreedoms, he is doing so with an eye toward the ways in which they contribute to individuals’ levels of overall freedom. In his view, then, overall freedom and particular freedoms and unfreedoms are all value-​dependent phenomena even though they are all value-​neutral phenomena. Their value-​neutrality, indeed, is crucially derivative of their value-​dependence—​because, according to Carter, it is their value-​neutrality that renders them promotive of the ethical values which he invokes. In connection with justice as well, value-​independence and value-​neutrality are disseverable. A large majority of the philosophers who seek to expound the concept of justice recognize that the task of conceptual explication is value-​dependent. They recognize that the pertinent justifications for unpacking the concept of justice in some direction(s) rather than in others are ethical. Yet, while any theory of justice is value-​ dependent, there can be theories of justice that are at least partly value-​neutral. Suppose, for example, that a theory ascribes injustice to all distributive arrangements in which any person’s annual income is below or above a certain range of percentages of the gross national output of the country in which the person resides. If the measurement of a country’s gross national output does not itself have to draw on ethical judgments, and if the attribution of incomes to sundry people likewise does not have to draw on ethical judgments, then the implications of the theory just posited are value-​neutral. Without engaging in further ethical judgments, we can ascertain whether those implications are fulfilled or unfulfilled in any given situation. Thus, although such an account of injustice is not value-​independent, it is value-​neutral. That is, although any credible justification for adopting such an account will be grounded in ethical values, the congruity or incongruity between the account’s prescriptions and the distributive arrangements in this or that society can be gauged without recourse to ethical beliefs.

17.5  Two Other Kinds of Neutrality Before we proceed to examine some pronouncements by major theorists of justice on the matter of methodological austerity versus methodological engagedness, we should note two types of neutrality that are different from the value-​neutrality discussed above. Because each of these other types of neutrality is quite often invoked in debates over justice, the endeavor of distinguishing each of them from value-​neutrality will help to avert some potential misunderstandings. First, as has been evident in my discussion of Rawls, an explication of the concept of justice typically progresses from a level of very high abstraction through levels of greater and greater concreteness. A position taken at a very high level of abstraction is logically consistent with an abundance of positions at any of the more concrete levels of conceptual explication. In that respect, stances on the most abstract matters in one’s reflections about justice are neutral vis-​à-​vis a wide array of stances on any of the more

378   Matthew H. Kramer concrete matters in those reflections. Thus, for example, a philosopher who agrees with Rawls in his understanding of justice as the appropriate apportionment of benefits and burdens among the members of any community can quite coherently diverge from Rawls when addressing some more concrete issues such as the soundness of the difference principle or the design of the Original Position. The connections between Rawls’s abstract conceptual stance and his principles of justice are substantive moral links rather than logical entailments. As a matter of logical consistency, his abstract conceptual stance is neutral between his own principles of justice and quite a few rival principles. Of course, the neutrality of a highly abstract formulation of the nature of just­ ice—​that is, its neutrality in relation to more concrete understandings of justice—​is not comprehensive. A formulation that encapsulates the nature of justice (however abstractly) will have ruled out some possible conceptions of justice at more concrete levels. Still, although no explication of the basic concept of justice is comprehensively neutral in relation to the more concrete doctrines of justice that might be propounded, any such explication is neutral among a multifarious medley of those doctrines. Of central importance here is that this neutrality of the abstract vis-​à-​vis the concrete is markedly different from the value-​neutrality discussed in section 17.4. For example, whereas Rawls’s abstract concept of justice is logically consistent with many concrete conceptions—​and is therefore neutral among a host of such conceptions in the sense specified in this paragraph and the preceding paragraph—​it is not value-​neutral. To the question whether the state of affairs encapsulated in that concept is currently an actuality, any pertinent answer must involve evaluative and normative judgments. Also quite different from the value-​neutrality recounted in section 17.4 is the neutrality at issue in debates between perfectionists and political liberals such as Rawls. Political liberals maintain (usually on contractualist grounds) that a legitimate system of governance must remain neutral between reasonable conceptions of the good.7 That is, a legitimate system of governance in its policies or actions must not aim to favor any reasonable conceptions of the good over other such conceptions either as means or as ends. In opposition to political liberalism, perfectionists argue that a liberal system of governance is both morally permitted and morally obligated to endorse some reasonable conceptions of the good in preference to other such conceptions. Instead of remaining neutral among views about the valuableness of sundry ways of life, a system of governance should be seeking to promote the flourishing of individuals by steering them away from discountenanced modes of living. So the perfectionist opponents of political liberalism have contended.

7  The phrase “conceptions of the good” has been used in a number of different but usually overlapping ways by various philosophers. It is typically employed to refer to views about what is valuable in life (beyond the basic goods that are covered by principles of justice), especially insofar as those views are grounded in religious creeds or wide-​ranging ethical doctrines.

Conceptual Analysis and Distributive Justice    379 The neutrality advocated by supporters of political liberalism is not value-​neutral in the sense specified by section 17.4. (Nor, of course, is it value-​independent.) To the question whether a system of governance has endeavored to favor any reasonable conceptions of the good over others, a pertinent answer must involve some evaluative and normative judgments—​most notably some judgments that gauge the reasonableness or unreasonableness of various conceptions of the good. Hence, the points at issue in debates between perfectionists and political liberals are orthogonal to the matters under consideration when we inquire whether certain theories of justice are value-​neutral.

17.6  The Value-​Dependence of Theories of Distributive Justice: A Dissenting View As has been stated, the value-​dependence of theories that expound the concept of just­ice has been recognized by most contemporary political philosophers (including both Rawls and Dworkin). Very few such philosophers aspire to explicate that concept without relying on basic ethical tenets. However, before this chapter adduces some pronouncements by philosophers who are in the majority on this issue, it should briefly take note of the views of the most prominent contemporary theorist who is in the dissenting camp. Throughout his career, Hillel Steiner has sought to establish normatively rich conclusions on the basis of exiguously formal premises. In that respect, his ambitions place him in a broadly Kantian tradition of theorists who attempt to endow moral conclusions with the prestige of logic (though, of course, not all philosophers who have been influenced by Kant are in sympathy with that methodological orientation). Steiner aims to construct a detailed theory of justice through reasoning that is austerely focused on considerations of logical consistency. He believes that his theory will prevail if it avoids the incoherence that supposedly besets alternative conceptions of justice. In the opening pages of his book An Essay on Rights, Steiner proceeds to delineate the chief method which he will employ to answer questions of justice. He propounds what he designates as a “compossibility test”: A set of rights being a possible set is, I take it, itself a necessary condition of the plausibility of whatever principle of justice generates that set. Any justice principle that delivers a set of rights yielding contradictory judgements about the permissibility of a particular action either is unrealizable or (what comes to the same thing) must be modified to be realizable. (Steiner 1994: 2–​3)

380   Matthew H. Kramer Steiner declares that his compossibility test “does exemplary service in filtering out many candidate conceptions of justice,” and he remarks that “[o]‌ur aspiration, obviously, is to pass through the eye of this needle with at least one theory of justice still intact” (Steiner 1994: 3). Yet, far from being a stringent filter that singles out one theory of justice (or a very small number of theories of justice), Steiner’s compossibility test is decidedly undemanding. It is satisfied by virtually every theory of justice that has ever enticed any estimable philosopher. Hardly any such theory generates the conclusion that some act-​type or act-​token is both morally permissible and morally impermissible. Steiner takes a contrary view largely because of his belief that every morally obligatory course of conduct is morally permissible. He persistently asserts that “a duty to do an action implies a liberty to do it” (Steiner 1994: 86), and he likewise repeatedly insists that “obligatory actions form a [proper] subset of permissible actions” (Steiner 1998: 268 n55). He thinks that any theory of justice is incoherent if it leaves room for situations in which people are morally forbidden to engage in morally obligatory courses of conduct. He elaborates and applies his compossibility test on the basis of that belief. Hoping thus to rely exclusively on the compossibility test and a few other formal considerations for the vindication of his theory of justice, Steiner contends that a philosophical account of justice—​and of connected phenomena such as liberty and rights—​can and should be value-​independent (as well as value-​neutral). Such an account will explicate the concept of justice without having to draw upon moral and political considerations, or so Steiner thinks. Elsewhere (Kramer 2009b), I have subjected Steiner’s methodological stance to a lengthy critique and have thereby endeavored to show that any satisfactory analysis of the concept of justice must draw centrally upon considerations of political morality. Any such analysis is value-​dependent. Someone who professes to be able to explicate the concept of justice on the basis of purely formal constraints is pursuing a chimera. In particular, Steiner systematically conflates “P is morally obligated not to φ” with “P is not morally obligated to φ,” as he presumes that the former proposition contradicts “P is morally obligated to φ.” When his confusion on that point is dispelled, and when we recognize the possibility and quite frequent actuality of moral conflicts—​each of which involves a situation in which someone is under a moral duty-​to-​φ and simultaneously under a moral duty-​not-​to-​φ—​we are able to see that theories of justice can allow for the existence of conflicting moral duties while remaining logically impeccable. Of course, any account of justice that envisions the pervasiveness of conflicting duties (as opposed to their occasional presence) is problematic. However, the dubiousness of such an account lies not in logic but in substantive morality. As a substantive moral matter, principles of justice should not lead to the conclusion that people will very frequently be unable to avoid the perpetration of wrongdoing. If Steiner wishes to assail a theory of justice that carries such a conclusion, he will have to contest it on moral grounds. His hope of dealing with the issue on logical grounds is illusive.

Conceptual Analysis and Distributive Justice    381

17.7  The Value-​Dependence of Theories of Justice: Raz and Campbell Instead of recapitulating the arguments which I have presented elsewhere in support of my rejoinder to Steiner, this chapter will now look briefly at some statements by a couple of political philosophers who affirm the value-​dependence of theories of justice. Let us first examine some pronouncements by Joseph Raz, whose 1986 book The Morality of Freedom is a major contribution to liberal political philosophy. In the opening chapter of that book, Raz ponders whether the concepts of freedom and justice lend themselves to being analyzed without recourse to ethical considerations. Although in his jurisprudential work he has submitted that the concept of law can indeed be analyzed without recourse to such considerations, he takes quite a different position in relation to the concepts of political morality. In a subsection of his opening chapter entitled “The Inadequacy of Linguistic Analysis,” Raz articulates his methodological position robustly: [M]‌uch that has been written in articles and books purporting to define and elucidate or analyse the concept of freedom will be helpful to anyone interested in the issues to be explored here. Philosophers and political theorists are sometimes better than their word. Much that is presented as conceptual analysis is really much more and includes advocacy of principles of political freedom. Moral and political philosophy has for long embraced the literary device (not always clearly recognized as such) of presenting substantive arguments in the guise of conceptual explorations. One may even say that the whole purpose of [The Morality of Freedom] is to defend a concept of political freedom. It is only important to remember that that concept is a product of a theory or a doctrine consisting of moral principles for the guidance and evaluation of political actions and institutions. One can derive a concept from a theory but not the other way around. (1986: 15–​16)

This understanding of the role of conceptual analysis informs all the subsequent portions of Raz’s book. For example, when Raz discusses coercion, he writes that “[w]‌hichever view one takes, it is not to be justified on linguistic or conceptual grounds but by the soundness of the moral theory of which it is a part.” His conception of coercion “is not to be justified as a piece of ordinary language analysis but on the grounds that while in accord with the core meaning of coercion it fruitfully ties it to sound moral principles” (1986: 150, 151). As these quoted passages indicate, Raz does not deny that conceptual analysis focused on formal and linguistic considerations should play a role in the elaboration of a theory of justice. His point is simply that those considerations are markedly insufficient in themselves for the vindication of any such theory. Tom Campbell, in the introductory chapter of his well-​known book on justice (2001: 9–​13), similarly grasps that any adequate analysis of the concept of justice

382   Matthew H. Kramer is value-​dependent. Campbell does not distinguish between value-​independence and value-​neutrality (either in those terms or in other terms), and some of his remarks can be construed as referring to either of those properties. However, that distinction is less important in the present context than his recognition that an analysis of the concept of justice has to go beyond formal and linguistic concerns. His emphasis on the value-​dependence of conceptual analysis becomes especially evident when he expresses some wariness of the distinction between concepts and conceptions. Specifically, Campbell queries any understanding of the concept/​conception distinction which assumes that “analyses of the concept of justice tell us what justice is all about in a detached and philosophical manner, while analyses of the differing conceptions of justice state what justice is in concrete terms and so enter the disputed arena of contentious and ideological political debate (2001: 10).” Like Raz, Campbell accepts that conceptual analysis focused on linguistic and formal considerations is of importance in any theorizing about justice, but—​again like Raz—​he contends that any such approach is insufficient on its own. As he declares: Further, the concept/​conceptions distinction can be misleading if based on the assumption that there is a clear line of demarcation between a morally neutral, if highly general, concept of justice on the one hand and specific conceptions which embody substantive moral interpretations of the general concept on the other hand. This strategy does not allow for the possibility that the concept of justice itself represents a distinguishable moral point of view which puts limitations on what can reasonably count as a conception of justice. (2001: 11)

Having shown alertness to the value-​dependence (and also the value-​non-​neutrality) of the concept of justice, Campbell salutarily admonishes his readers against packing too much substantive content into the abstract concept: “On the other hand, there is a danger of arriving at a restrictive analysis of the concept of justice which excludes rival political or philosophical views on an arbitrary basis.” As he elaborates: [T]‌he linguistic usages to which we appeal in order to establish a very specific concept of justice may be tendentious or dated, reflecting the experience and bias of the philosopher and their community rather than the alleged neutrality of ordinary and typical discourse. Analyses of the concept of justice which dictate its specific meaning may often be no more than devices for putting one set of values beyond the scope of critical evaluation. (2001: 12)

As Campbell’s remarks suggest, the value-​dependence of any analysis of the concept of justice is thicker and thicker as the analysis proceeds from the abstraction of the general concept to the textured concreteness of a conception. Even at the level at which the

Conceptual Analysis and Distributive Justice    383 abstract concept is picked out, the identification of it has to rest on some very general moral values. As that abstract concept is then developed into a conception, the value-​ dependence of the whole enterprise becomes all the more manifest and more intricate. Though formal and linguistic concerns are operative as well, moral principles (at varying levels of abstraction and concreteness) are the indispensable matrix within which any theory of justice unfolds.

17.8  Gallie on Essentially Contested Concepts My discussion of value-​dependence in analyses of the concept of justice will close—​as indeed will this chapter as a whole—​with a brief examination of a highly influential article published in 1956 by W. B. Gallie. In that article, Gallie singled out a class of concepts that share certain characteristics: (1) any such concept is evaluative or normative; (2) the matters appraised by any such concept are internally complex; (3) because of the multifacetedness of the matters appraised by any such concept, they can each be characterized in numerous different ways through different degrees of emphasis on their sundry components; (4) the matters appraised by any such concept can be considerably modified in response to changes of circumstances that are often unforeseeable; and (5) any such concept is typically invoked by exponents in favor of one characterization of it and in opposition to competing characterizations. If a concept does partake of these properties, it is essentially contested. That is, intractable contestation over the correct application of the concept is an essential feature of its correct application. Gallie maintained that one of the concepts in the essentially contested class is that of social justice (1956: 187). He somewhat puzzlingly suggested that the concept of social justice does not fully exhibit the third of the five properties listed above—​as he contended that the debates over social justice are structured by a dialectic between individualistic claims of desert and collectivist-​egalitarian claims of solidarity—​but he wisely left that suggestion aside as he pondered the concept of justice together with some other essentially contested concepts. His reflections commendably emphasize the value-​dependence of theories of justice, by highlighting the ways in which those theories proceed from normative and evaluative assessments of the problems which they address. Although Gallie’s ruminations somewhat resemble Dworkin’s approach to conceptual analysis, they lack two prominent elements of Dworkin’s writings. Each of those elements can help to remedy the shortcomings in one of the weakest sections of the article by Gallie, where he endeavored to explain why an essentially contested concept is a unity rather than an array of concepts (1956: 175–​80). He maintained that the unity of

384   Matthew H. Kramer such a concept derives from its having initially been invoked to cover some exemplary feat or situation that is recognized as such by all the participants in contestation over the concept. Now, the chief drawback of an account of conceptual unity along those lines is that it provides few grounds for thinking that there have ever been any essentially contested concepts. Dworkin’s account of conceptual unity is more sophisticated (albeit not wholly unproblematic). First, as this chapter has already recounted, Dworkin took shared paradigms to be the cement that holds together the practices of theorizing about just­ ice. Instead of following Gallie in perceiving the paradigms as some mythical point of origin, he perceived them as contemporary matters about which there is unanimity or virtual unanimity among the participants in the medley of ongoing debates over just­ ice. Second, and even more important, Dworkin persistently differentiated between indemonstrability and indeterminacy.8 The phenomenon of essential contestedness which Gallie explored is an epistemic phenomenon; it resides in the intractability of divergences among people’s understandings of some complex evaluative or normative properties. That epistemic situation of tenacious wrangling is fully consistent with the existence of determinately correct answers to the questions over which the disputants are clashing. Related to that second point is a third point, with which this chapter can conclude. As has been discussed in section 17.3, Dworkin rejected criterialism in analyses of the concept of justice because he presumed that all the underlying criteria would have to be shared by all competent participants in debates over justice. He was right to reject any such version of criterialism, but a more modest and sensible version can survive his worries. Although the participants in disputes over the concept of justice do not all converge on exactly the same set of criteria in their analyses of that concept, each of them adheres to an array of criteria that significantly overlaps with the arrays of criteria to which the other participants respectively cleave. To be sure, the structuring role of the criteria might oft-​times be largely implicit; some participants might not be able to articulate many of the standards that guide them (just as numerous users of the English language would not be able to articulate most of the syntactic and semantic rules that underpin their utterances). Nonetheless, the implicit sway of overlapping sets of criteria is what a careful investigation of the aforementioned disputes should be able to disclose. The overlapping among those sets—​as much as the convergence of participants on certain paradigms—​endows the disputation with a unified focus.

References Campbell, T. (2001). Justice, 2nd edn. Basingstoke: Macmillan Press. Carter, I. (1999). A Measure of Freedom. Oxford: Oxford University Press. Dworkin, R. (1986). Law’s Empire. London: Fontana Press. 8 

See, for example, Dworkin (1996: 129–​39). See also Kramer (2009a: 56, 88–​99, 179–​81).

Conceptual Analysis and Distributive Justice    385 Dworkin, R. (1996). “Objectivity and Truth:  You’d Better Believe It.” Philosophy and Public Affairs 25: 87–​139. Dworkin, R. (2011). Justice for Hedgehogs. Cambridge, MA: Harvard University Press. Gallie, W. B. (1956). “Essentially Contested Concepts.” Proceedings of the Aristotelian Society 56: 167–​98. Kramer, M. (2009a). Moral Realism as a Moral Doctrine. Oxford: Wiley-​Blackwell. Kramer, M. (2009b). “Consistency Is Hardly Ever Enough:  Reflections on Hillel Steiner’s Methodology,” in S. de Wijze, M. Kramer, and I. Carter (eds) Hillel Steiner and the Anatomy of Justice. New York: Routledge, pp. 201–​13. Quong, J. (2011). Liberalism without Perfection. Oxford: Oxford University Press. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (1974). “The Independence of Moral Theory.” Proceedings and Addresses of the American Philosophical Association 47: 5–​22. Raz, J. (1986). The Morality of Freedom. Oxford: Oxford University Press. Steiner, H. (1994). An Essay on Rights. Oxford: Blackwell. Steiner, H. (1998). “Working Rights,” in M. Kramer, N. Simmonds, and H. Steiner (eds) A Debate over Rights. Oxford: Oxford University Press, pp. 233–​301.

Pa rt  I V

T H E A P P L IC AT ION OF DI ST R I BU T I V E J U ST IC E

Chapter 18

Gende r Anca Gheaus

18.1 Introduction On one influential view, distributive justice concerns the fair sharing of the burdens and benefits of social cooperation (Rawls 1971). Another mainstream view of distributive justice identifies it with the elimination of inequalities in how well people’s lives go for which individuals do not bear responsibility (Arneson 2000). In most societies women appear to have a lesser share of benefits than men, a greater share of burdens than men, or both: even in countries where women enjoy the same legal rights as men there are large disparities among women and men with respect to wealth, income, and the holding of desirable social positions such as well-​paid or high-status jobs and political functions. Worldwide, more women than men live in poverty (United Nations 2010; United Nations Development Programme 2014). The gender pay gap—​that is, the fact that women on aggregate earn less than men—​and women’s under-​representation in positions of responsibility and power appear especially problematic in those societies where women have started to outperform men educationally. On the side of burdens, the massive entrance of women into the labour market over the past decades has not been accompanied by a redistribution of domestic work between women and men on a similar scale; as a result, it has been argued that on average women—​and especially mothers—​work longer hours than men when both paid work and unpaid domestic labour are taken into account (Hochschild 1989; United Nations 2010). Some deny this is still true in Europe (Hakim 2010). These facts generate a strong prima facie case to think that gender—​that is, the implicit and explicit norms we associate with being female or male—​bears on matters of distributive justice. The general purpose of this chapter is to explore the relationship between gender and distributive justice and the difficulties of addressing the causes of inequalities between women and men. Which disparities in the benefits and burdens—​in short, ‘outcomes’—​of social cooperation between women and men, if any, one identifies as indicative of gender injustice will depend on the theory of distributive justice that one endorses. Theoretical

390   Anca Gheaus decisions concerning the distribuenda of justice (the goods whose distribution is relevant for justice) and the site of justice (institutions-​only or individual behaviour, too) will all influence how one thinks about gender justice. This chapter discusses gender in relation to the most influential current accounts of distributive justice. According to John Rawls, distributive justice mandates a fair distribution of primary goods: wealth, income, opportunities to obtain desirable social positions, and the social basis of self-​respect. The inequalities of outcome between women and men with respect to all these goods could be indicative of gender injustice, especially if gender turns out to negatively influence the women’s opportunities to secure desirable social positions. According to luck egalitarianism, inequalities in how well off people are can be just only if they result from individuals’ responsible choices. We do not choose whether to be born as women or as men, and therefore on a luck egalitarian account it is objectionable to be worse off merely because of one’s sex. Luck egalitarianism is a principle of distribution and is as such consistent with different views on what the distribuendum of just­ ice is. Depending on what one takes to be the relevant distribuendum, various facts in addition to those already mentioned can be identified as forms of gender injustice—​for instance, women’s suffering more social marginalization and exclusion than men, their objectification through pornography, or their getting less credibility as epistemic agents than men. An upfront methodological clarification is in order. Some philosophers distinguish between two forms of social justice: distributive justice and recognition-​based justice. The first concerns just distributions while the second concerns egalitarian relations between individuals—​that is, relations free from marginalization, domination, exploitation, violence, and powerlessness have important connections to gender (Young 1990; Bird 2018). Whether or not one form of justice can be reduced to the other, and whether or not one is more important than the other (Fraser and Honneth 2003), I assume here that kinds of injustices that are relational in nature also have an important distributive aspect. This is because, by dint of suffering from a form of relational injustice, one is also likely to be worse off in terms of one or more of the following: one’s well-​being, one’s overall resources, and one’s opportunities. Therefore, unless one thinks that justice is restricted to the distribution of some material resources (like money), one will have to acknowledge that relational injustices can represent a form of distributive injustice. In this chapter I address three main, and closely related, issues at the core of contemporary debates on gender justice. The first regards the nature of gender injustice. I discuss whether gender disparities can be unjust even if they reflect individual choices. The answer to this question is particularly important for liberal accounts of justice that require equality in individuals’ opportunity for welfare, or resources (understood broadly), or capabilities, or access to advantage, and view inequalities in outcomes along these dimensions as compatible with justice. According to such theories, the causes of gender disparities are important, since not all inequalities in outcome are unjust. Several issues are relevant here:  whether the inequality-​inducing choices that women make are non-​coerced and voluntary, and whether those choices that are in this sense free

Gender   391 nevertheless sanction gender inequalities, for instance by expressing preferences shaped by gender norms. Finally, whether gender injustices affect only women. The second question I consider is how to legitimately change the problematic facts that result in women and men having unequal access to positions of advantage. Assume that, as Rawls argued, justice is a virtue of some institutions only and that, to a large extent, the inequalities between women and men are perpetuated by the family and by interactions on the labour market. An account of legitimate change will turn on whether the family and interactions on the labour market ought to be regulated by principles of justice—​for instance, by abolishing sexist upbringing and the gendered division of labour. Even if interventions in the family and the work market are as such legitimate, there remains the question of finding adequate policies and institutions that can diminish the impact of gender on distributions; at least in some cases it is difficult to promote gender fairness efficiently without upsetting other important liberal desiderata such as freedom of individual choice. By contrast, some think that justice also applies at the level of individual behaviour (Cohen 1997)—​that, as the popular feminist slogan has it, the personal is political; I will also consider individual duties to change the gendered norms that result in inequalities. A third question concerns the existence and relevance of an unconscious, pervasive tendency to associate normatively laden characteristics with either women or men in ways that result in their discrimination. They are said to operate in spite of the fact that more and more individuals explicitly endorse gender egalitarianism. Theorists concerned with unjust distributions between women and men have not, so far, sufficiently considered the relatively recent research on implicit bias and stereotype threat. I employ the classical distinction between sex and gender:  ‘sex’ refers to the biological features of individuals—​such as chromosomes, sex organs, hormones, and other physical features—​while ‘gender’ represents the social meanings associated with sex, including gender norms. Sex and gender map onto each other—​albeit imperfectly—​and inequalities between women and men are most obviously problematic when they result from gender norms that nudge individuals to make disadvantageous choices, or otherwise encumber their access to valuable goods. It is important to note a recent proposal to equate the social fact of being identified as female with systematic subordination. Sally Haslanger proposed the following definition of ‘woman’ to refer to an individual’s gender rather than sex: S is a woman iff(df)1 S is systematically subordinated along some dimension (economic, political, legal, social, etc.), and S is ‘marked’ as a target for this treatment by observed or imagined bodily features presumed to be evidence of a female’s biological role in reproduction. (Haslanger 2000: 39) 1 

This means ‘by definition’.

392   Anca Gheaus This definition has the advantage of capturing what is wrong with many gender norms: the fact that they subordinate a class of individuals to another class of individuals, merely in virtue of (perceived) biological facts. However, I do not rely on it. Others have noted that in everyday language we often use the term ‘woman’ to indicate a person’s sex, rather than her gender (Saul 2006). Moreover, by building hierarchy in the very notion of gender, this definition obscures the fact that men, too, are systematically disadvantaged by gender norms (Benatar 2012). This is not to deny that women are overall more negatively affected by gender injustice. If reference to hierarchy is not necessary to see why gender norms are inimical to justice, it is better to employ a non-​moralized concept of gender.

18.2  Gender Norms, Individual Choice, and Justice According to most theories of justice, outcomes are not unjust if they reflect non-​ coerced and voluntary choices. If just distributions are sensitive to individual choice, the causes of inequalities of outcome between women and men will bear on whether or not they are unjust. A possible source of obvious gender injustice is legislation that openly discriminates against women. Historically, this was the case in most societies. But nowadays, at least in the so-​called developed countries, laws do not raise formal obstacles to women’s access to economic, political, or social opportunities. Yet, it is likely that in such countries some of the current disadvantages faced by women continue to stem directly from legislation which, while itself gender-​neutral, necessarily affects women and men differently, partly due to sexual difference. For instance, when laws ban abortion and encourage the teaching of abstinence as the best form of contraception (with little success) it is inevitable that women will get pregnant, be forced to carry to term, and therefore have to pay the costs, including the opportunity costs, of unwanted pregnancies, while men escape these costs. Laws can generate unfair outcomes for men, too—​for instance, the conscription of men-​only into the army is probably unfair (Benatar 2012). However, even legislation that does not engage in direct discrimination can only explain to a small extent the disparity in economic, social, and political outcomes between women and men. Explanations must be sought elsewhere, in the content and working of informal gender norms. Therefore, one must consider the processes most likely to explain the disparities in outcomes between women and men, in order to see if these disparities are indeed morally arbitrary, or whether they can be traced to individual responsibility.

18.2.1 Gender Norms Gender norms are the rules, formal or informal, explicit or implicit, concerning the way in which women and men ought to be and behave; they shape not only social

Gender   393 interactions but also individuals’ expectations and preferences (and hence, indirectly, our bodies too). In societies where laws do not discriminate against women, individuals’ micro-​ choices—​many of which reflect and shape gender norms—​explain to a large extent the disparities that have been described. Consider, for instance, interactions between men and women in the context of co-​parenting within nuclear families: gender norms, embedded in the working of institutions and/​or directly guiding individual behaviour, encourage and sometimes make it rational that women, rather than men, take long leaves from work and/​or settle for part-​time or otherwise less demanding jobs in order to rear children (Allen 2008). This is because women are more encouraged to specialize as care givers and men as breadwinners—​for instance through the gendered education and social expectations that women face as adults—​and because only women breastfeed, and in many countries parental leaves are available only to, or disproportionately to, women. Amongst the opportunity costs of these choices are significantly reduced lifetime earnings and other work-​related benefits for women, which often involve economic dependency on their spouse and less access to powerful or prestigious social positions. Individual interactions between women and men outside the family, too, play some explanatory role. For instance, professions that are more feminized also tend to pay less (United Nations 2010). Amongst professions that do not require higher education, some male-​dominated professions offer not only better pay but also more interesting jobs than other professions; but they can be very inhospitable to women. A  working-​class woman who wanted to work in, say, construction, may have to put up with significantly more hostility than if she was to become a cleaner. It is likely that expected interactions with potential co-​workers dissuade some women from seeking the jobs they would otherwise want; in an important sense, they have less opportunity than men to enter and remain in such professions. Men, too, may face comparable obstacles in entering some highly feminized professions (Benatar 2012), but jobs in such professions come with less pay and social status than jobs which welcome men. This means that two separate issues of justice are at stake here: (a) that the gender segregation of the labour market can leave women worse off than men in terms of outcomes due to differences in pecuniary and non-​ pecuniary payoffs between ‘feminine’ and ‘masculine’ jobs; and (b) that the gender segregation of the labour market can leave women worse off than men in terms of opportunities if there is a greater range of jobs from which women are excluded and if the obstacles women face to enter the masculine jobs are greater than those men face to enter feminine jobs. This means that, in effect, women on average enjoy less freedom of occupational choice. Choices in these examples are not legally coerced; does this mean that they do not raise issues of justice? Much of the debate around this issue is whether the choices in question reflect genuine preferences or whether they are made under the pressure of either gender norms or gender-​biased institutional arrangements.

394   Anca Gheaus Some believe that behaviour caused by gender norms that significantly harms women—​ for instance, breast implants—​ cannot be autonomous as long as the preferences that lead the choice are the result of socially constructed gender norms (Chambers 2004). The thought here is that, if preferences are socially formed, one cannot appeal to the value of individual autonomy in order to justify non-​intervention with individuals’ behaviour:  autonomy has already been compromised by the existence of gender norms. It is debatable whether the existence of gender norms precludes individual autonomy sufficiently to justify paternalist intervention. The real worry is that gender norms impose unacceptable costs on the pursuit of autonomous choices: for example, a woman’s choice to bear and rear children can be autonomous even if it partly results from gender socialization, but paying the price of foregoing her career to be able to parent is unfair (Mason 2000). Most gender norms contribute to gender unfairness, and most cases of gender unfairness are at least in part the result of gender norms. Plausible complaints of justice against gender norms are that they result in unequal capability sets for women and men (Robeyns 2007); and that gender norms are objectionable because they make women’s access to some valuable goods, and men’s access to other valuable goods unduly difficult or costly, thereby diminishing individuals’ valuable freedom merely due to their sex (Gheaus 2012). If gender norms encumber women’s freedom more than men’s, they introduce an unfair inequality of valuable freedom. It is important to note that some of the individual interactions that result in women’s disadvantage, and which are generated by gender norms, do not involve any choices made by the women who are affected. One example is statistical discrimination against women: cases when an employer decides against hiring a woman although she is as competent as any of her male counter-​candidates, because the employer thinks that the woman in question is more likely than a man to take on caring responsibilities in her family—​most usually, childcare—​making her a less resourceful employee in the long run. Both Rawlsian and luck egalitarians may be able to explain why outcomes resulting from statistical discrimination are unjust. Other examples involve biases against women, biases that are not conscious or fully conscious and which lead to evaluating women as less competent merely because they are female. For instance, some studies show that the very same CV gets a worse evaluation if it is attributed to a woman than if it is attributed to a man (Holroyd 2012). It is difficult to classify the effects of implicit bias as a form of injustice: because implicit bias is, by assumption, largely unconscious, it is not clear that there is any agent to whom we can attribute responsibility for any outcomes it generates. Also, it is not clear whether implicit bias can be entirely eliminated, and therefore whether anyone can have a duty to entirely eliminate it. A gender-​just society would be free of inequality-​generating gender norms. Yet, participants in public debates on gender justice often assume this cannot be so because they believe that gender norms reflect natural preferences—​for instance, that women on average genuinely prefer to perform more care work than men, whether at home or as paid work. They think that many differences between women’s and men’s behaviour can be explained by reference to inborn differences between them—​for instance, differences

Gender   395 between how women’s and men’s brains are naturally ‘wired’. On this view, women are more inclined to nurture, cooperate, and display empathy, while men are more inclined to display logical, systematic thinking. Some scientific research endorses the theory of innate differences between women and men (Baron-​Cohen 2003; Pinker 2008); but there are also scientists who criticize this research for employing spurious methodology, and explain how differences (including brain differences) could be instead attributed to nurture (Fine 2010.) This includes brain differences, because experience ‘creates neural activity that can alter the brain either directly or through changes in gene expression’ (Fine 2010: 236). Whatever the truth of this matter, it would be simplistic to assume that the different outcomes between women and men are morally unproblematic to the extent to which they result from natural factors determining choice.2 Some philosophers think that the likelihood of inborn behavioural differences between women and men can make some inequality in outcomes unproblematic, since all that matters for gender justice is that women and men be given equal opportunities to engage in social competition—​that is, to compete unencumbered by socially created obstacles related to them being female or male (Radcliffe-​Richards 2014). This, however, would be true only if (a) the scope of justice was restricted to regulating competition; and (b) the rules of social competition were, themselves, gender-​neutral. The first is an implausibly restrictive conception of justice: many think that justice is about more than regulating social competition—​that it requires, for instance, the improvement of the situation of the worst-​off, or distributions that are not sensitive to brute luck. Moreover, behavioural differences between women and men may have moral significance independently of whether they lead to competitive advantage; for instance, some of the ‘feminine’ behaviour likely results in the provision of goods essential to human survival and flourishing and so the differences between women and men are directly relevant to distributive justice. I do not discuss (a) further. I discuss point (b) in the next subsections. At least four distinct issues pertaining to the behavioural differences between women and men are relevant to thinking about justice; I address each of them in the next four subsections:  (1) whether their causes are innate or socially constructed (subsection 18.2.2); (2) what kinds of advantage they are conducive to (subsection 18.2.3); (3) whether all gendered behaviours are, in themselves, equally morally desirable (subsection 18.2.4); (4) whether or not it is possible to change them, and at what cost (subsection 18.2.5).

18.2.2 Are Differences of behaviour between Men and Women Innate or Socially Constructed? Three things are beyond doubt: that women and men behave differently, that there are biological differences between their bodies, and that there are social norms encouraging 2 

On the normative importance of ‘the natural’, see also Casal 2018 (Chapter 12 in this volume).

396   Anca Gheaus them to behave differently. Many gender norms are explicit and train individuals to behave differently according to their sex, starting from early childhood. If research on gender schemas is correct,3 much of the encouraging is also done unconsciously (Valian 1999). Psychoanalysts have argued that some of the gender norms that entail women’s domination are formed unconsciously, in psychological processes that take place in infancy (Benjamin 1988). Since it is hard to deny that gender norms exist, one question is to what extent gendered behaviour is natural—that is, inborn—and to what extent it is nurtured—that is, acquired. John Stuart Mill (1869/​1973) noted that we cannot determine the answer, since there is no social environment free of gender norms that would allow us to find out how much of the behavioural difference is inborn. It is important to note that in any case innate differences between women and men are unlikely to explain, alone, the full extent of outcomes related to economic advantage and political and social power. The most that the inborn theory purports to show is that women and men differ in their average levels of various abilities, and so it can explain only relatively small differences in skill at the higher end of the range. Even if, for instance, men have higher innate mathematical ability, this alone could only provide a partial explanation of the extent to which some desirable professions—​such as computer science—​are dominated by men. And only a minority of women express gender-​conservative preferences concerning the division of household labour and the combining of paid work and care work. What is the import of the nature–​nurture debate on gendered behaviour for justice? As we have seen, it is unlikely that we shall ever be able to determine precisely to what extent gendered behaviour is shaped by nature and to what extent it is shaped by gender norms. Fortunately, the source of individual preferences that lead to significantly inequality-​inducing choices does not seem to be very important: one cannot be any more responsible for one’s innate desires than one can be for socially constructed ones (at least assuming that people can equally identify with both kinds of desires). And if some desires are sufficiently powerful to qualify as urges (for instance, to have children) they diminish individuals’ responsibility for acting on them independently of their source. If gendered preferences result in inequality-​inducing choices, they are objectionable independently of what causes them. The case for objecting to them does not depend on the resolution of the nature–​nurture debate. On a luck egalitarian conception of justice, the question of whether individual choices give legitimacy to disparities in outcomes turns on whether they are truly voluntary choices—​which in turn can be equally affected by their natural and social causes. Inborn differences between women and men can impact on the voluntary nature of individual choices: for instance, if some women were to truly experience over-​powering urges to procreate, it would be unfair to require them to bear the full responsibility 3  These are implicit assumptions that people make about gender differences, according to which small behavioural sex differences indicate differences between women’s and men’s characteristics (rather than socially constructed differences). Gender schemas are said to lead to different normative evaluations of women and men.

Gender   397 of their procreative decisions. On a Rawlsian conception of justice, the legitimacy of gendered disparities in outcome will depend on whether they occur against fair background, institutional conditions. But on any account, the question of whether or not those gendered disparities of outcome which are unjust are changeable will be very important.

18.2.3 What Kinds of Advantage Result from Behavioural Differences between Women and Men? The answer to this question will depend on the choice of a metric of justice and on whether justice requires individuals to be equally well off globally, or, more demandingly, to be equally well off in an irreducible plurality of respects. On the first issue, it seems important to look beyond the usual currencies of just­ ice—​such as money and opportunities for desirable positions—​in order to see how gender can impact on differences in women’s and men’s access to relational goods, such as credibility or social respect. This inquiry blurs the line between theories of distributive justice and theories of relational equality that are primarily concerned with how individuals relate to each other as equals. Some feminist theorists insist that oppression and lack of recognition, which comes in different expressions, are forms of injustice towards women at least as significant as unfair distributions of material resources (Young 1990). But, as already explained, it is likely that non-​materialist versions of luck egalitarianism can accommodate these concerns as matters of distributive justice. Recently, important feminist work has been focussing on the marginalization of women as knowers. According to Miranda Fricker (2007) women are subject to both testimonial injustice, when they are being wrongly treated as unreliable sources of information due to the hearers’ prejudice, and hermeneutical injustice, when their epistemic disadvantage results from insufficient collective interpretive resources—​for instance, when a woman’s complaint that she has been sexually harassed is incomprehensible because the concept of ‘sexual harassment’ is lacking. These phenomena are interesting, both as explanations of why women and men achieve different outcomes in spite of gender-​neutral legislation and, possibly, as forms of injustice in themselves. As already noted, that an inequality is the result of gender norms is neither sufficient nor necessary to be unfair. Take, for instance, the norm that says that girls ought to wear pink and boys ought to wear blue. In itself the inequality introduced by this norm seems innocent; however, the problem with the norm is that it marks individuals depending on their sex for different treatment with respect to things other than the colour of dress; the gendered colour scheme is part of a convention that shapes individuals’ different responses to baby girls and baby boys, reflected for instance in different choice of presents or praise. Furthermore, it is not necessary that unfair inequalities be the result of gender norms; men’s lower life expectancy, to the extent to which it is not determined by any (gender) norm, can nevertheless be deemed unfair by a luck egalitarian theory of justice.

398   Anca Gheaus Not only do women have a lesser share of economic and status-​related benefits and may work more hours than men, they also carry most of the procreation-​related burdens—​the health-​related burdens and most of the emotional burdens of pregnancy and breastfeeding. (Moreover, to different extents in different countries, they are also expected to pay for pregnancy-​related healthcare.) But note that women’s choices—​to bear children, to take time off paid work and settle for less attractive jobs in order to care for them, or to get employed in caring professions—​do not result in disadvantages independently from how institutions reflect gender norms. The disadvantage is partly due to the institutional set-​up which does not compensate women for bearing the costs of childbearing, which attaches financial and other penalties to part-​time or flexible jobs, and which does not reward feminized jobs as well as non-​feminized jobs. Identical choices by women made against a different institutional set-​up—​one that rewarded formal and informal care work better—​would not result in (the same level of) disadvantage. Because the payoffs of particular choices are determined by the context of choice, even if women’s choices were to express their genuine preference for caring activities this could not alone justify the different outcomes between women and men (unless men incurred a similar degree of overall disadvantage due to their being men).4 One way to argue that current institutional arrangements are just is to say that there is a general and fair division of labour between women and men, in which women as a group are mostly responsible for reproductive and caring work, and men as a group are mostly responsible for productive work and self-​defence. As already mentioned, men, too, are in some contexts disadvantaged qua men. For instance, there is less sympathy for male victims of (especially sexual) violence, and men are expected to repress their emotions to a greater extent than women (Benatar 2012). Many of the disadvantages that men incur qua men can be traced to women specializing in care-​ giving and men being the breadwinners. Most obviously, if men on average spend more time in paid employment and less time caring, they are also likely to miss out to some extent on the valuable goods of hands-​on parenting (Brighouse and Wright 2008). Caring for a dependent person whom one loves entails both burdens and benefits. Political theorists usually discuss care-​giving in the context of fair sharing of burdens, but some have recently turned their attention to how being an involved parent can play an important role in people’s flourishing (Brighouse and Swift 2014). More generally, to the extent to which they give less weight to intimate relationships in their lives, men are also less well positioned to enjoy their value; but if intimate relationships objectively contribute to well-​being, men are, in one respect, worse off than women. Thus, some think that if we look at the overall rather than local distribution of burdens and benefits in liberal democracies, women are not on balance worse off than men (Benatar 2012).

4 

For a similar line of reasoning, see Olsaretti (2009).

Gender   399 Here is a way to challenge this argument. Many, although not all disadvantages that gender norms impose on women and on men may be thought to result from a traditional gendered division of labour. But such a division of labour could be itself fair only if (i) women and men really were to be the subject of overall equal disadvantage; and, at the same time, (ii) such a division of labour was necessary in order to give all individuals an equal chance to lead maximally good lives. This picture may have been plausible at a time when people believed there were large inborn differences between women and men, making each uniquely fit for some kind of activities, and when frequent wars imposed very high burdens on men. Yet, given current knowledge about women and men’s abilities and current social circumstances, it is unlikely that any of the two conditions are met—​let alone both: first, in this day and age it is more likely that, on the whole, being a woman is more disadvantageous than being a men, even in liberal democracies. And second, some women express preferences for doing little or no care work (whether formal or informal); presumably, at least some men would prefer to do more care work if the institutions that regulate paid work gave them better opportunities to combine that with earning a living. The traditional gendered division of labour does not seem to be either efficient or fair. Assume it is true that both women and men are, in different contexts and at least partly due to gender norms, worse off than members of the opposite sex. There are several important questions that pertain to the relevance of this fact for distributive justice. First, does the fact that men, too, suffer some disadvantages qua men, make things overall worse or better from the point of view of injustice? Does justice require the elimination of each kind of disadvantage one by one, or the equalization of the overall amount of (dis)advantage that one can expect in virtue of being a woman and, respectively, a man? Whether we should take an aggregationist or a non-​aggregationist view on what constitutes unjust disadvantage is a difficult and, at least in the context of gender justice, under-​researched question. Similarly, any division of labour whereby some people do significantly more care work than others while others spend more time in paid employment will entail unequal burdens and benefits for people in both groups. Given this fact, some think that justice requires the abolition of gender norms such that, for all individuals, the costs of a gender-​neutral lifestyle—​understood as one combining care-​giving and paid work—​are no higher than the costs of leading a gendered lifestyle (Gheaus 2012). Others believe that justice, understood as comprehensive equality of opportunity, is compatible with a type of gendered division of labour that would truly promote equal overall life prospects for women and men (Arneson 1998). Gender norms are in part responsible for various disadvantages accruing to women and, sometimes, to men—​that is, for gender unfairness. Does this fact make all gender norms that create unfairness equally objectionable? The answer to this question must depend on the independent question whether all gendered behaviours promoted by various gender norms are, in themselves, equally morally desirable. I explain this point in subsection 18.2.4.

400   Anca Gheaus

18.2.4  Are all Gendered Behaviours, in Themselves, Equally Morally Desirable? Some of the gender norms that regulate women’s and men’s behaviour seem to be, in themselves, morally neutral:  for instance, those related to dress, appearance, or courtship codes. (This is not to say that it cannot be harmful to try to attain some ideals of feminine beauty, or that it is fair to expect women to invest more in their appearance than men in order to be socially acceptable.) When such gender norms disadvantage individuals by making their access to goods more or less difficult depending on their sex they are unfair and there is a strong pro tanto case for abolishing them if we can. Other gender norms have moral content. Women are expected to be more nurturing and caring than men. The (contested) belief that they do in fact tend to think in terms of individual needs and relationships, rather than rights and competing interests, has generated an ethical tradition—​the ethics of care (Gilligan 1982). This tradition builds on the insight that care in face-​to-​face encounters is essential for creating new human beings by nurturing and socializing them (Ruddick 1989) and instilling a sense of morality in them (Held 1993). Some have argued that personal care is an all-​purpose-​means, making it akin to a Rawlsian primary good, and hence a matter of direct concern for just distributions (Brake 2010; Kittay 1999). It is not clear how feasible it is to distribute equally the affective component of care, nor how desirable it is to attempt to do so (Gheaus 2009; Lynch, Baker, and Lyons 2009); yet this is not to deny that some affection is necessary for all children’s and most adults’ ability to lead flourishing lives, and hence a matter of distributive justice. At the same time, many feminists are critical of the ethics of care tradition because it makes those who endorse it vulnerable to exploitation and is  likely to entrench their subordination (Card 1990; Held 2018). If women have indeed been by and large subordinated to men, it seems that an ethics of care is a particularly counterproductive set of norms to be endorsed by women. As long as most goods are distributed through competition, and social success is more available to those who promote their own, rather than others’, interests, being nurturing and caring comes at a prudential cost. Conversely, social norms encourage men, but not women, to be highly competitive, self-​assertive, or even aggressive; these dispositions are also likely to advance the competitive advantage of individuals who display them, but it is far from clear that they are morally desirable. In the case of gender norms encouraging behaviours that are in themselves morally desirable, it is particularly difficult to arrive to an all-​things-​considered judgement about their desirability. Because they encourage morally praiseworthy behaviour, there is a strong prima facie case for universalizing, rather than abolishing, them. But this may be impossible as long as background conditions remain unchanged—​that is, as long as competition regulates access to most goods, and self-​assertiveness, rather than being nurturing, improves one’s chances of social success. Such conditions

Gender   401 make it very unlikely that feminine norms of nurture and care can be universalized and promoted with equal success amongst men. On the one hand, someone concerned with gender justice will want women to have the same chances to acquire social success as men, and hence will see the case for socializing women in the same highly competitive spirit, rather than in the spirit of care. On the other hand, there is the problem of weighing a gain in gender justice against a loss in morally valuable kinds of human relationships. The only way out of this dilemma is to reshape the social world such that the traditionally feminine virtues of care do not entail competitive disadvantage; this may be a long shot. Here is another example of how non-​ideal social conditions—​a failure to train men to competently care for babies—​can generate conflicts between gender justice and other moral goals in relation to the abolition of gender norms. Assume that men could be adequate care-​givers for babies, but they are either not helped to or outright discouraged from being so. If being able to form intimate relationships with babies is a kind of advantage, this norm is perpetuating gender injustice because it makes it harder for men to become care-​givers, as well as increasing the amount of morally required care that women must perform. In particular, this creates unfairness for those men who, in spite of gender norms, are willing to care for babies and who would make very good care-​givers. But, at least as long as men, on average, really are significantly less capable than women to provide competent care to babies, these norms also help protect children—some of the most vulnerable individuals—from neglect or abuse. Again, it is not clear that, all things considered, it is desirable to abolish these norms until and unless we rectify the underlying problem: significant average disparities in women’s and men’s competence to care for babies, if indeed such disparity exists.

18.2.5 Are Inequalities of Preferences between Women and Men Changeable? If some gendered behaviours lead to disadvantage, many will think that they can be identified as objectionable injustices (rather than mere unfortunate facts) only if they can be changed.5 A  related question is at what material and moral cost they can be changed. This is a very different question to that of the relative contribution of nature and nurture to these differences. Some facts that are (co-​)determined by innate features can be easily changed (like short-​sightedness.) Other features that are entirely the result of nurture are very resistant to change (for instance, one’s ability to hear and reproduce some of the sounds of languages to which one has not been exposed as a child). Similarly, some differences between women and men may be due to natural factors—​ for instance, differences in life expectancy or susceptibility to particular diseases—​but changeable—​for instance, through medical or technological progress. 5 

I dispute that only facts that we can change can constitute injustices (Gheaus 2013).

402   Anca Gheaus For the time being it is likely that there exist both biological and social facts which play a role in some inequalities of outcome between women and men and which we do not know how to change. The fact that only women can bear and breastfeed children goes some way towards explaining why women and men start to specialize in, respectively, care-​giving and bread-​winning and often create a path-​dependent development towards even greater specialization (Allen 2008). Some feminists have argued that full equality between women and men will not be possible before we break the connection between sex and childbearing (Firestone 1971). Whether it would be, all things considered, desirable to break this connection—​for instance, by artificial gestation—​is itself an under-​researched question that may soon become timely. Other causes of discrepancies of outcome between women and men that we do not yet know how to change are implicit biases and susceptibility to stereotype threats; to the extent to which they are the sediments of a long history of explicit sex-​based discrimination and prejudice they are unlikely to be easily shattered. But even if is impossible to eliminate all the causes of discrepancies of outcome between women and men, it is possible to limit their influence (for instance, anonymizing job applications and exams before evaluating them keeps some biases under check) or try to compensate individuals for unfair disadvantage. And it may be possible to compensate for those inequalities of outcome which we don’t know how to avoid. The four questions discussed in sections 18.2.2–​18.2.5—​concerning the socially constructed nature of gender norms, their relation to advantage, their independent moral desirability, and their immutability—​are different from each other; the last three seem more important than the first. This is because the extent to which innate factors contribute to behavioural differences between women and men does not determine either the possibility or the desirability of changing them. Section 18.3 discusses the legitimacy of rectifying gender-​based disadvantages.

18.3  Gender, Justice, and Institutional Change Many of the disadvantages that accrue to people in virtue of being female (or male) result from interactions in the private sphere, within institutions such as the family or economic enterprises. If one believes that justice does not regulate individual behaviour it is difficult to classify such disadvantages as forms of injustice and to attempt to legitimately rectify them through policy. According to Rawls (1971), the just­ ice of a society is determined by the rules that regulate its basic structure, which is society’s major political, economic, and social institutions. These institutions—​rather than individual behaviour within voluntary associations—​are the proper site of justice. It is not clear whether interactions within the family and the workplace come

Gender   403 under the purview of justice; for instance, is it unjust for parents to encourage their children to internalize gender norms (Macleod 2018)? And is it unjust for spouses to engage in a gendered division of labour (Neufeld 2009; Schouten 2013)? A main aim of feminist thinking about justice is to criticize the belief that private interactions are outside the scope of justice and that policies attempting to regulate them are therefore illegitimate. Much of this work engages with Rawls’s thought, and most feminist proposals to regulate private interactions in the name of gender justice concern at once family life and working life. These proposals are about policies promoting particular kinds of parental leaves, flexible work schedules for parents, and subsidized, high-​ quality care for small children.

18.3.1 Gender and Justice in the Family Systematic unjustified inequalities can arise between different groups of individuals (for instance, physically attractive people may have better chances on the job market). Some of these are particularly troubling because, like in the case of ethnicity or gender, they result from a long history of active discrimination and their endurance may indicate bigotry. But gender poses a unique challenge to distributive justice because a large number of women and men, perhaps a majority, want to share their lives substantially with one or more members of the opposite sex—​that is, set up house and raise children with them, and allow the kind of intimacy which, over time, shapes one’s self as well as one’s life. In liberal societies women are legally free to not enter, and to exit, relationships with any particular man. But the subjective costs of foregoing all close relationships with a member of the opposite sex, and in particular foregoing co-​parenting in the context of a heterosexual relationship, are very high. This makes it unreasonable to expect most women to lead, as a price for their liberation, a life devoid of close relationships with men. In a sense, women and men are locked in mutual relationships. This defines a special negotiation dynamic between women and men, and imbalances of power may easily get reproduced and magnified over time. The expectation that women perform most of the housework, including childrearing, significantly weakens their position on the market—​especially after they have children. Yet, as long as this expectation is widespread, it affects women’s chances to find partners who are ready to engage in an egalitarian division of labour. Moreover, as discussed in subsection 18.3.2, many existing policies and institutions encourage the members of a couple to specialize in either care-​giving or bread-earning; and the biological fact that women bear and breastfeed children, together with men’s typically higher earning power, often leads couples into a traditional division of labour, even when its members endorse an egalitarian division of labour (Gerson 2010). Another special issue of gender justice raised by the family is that most children are being brought up within (more or less stable) heterosexual relationships. Through processes of socialization children are likely to internalize and then reproduce

404   Anca Gheaus imbalances of power between women and men. A traditional gendered division of labour makes these imbalances very obvious: because women’s position on the market is worsened by their taking on caring responsibilities, they are less likely to economically afford to exit the relationship with their partner, absent special regulations to enable them to do so. As Susan Moller Okin noticed, women engaged in unequal marriages suffer disadvantage not only while married but also if they divorce, in which case women and their resident children, but not men, tend to be worse off economically (Okin 1989). These disadvantages, in turn, undermine women’s negotiation power within the couple and can generate everyday subordination. To what extent women’s direct involvement in care-​giving can also be a source of negotiation power over one’s (male) spouse is an interesting question, not yet explored by political theorists. Such power may stem from men’s interest in continuing to be an involved parent and it depends on legal provisions for deciding custody and visitation rights after divorce. For these reasons, the family—​or, at least, the heterosexual family—​is a main site of gender (in)justice. Because the ideal family is a cooperative venture freely entered into, grounded in affective ties, and supposed to last for a lifetime, it is easy to assume a unity of interest amongst its members and thus to bracket the question of its internal justice. Classical liberals as well as Rawls (1971) assumed that justice only regulates the public relationships between (heads of) families, not relationships between family members. But this seems wrong. First, as already mentioned, the subjective costs of avoiding all heterosexual family life are, for most individuals, very high, which qualifies the voluntary nature of a decision to avoid all heterosexual family life.6 Second, the existence of affective ties does not preclude conflicts of interest between family members: if men have an opportunity to combine a career and family life, women must have the same opportunity to do so. And if it is unavoidable that one parent makes more career compromises in order to raise children than the other parent, then combining a career and family life is a scarce good to which both female and male spouses have a claim. Finally, affective ties wither, and many families disintegrate, and therefore it is important to ensure that participation in the family does not systematically leave one group of people worse off than others in case of divorce. All these considerations, distilled in the slogan that ‘the private is political’, ground the conclusion that a credible theory of distributive justice cannot leave out the family. To what extent can the Rawlsian framework of thinking about justice accommodate this feminist insight? Or, in the words of Susan Lloyd, ‘in what way must families be just in order for society to be just’ (Lloyd 1995)? It is not difficult to show that the family is part of the basic structure, that is an institution which distributes the main benefits and burdens of social cooperation—​for instance, through the pervasive influence it has on people’s opportunities. In addition, both Rawls and Susan Moller Okin (1989)—​his main feminist critic—​agree that the family is the

6 

At least on the view according to which a choice is voluntary if and only if it is not only or mostly motivated by a lack of acceptable alternatives (Olsaretti 2004).

Gender   405 context where children—that is, future citizens—form a sense of justice. This last fact may be enough to rule out some forms of injustice within the family as illegitimate. For instance, the denial of basic freedoms to women, and perhaps gender (or other) hierarchies that are enduring and humiliating are inconsistent with cultivating a sense of justice in the children who are reared within such families. Moreover, as Elizabeth Brake has recently noted, such overt forms of sexism can undermine girls’ access to self-​respect, which Rawls himself deemed the most important of the primary goods (Brake 2015). However, recognizing the family as part of the basic structure does not entail that it has to be internally structured by (the Rawlsian) principles of justice—​for instance, that it has to observe fair equality of opportunity. Okin (1994) suggests that a sexist upbringing can interfere with children’s acquisition of a sense of justice to the point of threatening the stability of a well-​ordered society. Yet it is implausible that all kinds of sexism can threaten stability and therefore that no forms of sexism should be tolerated within the family. In particular, the traditional gendered division of labour is not as such an overt form of injustice that can undermine children’s moral development. This is in part because—​as I go on to explain—​individual endorsement of gender norms is not necessary to explain why the traditional gendered division of labour endures. Sometimes it is a rational response to institutional and biological constraints; at other times it is genuinely and equally conductive to both spouses’ well-​being.7 But even when it is a result, and an expression, of sexist socialization of women and men, the endorsement of a traditional division of labour is one among the many conceptions of the good life over which citizens can reasonably disagree. According to Rawls’s later thinking about just­ ice, principles of justice and state interference cannot be justified by appeal to any particular, contestable, conception of the good life (Rawls 1993). Rawls’ response to Okin’s criticism was that, to ban all sexism from the family on pain of injustice, one would have to show that children who grow up in sexist families are thereby unable to choose a feminist, or gender-​egalitarian lifestyle. Individuals brought up in sexist families are likely to find it psychologically more difficult to pursue a non-​sexist conception of the good as an adult. But this observation pertains to the adoption of any conception of the good inimical to that in which one has not been socialized (Lloyd 1995).8 The only way out of this problem would be to show that sexism in general—​or at least norms encouraging the traditional gendered division of labour—​cannot be part of a reasonable conception of the good. It is more plausible that, just like other institutions that are part of the basic structure, the family need not have its internal workings regulated by the principles of just­ ice. Recognizing the family as part of the basic structure does not entail that it has to be 7 

Even then it may be objectionable because the aggregated effect of many couples engaging in it perpetuates gender norms and makes it more costly for individuals to choose gender-​egalitarian arrangements. 8  Some believe that Rawlsian justice bans the enrolment of children in any conception of the good (Clayton 2018) (Chapter 20 in this volume)

406   Anca Gheaus internally structured by (the Rawlsian) principles of justice—​for instance, that it has to observe fair equality of opportunity. It seems legitimate for families to make decisions about how to distribute goods within families in ways that fail to observe the difference principle. An anti-​perfectionist theory of justice may be unable to accommodate the feminist criticism of the gendered division of labour in the family as a matter of just­ ice and as an object of legitimate state intervention. All of the likely harms it entails—​ the socialization of children into gender norms, the way in which it affects women’s opportunities on the labour market, the unequal negotiating power between spouses—​ are disputed, qua harms, by some reasonable members of society. Better allies to the feminist project may have to be sought amongst theories of justice that (a) consider individual behaviour a proper site of justice; and/​or (b) can account for the injustice of these harms in terms of how they impact on individuals’ opportunity to welfare, access to advantage, resources, or capabilities. Subsection 18.3.2 discusses prominent policy proposals meant to address gender injustices on the assumption that such interventions can be legitimate.

18.3.2  Gender, Justice, and the Life–​Work Balance Even if the family was a fully voluntary association, its interface with the wide society is shaped directly by coercive legislation governing marriage, divorce, inheritance, and tax law; on the other side of the interface, welfare policies regarding safety nets for the worst-​ off, institutional childcare, and employment regulation all function on the assumption that domestic and care work is largely performed informally by women (Neufeld 2009). Some feminists think that marriage itself is incompatible with gender justice; others have recently argued for retaining marriage but radically reforming it (Brake 2010). I leave this matter largely aside since I assume that, even if marriage was abolished or radically reformed, women and men would continue to co-​parent and co-​habit, thus perpetuating some kind of division of labour amongst them. Married or not, women who dedicate too much time to the provision of domestic services and care at home are economically vulnerable in case of separation; as a solution, Okin (1989) suggested that both spouses should have a legal entitlement to their combined earnings, by having employers pay split salaries to the employee and their partner who provides free domestic services. One worry with proposals like Okin’s ‘housewife wage’ is that they encourage women to stay out of the job market.9 This deprives them of the non-​monetary advantages of being employed, such as opportunities for meaningful work outside the home, for social respect, for a proper social life, and for political participation. Clearly, this objection has more force in the case of better-​off women who could have real opportunities for such goods in the first place. Women’s lower participation in the labour market is also

9 

The payment of a universal basic income to all citizens is another policy suspected to have a similar effect (Robeyns 2001).

Gender   407 likely to trap them in their gender roles, if only by entrenching statistical discrimination against women on the labour market (although the introduction of generous ‘housewife wages’ may allow those women who prefer to entirely avoid the labour market to do so, making statistical discrimination less rational). Other feminists think that social policy should instead make it possible for women to compete on the market on equal terms with men. This can be achieved, for instance, through widescale early, state-subsidized,10 socialization of childcare (Bergmann 2004). To decide whether this institutional solution is just, one would first have to settle the contentious questions regarding what kind of institutional care, if any, is good for children. Whatever the answer, it is unlikely that it is desirable or possible to outsource all care work. Therefore, if women continue to do whatever care work will have to be done within the family, this solution is only partial. Another possible difficulty with this proposal is that it will go against the expressed and persistent preferences of a significant group of women, who want to work on a part-​time basis and spend the rest of their working time in the household (Hakim 2000). Policies incentivizing women’s and men’s equal participation in the labour market therefore seems to sponsor a particular, contestable conception of the good—​in which case it could be, at least by liberal non-​perfectionist standards, illegitimate. Finally, some think that the best way to bring about gender justice is to eliminate the institutional incentives that maintain the traditional gendered division of labour (Fraser 1994; Williams 2000). They argue that at present the workplace is structured by the assumption that an ideal worker—​the one to be employed in the best jobs—​is free from caring responsibilities, including childcare, and free from domestic work: that is, an individual with a partner who takes on this kind of work. The solution they see involves a structural transformation of both the workplace and childcare institutions (Gornik and Meyers 2003), one that would make it possible for both spouses to combine similarly good jobs and childcare. This is what Nancy Fraser calls the model of the universal care-​giver (Fraser 1994) and would require a restructuring of the labour market to include flexible working hours and special leaves allowing parents to look after their sick children, the creation of part-​time jobs that pay the same pro rata wages as full-​time jobs (and come with similar benefits), and generous and gender-​egalitarian parental leaves, as well as adequate and affordable childcare. Leaving aside questions of feasibility, there are several matters of desirability concerning this proposal, too. The largest is about the legitimacy of state interference with the structure of jobs. Then, like in the case of other proposals, there is the worry that a liberal neutral state ought not to sponsor particular views of the good life, such as a gender-egalitarian view. Moreover, allowing parents to work flexible hours can raise the objection that it discriminates against non-​parents who also have an interest in a more harmonious combination of working lives and life outside the workplace. This is a

10 

If justice requires all members of society to pay for childcare, such policies ought to be funded by universal taxes; if not, by taxing only parents.

408   Anca Gheaus complaint of justice if one understands childrearing to be a lifestyle amongst others, for which parents are not entitled to special support (Casal and Williams 2004; Rakovsky 1991; Vallentyne 2002). That child-​rearing is a mere lifestyle is of course contested (Olsaretti, 2013; Gheaus, 2015). Debates surrounding parental leaves in particular illustrate the difficulties of compensating women for their unpaid care work while at the same time undoing the traditional gendered division of labour. They also illustrate the difficulties of realizing gender justice by legitimate means—​that is, without failing to show respect to individual freedom and state neutrality. A first issue, having to do with the length of the leaves, concerns their impact on women’s participation in the labour market. The choice is between generous (long and paid) parental leaves and less generous leaves (only long enough to allow women to recover properly from childbirth plus a few additional months). The first are recommended by gender justice because they acknowledge and support women who look after children in their first year or first few years of life; however, these very features of long leaves entrench the gendered division of labour: they help perpetuate gender norms and keep women away from paid jobs for years. The second is women-​friendly because it creates less disincentive for employers to hire women, and in general it is less likely to keep women out of the job market for long. The first type of policy makes it easier for women to combine career and family life (given the current structure of jobs and a shortage of affordable quality childcare); the second makes it easier for them to compete on equal terms with men in their career. Yet, for reasons already explained, it is not clear that either of these proposals can be legitimized by appeal to gender justice. A second issue concerns the impact of parental leaves on the dismantling of gender norms and has to do with who takes them up. Maternal leaves, which exclude fathers, are obviously unjust. But experience shows that gender-​neutral parental leaves are disproportionately taken by women. Some countries try to combat this by making at least part of the leave non-​transferable to the other parent (the ‘use-it-or-lose-it’ model) and some theorists proposed to take a step further and to make the availability of the leave to one parent dependent on the other parent taking his or her leave (Brighouse and Wright 2008). Yet, in cases of couples that lead gender-​traditional lifestyles this proposal could backfire: fathers could refuse to take the leave, forcing mothers to quit their jobs. Moreover, if parents are owed parental leaves, it is unclear that conditioning one spouse’s entitlements on the choice of the other spouse is consistent with respect for individual rights (Gheaus and Robeyns 2011). In any case, men’s uptake of parental leaves on a par with women does not guarantee that the time of the leave will actually be spent on doing childcare. It seems that efforts to dismantle the effects of the traditional gendered division of labour by institutional means will, to some extent, run against the difficulty of entrenched gender norms. This leads some theorists to think that at least some responsibility for change lies with individuals.

Gender   409

18.3.3  Is There a Duty to Resist Gender Injustice by Individual Behaviour? According to Ann Cudd (2006) the oppressed have a duty to resist their own oppression. In the case of gender-​related injustices, this means that women ought not to collaborate in reproducing the gendered division of labour, because it entrenches gender stereotypes. They ought to engage in household strikes: that is, refuse to be housewives and instead participate in the labour market (Cudd 1998). The question of what individual duties people have to change the unjust circumstances in which they live has been attracting increasing attention, and much work remains to be done on this topic in general and with respect to gender justice in particular. Here are a few problems raised by Cudd’s suggestions. First, there is the usual concern that individuals cannot have duties to do what they cannot do and, since the change of gender norms and behaviour is only achievable through collective action, separate individuals cannot have duties in this respect before they start coordinating with each other. Cudd acknowledges this problem, but thinks that women who do not resist their own oppression are nevertheless guilty of harming other women by strengthening gender norms. It is not clear this line of reasoning is successful, unless perhaps one can show that every instance of women’s resisting gender norms contributes something to the advancement of gender egalitarianism: for instance, women leading gender-egalitarian lifestyles may represent good role models for other women and men. A second worry—​which Cudd’s theory may be able to accommodate—​is that, in many women’s case, a duty to resist the traditional gendered division of labour seems overly demanding. Presumably, some women who choose to be full-time home-​ makers could only take jobs that are themselves exploitative, highly stressful, and meaningless. In comparison, being a home-​maker, especially if one raises children, can be a meaningful and fulfilling, even when hard and exploitative, type of work (Baker 2008). Gender injustices intersect with other kinds of injustice. Would it not be more plausible to expect only well-​off individuals—​that is, women and men who do not suffer, at the same time, injustices related to their class, ethnicity, or disability—​to take on the burdens of undermining gender norms? Third, a duty to refuse to do more than one’s fair share of household work must depend on how it impacts on innocent third parties, especially parties towards whom the ‘striker’ has duties of care. For instance, striking becomes problematic once children enter the picture. A woman’s duty to participate in the labour market may be incompatible with her duty to provide adequate care to her children in cases when her partner refuses to engage in domestic and care work (and so she has to pick up his slack.) This, again, makes the duty to hold a job appear overly demanding.

410   Anca Gheaus

18.4  The New ‘The Personal Is Political’: Implicit Bias, Stereotype Threat Much gender injustice is relational in nature—​that is, it consists of relationships that give women less recognition for their work, or marginalize, exploit, or disempower them. These injustices have a distributive aspect—​to the extent that they suffer from them, women are worse off than men. Women enjoy certain important relational goods to a lesser extent than men do; this is the first distributive aspect of relational injustice. In addition, these relationships impact on women’s opportunities to various goods they may pursue—​for instance, career advancement or positions of authority and power—​ and therefore can be at the root of further unfair distributions. It is debatable whether feminists have traditionally been interested exclusively in intentional relational injustices; but in any case, they focussed on relational injustices that tend to spring from individuals’ conscious attitudes—​typically, from explicit, formal, or informal gender norms endorsed by both women and men. Recently, however, much attention has been devoted to gender norms that are not conscious and that are said to shape the evaluation of women in all walks of life (Staats and Patton 2013). They are personal in the sense that, being unconscious, they are never formal. At the same time they are political, because they have a pervasive impact on how women are being treated by others, how fairly they are being rewarded, and eventually on their economic, social, and political opportunities. Even bracketing the disadvantages that flow from these implicit norms, it is plausible that it is in itself a disadvantage to be the recipient of unconscious discrimination, provided one has a right to be perceived in a particular way. The latter is a contentious claim which connects, once again, gender justice to issues of recognition (Bird 2018). The most widely discussed mechanisms through which unconscious gender norms operate are implicit bias and stereotype threat. Implicit bias refers to the attitudes or stereotypes concerning other individuals that people hold unconsciously and which affect their thinking and behaviour. Not only is implicit bias unconscious: often its content is significantly different from the content of the explicit attitudes and beliefs held by that same individual. Implicit bias is robust, pervasive, and activated automatically (Staats and Patton 2013). Stereotype threat refers to people’s tendency to confirm negative stereotypical expectations that others typically have of them, especially in social contexts that make the stereotypes salient. The content of both implicit biases and stereotype threats seem to be highly gendered: as an example of the first, women’s achievements are evaluated as less impressive than men’s exact same achievements. As an example of the second, girls perform worse at maths if, just before an exam, they have been explicitly reminded of their gender. Some authors explain how these mechanisms are partly responsible for women’s failure to be recruited, retained, and promoted in some attractive jobs (Saul 2013).

Gender   411 If empirical research on implicit bias, stereotype threat, and gender is correct, it raises several fascinating issues concerning individuals’ duties of justice. Even those who believe that justice evaluates individual behaviour (and not merely laws and institutions) will want to say that someone can behave unjustly only to the extent to which they are responsible for their action. But it is difficult to attribute responsibility for unconscious beliefs. If implicit bias really is universal, does this mean that we all should assume that we are going to act on it, and be held responsible to minimize its influence on our behaviour as well as we can? We know that some procedures can block implicit bias or minimize its effects. For instance, justice might require that evaluators of exams or job applications do not know whether the applicant is a woman or a man (Holroyd 2012). And if we know that people are susceptible to stereotype threats related to their gender (or ethnicity), which is likely to negatively affect their performance, does this mean that we have a general duty to avoid activating stereotype threats? Is it objectionable, for instance, to praise a woman’s kindness just before she has to exercise a technical skill? These questions indicate new directions of research on gender justice. Finally, what, if anything, does the existence of implicit bias mean for appeals to state neutrality in discussions about gender justice? We believe that states ought to be neutral with respect to their citizens’ conceptions of the good and this means that they ought to allow explicit discrimination within voluntary associations to the extent to which individuals themselves endorse such discrimination. For example, if people endorsed a traditional division of labour within a couple, policies that nudge them into genderegalitarian roles could compromise state neutrality. But if we hold implicit biases which contravene the ideals we explicitly endorse, this means we are also discriminating against some individuals against our best judgement. It is not clear that the appeal to state neutrality delegitimizes political interventions meant to mitigate such discrimination. Assume that a very large proportion of the population explicitly rejects the gendered division of labour,11 yet perpetuates it via implicit biases and susceptibility to stereotype threats. It is an interesting topic for future research to consider what a neutral state should do in such cases.

Acknowledgements I am grateful to Daniela Cutas for comments on previous drafts, to James Lenman for a helpful discussion, and to Serena Olsaretti for much feedback and unfailing editorial support.

References Allen, A. (2008). ‘Rationalising Oppression’. Journal of Power 1(1): 51–​65. Arneson, R. (1998). ‘What Sort of Sexual Equality should Feminists Seek?’ Journal of Contemporary Legal Issues 9: 21–​36. 11 

There is some evidence that this is the case in the United States (Gerson 2010).

412   Anca Gheaus Arneson, R. (2000). ‘Luck Egalitarianism and Prioritarianism’. Ethics 110(2): 339–​49. Baker, J. (2008). ‘All Things Considered, Should Feminists Embrace Basic Income?’ Basic Income Studies 3(3): 1–​7. Baron-​Cohen, S. (2003). The Essential Difference. London: Allen Lane. Benatar, D. (2012). The Second Sexism:  Discrimination against Men and Boys. Oxford: Wiley-​Blackwell. Benjamin, J. (1988). The Bonds of Love:  Psychoanalysis, Feminism and the Problem of Domination. New York: Panthenon. Bergmann, B. (2004). ‘A Swedish‐Style Welfare State or Basic Income: Which should Have Priority?’ Politics and Society 32 (1): 107–​18. Bird, C. (2018). ‘The Theory and Politics of Recognition’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 235–55. Brake, E. (2010). Minimizing Marriage. Oxford: Oxford University Press. Brake, E. (2015). ‘Rereading Rawls on Self-​Respect: Feminism, Family Law, and the Social Bases of Self-​Respect’, in R. Abbey (ed.) Re-​Reading the Canon Series: Feminist Interpretations of Rawls. University Park, PENN: Penn State University Press, pp. 57–​74. Brighouse, H. and Swift, A. (2014). Family Values: The Ethics of Parent–​Child Relationships. Princeton, NJ and Oxford: Princeton University Press. Brighouse, H. and Wright, E. O. (2008). ‘Strong Gender Egalitarianism’. Politics and Society 36: 360–​72. Card, C. (1990). ‘Caring and Evil’. Hypatia 5(1): 101–​8. Casal, P. (2018). ‘Distributive Justice and Human Nature’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 259–82. Casal, P. and Williams, A. (2004). ‘Equality of Resources and Procreative Justice’, in J. Burley (ed.) Dworkin and His Critics. Malden, MA: Blackwell, pp. 150–​69. Chambers, C. (2004). ‘Are Breast Implants Better than Female Genital Mutilation? Autonomy, Gender Equality and Nussbaum’s Political Liberalism’. Critical Review of International Social and Political Philosophy 7(3): 1–​33. Clayton, M. (2018). ‘Education’, in Olsaretti, S. (ed.) The Oxford Handbook in Distributive Justice. Oxford: Oxford University Press, pp. 438–59. Cohen, G. A. (1997). ‘Where the Action Is: On the Site of Distributive Justice’. Philosophy and Public Affairs 26(1): 3–​30. Cudd, A. (1998). ‘Strikes, Housework, and the Moral Obligation to Resist’. Journal of Social Philosophy 29(2): 20–​36. Cudd, A. (2006). Analysing Oppression. New York: Oxford University Press. Fine, C. (2010). Delusions of Gender:  How our Minds, Society and Neurosexism Create Difference. New York and London: W. W. Norton & Co. Firestone, S. (1971). The Dialectic of Sex: The Case for Feminist Revolution. New York: Bantam Books. Fraser, N. (1994). ‘After the Family Wage: Gender Equity and the Welfare State’. Political Theory 22(4): 591–​618. Fraser, N. and Honneth, A. (2003). Redistribution or Recognition? A  Political-​Philosophical Exchange. London: Verso. Fricker, M. (2007). Epistemic Injustice:  Power and the Ethics of Knowing. Oxford:  Oxford University Press. Gerson, K. (2010). The Unfinished Revolution:  How a New Generation is Reshaping Family, Work, and Gender in America. New York: Oxford University Press.

Gender   413 Gheaus, A. (2009). ‘How Much of What Matters can We Redistribute? Love, Justice and Luck’. Hypatia 24(4): 63–​83. Gheaus, A. (2012). ‘Gender Justice’. Journal of Ethics and Social Philosophy 6: 1–​24. Gheaus, A. (2013). ‘The Feasibility Constraint on the Concept of Justice’. Philosophical Quarterly 63(252): 445–​64. Gheaus, A. (2015). ‘Could There Ever be a Duty to Have Children?’, in S. Hannan, S. Brennan, and R. Vernon (eds) Permissible Progeny: The Morality of Procreation and Parenting. Oxford: Oxford University Press, pp. 87–​106. Gheaus, A. and Robeyns, I. (2011). ‘Equality-​Promoting Parental Leave’. Journal of Social Philosophy 42(2): 173–​91. Gilligan, C. (1982). In a Different Voice:  Psychological Theory and Women’s Development. Cambridge, MA: Harvard University Press. Gornick, J. C. and Meyers, M. K. (2003). Families That Work: Policies for Reconciling Parenthood and Employment. New York: Russel Sage Foundation. Hakim, C. (2000). Work–Lifestyle Choices in the 21st Century:  Preference Theory. Oxford: Oxford University Press. Hakim, C. (2010). ‘(How) can Social Policy and Fiscal Policy Recognise Unpaid Family Work?’ Renewal 18(1/​2): 23–​34. Haslanger, S. (2000). ‘Gender and Race:  (What) are They? (What) Do We Want Them to Be?’ Noûs 34(1): 31–​55. Held, V. (1993). Feminist Morality:  Transforming Culture, Society and Politics Chicago, IL: University of Chicago Press. Held, V. (2018). ‘The Ethics of Care’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 213–34. Hochschild, A. R. with Machung, A. (1989). The Second Shift:  Working Parents and the Revolution at Home. New York: Viking Penguin. Holroyd, J. (2012). ‘Responsibility for Implicit Bias’. Journal of Social Philosophy 43(3): 274–​306. Kittay, E. F. (1999). Love’s Labor:  Essays on Women, Equality, and Dependency. New  York: Routledge. Lloyd, S. A. (1995). ‘Situating a Feminist Criticism of John Rawls’s Political Liberalism’. Loyola of Los Angeles Law Review 28: 1319–​44. Lynch, K, Baker, J., and Lyons, M. (2009). Affective Equality: Love, Care and Injustice. London: Palgrave. Macleod, C. (2018). ‘The  Family’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 415–37. Mason, A. (2000). ‘Equality, Personal Responsibility, and Gender Socialisation’. Proceedings of the Aristotelian Society 100(3): 227–​46. Mill, J. S. [1869] 1973. ‘The Subjection of Women’, in A. S. Rossi (ed.) Essays on Sex Equality. Chicago, IL: University of Chicago Press, pp. 123–​242. Neufeld, B. (2009). ‘Coercion, the Basic Structure, and the Family’. Journal of Social Philosophy 40: 37–​54. Okin, S. M. (1989). Justice, Gender, and the Family. New York: Basic Books. Okin, S. M. (1994). ‘Political Liberalism, Justice, and Gender’. Ethics 105(1): 23–​43. Olsaretti, S. (2004). Liberty, Desert and the Market. Cambridge: Cambridge University Press. Olsaretti, S. (2009). ‘Responsibility and the Consequences of Choice’. Proceedings of Aristotelian Society CIX, 165–​89. Olsaretti, S. (2013). ‘Children as Public Goods?’ Philosophy and Public Affairs 41(3): 226–​58.

414   Anca Gheaus Pinker, S. (2008). The Sexual Paradox: Extreme Men, Gifted Women and the Real Gender Gap. New York: Scribner. Radcliffe-​Richards, J. (2014). ‘Only X%: The Problem of Sex Equality’. Journal of Practical Ethics 2(1): 44–​67. Rakowski, E. (1991). The Price of Justice. Oxford: Clarendon Press. Rawls, J. (1971). A Theory of Justice. Oxford: Oxford University Press. Rawls, J. (1993). Political Liberalism. New York: Columbia University Press. Robeyns, I. (2001). ‘Will a Basic Income do Justice to Women?’  Analyse und Kritik 22(2): 88–​105. Robeyns, I. (2007). ‘When will Society be Gender Just?’, in J. Browne (ed.) The Future of Gender. Cambridge: Cambridge University Press, pp. 54–​74. Ruddick, S. (1989). Maternal Thinking: Towards a Politics of Peace. Boston, MA: Beacon Press. Saul, J. (2006). ‘Philosophical Analysis and Social Kinds: Gender and Race’. Proceedings of the Aristotelian Society Suppl. vol. 80: 119–​44. Saul, J. (2013). ‘Implicit Bias, Stereotype Threat, and Women in Philosophy’, in K. Hutchison and F. Jenkins (eds) Women in Philosophy:  What Needs to Change. New  York:  Oxford University Press. Schouten, G. (2013). ‘Restricting Justice: Political Interventions in the Home and in the Market’. Philosophy and Public Affairs 41(4): 357–​88. Staats, C. and Patton, C. (2013). Implicit Bias Review 2013. Kirwan Institute for the Study of Race and Ethnicity, Ohio State University. United Nations (2010). The World’s Women: Trend and Statistics, 5th ed. New York, NY: The United Nations. United Nations Development Programme (2014). ‘Gender Inequality Index’. http://​hdr.undp. org/en/composite/GII (accessed 18 November 2017). Valian, V. (1999). Why So Slow? The Advancement of Women. Cambridge, MA: MIT Press. Vallentyne, P. (2002). ‘Equality and the Duties of Procreators’, in D. Archard and C. Macleod (eds) The Moral and Political Status of Children. Oxford: Oxford University Press, pp. 195–211. Williams, J. (2000). Unbending Gender: Why Family and Work Conflict and What to Do about It. New York: Oxford University Press. Young, I. M. (1990). Justice and the Politics of Difference. Princeton, NJ: Princeton University Press.

Chapter 19

The Fam i ly Colin Macleod

The most influential contemporary theories of distributive justice have been developed and discussed with little sustained attention to the distinctive challenges that families and children pose for the articulation of a satisfactory comprehensive theory of justice. This neglect is largely attributable to a widely shared but problematic assumption that the principal problem a theory of distributive justice addresses concerns the distribution of the benefits and burdens of social cooperation amongst responsible adults. This assumption is reflected both in the kinds of topics that have received most attention by prominent theorists and in the theoretical devices and strategies commonly employed in the development and justification of principles of justice—​for example, hypothetical contracts, or  idealized markets. Indeed, the unsuitability of such argumentative strategies for capturing and illuminating the distinctive claims of children raises more general issues about the methodological adequacy of a great deal of contemporary theorizing about justice. Although few theorists would endorse Gauthier’s claim that “children and the disabled fall beyond the pale of morality tied to mutuality” (Gauthier 1986), a great many influential theorists assume that distributive justice primarily regulates the claims that mature agents have to the resources and opportunities created via complex forms of social cooperation. To a large degree, the family is conceived as an important voluntary association that many persons value, but which does not pose unique or especially pressing theoretical problems for distributive justice. The fact that children are non-​voluntary members of families, and that children’s access to resources and opportunities is highly influenced by family membership has not been explored in depth by leading theorists such as Rawls (1993, 1999, 2003), Dworkin (2000), Nozick (1974), and G. A. Cohen (1989), whose work dominates the philosophical discussion of distributive justice.1 1  Okin (1989) offers an important feminist critique of the neglect of the family by mainstream theorists but says surprisingly little about distributive matters. Some recent feminist work that addresses issues related to the distribution of responsibilities for caring for children include Kittay (1999), Robeyns (2010), Brake (2010).

416   Colin Macleod In what follows, I will consider ways in which contemporary theorists have begun to address issues about the family and distributive justice that many influential theorists have largely neglected. In one way or another, all the issues I raise bear upon the resources and opportunities to which adults and children have access, and the considerations that are relevant to determining how resources and opportunities should be distributed in and between families. Most of the discussion will take as granted that the family is an important social institution for the rearing and nurturing of children. As I will treat it, . . . the family is a social group comprised by one or more adults and one or more children who are linked together by a special history (e.g., as biological kin or adoption) and by sentiments of mutual affection. Some of the adults in such a group have a socially and legally recognized status as parents or guardians that confers upon them a special measure of authority over and responsibility for the children in the group. (Macleod 2002: 213–​14)

How families are internally structured as well as the recognition and the support they receive socially can affect the resources and opportunities to which both children and adults have access. It is for this reason that the family has great general significance from the point of view of distributive justice—​it can profoundly affect the life prospects of its members. But the existence of families also has implications for the distribution of resources and opportunities between families and between members of the community who elect not to form families in this sense. For example, the resources and opportunities to which childless couples have access will be influenced by the degree to which families and the opportunity to create a family are supported by the state. The overview I supply is in many ways selective. I do not address all the theoretical perspectives and issues that have received attention in the burgeoning literature in the field.2 Similarly, although I do not defend any comprehensive conception of justice, the characterization of various issues I provide does rest on a few abstract but substantive claims about justice. I assume, for instance, that justice requires recognizing the equal moral status of persons—​both adults and children. This in turn provides the impetus for a broadly egalitarian conception of justice in which: (a) the interests of all persons are given equal consideration; and (b) importance is attached to eliminating or mitigating the ill effects that morally arbitrary factors such as race, sex, and class have on persons’ access to important resources and opportunities. There is room for debate both about what equal consideration of people’s interests entails and how concern for the elimination of arbitrary sources of inequality is best understood. However, for the purposes of this chapter, I assume that debate about matters of justice concerns interpretation of these abstract egalitarian convictions rather than about the soundness of the convictions themselves (Kymlicka 2002: 3–​4). 2 

Some influential books and edited collections in the field include Archard (2004), Archard and Macleod (2002), Brighouse and Swift (2014), Dwyer (2006), Macedo and Young (2003), and Richards (2010).

The Family   417

19.1  Justice and Opportunities to Parent For many people creating a family and raising children is a profoundly valuable life project that assumes an importance in their lives that dwarfs other commitments. It often serves as the centerpiece of people’s lives around which other valued projects are oriented. The nature of the value for parents of assuming responsibility for nurturing children and exercising a measure of authority over them is complex, as is the value children can derive from their participation in family life. I will return to some issues that arise about characterization of the value of families in moment. But for the time being I  shall simply assume that many adults reasonably place a great deal of importance on having and rearing children. Although for many adults founding a family is a relatively straightforward matter of finding a partner, engaging in procreation, and assuming responsibility for raising the resulting offspring, natural and social conditions are not always hospitable to those who wish to parent. Some would-​be parents are in relationships in which at least one person is infertile, others are prevented by law or social custom from becoming parents. Until relatively recently, gay people in most states were legally barred from having families and openly raising children as parents. Even where gay marriage has been recognized along with the rights of gay couples to form families with the same custodial rights over children as heterosexual couples, homophobic prejudice persists and unfairly denies gay parents access to family life. Economic inequalities between adults also affect the opportunities that adults have to parent. Sexist social norms and practices (frequently rooted in influential religious and cultural beliefs) about the social division of child-​ rearing responsibilities are significant too. The opportunity to parent and enjoy the goods of family life is not relevantly equal between men and women, if women face dramatic trade-​offs in their career and other life aspirations by a decision to become a parent that men do not routinely face.3 Finally, some people who wish to parent may face obstacles rooted in their physical or mental disabilities that cannot be overcome without significant assistance from others.

19.1.1 Anti-​Natalism Against this background one general issue concerns what justice requires with respect to the distribution of the opportunity to become a parent. This issue has many facets. On strong anti-​natalist views according to which it is wrong for humans to procreate in the first place, the significance of securing the opportunity to parent is arguably not

3 

See Gheaus 2018 (Chapter 18 in this volume).

418   Colin Macleod a pressing demand of justice. If, as David Benatar maintains, “coming into existence is always a serious harm” (Benatar 2006: 1), then it would seem important to prevent or at least strongly discourage human procreation. And this would suggest that securing opportunities for adults to procreate and become parents is not something that justice requires. The strong anti-​natalist view is compatible with recognizing that adults strongly value parenting and it is also compatible with the observation that the distribution of the opportunity to become a parent (e.g., through natural procreation) can be unfairly distributed. It may be unfair that fertile couples have an opportunity to procreate that infertile couples lack. Yet the significance of this unfairness will be diminished if acting on the opportunity to procreate is itself ethically dubious. There is a lively debate about the merits of anti-​natalism in general (Overall 2012) as well as more fine-​grained debate on more specific cases in which procreation may be objectionable (Shiffrin 1999).4 Important though these debates are, I will set them aside and proceed on the assumption that procreation is morally permissible in a wide variety of cases.

19.1.2 Opportunities to Procreate Although most procreation occurs without the need for elaborate medical intervention, there have been a variety of remarkable developments in reproductive technology (e.g., in vitro fertilization) that both address problems of infertility and that broaden the range of people who can have children to whom they are biologically related. These developments raise interesting questions about the significance of having a biological relation to one’s children as well as questions of distributive just­ ice concerning access to reproductive technology. It is implausible to think there is a necessary condition between realizing the value of parenting and being the genetic parent of the children in one’s family. Parents who adopt children can readily realize the goods of family life. Nonetheless, many people place importance on having children to whom they are biologically related (Velleman 2005). If this is a weighty interest and if it is important to eliminate arbitrary barriers between people with respect to having children who are their biological kin, then justice would seem to require that those unable to procreate ‘naturally’ have ready access to reproductive technologies. Securing such access would mitigate arbitrary differences between people in their opportunities to procreate. However, if little real value attaches to being a biological parent per se, then the importance of ensuring ready access to reproductive technologies is greatly diminished. Moreover, ensuring that people have ready access to reproductive technologies can have a deleterious effect on efforts to secure parents for the many children who lack parents.

4 

Shiffrin’s position is associated with strong anti-​natalism but she does not argue that procreation is always wrong (1999: 139).

The Family   419 The valorization of the biological families that support for reproductive technologies generates arguably creates a disincentive for couples wishing to have a family to adopt children (Ahlberg and Brighouse 2010). Given that children in need of adoption are especially vulnerable, it is important that we adopt social policies that make adoption feasible and attractive to potential parents. If, as seems plausible, adoption can be a viable route through which the opportunity to parent can be made available, then, at least in present circumstances in which many existing children lack parents, justice does not require that adults have a (roughly equal) opportunity to become biological parents. A successful objection to this claim would have to demonstrate that biological parenting is both a source of special value and that its importance can outweigh the claims of existing parentless children to have parents.5

19.2  Fair Access to Parenting As I have noted, the way in which the opportunity to parent is available to people can be affected by social and material circumstances. Parenting brings with it significant responsibilities and costs. Justice in the provision of this opportunity must be sensitive to the various ways in which access to the opportunity can be impeded or is available to some only on unfair terms. I have already noted the way in which the gendered division of labour in families and homophobia can unfairly impede access to the opportunity to parent. On the assumption that justice requires eliminating or at least significantly mitigating sources of disadvantage rooted in gender or sexual orientation, justice requires reconfiguring social practices and institutions so as to ensure that the opportunity of women and gay people to parent is relevantly similar to the opportunities of men and heterosexual people. Resistance to this suggestion is sometimes advanced by appeal to the responsibilities parents have to nurturing their children properly. For example, it is sometimes alleged that special maternal care that infants need can only be supplied by women (Frailberg 1977). As a consequence, women, but not men, who choose to become parents must be prepared to limit their pursuit of other goods (e.g., rewarding but demanding careers). On this view, gendered inequality between parents that affects the character of the opportunity to parent is justified because it is rooted in biological differences in the capacities men and women have to meet the needs of their children (e.g., only women can breastfeed infants). In a different but related vein, some commentators think it is appropriate to deny or severely circumscribe the rights of samesex couples to parent on the grounds that children have a right to be raised by both a 5 

This does not mean that parents do not have a powerful claim to rear their biological offspring. Anca Gheaus argues that the intimate bonds that are formed between a pregnant woman and her fetus ground a strong (but defeasible) right to rear one’s biological offspring (Gheaus 2011). For further discussion of the significance of biological relatedness to the right to rear children, see Hall (1999), Haslanger (2009) and Liao (2006).

420   Colin Macleod mother and a father (Sommerville 2007). On this view, same-sex couples may be denied an equal opportunity to parent because same-sex parenting falls below the threshold of adequacy to which children are entitled. The particular claims about parental competency on which these arguments rest are highly dubious. While it is true that some biological differences between men and women strongly influence the provision of some facets of nurturing to infants and fetuses, the differences are relevant only for a relatively short period (e.g., during pregnancy and early infancy) and there is no reason why the many other facets of nurturing young children cannot be distributed between men and women so as to compensate for the fact that women are temporarily uniquely placed to meet some of the responsibilities of child rearing. Biology need not result in an overall inequality between sexes with respect to the character of the opportunity to parent. Similarly, despite its popularity in right-​wing political discourse, there is no evidence that same-sex couples, simply in virtue of being a couple composed of two people of the same sex, cannot love and nurture children properly (Woodcock 2009). More generally, there is no empirical evidence that same-sex parenting has an adverse impact on children’s well-​being (Crouch et al. 2014).

19.3  Parental Competency and the Right to Parent Misguided though these reasons for restricting opportunities to parent are, they do point to an important general issue about the relation between the opportunity to parent and competency to parent. It is widely recognized that the right to parent is conditional on would-​be parents satisfying a standard of adequacy. Abusive or grossly negligent adults can lose or be denied the right to parent on the grounds that they are unable or unwilling to protect and suitably promote the basic interests of children. So distribution of the opportunity to parent is sensitive to some considerations of competency. Establishing the relation between parental competency and the right to parent has at least three dimensions: (a) how to identify the elements of parental competency; (b) how to ensure that parents are competent; and (c) the nature of the justificatory basis of the right to parent. First, there is the question of what parental competency consists in. This is a complex matter, both because there is disagreement about what duties parents owe their children and because whether or not would-​be parents can satisfy a standard of competency will depend not only on their own attributes and capacities but also on the social division of moral labour in society in which they live. Although it is clear that competent parents must not physically abuse or grossly neglect their children, there can be controversies about whether certain practices (e.g., circumcision, corporal punishment) constitute abuse or neglect. Good parenting obviously requires more than refraining from abuse

The Family   421 and neglect, so it seems reasonable to set the threshold of competency requisite to having the right to parent at a higher level. One way to address the issue is to set a standard of competency that tracks the justice-​based entitlements of children according to which competent parents must be able to safeguard their children’s basic rights and furnish them with the resources and opportunities to which, as a matter of justice, they have a right (Macleod 2015). So if children have a right to be loved (Liao 2006) or a right to an autonomy-​facilitating upbringing6 (Brighouse 2000, Clayton 2006, Levinson 1999, Macleod 1997), then parents who cannot or will not love children or seek to indoctrinate children will not have a right to parent. Matters are complicated, however, because whether or not would-​be parents fulfil a standard of parental competency grounded in the entitlements of children will depend partly on the background division of moral labour. So a satisfactory account of the relationship of parental competency to the right to parent cannot focus solely on the attributes of would-​be parents. To see this, consider a simple example. Suppose, as seems plausible, that children have a right to education and healthcare. In a society in which responsibility for meeting the basic educational and healthcare needs of their children is assigned solely to parents, poor adults who are unable to afford adequate education or healthcare for children will not be considered competent because they will be unable to reliably secure adequate education and healthcare for their children. Similarly, the presence or absence of social support for people with disabilities can determine whether or not disabled persons qualify as competent parents. The point here is that the character of background social and economic institutions can influence whether people are able to meet the needs of children in their care and hence whether they can be considered competent. If the opportunity to parent is extremely valuable, then we have reasons of justice to ensure that institutions are arranged so as to ensure that those who wish to parent can readily do so in a way that is compatible with meeting the needs of children. To some important degree, the character of social institutions can enable parental competency both by ensuring that children have adequate access to crucial resources and opportunities and by helping parents develop good parenting skills. For instance, the opportunity to parent competently will be more widely available if there is public provision of excellent healthcare and education and if workplaces are sensitive to the needs of parents (e.g., through policies of parental leave and subsidized daycare). Similarly, it makes sense to have educational programs that provide information about good parenting and help people develop the skills and traits requisite to good parenting. In short, the value that is assigned to the opportunity to parent can have significant implications for the distribution of resources in society. A second issue related to parental competency and the right to parent concerns the manner in which a just state should try to ensure that parents fulfil the requisite criteria of competency. What may the state do in order to determine whether parents or would-​be parents are competent? LaFollette (1980, 2010; see also McFall 2009) has 6 

See Clayton 2018 (Chapter 20 in this volume).

422   Colin Macleod argued that since parenting is an activity that poses significant risks to children and since bad parenting imposes enormous costs on society, it is appropriate for states to license parents. Advocates of licensing think that a well-​designed scheme could identify some people who clearly lack the attributes necessary for successful parenting and could filter out many adults who are likely to be egregiously bad parents. This is a controversial proposal. There are concerns about whether a licensing program would reliably screen out bad parents and whether other less intrusive social policies aimed at facilitating competent parenting are more effective (De Wispelaere and Weinstock 2012). Yet it is routinely assumed that it is reasonable to subject adoptive parents to extensive and expensive screening in the name of ensuring that they meet some threshold of competency. Opponents of parental licensing who nonetheless favour the screening of adoptive parents must explain why such asymmetrical treatment of potential parents is justified (Botterell and McLeod 2014). A third less practical but important theoretical problem concerns the justificatory basis of the right to parent. The right to parent is an unusual right since it involves the right to exercise authority over another person. Most theorists accept that the right must be grounded in important interests it serves. But whose interests are relevant to justifying the right to parent? An important issue here is whether the justification of the right to parent is wholly grounded in the interests of children or whether independent interests that adults have in becoming parents play a role in shaping and grounding the right to parent. On so-​called ‘child-​centered’ views, the right to parent is grounded in the needs that children have to protection and promotion of their interests (Hannan and Vernon 2008). As vulnerable and highly dependent beings, children are not able to secure their own interests reliably and they need to be assigned to guardians who are able and motivated to promote children’s interests. A common interpretation of this view holds that children’s interests are paramount in the sense that parents have a fiduciary responsibility to promote the best interests of children under their care. However, if the role of parent is grounded solely in promotion of children’s interests, then it seems that children should be assigned to the parents who are best able to promote their interests. In principle this means that adults who would be good parents might be denied the opportunity to parent children because other adults would be better parents. The child-​centered view can acknowledge that adults have an interest in being parents and it is compatible with furnishing people with the resources and education requisite to being excellent parents. However, it holds that adults may be legitimately denied the opportunity to parent simply in virtue of not being the best available parents for particular children. So the child-​centered view seems vulnerable to the objection that it would authorize the redistribution of children from good parents to better parents. Dual-interest views offer a way of resisting this awkward implication by holding that the interests that adults have in being parents play a role in grounding and shaping the character of the right to parent (Brighouse and Swift 2006; Clayton 2006; Macleod 2015). Dual-interest views do not deny that parents have stringent duties to protect and promote their children’s interests. So the right to parent is conditional on would-​be parents

The Family   423 being able to nurture children adequately. The standard of parental adequacy can be high but does not require that parents be the best. On dual-interest views, the goods of parenting for parents are enormously valuable and the important of facilitating access to these goods by would-​be adults helps both to justify the right to parent and to delineate some of the prerogatives of the right to parent. Development of a dual-interest view naturally focuses on the distinctive value of relationships between parents and children in families. Brighouse and Swift give special emphasis to the unique and highly valuable character of intimacy between parents and children. The close affective ties that parents have with their young children are special and powerful (Brighouse and Swift 2006; see also Schoeman 1980). Caring for, loving, and being loved by a child to whom one has special nurturing responsibilities as well as a complex relationship of authority and respect is, for Brighouse and Swift, a profound and unique form of human flourishing. For many, perhaps most adults, the opportunity to realize this kind of flourishing matters immensely and it provides us with a reason to extend the right to parent to all competent would-​be parents. In addition to the interest in intimacy, some theorists locate the special value of parenting in the powerful interest we have in expressing our conception of the good to other people and having them appreciate and perhaps participate in it with us. Macleod calls this an interest in creative self-​extension (Macleod 2012, 2014). Because children initially lack their own conception of the good and are receptive to the ideals and values of those who raise them, parenting presents especially rich but complex opportunities for creative self-​extension. Parents have the opportunity to shape the identity of children by sharing valued commitments, ideals, and projects with their children. Thus parents often seek to include their children in specific religious or cultural practices and thereby to express their own values and enthusiasms with their children. Such activities are bound up with intimacy, but arguably they have an importance to parents that is distinguishable and independent of the value of intimacy.7 Recognizing the importance of creative self-​extension supplements the intimacy-​ based rationale for the right to parent and helps explain why the right to parent is widely thought to include special parental prerogatives to privilege their conceptions of the good (e.g., religious or cultural traditions) in rearing children. It also amplifies the value and importance of the opportunity to parent. How strong parental prerogatives are to promote a particular conception of the good is itself a matter of some theoretical and practical controversy. If, as many theorists argue, children have a right to an autonomy-​ facilitating upbringing, such prerogatives must be exercised in a way consistent with the acquisition of meaningful autonomy by children. Resolution of controversies in this domain depend on a satisfactory articulation of the kind of autonomy to which children have a right, as well as an account of the social and educational circumstances conducive 7  Some dual interest theorists reject the suggestion that parental interest in sharing cultural or religious traditions with children can ground any facet of the right to parent. See Clayton (2006), Brighouse and Swift (2014).

424   Colin Macleod to the acquisition of autonomy (Callan 2002; Clayton 2006; Mullin 2007; Bou-​Habib and Olsaretti 2014). Ensuring that children are furnished with the circumstances conducive to autonomy acquisition will in turn have implications for the distribution of resources and opportunities.

19.4  Distributing the Costs of Raising Children So far we have seen why the opportunity to form a family and raise children is highly valuable, and how social and material circumstances can generate unfair inequalities in access that people have to this opportunity. Ensuring that people have fair access to the opportunity to parent requires policies both that mitigate unfair obstacles to becoming a parent and that adequately prepare would-​be parents for the demanding but rewarding project of raising a family. As valuable as parenting can be it is arguably not a necessary part of a flourishing human life. There are valuable life projects that do not involve raising children. People interested in pursuing other valuable projects have an interest both in having opportunities to do so and to having a fair share of resources with which to pursue their projects. On many accounts of distributive justice the share of resources one has to devote to one’s valued projects can be determined without evaluating the relative value of the projects that people wish to pursue. Rawls’s theory, for instance, focuses on the distribution of social primary goods that are assumed to have value for people irrespective of the particular conception of the good they hold. Thus the share of income to which people are entitled in a just society is not supposed to be influenced by consideration of the kinds of projects they wish to pursue. For instance, a lover of gourmet food who wishes to dine on caviar and drink champagne is arguably not entitled to a greater share of income than the person with more modest tastes simply on the ground that the gourmand requires a higher income in order to pursue his conception of the good successfully. Providing just background conditions obtain,8 people are expected to assume responsibility for the projects they chose to pursue and to bear the costs attendant to those projects. Justice does not require some people to subsidize the expensive tastes of others. Raising children is in many ways a costly project—​time, energy, and resources must be devoted to the care of children. In light of the concern that a person’s expensive tastes do not ground an entitlement to extra resources, it might be thought that justice requires those who voluntarily elect to have children to bear all the costs of raising them. Yet it is common for there to be general public provision of some of the resources parents need 8  The precise character of just background conditions in this context is open to debate. But they include the provision of resources and opportunities relevant to informed and reflective deliberation about the projects and commitments one might consider adopting and pursuing.

The Family   425 to raise children successfully. For example, the funding for education and healthcare of children is typically derived from schemes of taxation that do not distinguish between parents and non-​parents. Similarly, in many communities, daycare is publicly subsidized. These forms of public provision spread some of the costs of raising children to people who are not parents themselves, and thus it might seem that they unfairly deprive non-​parents of resources that they would prefer to devote to their own projects. This raises the question of how the costs of raising children should be shared between parents and non-​parents. Does justice require or permit non-​parents to bear some of the cost of raising children? For some theorists, the answer is clearly no. Rakowski (1993) and Vallentyne (2002) maintain that any and all costs that children create must be absorbed by the procreators who are responsible for bringing new persons into existence (see also Casal and Williams 1995). On their view, it is unfair for those who wish to have a family to expect others with different projects to underwrite, even partly, the project of raising children. The issue is complicated. To begin with, many of the resources that go to raising children are resources to which children have a justice-​based entitlement. So although it is true that children stand in a special normative relation to their parents, children are also independent members of the moral community whose claim to a fair share of resources is arguably not dependent on the capacity of their parents to secure it for them. This means that non-​parents must be prepared to bear some costs of raising children. If, for instance, a child’s parents are killed when the child is an infant, other members of the community must take responsibility for meeting the needs of the child even if none of them antecedently sought to have children. In this respect, the project of having children is very different from other projects: the creation of a child with rights, including rights to a fair share of resources, implicates non-​parents in the project because they must stand ready to assist with raising children if the parents are unable or unwilling to do so. The same does not hold true of most other projects: non-​religious people are not required to sustain religious rituals if the religious people who value them are unable to do so.9 Second, it is frequently difficult to distinguish sharply between those who are responsible for introducing children to the community and those who are not. People who voluntarily engage in procreation are the most obvious candidates to whom responsibility for creating children can be assigned, and thus it might seem that, at least in principle, costs of bearing children should be borne by procreators. However, in many contexts responsibility for bringing children into existence is more diffuse than the distinction between procreators and non-​procreators suggests. States and communities often actively encourage procreation and provide material and social incentives for people to have children. These policies and social norms 9  In this sort of case, the duty of non-​procreators to secure the entitlements of orphaned children might be characterized as duty of non-​ideal justice. Whether that is so or not is debatable, but the basic point is that children cannot be viewed merely as private projects of their parents. This means we should be wary of the suggestion that the project of having children can be viewed as an expensive taste.

426   Colin Macleod arguably reflect collective decisions to engage in social reproduction and they put pressure on a sharp distinction between procreators who must bear the costs of having children and non-​procreators who are free from such costs. To the degree that social reproduction via procreation is a collectively sanctioned project, it is appropriate for many of the costs associated with such a collective enterprise to be widely shared precisely because responsibility for the existence of children cannot be located solely in those who engage directly in procreation. This is not to deny that procreators may reasonably face extra burdens associated with rearing children that non-​procreators do not. But in many instances, non-​procreators may be reasonably expected to underwrite some costs of child rearing because they have actively encouraged others to procreate. The fact that children typically become productive members of communities further complicates the question of how the costs of rearing children should be distributed. One strategy for arguing that the costs of rearing children should be shared focuses on the idea that rearing children generates a public good from which non-​parents benefit. Children become customers and workers in the general economy in ways that confer benefits (e.g., taxation revenue) on parents and non-​parents alike. According a principle of fair play that gets invoked in this context, those who benefit from a public good created through the voluntary activity of others have a duty to contribute to the costs associated with production of the good (Rawls 1999: 93–​8). The public goods argument for sharing the costs of child rearing between parents and non-​parents has elicited a variety of responses. Some reject the principle of fair play on which it rests altogether. Others allow that fair play considerations can ground a requirement to share in the costs involved in the provision of public goods, but argue that the manner in which parents benefit non-​parents by having children does not satisfy the requirements of the fair-​play principle when properly interpreted. Casal and Williams (1995) argue that fairness only requires sharing the costs of rearing children if a costliness requirement and an intentionality requirement are met: becoming parents must be a net cost to parents, and parents must deliberately aim to benefit non-​parents by having children. Casal and Williams contend that both these conditions do not typically obtain, and hence the public goods argument fails. Olsaretti (2013) allows that Casal and Williams’ insistence on the costliness and intentionality requirements is reasonable but argues that their interpretation of these requirements is too stringent. On a looser and more plausible interpretation, the relevant requirements are typically met by those who deliberatively decide to have children. Nonetheless, Olsaretti rejects the public goods argument because she thinks that it rests on a mistaken assumption that the goods that children generate are genuine public goods in the technical economic sense. Olsaretti argues that the case for sharing the costs of child rearing between parents and non-​parents is more credibly grounded in the idea that the benefits that arise from having children are “socialized goods.” On this view, sharing the costs of rearing children rests on the (contingent) fact that social institutions have been deliberately structured so as to ensure that children create benefits for all members of society.

The Family   427

19.5  Justice-​Based Entitlements of Children Determining how the costs for raising children are to be distributed between parents and non-​parents is related to another, arguably more fundamental issue, of what the justice-​based entitlements of children are in the first place. Whether or not parents should bear all or most of the costs associated with rearing children, it is clear that parents alone cannot effectively meet all the needs of their children. In even a moderately complex society, there will be a division of labour involved in meeting the needs of children, and some non-​parents will be better placed to meet the needs of children than parents. Thus to a large degree, the effective provision of education and healthcare to children is best secured by professional teachers and healthcare professionals rather than parents themselves.10 However, some important goods for children such as love and intimacy, as well as many opportunities for play and participation in cultural activities are best secured for children by families. Just what the most appropriate distribution of moral labour is depends both on the justice-​based entitlements of children to goods and the character of those goods themselves. For instance, the kind of love children need can only be provided by adults with whom children have a close, intimate, and trusting relationship. So it makes sense to assign the special responsibility of loving children to parents and family members. By contrast, if the acquisition of meaningful autonomy requires exposing children to diverse conceptions of the good and affording them the opportunity to reflect critically on different ideals and perspectives, then it is likely that important components of education must be provided by teachers who can offer a more impartial perspective on such matters than parents. The point here is not to settle the question of precisely who should assume the responsibility for meeting the different justice-​based claims of children. Rather I wish to point out that the appropriate division of moral labour will depend on the precise character of those claims, and that the overall structuring of social institutions needs to be sensitive to accommodation of those claims. We can turn now to consideration of the central issue of determining what the justice-​ based entitlements of children are. Two related challenges arise here. First, we must identify a suitable metric for assessing and comparing the benefits to which children have access. Second, we must ask what considerations should determine the kind of access that children have to valuable resources and opportunities. This is a matter of identifying

10  The point here does not concern who should bear the costs of securing the entitlements of children. Rather the point is to distinguish between entitlements of children that cannot be effectively secured by parents, and entitlements that parents are usually best placed to secure. Parents who fully pay for their children’s education bear the cost of meeting their children’s entitlement to education, but by hiring teachers they give recognition to the fact that (some) education is best provided by professional educators and not parents themselves.

428   Colin Macleod sound distributive principles and determining how the distinctive distributive claims of children interact with those of adults. This second task gives rise to important issues concerning the interpretation of equality, the domains in which distributive equality is relevant, and the relationship between equality and parental partiality.

19.5.1 A Metric of Distributive Justice As I noted at the outset, many contemporary theories of distributive justice assume that principles of distributive justice regulate the distribution of the benefits and burdens of mutually beneficial social cooperation amongst mature adults. However, even within this literature, controversy arises over what considerations are relevant to gauging and comparing the distributive shares of individuals. When, for the purposes of distributive justice, is one person more advantaged than another? What metric should we employ to gauge or measure relative advantage? Should justice be concerned with the distribution of welfare, resources, or capabilities?11 The choice of a metric affects the judgments of distributive justice we make. If, for instance, justice is ultimately concerned with the distribution of welfare and we adopt equality as a principle of just distribution, then a distribution in which income is equally distributed yet some people experience much higher levels of welfare than others will be unjust. If, by contrast, we adopt income as the metric, then the same distribution will fulfil the demands of egalitarian justice. A good deal of the mainstream debate between sponsors of different positions on the metric issue assumes that we need only articulate a metric suitable for comparing the shares or standing of mature adults. This assumption is problematic. Mature agents have the capacities to form and implement their own conception of the good and to take responsibility for the consequences of their choices about the conduct of their lives. The assumption that a theory of distributive justice is to be framed in relation to agents so conceived appears to provide a rationale for adopting a metric that does not speak adequately to children’s interests. Consider, for instance, Rawls’s focus on social primary goods such as income and basic political liberties as the metric of distributive justice. Primary goods are supposed to be valuable for persons irrespective of the particular conception of the good they have adopted and wish to pursue. Equipped with a fair share of primary goods, adults may determine how best to lead their lives according to their own lights.12 For children, however, access to generic resources such 11  See Fredman (2018), Arneson (2018), and Robeyns (2018) (Chapters 1, 2, and 5 in this volume, respectively) for further discussion of these different proposals. 12  The capabilities approach (Nussbaum 2006; Sen 1999) and welfarist metrics suffer from related difficulties. Capabilities are opportunities to achieve valuable functionings. The metric assumes that persons can responsibly choose whether or not to pursue certain functionings and thus most obviously applies to adults and not to children (Macleod 2010a). Welfarist metrics generally define welfare as consisting in preference satisfaction of rational mature agents. A preference satisfaction model of welfare cannot capture children’s welfare adequately, since children lack the mature cognitive and emotional capacities on which preference satisfaction accounts of welfare are predicated.

The Family   429 as income, basic liberties, and career opportunities is not very important per se.13 This is both because they lack the wherewithal to use these resources in a sensible fashion and because some of them cannot be meaningfully used by children to realize goods in their lives. Thus a child’s interest in leading a good life is not secured by providing them with an income to spend as they wish. Similarly, the right to hold political office is not a right exercise of which could be valuable to an infant. Precisely because children are not mature agents who can assume responsibility for forming and pursuing life plans as they see fit, a child-​sensitive metric of distributive justice must identify goods that have special salience for children. Some of the dimensions of such a metric are obvious enough. For instance, from the point of view of justice, it matters not whether children have an opportunity to access education but that they actually receive a good education. Whereas adults are generally in a position to determine for themselves what educational opportunities to pursue, decisions about the appropriate character of education have to be made on behalf of children. Similarly, we cannot assume that the full extension of basic liberties that serve adults’ interests to children constitutes a benefit to them that is significant from the point of view of justice. One natural way to address this issue is to link a child-​sensitive metric to the material and social circumstances conducive to the development of mature agency. A  metric developed in this way will have some obvious elements that are relevant to matters of justice affecting children. For example, access to educational and healthcare resources is salient partly because it contributes to the development of agency. Although some features of a child-​sensitive metric identified in this way closely parallel an adultoriented metric, that metric will also have features that typically do not figure in an adult-​oriented metric. Consider the example of love. Parental love plays a crucial role in the well-​being and normal psychological development of children, and arguably is important to development of the moral capacities that responsible adult citizens are expected to have (e.g., Rawls’s two moral powers). Given this connection, we can argue that justice must be concerned with the provision of love to children by their parents. This is because parental love helps to facilitate the development of mature agency. By contrast, ensuring that adults are loved—​either by their parents or others—​arguably does not figure in a metric of distributive justice concerned with the claims of adults. A possible limitation of developing a child-​sensitive metric in this way is that it will be insensitive to the importance, from the point of view of justice, of the intrinsic goods of childhood (Macleod 2010a; Brennan 2014; Gheaus 2014). These are goods that have a significant bearing on the quality of children’s lives, but which need not play a crucial role in development of mature agency. Consider the access that children have to opportunities for imaginative, carefree play and adventure both at home and at school. 13 

For adults, these generic resources can have great instrumental value in deliberating about, devising, and implementing conceptions of the good. But this instrumental value is predicated on the possession of mature agential capacities that children lack. Of course, children often need what some of these resources can secure. For instance, income can buy nutritious food for an infant. But an infant’s interest in eating well cannot be secured simply by giving her money for food.

430   Colin Macleod While some access to opportunities for play probably plays a role in the development of agency, it is likely that there can be enormous differences in children’s opportunities for play that do not affect the development of agency, yet which can have significance from the point of view of justice. For example, the recreational facilities at schools often differ dramatically, and those differences have an impact on the goodness of children’s lives—​the degree to which they flourish as children—​even if they are not causative of unjust economic or social inequalities of adults. Such inequalities matter because they bear upon children’s access to intrinsic goods of childhood. A distinction can be made between stronger and weaker versions of this claim. On the weaker claim, access to the intrinsic goods by children matters from the point of view of justice but there need not be a justice-​based entitlement to equal enjoyment of such goods by children. On the strong claim, there is a justice-​based entitlement to equal access to the intrinsic goods of childhood. The point here is not to adjudicate between these claims but rather to identify an issue that merits further consideration. If there are significant intrinsic goods of childhood of this sort, then a child-​sensitive metric of distributive justice should give recognition to them, even if doing so adds further complexities to theories of justice.14

19.5.2 Interfamilial Justice The challenge of developing a metric for distributive justice that is sensitive to the distinct interests of children is largely a matter of determining what the appropriate distribuendum of justice is. To the degree that a just distribution of the resources and opportunities that have special salience for children is one that extinguishes or mitigates the influence of morally arbitrary factors on children’s access to relevant goods and opportunities, there is reason to favour a strongly egalitarian account of distribution. Children do not choose and cannot be held responsible for the character of the families into which they are born. So children’s access to important goods and opportunities should not be determined or influenced by the wealth, social standing, or ideological commitments of their parents. Thus in the domain of healthcare and education, justice seems to require that all children enjoy access to equally good healthcare and equally good education. In a parallel way, justice seems to require that children have equal access to the social conditions conducive to the development of autonomy. Similarly, if there are important intrinsic goods of childhood, then children should be able to equally enjoy them irrespective of the social or economic background of their family. Those who wish to resist the strong presumption of this strongly egalitarian approach must explain why the life prospects of children (both those they experience as children and those to which they have access as adults) should be significantly influenced by children’s family 14  One issue that will arise concerns the relation between securing the intrinsic goods of childhood and securing goods for later stages of life. There could be trade-​offs that arise within and not solely across people’s lives. For example, the provision of fewer opportunities for childhood play could result in the provision of a greater range of valuable opportunities for adults to pursue.

The Family   431 background. Meeting this justificatory burden looks especially difficult if we acknowledge that a justification must address the claims of children directly and not simply subsume them under the claims of their parents.

19.5.3 Equality and Parental Partiality I have suggested that a conception of justice that seeks to eliminate or mitigate the effects that morally arbitrary factors have on people’s access to important resources and opportunities, favours a strongly egalitarian account of children’s distributive entitlements. However, the theoretical picture becomes more complicated once we determine how the distributive entitlements of adults are to be integrated with the entitlements of children. Two factors are especially salient in this context: (a) permissible adult resource inequality 15—​some economic inequalities between adults can be just—and (b) parental partiality—​parents may legitimately display greater concern for the interests of their own children than for the children of others. Although the basic issue about the relation between equality and partiality can arise in relation to a variety of theories of justice, it is helpful to consider it in relation to responsibility-​sensitive accounts of distributive justice. On these accounts, the responsible choices that adults make about how to conduct their lives can give rise to differential shares of resources. Providing suitable background conditions are met (e.g., absence of discrimination, existence of fair competitions, suitable access to education and healthcare), the choices that responsible adults make can give rise to resource inequalities that are not morally arbitrary. A  great deal of debate surrounds how the distributional effects of responsible choice are to be tracked and what background conditions must obtain before differences in choices confer justification on unequal holdings between people. But for our purposes, the main point is that economic inequalities between adults that are compatible with justice can give rise to interfamilial inequality that affects the prospects of children in troubling ways. Imagine that the difference between the holdings of rich and poor parents is entirely the upshot of responsible choices on the part of each. We may suppose that in all the domains in which children have an equal entitlement to goods, affluent parents can afford to supply their children with additional goods that benefit their children. Affluent parents can provide their children with education, healthcare, holidays, recreational activities, and aesthetic experiences that are beyond the means of poor parents. Rich parents are motivated to do so because they love their children and seek to promote their children’s well-​being. Moreover, we may suppose that the partiality that the 15 

Stipulating resource inequality here is not intended to indicate that a proper resolution of the debate over a proper metric of justice favours some variety of resourcism. The problem of parental partiality can arise with other metrics providing we assume that parents differ in their capacity to confer advantages on their children. However, for expository purposes it is easiest to focus on permissible resource inequalities between adults.

432   Colin Macleod affluent parents wish to express to their children is tightly bound up with realization of the special goods of family life such as intimacy and creative self-​extension that ground the special value of the affective family in the first place. But in this scenario, the expression of seemingly laudable parental partiality disrupts the justice-​based entitlement of children to equal provision of goods and opportunities (Brighouse and Swift 2006, 2009; Macleod 2002). The children of the rich are not entitled to a better education than the children of the poor simply in virtue of the fact that the rich are able to confer advantages on their children that the poor cannot.16 One possible, though not very attractive, way of addressing the problem of parental partiality is simply to abolish the affective family and replace it with communal forms of child rearing that do not permit the expression of equality-disrupting parental partiality.17 Aside from the fact that there are serious doubts about the feasibility of raising children collectively, a principal objection to the abolition of the affective family is that abolition would deny both children and adults access to the great goods of family life. A less radical and more attractive approach involves distinguishing between legitimate and illegitimate expressions of parental partiality by considering the relation in which different kinds of partiality stand to the realization of the goods of the family. Consider, for instance, the good of familial intimacy. Arguably its realization depends, among other things, on parents having and displaying special concern for their own children. Loving parents are especially attentive to the well-​being of their children and they engage in common activities that help to establish deep and rewarding emotional ties between parents and children. Although the realization of the good of familial intimacy requires that parents manifest greater concern for the interests of their own children than others, not all ways of expressing that greater concern are required to generate the good. For example, although affluent parents may be motivated to express their love and concern for their children by sending them to elite private schools, having a rich and rewarding relationship with their children does not depend on expressing partiality in this way. By contrast, the activity of reading bedtime stories to children is more integral to parents having rewarding relationships with children. Observing the connection between different expressions of partiality and the realization of familial goods facilitates a distinction between legitimate and illegitimate forms of parental partiality. Legitimate forms of parental partiality are those that are integral to the realization of important familial goods. Illegitimate forms of parental partiality are those that are not integral to the realization of familial goods and that would disrupt a just distribution between children. Of course, parents 16 

Here the problem of parental partiality arises at the level of ideal theory. It concerns how the affective family and the parental partiality on which it seems to depend can be harmonized with the strict distributive equality to which children are entitled in the face of a just but unequal distribution of resources between parents. A related issue concerns the degree to which parents can confer advantages on their children when the background distribution of resources and opportunities is itself unjust. For discussion of variants of this problem, see Swift (2003) and Macleod (2010b). 17  Rawls notes that the family can frustrate full attainment of fair equality of opportunity. He identifies and quickly rejects abolition of the family (1999: 448). See also Munoz-​Darde (1999).

The Family   433 cannot be expected to constantly monitor the partiality they display to their children with a view to distinguishing legitimate and illegitimate forms of partiality. Arguably such efforts would foster alienation between parents and children and would be corrosive to familial values. However, whether or not certain illegitimate forms of parental partiality can be expressed depends on the institutional arrangements that are in place. So the expression of some forms of illegitimate partiality can be blocked through suitable institutional design (e.g., elite private schools can be prohibited, tight limits on inheritance can be established). Against a suitable institutional background parents can express their partiality for their children and be vigorous advocates for their own children’s interest without disrupting the distributive entitlements of children significantly (Macleod 2002). Nonetheless, if we allow that the resource entitlements of adults can vary significantly, there will still be some non-​trivial respects in which the children of some families will have access to more advantages than the children in other families. For example, the holidays, recreational and extracurricular activities available to the children of affluent parents may be significantly better than those available to the children of poor families. To address these kinds of unjust inequalities between children it is arguably important to ensure that there is generous public provision of resources and opportunities that bear upon the quality of children’s lives. Public provision of excellent public parks, recreational facilities, and aesthetic programmes can play a role in mitigating some of these inequalities. The combined strategy of using institutional design to block the expression of illegitimate forms of parental partiality and generous public provision of resources that contribute to the flourishing of children’s lives provides a reasonable, but not perfect, way of harmonizing parental partiality with the demands of distributive justice.

19.6  Relational Egalitarianism Much of the foregoing discussion has been situated against the assumption that just­ ice is concerned with eliminating or mitigating arbitrary sources of inequality in the life prospects of people. However, brief discussion of a somewhat different approach to egalitarian justice is in order. On a “relational egalitarian” theory of the sort developed by Elizabeth Anderson (1999) the status of arbitrary inequalities in the life prospects of children rooted in their family background is unclear. Relational egalitarianism holds that the focus of egalitarian justice should not be on distribution per se. Rather, egalitarian justice is fundamentally concerned to ensure that all citizens have access to the social and material conditions that secure the basic dignity of persons and permit equal participation in fair democratic processes. Anderson’s influential articulation of relational equality does not broach the question of the justice-​based entitlements of children in much detail. However, in the domain of educational justice, Anderson (2007) and Debra Satz (2007) endorse a sufficientarian account of educational justice. On this

434   Colin Macleod view, all children are entitled to an education that is sufficient to prepare them for entry into college and that equips them with capacities to be democratic citizens. Although some forms of class inequalities are incompatible with facilitating mutually respectful relations between equal citizens, both Anderson and Satz allow that wealthy parents can confer advantages on their children (e.g., access to better schools and universities) that are beyond the means of poorer parents. So against the background in which all children have access to decent schools, relational egalitarians do not object to forms of parental partiality that create significant inequalities in the opportunities and life prospects of children from different family backgrounds. The relational conception of equality does provide an alternative reading of egalitarian concern that animates justice and I cannot adequately discuss it here. However, if we focus on the character of relations between children, it is not clear why substantial (and avoidable) inequalities between the life prospects of children are compatible with the ideal of mutual respect that animates relational equality. The child of a poor family whose life prospects are much worse than the child of a rich family may reasonably ask whether the rich child can offer a compelling, mutually respectful, justification for her entitlement to better life prospects (Macleod 2012).

19.7 Conclusion The foregoing discussion is not an exhaustive survey of the problems of distributive just­ ice that arise in relation to the family. There are many interesting questions that arise concerning the interpretation and extension of diverse theories of distributive justice to the family. I have identified some obstacles some of these theories face in grappling with the interests and distributive entitlements of children. But I have not considered in any detail how theories such as Rawls’s justice of fairness, contemporary libertarianism, luck egalitarianism, Hobbesian contractarianism, or various feminist theories might be developed so as to address the problems concerning distributive justice and the family. In the domain of distributive justice and the family, many interesting issues, both theoretical and practical, merit further exploration.

References Ahlberg, J. and Brighouse, H. (2010). “An “Argument against Cloning,” Canadian Journal of Philosophy 40(4): 539–​66. Anderson, E. (1999). “What is the Point of Equality?” Ethics 109: 287–​337. Anderson, E. (2007). “Fair Opportunity in Education: A Democratic Equality Perspective,” Ethics 117: 595–​622. Archard, D. (2004). Children: Rights and Childhood, 2nd edn. New York: Routledge. Archard, D. and Macleod, C. (eds) (2002). The Moral and Political Status of Children: New Essays. Oxford: Oxford University Press.

The Family   435 Arneson, R. (2018). “Dworkin and Luck Egalitarianism: A Comparison,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 41–64. Benatar, D. (2006). Better Never to Have Been: The Harm of Coming into Existence Oxford: Oxford University Press. Botterell, A. and McLeod, C. (2014). “‘Not for the Faint of Heart’: Accessing the Status Quo on Adoption and Parental Licensing,” in F. Baylis and C. McLeod, (eds) Family-​Making. Oxford: Oxford University Press, pp. 151–67. Bou-​Habib, P. and Olsaretti, S. (2014). “Autonomy and Children’s Well-​Being,” in A. Bagattini and C. Macleod (eds) The Wellbeing of Children in Theory and Practice. Berlin and New York: Springer, pp. 15–​34. Brake, E. (2010). “Willing Parents: A Voluntarist Account of Parental Role Obligations,” in D. Archard and D. Benatar (eds) Procreation and Parenthood. Oxford: Oxford University Press, pp. 151–​77. Brennan, S. (2014). “The Goods of Childhood, Children’s Rights, and the Role of Parents as Advocates and Interpreters,” in F. Baylis and C. McLeod (eds) Family-​Making: Contemporary Ethical Challenges. Oxford: Oxford University Press, pp. 29–​48. Brighouse, H. (2000). School Choice and Social Justice. Oxford: Oxford University Press. Brighouse, H. and Swift, A. (2006). “Parents’ Rights and the Value of the Family.” Ethics 117(1): 80–​108. Brighouse, H. and Swift, A. (2009). “Legitimate Parental Partiality.” Philosophy and Public Affairs 37(1): 43–​80. Brighouse, H. and Swift, A. (2014). Family Values. Princeton, NJ: Princeton University Press. Callan, E. (2002). “Autonomy, Child-​Rearing, and Good Lives,” in D. Archard and C. Macleod (eds) The Moral and Political Status of Children. Oxford: Oxford University Press, pp. 118–​41. Casal, P. and Williams, A. (1995). “Rights, Equality and Procreation,” Analyse & Kritik 17: 93–​116. Clayton, M. (2006). Justice and Legitimacy in Upbringing. Oxford: Oxford University Press. Clayton, M. (2018). “Education,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 438–59. Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 99: 906–​44. Crouch, S. R., Waters, E., McNair, R., Power, J., and Davis, E. (2014). “Parent-​Reported Measures of Child Health and Wellbeing in Same-​Sex Parent Families: A Cross-​Sectional Survey.” BMC Public Health 14: 635. De Wispelaere, J. and Weinstock, D. (2012). “Licensing Parents to Protect our Children?” Ethics and Social Welfare 6(2): 195–​205. Dworkin, R. (2000). Sovereign Virtue. Cambridge, MA: Harvard University Press. Dwyer, J. (2006). The Relationship Rights of Children. Cambridge: Cambridge University Press. Freeman, S. (2018). “Rawls on Distributive Justice and the Difference Principle,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 13–40. Frailberg, S. (1977). Every Child’s Birthright: A Defense of Mothering. New York: Basic Books. Gauthier, D. (1986). Morals By Agreement. Oxford: Oxford University Press. Gheaus, A. (2011). “The Right to Parent One’s Biological Baby.” Journal of Political Philosophy 20(4): 432–​55. Gheaus, A. (2014). “The Intrinsic Goods of Childhood and the Good Society,” in A. Bagattini and C. Macleod (eds) The Wellbeing of Children in Theory and Practice. Dordrecht: Springer, pp. 35–​52.

436   Colin Macleod Gheus, A. (2018). “Gender,” in Serena Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 389–414. Hall, B. (1999). “The Origin of Parental Rights.” Public Affairs Quarterly, 13(1): 73–​82. Hannan, S. and Vernon, R. (2008). “Parental Rights:  A Role-​Based Approach.” Theory and Research in Education 6: 173–​89. Haslanger, S. (2009). “Family, Ancestry and Self: What is the Moral Significance of Biological Ties?” Adoption & Culture, 2(1): 91–​122. Kittay, E. F. (1999). Love’s Labor:  Essays on Women, Equality, and Dependency. New  York: Routledge. Kymlicka, W. (2002). Contemporary Political Philosophy:  An Introduction. Oxford:  Oxford University Press. LaFollette, H. (1980). “Licensing Parents.” Philosophy & Public Affairs 9(2): 183–​97. LaFollette, H. (2010). “Licensing Parents Revisited.” Journal of Applied Philosophy 27(4): 327–​43. Levinson M. (1999). The Demands of Liberal Education. Oxford: Oxford University Press. Liao, M. (2006). “The Right of Children to be Loved.” The Journal of Political Philosophy 14(4): 420–​40. Macedo, S. and Young, I. (2003). Children, Family and the State. Nomos XLIV. New  York: New York University Press. Macleod, C. (1997). “Conceptions of Parental Autonomy.” Politics and Society 25(1): 117–​40. Macleod, C. (2002) “Liberal Equality and the Affective Family,” in D. Archard and C. Macleod (eds) The Moral and Political Status of Children:  New Essays. Oxford University Press, pp. 212–​30. Macleod, C. (2010a). “Primary Goods, Capabilities and Children,” in H. Brighouse and I. Robeyns (eds) Measuring Justice: Primary Goods and Capabilities. Cambridge: Cambridge University Press, pp. 174–​92. Macleod, C. (2010b). “Parental Responsibilities in an Unjust World,” in D. Archard and D. Benatar (eds) Procreation and Parenthood: The Ethics of Bearing and Rearing Children. Oxford: Oxford University Press, pp. 128–​50. Macleod, C. (2012). “Justice, Educational Equality and Sufficiency,” in Colin Macleod (ed.) Justice and Equality. Calgary: University of Calgary Press, pp. 151–​75. Macleod, C. (2014). “Agency, Authority and the Vulnerability of Children,” in A. Bagattini and C. Macleod (eds) The Nature of Children’s Well Being: Theory and Practice. Dordrecht: Springer, pp. 53–​64. Macleod, C. (2015). “Parental Competency and the Right to Parent,” in R. Vernon, S. Hannan, and S. Brennan (eds) Permissible Progeny. New York: Oxford University Press, pp. 227–​45. McFall, M. (2009). Licensing Parents:  Family, State, and Child Maltreatment. Lanham, Md: Lexington Books. Munoz-​Darde, V. (1999). “Is the Family to be Abolished Then?” Proceedings of the Aristotelian Society 99: 37–​56. Mullin, A. (2007). “Children, Autonomy and Care.” Journal of Social Philosophy 38: 536–​53. Nozick, R. (1974). Anarchy, State and Utopia. New York: Basic Books. Nussbaum, M. (2006). Women and Human Development:  The Capabilities Approach. Cambridge: Cambridge University Press. Okin, S. M. (1989). Justice, Gender and the Family. New York: Basic Books. Olsaretti, S. (2013). “Children as Public Goods?” Philosophy and Public Affairs 41(3): 226–​58. Overall, C. (2012). Why Have Children? The Ethical Debate. Cambridge, MA: MIT Press. Rakowski, E. (1993). Equal Justice. Oxford: Clarendon Press.

The Family   437 Rawls, J. (1993). Political Liberalism. New York: Columbia University Press. Rawls, J. (1999). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (2003). Justice As Fairness: A Restatement. Cambridge, MA: Harvard University Press. Richards, N. (2010). The Ethics of Parenthood. New York: Oxford University Press. Robeyns, I. (2010). “Gender, Care and Property-​Owning Democracy,” in M. O’Neill and T. Williamson (eds) Property-​Owning Democracy:  Rawls and Beyond. London: Wiley-​ Blackwells, pp. 163–​79. Robeyns, I. (2018). “The Capability Approach,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 109–28. Satz, D. (2007). “Equality, Adequacy and Education for Citizenship.” Ethics 117: 623–​48. Schoeman, F. (1980). “Rights of Families: Rights of Parents, and the Moral Basis of the Family.” Ethics, 91: 6–​19. Sen, A. (1999). Development as Freedom. New York: Knopf. Shiffrin, S. (1999). “Wrongful Life, Procreative Responsibility, and the Significance of Harm.” 5 Legal Theory: 117–​48. Sommerville, M. (2007). “Children’s Human Rights and Unlinking Child–Parent Biological Bonds with Adoption, Same-​Sex Marriage and New Reproductive Technologies.” Journal of Family Studies 13: 179–​201. Swift, A. (2003). How not to be a Hypocrite: School Choice for the Morally Perplexed Parent. London: Routledge. Vallentyne, P. (2002), “Equality and the Duties of Procreators,” in D. Archard and C. Macleod (eds) Children and Political Theory. Oxford: Oxford University Press, pp. 195–​211. Velleman, D. (2005). “Family History.” Philosophical Papers 34(3): 357–​78. Woodcock, S. (2009). “Five Reasons why Margaret Somerville is Wrong about Same-​Sex Marriage and the Rights of Children.” Dialogue 48: 1–​21.

Chapter 20

Edu cat i on Matthew Clayton

The importance of education—​which here includes not just the activities of institutions of formal schooling but also other interactions that improve individuals’ skills, knowledge, and understanding, and shape their ambitions and beliefs—​within a conception of distributive justice is multifaceted. In the first place, in almost all economies, educational success gives individuals access to other advantages, such as more fulfilling and better paid jobs; for that reason, the distribution of educational opportunity has attracted the attention of theorists of justice. Second, education of a certain sort might facilitate a just society or world, if, for example, it cultivated individuals to be disposed to support and comply with just institutions. Accordingly, a significant issue is whether it is morally permissible or required for others to shape individuals’ political convictions or motivation in that way. Third, in many societies, the primary recipients of education are children whose beliefs, desires, and opportunities are shaped by their parents and in schools. Because young children have no choice over whether they are educated, questions arise concerning what parents and adults are morally permitted to do to, or for, them. And, finally, because education is practised by several different agents—​ parents, the state, and associations such as religious institutions—​an adequate conception of educational justice must provide an account of the relevance and strength of their respective claims to determine the content and distribution of education.

20.1  The Distribution of Educational Opportunity The problem of how to distribute educational opportunity is normally framed by reference to a few relevant background socioeconomic facts. First, in most societies there is significant occupational inequality—​inequality with respect to the value of different jobs in terms of both the desirability of the work they involve and the income paid for

Education   439 doing them. Second, occupational success often tracks educational success:  because jobs are typically offered to the best-​qualified candidates, understood as individuals who have the wherewithal to perform best in the job, the education individuals receive is an important determinant of their occupational success. Third, education may enhance society’s productive capacity in at least two ways: it can increase the human capital available within the economy by improving individuals’ understanding and skills, and it provides valuable signals of workers’ productive potential, which educational qualifications are thought to track to some degree. Granted these facts, there are two central distributive questions concerning education justice. First, how large should the education budget be: how many resources should be devoted to education rather than, say, healthcare, the police, or social security? Second, what is a just distribution of educational resources?1 The first question about the size of the education budget has, despite its political importance, received comparatively little attention from political philosophers.2 This neglect might be explained by the complexities of identifying the costs and benefits of differently sized educational budgets. The question is made even more difficult to answer if, as many argue, how much we ought to spend on education in aggregate terms depends on whether, and the extent to which, we have a duty of justice to compensate individuals for their poor educational or occupational prospects. The second question has received considerable attention, and the most widely endorsed answer appeals to the ideal of equality of educational opportunity. Like other interpretive concepts, however, how best to elaborate this ideal is a matter of dispute, with different theorists articulating and defending different conceptions of equal opportunity.3 A common point of reference in these debates is John Rawls’s conception of Fair Equality of Opportunity (Rawls 1999). Rawls proposes at least two different conceptions of equal opportunity, which have different implications for educational institutions and have given rise to different debates: the first is a meritocratic conception under which inequality of educational attainment is just only if it does not reflect differences in class origins; the second is a conception that permits educational inequalities that are beneficial to less advantaged citizens.4

1  Given the broad conception of education adopted at the outset, educational resources are not exhausted by the income enjoyed by formal schools. A part of the healthcare budget—​the funding of glasses for children, for example—​and other budgets should be viewed as educational resources to the extent that they affect the skills, understanding, and motivation of individuals. The resources parents spend on their children for these purposes should also be included. 2  For a preliminary attempt to address the question, see Clayton (2015). 3  The idea of interpretive concepts and how they differ from criterial concepts is set out in Dworkin (2011). See Jencks (1988) for a nice statement of how, while the concept is widely endorsed, the nature and implications of equality of educational opportunity are disputed. 4  For discussion of different conceptions of equality of opportunity that can be identified in A Theory of Justice, see Pogge (1989: 166–​9).

440   Matthew Clayton

20.1.1 Meritocratic Equality of Opportunity The meritocratic conception of equality of opportunity that Rawls appears to defend (Rawls 1999: 57–​65, 73–​8)—​call this MEO—​holds that the educational and occupational achievement of individuals with the same biological endowment and life plans should not be unequally affected by differences in their respective class origins—​naturally bright and ambitious children from working-​class backgrounds should achieve no less than similarly able and motivated children from middle-​class backgrounds.5 Rawls argues that to achieve equal opportunity it is not enough to ensure that individuals are not the victims of racial or sexual discrimination in hiring for employment: justice also requires everyone to enjoy a fair chance to acquire the abilities that make them suitable for the jobs in question (see also Mason 2006). But what counts as a fair chance? Rawls’s fundamental intuition is that an individual’s access to an occupation should not be worse than another’s due to ‘morally arbitrary’ factors (Rawls 1999: 63). One such factor is the social environment in which one grows up. It is unfair for an individual’s occupational opportunities to be worse than another’s when that inequality is explained by differences in their respective families or social milieux. Accordingly, MEO asserts that a just political community would provide additional educational resources to those born into relatively deprived social environments, ideally up to the point at which the educational and occupational achievement of individuals with similar biological endowments and ambitions is equal, and the differential effects of class are eliminated. Before moving on to Rawls’s second conception of equal opportunity, it is worth pausing to consider a puzzling feature of MEO that some argue renders it implausible as an account of educational justice. MEO asserts that justice forbids inequality of educational and occupational opportunity that reflects social inequalities, such as family background, but it permits such inequality when, subject to certain requirements, it reflects natural differences, such our different genetic endowments. The puzzle is this: Rawls claims that we have a reason to eliminate or reduce inequalities caused by the social lottery, because social differences are merely matters of good or bad luck that do not justify the inequalities they produce. Since the genetic lottery is just another distribution of good and bad brute luck, surely the same reason applies: justice requires us to eliminate or reduce inequalities that are the product of individuals’ different genetic inheritances. In short, if it is unjust to allow ‘silver spoons’ to confer relative educational and occupational advantage, why isn’t it also unjust to allow ‘golden genes’ to confer it?6

5 

The term ‘meritocracy’ was coined by Michael Young (1958), who was a critic of its vision of society. Young equated merit with IQ plus effort. Rawls’s conception is somewhat different, because he replaces IQ with a broader range of natural abilities—​indeed, any talents that are marketable—​and effort is replaced with ambition or willingness to use one’s talents. In addition, Rawls embeds meritocracy within a larger social democratic conception of a just society that sidesteps many of Young’s worries. 6  See Nagel (1997) for this nice way of stating the issue. For discussion of this problem with MEO in Rawls’s conception of justice see Arneson (1999) and Clayton (2001).

Education   441 Acknowledging the force of this problem, some have entertained the idea of abandoning MEO in favour of a radical or extreme conception of equal opportunity.7 According to the radical conception, occupational opportunity should be equal for those who are similarly motivated or ambitious:  it seeks to eliminate or reduce inequalities in education or occupational achievement that reflect either class-​based or genetic differences. Taking a step further, the extreme egalitarian conception proposes that we have a reason to equalize educational and occupational achievement full stop, on the grounds that children—​at least young children—​can’t be held responsible for having or lacking the ambition or motivation to succeed educationally. A different response to the puzzle is to insist that, even though they are both distributions of brute luck, there is a morally relevant difference between the natural and social lotteries that supports MEO. One suggestion is that justice requires us to compensate individuals who suffer relative disadvantage only when, or specially when, we are responsible for the inequality in question; and we as a collective are responsible for the social lottery, but not, or less so, for the genetic lottery.8 If an argument of that kind were sound it might form the basis of a defence of MEO because it would support moves to equalize educational achievement between individuals born into different social settings, but would not require us to equalize educational achievement between individuals who are differently endowed by nature. The appeal to responsibility to ground the distinction between social and natural causes of inequality that is central to MEO is controversial. Some question whether it is possible to identify more or less advantageous genes independently of facts about the social environment and, therefore, the central distinction on which MEO rests is scientifically untenable (Fishkin 2014). Others have argued that, although the distinction can be made, it lacks moral relevance (Lippert-​Rasmussen 2004). One way of pressing the latter claim is as follows. The differential responsibility claim rests on a counter-​ intuitive view of responsibility. Suppose that the distribution of good and bad genetic luck is beyond our control. If we believe that injustice requires wrongdoing, it follows that that distribution is not unjust—​Rawls agrees and describes such a distribution as ‘neither just nor unjust’ (Rawls 1999: 87): social justice and injustice refer only to how social institutions distribute benefits and burdens within their control. Still, even though the distribution of genes is not unjust, because collectively we do not have control over it, the way in which social institutions respond to genetic inequality is within our control. We might take steps to ensure that genetic disadvantage does not translate into educational or occupational disadvantage by, for example, devoting additional resources to the schooling of the genetically disadvantaged so that they have the skills

7  See Brighouse and Swift (2014a), who argue that there are prima facie reasons to abandon MEO in favour of the radical or extreme conceptions to the extent that equality is valuable. However, they go on to suggest that the value of equality is very often overridden by the value of benefiting the least advantaged. 8  For an elaborate version of this kind of argument, which cannot be given the attention it merits, see Nagel (1997).

442   Matthew Clayton and understanding that the genetically advantaged have. If collectively we fail to do that, it seems appropriate to attribute to us responsibility for the fact that the genetically disadvantaged have less access to educational and occupational goods than the genetically advantaged. Thus, even if, in contrast to the social lottery, the genetic lottery is not of our making, the consequences of the genetic lottery with respect to how it affects access to education and employment are within our control and, accordingly, the simple appeal to responsibility to rescue MEO fails. A different way of responding to the objection that the distinction between natural and social causes of inequality lacks moral relevance is to argue that we ought to measure individuals’ advantage not by reference to how well they achieve in education or employment, but how well they realize their potential. If that idea is attractive, then we have a principled reason to favour meritocratic over radical or extreme conceptions of equal opportunity. To see how this argument works a few distinctions are required. One of the central problems in distributive justice is how to specify its ‘currency’, the fundamental items that we employ to identify how advantaged different individuals are and how much they gain or lose from different policies. In recent political philosophy there has been considerable debate about which set of items—​‘primary goods’, ‘resources’, ‘welfare’, or ‘capabilities’—​constitutes the right currency. But the idea of realization of potential suggests a further dimension, and one that seems particularly relevant within education, not least because teachers, educationalists, and politicians often deploy it as the yardstick by which to evaluate education policy.9 Realization of potential might be defined in terms of the difference between one’s actual achievement and one’s maximum feasible achievement. Consider a very simple example. Amy could be a great chess player; Ben only a moderately good chess player. However, Amy lacks access to competitors with whom she might play and, thereby, hone her skills. Ben competes with and learns from others on a daily basis. If they played each other, Amy would easily defeat Ben. However, Ben realizes his chess-​playing potential to a greater degree than Amy realizes hers: in terms of achievement simpliciter, Amy achieves more than Ben; but Ben outscores Amy in terms of achievement relative to possible achievement. As Vallentyne (2007) has observed, there are several ways of modelling the difference between actual and possible achievement to generate a measure of realization of potential. Perhaps the most intuitive measure is the ratio of actual achievement to possible achievement, with complete realization of potential as 100 per cent and less complete realization expressed as percentages less than 100. In the chess case, for example, suppose that Amy’s maximum possible achievement can be expressed numerically as 50 and Ben’s as 20. Given their circumstances, while Amy’s actual achievement is 25, Ben’s

9  For example, in an article setting out his vision for education in the twenty-​first century, the then UK Prime Minister, Gordon Brown, proposed ‘a personalized education tailored to realize the potential of every child and transform their life chances’ (Brown 2009).

Education   443 is 15. Accordingly, the ratio view of realization of potential identifies Amy’s realization as 50 per cent, which is considerably lower than Ben’s 75 per cent.10 Putting issues of how to measure realization of potential to one side, let us turn to the question of how it relates to MEO. When teachers or politicians invoke the idea of realization of potential as the aim of education policy, they typically appeal to a particular baseline to evaluate individuals’ realization. It is how much an individual might achieve at conception, or given her genetic endowment, that represents the reference point identifying ‘possible achievement’ which is used to gauge the extent of her realization of potential. If, at some later point in her life, her maximum possible achievement is diminished because of the lack of parental investment she has suffered, that counts as a loss of realization of potential on this view.11 MEO might, then, be defended by appealing to the ideal of equal realization of different individuals’ natural potential, where that is understood as what different individuals would achieve at conception if available resources were devoted to their achievement. Although the details of such a conception would need spelling out in more detail to make the view determinate, the connection between realization of potential and meritocracy is sufficiently clear for present purposes. Nevertheless, despite its popularity, the realization of potential argument for MEO needs further defence, for it is not clear why we should adopt realization of natural potential as the appropriate goal. Why should we treat feasible achievement given our talents provided by nature as the appropriate baseline for judging realization, rather than achievement given our abilities that have been, perhaps unequally, enhanced by our respective parents after some period of time, for example? It is not obvious why it makes a difference whether the talents we have, and by which we measure realization of potential, were given to us by some natural process, such as the genetic lottery, or some social process, such as our parents paying for gene therapy. Thus, for all its initial promise, the realization of potential argument for MEO faces the same normative puzzle as Rawls’s initial defence.12 Finally, it is worth mentioning the argument that MEO is instrumentally valuable. According to this view, because it is good for society to have more, rather than less skilled workers available to the economy, we ought to prioritize the schooling of 10 

An alternative way of characterizing realization of potential is to calculate it as the absolute shortfall from maximum possible achievement. On this view, given the figures as represented, Amy’s loss of realization is 25 while Ben’s is 5: for their realization to be equal, Amy’s actual achievement would need to be raised to 45. Using the absolute shortfall view to determine equal opportunity for realization of potential seems objectionable, however, because it is too favourable to those whose potential to achieve is greater. 11  There might be further issues that require attention in this account. For example, an individual’s potential at conception might be affected by the activities of her procreative parents—​their smoking, for example—​which might be problematic from the point of view of egalitarian justice. I shall not consider further issues of that kind here. 12  There are further questions about the relevance of genetic facts to distributive justice, which are beyond the scope of this chapter. See Casal (2018) for a wide-​ranging discussion of the several ways in which principles of morality and justice might be responsive to facts about human nature.

444   Matthew Clayton individuals whose skills can be enhanced most with the available educational resources. Suppose Celine is lucky enough to have been born with talents that enable her to convert educational resources into marketable skills at a high rate, whereas Donald lacks those talents. Other things equal, a society wanting to maximize its human capital would devote more educational resources to Celine. Although this argument appears to support MEO, in fact it does so only under certain circumstances. Consider a different case. Eric was born with few talents but, as a product of his parents’ investment in high-​quality child care and pre-​school education, he now converts educational resources into skills at a high rate; by contrast, Frank, who was born with a favourable biological endowment, has not had the advantages Eric enjoyed and now acquires skills from educational resources at a low rate. If the efficient production of human capital is our aim, then in this case we must give priority to Eric in educational spending and, thereby, abandon MEO. Of course, it might be replied that Eric’s parents are not morally permitted to give him a high-​quality pre-​school environment and, when distributing schooling, we ought to be responsive to individuals’ attributes that are not the product of injustice or wrongdoing. But it is hard to believe that every parental activity that has the effect of improving an individual’s capacity to learn is morally impermissible (Brighouse and Swift 2008, 2014b). (We shall return to this issue later in the chapter.) One general concern with equality of opportunity accounts of educational just­ ice—​which seems to apply to MEO, but clearly applies to the radical and extreme conceptions—​is that they run into ‘bottomless pit’ problems. They seem to assert that we have reason to devote potentially huge amounts of resource slightly to improve the educational achievement of those who are cognitively impaired or socially disadvantaged, even when that is at the costs of significant reductions in the educational achievement of the cognitively lucky or socially privileged, which is bad for both the latter and others in virtue of lost human capital. In response to this problem, it might be claimed that the relevant conception of equal opportunity is the right interpretation of educational fairness, but that educational fairness is not the only relevant ideal: fairness needs to be weighed in the balance with other goods, such as enhancing the productive potential of society, or promoting human well-​being, or attending to the overall, rather than educational, advantage of the most unlucky, and these other considerations may override our reason to promote fairness in a way that avoids the bottomless pit (Brighouse and Swift 2014a). However, once this move is made, it is not clear what practical role, if any, equality of opportunity plays within an account of educational justice, because it is not obvious that it makes any difference to what we ought to do.

20.1.2 Opportunity for the Least Advantaged We have reviewed some debates about MEO, a particular conception of equal educational opportunity that Rawls appears to defend in Ch. II of A Theory of Justice. It is worth noting, however, that when he presents his Two Principles in their final form,

Education   445 Rawls asserts that justice permits certain inequalities of opportunity—​those inequalities that ‘enhance the opportunities of those with lesser opportunity’ (Rawls 1999: 266). The conception of educational justice suggested by Rawls’s final statement represents a departure from simple egalitarian views of equal opportunity that have as their aim the reduction or elimination of inequality of occupational opportunity. I shall call this view Opportunity for the Least Advantaged (OLA). Whereas MEO supports educational measures to nullify the unequal impact of social privilege, OLA leaves open the possibility of a stratified class system with little social mobility, provided that such a society generates better education or employment for those with least opportunity. It is important to note that, for Rawls, it is not enough to justify more unequal education to show how the least advantaged are better off in monetary terms than they would be if educational opportunity were less unequal (1999: 265); the inequality is permissible only if their opportunities are enhanced. Unfortunately, Rawls provides us with little indication of what he means by ‘opportunity’. It is, however, worth exploring how the general idea of OLA might be elaborated as a conception of educational justice. In the first place, we need to clarify how we should understand ‘Opportunity’: the central question is ‘Opportunity for what?’ Different versions of OLA might be distinguished by how they understand the goods access which ought to be our focus. The discussion so far has referred to educational opportunity as a means of gaining access to desirable occupations. A natural move, then, is to assert that we may allow inequalities in access to occupations provided that they improve the quality of jobs available to those with least opportunity. As an illustrative example, consider a society in which there exist inequalities of wealth and income and it is legally permissible for the rich to buy excellent education for their children. Let us assume that their children receive better education and, in virtue of gaining more or better qualifications, they have an advantage in the competition for jobs compared to those with poorer parents. Preventing the rich from purchasing better education would equalize opportunity, but it may not enhance the educational or job opportunities available for those with the fewest. It might be that allowing parents to invest in their children’s education would (a) generate incentives for individuals to be more productive (because they would not work as hard if denied the opportunity to spend their augmented income on their children’s education); and (b) enhance the human capital available within the economy (because in addition to the taxes they pay to fund state-managed schooling they invest further sums to improve the human capital of their own children). If that were the case—​I shall not address whether those facts obtain—​the economy would be richer and might benefit the least advantaged in several ways. Given an appropriate tax regime, the boost in tax receipts gained from greater productivity might enable the government to invest in education and, thereby, improve the educational experience of everyone compared to that offered in an egalitarian society. It might also have the resources to make work more attractive for those for whom it is least attractive. That sketch of how OLA might countenance inequality to improve the educational and occupational opportunities of the least advantaged depends on the truth of several empirical facts, and trades on certain hidden normative assumptions. For example, some

446   Matthew Clayton deny that allowing the rich to pay large sums for their children to receive better schooling always benefits the least advantaged, because education is a competitive positional good (Swift 2003; Brighouse and Swift 2006a). Competitive positional goods are those whose value increases as others receive less of them. In the case of education, part of the value of a good schooling is that it improves one’s chances of securing scarce opportunities for university education and desirable employment, but that competitive advantage is boosted if others receive a less good schooling. Similarly, even if Paula receives a decent education, if Rachel, the child of rich parents, receives a better schooling than Paula and is thereby better placed in the competition for a desirable job, the effect of that purchase is to worsen the occupational opportunities of Paula, who might have been selected had Rachel’s parents not bought the better schooling. To the extent that we focus on the positional aspect of education, then, OLA does not support allowing some to pay for educational advantage, because such choices merely redistribute educational opportunity, without improving it for those with fewer opportunities. Nevertheless, it should be noted that the value of education and schooling is only partly positional in nature: a different valuable aspect of education is that it has the potential to increase the amount of human capital in society. Other things equal, if among the available options, allowing the rich to buy better education for their own children most efficiently enhances human capital, and improved human capital translates into improved education or employment opportunities for the least advantaged, then OLA would support such a policy, even if particular individuals’ opportunities are diminished by it.13 OLA’s implications for education policy turn on further empirical facts. It asserts that a policy allowing the rich to purchase expensive advantage-​conferring schooling for their own children might be just, but it is just only if it serves to benefit the least advantaged in terms of education and/​or occupation. Accordingly, that permissive policy can only ever be conditionally just, because its justice depends on other socio-​ economic arrangements being in place that transform greater societal productivity into opportunities for the least advantaged. Classical liberals tend to believe that a relatively unregulated market society maintains those arrangements (Tomasi 2012); others are more sceptical and would support the permissive policy only if suitable tax and spend policies are in place (Brighouse and Swift 2014a). One normative assumption of the OLA-​based argument for educational inequality is that the productive rich are permitted by justice to work less hard in the absence of the opportunity to buy relatively more educational advantage for their own children. If that assumption is denied, as it is by some followers of G. A. Cohen (2008), then the gains to the least advantaged that come from a more productive economy might be realized without allowing educational inequality. Thus, again, the implications of OLA 13 

The reference to educational opportunity alerts us to the fact that there might be various ways in which improved skills and understanding might be valuable. Some believe that being better educated is intrinsically valuable. Others might point to the instrumental benefits for us as citizens from having a better facility with foreign languages or a greater appreciation of the natural and social sciences. At least some goods of education are non-​positional in character.

Education   447 for educational policy might vary according to the larger conception of distributive just­ ice in which it is nested, and more work needs to be done precisely to identify the educational implications of OLA once it is embedded in the most plausible general account of justice.

20.1.3 Educational Adequacy OLA shares one important feature with egalitarian accounts of educational justice. It asserts that, although departures from equality are permitted if they enhance the opportunities of the least advantaged, equal opportunity is the appropriate initial baseline from which to judge improvements from the point of view of justice. An alternative to both OLA and egalitarian accounts, which has attracted some support in recent years, is a sufficientarian account that claims that educational justice is satisfied if everyone has an adequate education. In a society in which every individual receives enough education, parents and others are entitled to devote additional resources exclusively to the education of their own children. Adequacy views can take various forms, but they typically distinguish between different goods that education delivers, and pick out a satiable subset of those goods as ones that we are required by justice to promote. Consider some different goods education might deliver: it enhances an individual’s knowledge and understanding of the world, their economic productivity, their ability to live autonomously or well (according to some account of human flourishing), and their ability to relate to others on equal terms or to participate as a democratic citizen, and, in some circumstances, their competitive advantage with respect to securing occupational opportunity. Educational adequacy theorists pick out a subset of these goods and assert that educational justice is satisfied if those goods are realized. For example, White argues that education’s role is to enable every individual to lead an autonomous life (1994: 177). For most individuals, there is an amount of education, which varies between individuals depending on their different needs, that is enough to enable them to live autonomous lives. The satisfaction of education for personal autonomy is, therefore, consistent with some being given the opportunity to understand the world better or to develop their talents more than others because such knowledge or realization of potential is not necessary for personal autonomy. Anderson (2007) and Satz (2007) theorize justice in education in terms of giving every citizen the wherewithal to relate to each other as democratic equals. Although that aim has different institutional implications for educational institutions and the school curriculum compared to the promotion of autonomy view, the vision is similarly compatible, with some parents buying extra schooling to give their children competitive advantage in the job market. It is puzzling, however, why the remit of justice should be restricted to a subset of educational goods. MEO and OLA both operate on the assumption that, other things being equal, more rather than less occupational opportunity is valuable for individuals, because more opportunity enables them to choose between a wider set of lifestyles or to

448   Matthew Clayton take on more fulfilling work. Principles of justice, they claim, should accordingly be responsive to individuals’ claims for greater occupational opportunity. True, occupational inequality or disadvantage does not necessarily mean that individuals are unable to lead autonomous lives or to relate to each other as equals. But our social and political institutions should acknowledge that individuals have a legitimate interest in having the opportunity, including occupational opportunity, to pursue their own conceptions of the good. Since that is the case, we need principles of justice, including principles for the regulation of educational institutions, which are normally an important vehicle for occupational success, to adjudicate between individuals’ competing claims for occupational opportunity. Education for democratic citizenship or personal autonomy is not enough.14

20.2  Justice and the Content of Education Even if their ideals are insufficiently demanding, adequacy theorists such as Anderson and Satz highlight an important, though not uncontroversial, truth about educational justice: that one very important role of educational institutions is to help to create a just society by giving individuals the wherewithal and motivation to participate as equals in a democratic society. The view that educational institutions, understood broadly to include parental conduct as well as sites of formal schooling, are permitted intentionally to shape children’s convictions and ambitions is not universally held. Indeed, the question of whether and when it is permissible to shape the beliefs and desires of children, as opposed to imparting them with certain knowledge, skills, and understanding, has generated considerable debate in recent political and educational philosophy. To navigate our way through that debate it is worth drawing a couple of distinctions: first, between shaping political and ethical convictions; second, between whether the state or parents are the shapers of children’s beliefs and desires.

20.2.1 Education for Justice The distinction between political and ethical convictions mirrors Rawls’s distinction between a set of ideals for the regulation of our major legal, political, and socioeconomic institutions and those that inform individuals’ beliefs about their place in the universe, religion, and the virtues and goals that are worthy of pursuit given the legal options they 14  In this respect, the critique of educational adequacy follows the critique of sufficientarianism in general (see Casal 2007). For further criticisms of educational adequacy, see Brighouse and Swift (2009, 2014a) and Macleod (2010).

Education   449 enjoy. Consider political convictions. Is the state’s use of educational institutions to encourage children to become committed to particular political ideals ever permissible? One concern about the moulding of political convictions is expressed by Brighouse (2000). A set of political arrangements is legitimate, he claims, only if it enjoys the free, informed, and reasoned consent of at least most of its citizens. Accordingly, the state’s role in political education should be restricted to ensuring that its citizens understand the legal and political principles and institutions that govern them, have an appreciation of and ability to engage with the competing conceptions of political morality, and have the skills to participate in different kinds of political action should they so desire. It must desist from manipulating citizens’, particularly children’s, political convictions, because that is inconsistent with citizens giving free, informed, and reasoned consent to the state. If Brighouse’s consent-​based conception of political legitimacy is rejected, then his critique of the shaping of political convictions can be challenged. Consider, for example, Rawls’s conception of legitimacy, which trades on his claim that there is a natural duty of justice (1999: 99–​100, 293f). Each of us is duty-​bound to promote just institutions where they do not exist and to comply with just, or nearly just, institutions where they do exist, and that duty applies to us irrespective of whether we consent to those institutions. If that view is sound, then educational institutions have an important role to play in enabling individuals to do what they are duty-​bound to do, by encouraging children to form the ambition to live in a society that treats everyone as a free and equal person, as having certain interests in the enjoyment of the freedoms of conscience, expression, and association and in the possession of socioeconomic goods that enhance their ability to pursue their ethical goals. On this view, because we have a duty to support political institutions that treat us as equals, our education should, where feasible, encourage us to recognize that duty. If the moulding of political convictions is sometimes permissible, further questions arise about the limits of permissible shaping. If we adopt Rawls’s view, then educational institutions might encourage the democratic virtues of toleration, non-​discrimination, open-​mindedness, deliberation, and courage to challenge perceived injustice, and discourage political attitudes that are more congenial to authoritarian regimes. More controversially, his conception supports education that discourages certain extreme forms of libertarian thought, such as Nozick’s (1974), which do not acknowledge each citizen’s valid claim to a share of wealth and income that enables them to pursue their goals. However, there are limits to the state’s power to shape citizens’ convictions in pursuit of justice. In Rawls’s conception, political autonomy is a good: it is valuable that free and equal citizens are able to affirm the principles that govern them (1996:  66–​7 1). Even free and equal citizens disagree about the details of the rules that should govern them, and the good of political autonomy requires the state not to side with any particular side within those disagreements about how free and equal institutions ought to be regulated. If it adopted a particular controversial political conception and asked its schools to promote it, then those citizens who disagree would be deprived of the opportunity to endorse the state. In the Rawlsian view, then, the limit of the state’s permission to shape citizens’ political convictions in a Rawlsian society aligns with the

450   Matthew Clayton point at which ‘reasonable pluralism’ emerges. In any society, there will be disagreement between individuals who share the ideals of freedom and equality, but interpret them in somewhat different ways. The disagreement between the three distributive views above—​MEO, OLA, and educational adequacy—​is a case in point. Even if MEO were true, it would not be within the remit of the state to try to encourage children to affirm it. Other political conceptions that permit the shaping of political convictions might distinguish between legitimate and illegitimate political education in different ways. Rawls’s account has both procedural and substantive components because it encourages citizens’ allegiance to both the democratic process and to the pursuit of particular laws and policies within that process (Rawls 1996: 428). Those who reject Rawls’s view that legitimacy depends to some extent on the satisfaction of certain fundamental socioeconomic interests, for example, might support a citizenship education curri­ culum with the more limited aims of encouraging a commitment to democratic decision procedures.

20.2.2 Perfectionist or Anti-​Perfectionist Education? I turn now to the issue of ethical education. Is it permissible for the state to use educational institutions to encourage particular religious beliefs or convictions about how to live well? The answer to that question may depend on whether political perfectionism or anti-​perfectionism is sound. Perfectionists defend the view that if state-​regulated educational institutions are capable of promoting the right view of our place in the universe and imparting sound convictions about how to live well, then they are permitted to exercise that role (Raz 1986). The implications of perfectionism for education turn on what it means to live well. If, as is widely believed, personal autonomy is valuable and there are countless different ways of leading a good life, then schools have a responsibility to equip individuals with the understanding of the range of valuable goals and relationships, the skills to deliberate about which goals they are particularly suited to, and the intellectual means of pursuing the goals they adopt. But, importantly, perfectionists assert that if schools can successfully steer children away from goals and relationships that make people’s lives go worse, then they ought to do that: if living well requires religious commitment, for example, then there may be reason for schools to have a religious studies syllabus that discourages atheism. Anti-​perfectionists, like Rawls, appeal to political autonomy in defence of the view that it is impermissible to use state-​run schools to promote particular conceptions of religion or of living well—​Rawls groups these under the heading of ‘comprehensive doctrines’. Political autonomy requires that citizens endorse the rules that govern them. In a free society that protects freedom of conscience, expression, and association, disagreement about comprehensive doctrines is a permanent fact of life. Thus, if the state used its educational institutions to promote particular religious goals or specific kinds of relationship, some citizens would inevitably reject the laws and policies that constrain their conduct with respect to their children, for example. For the stake of the political

Education   451 autonomy of its citizens, then, the government must not ask its schools to favour particular comprehensive doctrines (Rawls 1996).15 Further questions arise for anti-​perfectionism. One concerns whether the state may use schools to encourage personal autonomy, the ability and desire to decide for oneself how to live one’s life, in preference to non-​autonomous lifestyles in which, for example, children follow parents in their ethical ambitions, without serious reflection of the merits of doing so or significant exposure to alternative lifestyles. While some argue that the state must not take a stand even on that question, the dominant anti-​perfectionist view draws a distinction between the details of the final ends that individuals might pursue, about which the state remains neutral, and the way in which people hold their comprehensive ends, which it is proper for the state to address. Feinberg (1992) describes this commitment in terms of ‘the child’s right to an open future’. Even though the state remains anti-​perfectionist with respect to religion, it ought to give every individual the wherewithal critically to reflect on the various comprehensive doctrines that are available and to choose and follow one without legal or social pressure from others.

20.3 Parents’ Rights The previous two sections reviewed some debates about the principles that should guide the government in its education policy. But principles of justice also apply to individuals and, particularly with respect to children, it is important not to overlook questions concerning the rights and duties of parents, because, like governments, parental conduct makes a significant difference to children’s lives. Issues of parental justice regarding education might be divided into three broad categories. What are the appropriate principles for the allocation of authority to make decisions that affect children’s lives, including choice of school? Under what circumstances is it morally permissible or required for parents to advantage their children? And, under what circumstances are parents morally permitted or required to shape their children’s values?

20.3.1 The Allocation of Educational Authority Should parents have the legal right to decide how their child is educated—​whether at home, in a private school, or among the various state-​financed schools that are available? One approach to answering that question is educational consequentialism: legal rights over children’s education should be allocated in a way that produces the best 15  This objection to the promotion of particular comprehensive doctrines in state-​run or state-​funded schools is compatible with the promotion of a sense of justice in children, outlined earlier, because the reason to encourage just conduct is that everyone is duty-​bound to treat others justly. We are under no duty to others to live well.

452   Matthew Clayton educational outcome. The best outcome here need not be interpreted as ‘best for that particular child’, because no one is entitled to an education that is best for her—​if, for example, a side effect of that upbringing is that others are disadvantaged. A consequentialist might appeal to a principle of educational justice such as OLA, MEO, or educational adequacy and propose that educational decisions ought to be made by those agents who are most likely to realize the chosen principle. That said, one prominent consequentialist argument for parental choice over schooling is that parents are the best judge of their child’s interests. It might be appropriate for the government to require that any education that is offered is sufficient to bring the child up to a certain level of competence in various disciplines, but state control of education is likely to produce a more uniform school system that is less capable of improving itself than privately run schools that are free to experiment in what and how they teach (Mill 1859: Ch. V, paras 12–​14). This argument for parental choice needs more work, however, because parental freedom with respect to educational choices might produce unjust inequalities between the parents’ own child and other children—​if, for instance, rich parents devote considerable resources to their child’s education that, as a side effect, unjustly worsens other children’s occupational prospects. Many people believe that educational consequentialism is mistaken because parents’ rights over children should not be viewed merely as a means of producing good outcomes for the child or others: it overlooks the fact that parents have a fundamental claim to take certain decisions for children regardless of whether their taking those decisions would produce desirable outcomes (Brighouse and Swift, 2006b, 2008, 2014b; Clayton 2006: 48–​61). In effect, this position holds that parents have the right to take certain decisions with respect to their children’s education, even when others taking them would benefit their child more or better realize just social arrangements. If there is a fundamental parental right to decide certain aspects of the child’s education, the nature and extent of that right is disputed, because there is disagreement about what interest or claim grounds the right. Galston (2002) equates the denial of parental choice over schooling with the denial of parents’ ‘expressive liberty’—​the interests of individuals to live in accordance with their own conceptions of what gives life meaning—​because, while there are constraints on it to protect the child from dependency and vulnerability, the freedom to educate one’s own children according to one’s own values is an important part of the freedom to pursue one’s own values. But this argument requires further defence. It is a standard feature of liberal political morality that individuals are not morally permitted to use political power to make others pursue a particular religion, even if the conversion of others is part of their worldview; if that is accepted, it is puzzling why that prohibition does not also serve to forbid individuals using parental power to pursue their ethical values by making the child over whom they have custody practise their religion. A more plausible defence of fundamental parental rights cites our interest in securing familial relationship goods: it argues that parents have an interest in developing and maintaining an intimate relationship with particular children (Macleod 2002;

Education   453 Brighouse and Swift 2014b). The maintenance of such a relationship demands that parents should be permitted to engage in certain activities, such as cooking for or reading to their child, and should have discretion over certain aspects of the child’s life, such as whether he or she joins them at religious services, for example. Importantly, however, this account does not generate significant parental rights with respect to school choice. In the first place, the relationship goods that are picked out as relevant are threatened by certain kinds of school choice, as when parents choose to send their child to a boarding school. Second, parents not having the right to choose their child’s school need not jeopardize the value of the family, because there is sufficient opportunity to pursue loving familial relationships out of school hours. True, the relationship good account limits the extent to which certain distributive ideals can be realized, because the way in which different families are organized, and the unequal resources different parents bring to their role, can set back the pursuit of particular distributive ideals. Nevertheless, as the advocates of familial relationship goods acknowledge, the account does not give parents a fundamental right to send their child to an advantage-​conferring private school or a religious school. On this view, the regulation of school choice reverts back largely to a consequentialist calculation.16

20.3.2 Parental Partiality We turn, finally, from a concern with the legal rights parents should enjoy to the moral options that are available to them. Under what circumstances is it morally permissible or required for parents to advantage their children educationally? In the UK, about 7 per cent of children are educated privately, which involves parents spending sometimes considerable sums on education in addition to their tax liability. In other countries, educational advantage is obtained by parents buying more expensive houses in educationally well-​ resourced districts. Are rich parents morally required to advantage their children in these ways? Are they morally permitted to do so? A view that is sometimes expressed in public discourse is that parents are duty-​bound to give their children the best education they can afford. Although popular, however, the view is unsound, because such a moral requirement places too many demands on parents. Parents are also individuals with their own lives to lead; the moral requirement maximally to benefit their children educationally would require parents to make potentially massive sacrifices with respect to the pursuit of their non-​parental goals if that were necessary to achieve even merely modest improvements in their child’s education. That demanding moral rule can reasonably be rejected even by the child who, looking at her life ahead, should understand that morally required educational benefits in

16 

For further discussion of different conceptions of parental authority, including his own ‘creative self-​extension’ account, see Macleod (2018).

454   Matthew Clayton childhood come at the price of having very costly duties should she become a parent in the future.17 A different view, which is widely endorsed in popular opinion, is that, although parents are not morally required to give their child educational advantages that other children do not receive, they are morally permitted to confer such benefits. But, again, this view is difficult to sustain. Suppose that a parent wants to educate his child about the wildlife in their local wood, but he knows that a side effect of doing so is that thousands of wasps will become agitated and sting children playing in a neighbouring housing estate. If the stings are sufficiently numerous and painful, he is not morally permitted to benefit his child by taking him to the wood. Similarly, if buying advantage-​conferring education for one’s child has negative consequences for others, then it may not be morally permitted. That would turn on whether the gift is consistent with the right principle for distributing educational and occupational opportunity between citizens, which we reviewed in section 20.1. MEO and other egalitarian conceptions of educational opportunity are likely to be more restrictive with respect to permissible parental partiality compared to adequacy views, for example. The relationship good account provides a way of permitting some parental partiality. In this view, although parents are not required to confer relative advantage to their child, they are often permitted (perhaps required) to sustain an intimate loving relationship with them, which may confer such advantage as a side effect (Macleod 2010; Brighouse and Swift 2008, 2014b). One feature of this view that might be questioned is its perfectionism. Suppose that Ailsa expresses her love for her child by reading her bedtime stories, which she executes in a particularly engaging way; by contrast, Beatrice, who knows that she is particularly bad at reading to her child, employs someone to perform that task. On the relationship good view, Ailsa’s conferral of advantage is protected because it is a by-​product of her having the right kind of relationship with her child; Beatrice’s employment of a tutor is not protected even though it appears to be just a different way of expressing her love for her child. The question is whether the permission to confer advantage or not should turn on a particular, controversial, conception of the right way to practise parental love. Anti-​ perfectionists prefer an account of parental partiality that does not depend on a controversial account of the significance of different kinds of giving for human flourishing (Clayton 2012b). It is also an open question how much parental partiality the relationship good account protects. It is uncontroversial to observe that the children of well-​resourced parents receive advantages denied to others: merely living with well-​educated parents enhances one’s development in countless ways. However, consider a case of avoidable partiality that might be advanced by an advocate of MEO: Matthew and Katherine share 17  This short rejection of the moral requirement maximally to benefit one’s child educationally trades on an important fact about justice between age groups. When identifying the best principles, we need to attend to our interests as children—​as the beneficiaries of good education—​but also to our interests as parents and adults in not being liable to pay for education (Clayton 2015).

Education   455 the routine bedtime parental tasks of making a drink and reading a story by alternating the jobs. However, Katherine’s story-​telling confers more advantage on their children than Matthew’s, because Katherine introduces their children to a wonderful range of regional accents and reads with passion and verve; Matthew’s story-​telling is drab. Given that they can rearrange the nightly tasks in a way that upsets equality less—​by Matthew always reading the story and Katherine always making the cocoa—​aren’t they under a duty of justice to do so? Brighouse and Swift might resist the claim that Katherine is not morally permitted to read when Matthew’s dreary reading would upset equality less, by pointing out the importance for the relationship good of being free to act spontaneously and without excessive self-​monitoring. But, as in other moral contexts, egalitarians might insist that issues of proportionality must be considered. If Katherine’s stories confer considerable advantage it is surely the case that self-​monitoring is morally required. It may be that, because the relationship good can be realized by sharing cocoa over a drab story, the loss of spontaneity and loss of freedom from self-​monitoring is outweighed by the gain in equality. When considering the moral permissibility of their actions, then, egalitarian parents should conform to certain rules of thumb to ensure that their relationship-​realizing conduct is not disproportionately inegalitarian. On this view, the relationship good account might not protect much inequality-​generating parental partiality.

20.3.3 The Comprehensive Enrolment of Children Even if parents lack the right to choose the school that their child attends, there is a separate question of whether they are morally permitted to use their custodial rights to shape her comprehensive values—​her religious commitments, for example. Most people believe that they are. In defence of that view, some appeal to parents’ expressive liberty as we have discussed. A simpler defence claims that parental shaping of the child’s religious values is harmless if the child is also given an education that equips her with the powers to make informed, reasoned choices for herself when she is an adult (the open future account). Whether the common view is right depends on whether the child is wronged by being enrolled in a religious view before she has developed an autonomous will of her own. One argument for the view that children are wronged by comprehensive enrolment appeals to Rawls’s political autonomy argument for anti-​perfectionism (Clayton 2006, 2012a, 2014). If anti-​perfectionists are right that unchosen political arrangements should be regulated by principles that can be endorsed by citizens regardless of their distinctive comprehensive convictions, then, because the upbringing children receive is similarly unchosen, the activities of parents should be similarly regulated. Parental conduct should be guided by principles that cannot reasonably be rejected by their children. Of course, children, particularly young children, do not possess the mental faculties to accept or reject the principles that guide their upbringing; but in the normal

456   Matthew Clayton course of events they will later have such capacities and can, therefore, retrospectively reject or accept the principles that guided their parents. Since, in a free society, parents cannot tell which particular comprehensive views their children will endorse, the need for retrospective consent supports parenting regulated by principles that serve their children’s interests as free and equal persons, but which refrain from promoting controversial conceptions of religion. If adopted, this view may also have distributive implications. One source of socioeconomic inequality in our society is the different attitudes towards work, and occupational and financial success that people are encouraged to develop in childhood. Other things equal, those who are encouraged to embrace a work ethic are more likely to succeed in education and work than those whose parents impart to their children scepticism about the value of work or take a more relaxed attitude to the alleged virtues of hard work. If parental anti-​perfectionism is sound, then parents wrong their child by inculcating a work ethic in their children as much as if they try to make them devout;18 our attitudes to work, at least in societies sufficiently rich to maintain just institutions, should be our own choice. If parents abided by that restriction, then the pattern of inequality in Western societies, which presently exhibits very low relative mobility between classes, might change somewhat in character. Parental anti-​perfectionism is a controversial view. Some reject the view that imparting to children the right view of religion, for example, violates children’s autonomy because respect for autonomy is possible only if it interferes with the individual’s development or exercise of the capacity to choose for herself in the right way, which young children lack (Cameron 2012; Bou-​Habib and Olsaretti 2014; Fowler 2014). Others claim that respect for autonomy, even for adults, is compatible with certain kinds of paternalism or education that improves people’s well-​being (Raz 1986). Even if perfectionism is embraced, however, it is not clear that it always permits parents to enrol their children in the religion of their choice, because that religion might not be worth pursuing and may, therefore, set back the child’s interests. One way of trying to avoid that conclusion is to appeal to the ideal that it is intrinsically valuable for parents and children to share an ethical life together or to share comprehensive enthusiasms (Brighouse and Swift 2014b). But it is hard to believe that sharing enthusiasms is always unconditionally intrinsically valuable. If it does not improve one’s life to pursue comprehensive goals that are not worth pursuing, it is not clear how pursuing those goals with one’s parents is any—​or sufficiently—​different. So, the rejection of anti-​perfectionism would not rescue parents’ moral rights to enrol their children into ethical traditions they endorse. It supports the different conclusion that parents are morally permitted to enrol their child only into comprehensive practices that enhance the child’s well-​being.

18  I have in mind a hard work ethic that goes beyond a work ethic that is necessary for an individual to attend to their duties to others and to operate as an autonomous agent. Inculcation of the latter seems permissible and, perhaps, required if the argument of §20.2.1 is right.

Education   457

20.4 Conclusion In this chapter I have tried to explain how issues concerning both the content and distribution of education are connected to wider debates about justice, and how alternative conceptions of political morality have different implications for educational justice. One notable feature of the existing literature is its preoccupation with different goods and how different kinds of education might deliver valuable items to children, parents, and others. Less work on educational justice has been done by non-​consequentialist theorists of morality who argue that individuals enjoy moral options but are subject to moral constraints; in other words, they claim that individuals are often morally permitted to pursue their own goals, even when that pursuit is bad for them or bad for others, and they are sometimes forbidden from treating others in certain ways even if doing so is good for them or good for others.19 Because education involves children who cannot consent or withhold consent to the education they receive, it is not straightforward for non-​consequentialists to identify the options and constraints that apply to educators, be they parents or teachers. Debates about the nature, significance, and implications of autonomy within education have made some progress in addressing these issues; but debates about the degree of partiality parents are permitted to display towards their own children, and issues concerning the shaping of individuals’ beliefs and ambitions, might profit from further engagement with non-​consequentialist conceptions of morality and politics.

Acknowledgements For helpful comments and conversations about this chapter, I am grateful to Paula Casal, Serena Olsaretti, Tom Parr, Adam Swift, and Andrew Williams.

References Anderson, E. (2007). ‘Fair Opportunity in Education:  A Democratic Equality Perspective’. Ethics 117(4): 595–​622. Arneson, R. (1999). ‘Against Rawlsian Equality of Opportunity’. Philosophical Studies, 93(1): 77–​112. Bou-​Habib, P. and Olsaretti, S. (2014). ‘Autonomy and Children’s Well-​Being’, in A. Bagattini and C. Macleod (eds) The Nature of Children’s Well-​Being. Dordrecht: Springer, pp. 15–​33. Brighouse, H. (2000). School Choice and Social Justice. Oxford: Oxford University Press. Brighouse, H. and Swift, A. (2006a). ‘Equality, Priority and Positional Goods’. Ethics, 116(3): 471–​97. 19 

For discussion of the differences between consequentialist and non-​consequentialist conceptions, see Kagan 1989.

458   Matthew Clayton Brighouse, H. and Swift, A. (2006b). ‘Parents’ Rights and the Value of the Family’. Ethics, 117(1): 80–​108. Brighouse, H. and Swift, A. (2008). ‘Legitimate Parental Partiality’. Philosophy and Public Affairs 37(1): 43–​80. Brighouse, H. and Swift, A. (2009). ‘Educational Equality v Educational Adequacy: A Response to Anderson and Satz’. Journal of Applied Philosophy 26(2): 117–​28. Brighouse, H. and Swift, A. (2014a). ‘The Place of Educational Equality in Educational Justice’, in K. Meyer (ed.) Education, Justice and the Human Good. London: Routledge, pp. 14–​33. Brighouse, H. and Swift, A. (2014b). Family Values: The Ethics of Parent–​Child Relationships. Princeton, NJ: Princeton University Press. Brown, G. (2009). ‘I Want to Free Teachers to Perform Small Miracles’. Times Educational Supplement, 30 October. Cameron, C. (2012). ‘Clayton on Comprehensive Enrolment’. Journal of Political Philosophy, 20(3): 341–​52. Casal, P. (2007). ‘Why Sufficiency is not Enough’. Ethics 117(2): 296–​326. Casal, P. (2018).  ‘Distributive Justice and Human Nature’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 259–82. Clayton, M. (2001). ‘Rawls and Natural Aristocracy’. Croatian Journal of Philosophy, 1(3): 239–​59. Clayton, M. (2006). Justice and Legitimacy in Upbringing. Oxford: Oxford University Press. Clayton, M. (2012a). ‘The Case against the Comprehensive Enrolment of Children’. Journal of Political Philosophy 20(3): 353–​64. Clayton, M. (2012b). ‘Equal Inheritance:  An Anti-​Perfectionist View’, in J. Cunliffe and G. Errreygers (eds) Inherited Wealth, Justice and Equality. London: Routledge, pp. 98–​118. Clayton, M. (2014). ‘Anti-​Perfectionist Childrearing’, in A. Bagattini and C. Macleod (eds) The Nature of Children’s Well-​Being. Dordrecht: Springer, pp. 123–​40. Clayton, M. (2015). ‘How Much Do We Owe to Children?’,  in S. Brennan, S. Hannan, and R. Vernon (eds) Permissible Progeny. New York: Oxford University Press, pp. 246–​64. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Dworkin, R. (2011). Justice for Hedgehogs. Cambridge, MA: Harvard University Press. Feinberg, J. (1992). ‘The Child’s Right to an Open Future’, in J. Feinberg (ed.) Freedom and Fulfilment: Philosophical Essays. Princeton, NJ: Princeton University Press, pp. 76–​97. Fishkin, J. (2014). Bottlenecks:  A New Theory of Equal Opportunity. New  York:  Oxford University Press. Fowler, T. (2014) ‘Perfectionism for Children, Anti-​Perfectionism for Adults’. Canadian Journal of Philosophy, 44(3–​4): 305–​23. Galston, W. A. (2002). Liberal Pluralism: The Implications of Value Pluralism for Political Theory and Practice. Cambridge: Cambridge University Press. Jencks, C. (1988). ‘Whom must We Treat Equally for Educational Opportunity to be Equal?’ Ethics 98(3): 518–​33. Kagan, S. (1989). The Limits of Morality. Oxford: Clarendon Press. Lippert-​Rasmussen, K. (2004). ‘Are Some Inequalities more Unequal than Others? Nature, Nurture, and Equality’. Utilitas 16(2): 193–​219. Macleod, C. (2002). ‘Liberal Equality and the Affective Family’, in D. Archard and C. Macleod (eds) The Moral and Political Status of Children. Oxford: Oxford University Press, pp. 212–​30. Macleod, C. (2010). ‘Justice, Educational Equality, and Sufficiency’. Canadian Journal of Philosophy 40(sup1): 151–​75.

Education   459 Macleod, C. (2018). ‘The Family’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 415–37. Mason, A. (2006). Levelling the Playing Field: The Idea of Equal Opportunity and its Place in Egalitarian Thought. Oxford: Oxford University Press. Mill, J. S. (1859). On Liberty. Nagel, T. (1997). ‘Justice and Nature’. Oxford Journal of Legal Studies 17(2): 303–​21. Nozick, R. (1974). Anarchy, State and Utopia. Oxford: Blackwell. Pogge, T. (1989). Realizing Rawls. Ithaca, NY: Cornell University Press. Rawls, J. (1996). Political Liberalism, paperback edn. New York: Columbia University Press. Rawls, J. (1999). A Theory of Justice, revised edn. Cambridge, MA: The Belknap Press of Harvard University Press. Raz, J. (1986). The Morality of Freedom. Oxford: Clarendon Press. Satz, D. (2007). ‘Equality, Adequacy, and Education for Citizenship’. Ethics 117(4): 623–​48. Swift, A. (2003). How not to be A Hypocrite: School Choice for the Morally Perplexed Parent. London: Routledge. Tomasi, J. (2012). Free Market Fairness. Princeton, NJ: Princeton University Press. Vallentyne, P. (2007). ‘Of Mice and Men’, in N. Holtug and K. Lippert-​Rasmussen (eds) Egalitarianism: New Essays on the Nature and Value of Equality. Oxford: Clarendon Press, pp. 211–​38. White, J. (1994). ‘The Dishwasher’s Child: Education and the End of Egalitarianism’. Journal of Philosophy of Education 28(2): 173–​81. Young, M. (1958). The Rise of the Meritocracy. London: Thames and Hudson.

Chapter 21

Healt h Shlomi Segall

21.1  Why Justice in Health? Why care about justice in health? We might think, rather, that a theory of justice in health is redundant. After all, we don’t have a separate and special theory of justice for every commodity out there, not even for important ones, say housing. So, what’s so special about health? Why, in other words, a theory of justice in health? There may be practical as well as theoretical advantages to theorizing about justice in health. Health represents in many countries a significant share of GDP (in the US, well above 15 per cent). Given that it is such a major area of public policy, justice should surely have something to say about the just allocation of those large funds. But a second, and more theoretical reason is that health might offer us a good venue in which to test (general) theories of justice. It might be the case that there is something about health such that it puts to the test, in a unique way, our intuitions with regard to the various theories of justice. In short, we might be interested in justice in health because, first, it might teach us something about the ethical distribution of health, but also, second, because it might teach us something about justice. In what follows I shall explore different accounts of justice in health. I shall attempt to do so chronologically, tracing the evolution of theorizing (or lack of) about it by Rawls and his predecessors, continuing with the rival theories that have sprung up in more recent years. In studying and comparing these competing theories I shall focus on three questions in particular. These concern the issues of currency, pattern, and scope (or, the what, how, and who): • Currency—​What is the currency the re-​distribution of which achieves justice in health? • Pattern—​ How ought that currency be distributed? Along which patterns of distribution? • Scope—​What is the scope of justice in health? Among whom does it obtain?

Health   461 In this chapter I will attempt to examine justice in health by weaving both methods of investigation. I shall try and outline the competing accounts of justice in health by tracing the evolution of that field of research, and I shall try, en route, to outline the different responses they give to these three focal questions. (This is, of course, not to deny that there are other important questions to ask with regard to justice and health, and we shall try and touch also on those.)

21.2  Rawlsian Justice in Health Let us start at the beginning. And, in the beginning, of course, there was Rawls. In Rawls’s theory, as far as health is concerned, there is a lacuna, for in A Theory of Justice, he has bracketed the issue of health. Rawls did so by assuming that in the original position everyone is healthy and able-​bodied. Indeed, as far as one can judge, in A Theory of Justice nobody ever gets sick or dies! In constructing the original position in this way, questions concerning morbidity and mortality were thereby bracketed from the most important contemporary theory of justice. Rawls acknowledged this lacuna, of course, and indeed it would have been unreasonable to expect his comprehensive and path-​ breaking account to address every conceivable question. Still, it fell to Rawls’s successors to plug that important hole. The main puzzle for Rawls’s followers, then, was where exactly, within the Rawlsian scheme, ought healthcare to be placed. The obvious answer would have been that healthcare is a primary social good, the kind of goods that are to be distributed along the lines of the difference principle. (The difference principle says that inequalities, in primary social goods, are justified so long as they maximally improve the position of the worse off.) Primary Social Goods are goods that “every rational man is presumed to want.” These are things that have value “whatever a person’s rational plan of life.” In Rawls’s list these include “rights and liberties, powers and opportunities, income and wealth” (Rawls 1971: 62). This seems like a natural place for healthcare to fit in. But early critics such as Amartya Sen and Kenneth Arrow have observed the following unsavory implication of such a move (Sen 1980; Arrow 1973). Famously, the difference principle adopts a maximin pattern with regard to the bundle of social primary goods. (It decrees that inequalities are just [or not unjust] when they maximize the position of the worse-​off.) One obvious implication of including healthcare in the primary goods basket is that inequalities in it become acceptable, so long as they (maximally) improve the position of the worse-​off. Rawls’s scheme, in other words, is compatible with a two-​ tier (or indeed multi-​tier) system of healthcare. It is compatible with a public healthcare system that offers the poor second-​grade care. But that is not all. Given that it would be part of the bundle of social primary goods, healthcare may be used to improve the position of the worse-​off compared to better-​off individuals (or groups). In other words, healthcare would be prioritized to the (overall worse-​off) even when they happen to be healthier than those who are overall better off. Marking healthcare as a Rawlsian

462   Shlomi Segall primary good thus has the unhappy double-​barreled result of potentially discriminating in healthcare both against the poor, and in a different way, against the rich. Now, the target of Sen’s and Arrow’s criticism was not so much Rawls’s treatment (or lack thereof) of healthcare, but the more general use of primary goods as his currency of justice, as well as the employment of maximin (namely, maximize the minimal [or worse-​off] position in a given set) as his preferred pattern of distribution for these goods. We may note that this is perhaps the earliest incidence of something we have mentioned earlier, namely, the use of healthcare as a testing ground of general arguments about justice. Rawlsians, and in particular those wishing to offer a Rawlsian account of just­ ice in healthcare, thus had their work cut out for them. Healthcare cannot, unless one is willing to bite the two bullets that have just been mentioned, be plugged into the list of social primary goods. (Or, at least, it cannot be so plugged without some additional constraints.) Enter Norman Daniels. With the hindsight of more than thirty years, it is perhaps easy to forget how original Daniels’s theory was. His idea was to approach the Rawlsian dilemma with regard to healthcare from a completely different angle. Healthcare, reasoned Daniels, is no mere primary social good. Health (and healthcare) is more special than that because it has a significant contribution to our ability to carry out the life plans that we each devise for ourselves (Daniels 1985). Simply put, we need to be healthy in order to be able to pursue our respective life plans. Making sure that everyone is as healthy as could be would thus make a significant contribution towards equality of opportunity in individuals’ ability to pursue their life plans. Rather than placing healthcare under the auspices of the Difference Principle (DP), then, as one would have expected, Daniels’s argument allows for placing it under the lexically prior Fair Equality of Opportunity Principle (FEOP). This strategy has the happy outcome of avoiding the two problematic implications that have been mentioned before. For, in difference from the prioritarian DP, the FEOP mandates strict equality (with regard to the said opportunities). It does not, in other words, tolerate inequalities, in this case in access to healthcare. Daniels’s equality of opportunity (EOP) argument, in short, solves with one stroke both the problem of discrimination against the poor and that of discrimination against the rich, and ensures an equal and universal provision of healthcare. Recalling the three core questions of distributive justice, we can glean Daniels’s answer to the question of pattern. Linking health with the principle of FEOP has enabled him to justify an egalitarian, as opposed to prioritarian (as the DP would have mandated) pattern of distributing healthcare. But another important plank of Daniels’s account concerns the question of currency. This was, recall, the question of what it is that a just healthcare system distributes. Or, to put this more practically, what is included in the basket provided by the just healthcare system? (We may set aside the third question, that of scope, for the time being.) On Daniels’s account, his defense of egalitarian healthcare cannot be understood independently of his choice for the currency of justice in healthcare. Let me explain. Recall that the premise of the account was that we need to be healthy in order to pursue our respective

Health   463 life plans. Suppose we can identify a maximal level of health that humans may currently enjoy (say, the healthy life expectancy of the average Japanese woman, just to illustrate). And suppose it is possible to bring everyone, whether they are ill now or will be in the future, to that maximal level. In having done so, we will have made a significant contribution towards equalizing individuals’ opportunity to pursue their respective life plans (given, as we said, that health is crucial for individuals’ ability to do so). This argument, then, indicates not only the correct pattern of just healthcare (namely, strict equality), but also its currency. That is, justice in health, if we follow Daniels, requires bringing everyone up to maximal health. To the “what” question we posed earlier, namely what should a just healthcare system provide, Daniels’s answer is, whatever medical care is needed to bring everyone to maximal health. And, the latter he understands as the maximal normal functioning level typical of the human species. In specifying “health” to consist of “normal species functioning,” Daniels relies on Christopher Boorse’s “biological” (as opposed to social) account of health. In a nutshell, Boorse’s approach to defining health was for us to look, as biologists, at the human body and identify what is a normal function for it, and what, in contrast, constitutes pathology. It is to that level of normal species functioning, according to Daniels, that we ought to aim. The use of “normal species functioning” (NSF) as the currency of justice in health has many advantages. First, it allows us to give a rather determinate answer to the question of currency. Namely, a just healthcare system should confine itself to curbing pathologies (as opposed to enhancing biological capabilities beyond their normal level). Once we identify those, we would have a pretty good idea of the perimeters of healthcare. This answer, second, is not only quite determinate and easy to follow for policymakers; it is also intuitively rather compelling. We commonly think that what hospitals and healthcare systems generally do is, well, cure and prevent illnesses, that is, address shortfalls to normal species functioning. Third, Daniels’s NSF argument has the happy result of avoiding public healthcare being hijacked by quasi-​medical expenditures. Should publicly funded healthcare include vanity-​driven plastic surgeries, memory-​improving pills, and non-​therapeutic abortion clinics? A “social” or well-​being-based definition of health would have perhaps mandated precisely that, something against which the intuitions of most people probably rebel. Daniels’s alternative is simple and initially persuasive:  only those medical procedures that can be said to address pathologies should be included in the healthcare basket. Just to give one example, reconstructive breast surgery for women who have undergone mastectomy is a procedure that helps restore them to NSF, and should therefore be funded by public healthcare (something most of us find intuitive). In contrast, women who wish to undergo plastic surgery for aesthetic reasons do not deviate from normal species functioning, no matter how unhappy they are with their bodies.1 So the third, and perhaps most important, implication of Daniels’s NSF account is that it seems to accord with our intuitions 1  Notice that in this Daniels is taking NSF to be concerned not merely with “functioning” strictly speaking, but restoring patients to the normal form and shape of their bodies, which I think is a plausible expansion of the concept.

464   Shlomi Segall about what just healthcare should and should not include (this, again, in contrast to an account of justice in healthcare whose currency relies on a social, rather than a biological, understanding of health).

21.3  Problems with Daniels’s Rawlsian Account Daniels’s account remains the centerpiece of justice in health. Like Rawls’s theory itself, Daniels’s is the one theory to which all others must relate. It has therefore, deservedly, received much critical attention. Two of the criticisms the literature raises are particularly noteworthy. One concerns the confinement of the currency of justice in health to normal species functioning, whereas the second concerns Daniels’s anchoring of justice in health in Rawls’s FEOP and equality of opportunity more generally. Let us examine these in turn.

21.3.1 The Currency of Justice in Healthcare:  Treatment vs Enhancement Consider first Daniels’s delimiting of just health to addressing pathologies, that is, deviations to NSF. We said that such an account has clear advantages in terms of clarity and plausibility. But it has also been noted that the account may lead to arbitrary and counterintuitive policy recommendations. Daniels himself, in typical intellectual integrity, indicated as much by invoking the following real-​life case. Consider two equally very short six-​year-​old boys. Both are at the low fifth percentile of height for their age, say. There is one important difference between the two, though. Boy A suffers from a deficiency in the growth hormone, while boy B does not. The latter’s short stature is simply the result of having short parents (i.e., his genes). Boy A suffers from pathology (deficient growth hormone), while boy B does not. (To be clear, on the biological account, short stature does not, as such, constitute pathology. What constitutes pathology is the deficient growth hormone). We can easily see why this constitutes a problem for the NSF account: since he suffers from pathology boy A deserves, on that account, publicly funded growth hormone therapy, while equally short boy B, who suffers no pathology (but who stands to benefit from hormone therapy all the same), does not. This is counterintuitive, as Daniels himself was the first to admit (Daniels 2008: 119–​120). But the shortcomings of the NSF account are not confined to such rare examples. Holding Daniels to the letter of NSF theory implies no public funding for any medical condition that cannot properly be described as pathology. Consider this. Pregnancy, labor, menstruation, and old age are part of NSF. Even the pain

Health   465 that accompanies labor is not itself a deviation from the healthy functioning of the body. (It probably wouldn’t be called “labor” otherwise.) But nobody would seriously suggest that treatment of such pain, as well as medical care throughout pregnancy, pain relief for menstruation, and all manner of treatments to slow and alleviate the ills of aging should fall outside the purview of public healthcare. Daniels’s NSF account seems, in that respect, unduly narrow (Segall 2010: 128). On a closer look, then, the confinement of healthcare to addressing NSF is not as intuitive as it may initially seem. There is no reason, or at least we have yet to see one, to restrict the currency of justice in health to treatment rather than (also) enhancement (understood as medical procedures that boost our health beyond NSF). However, one reason that might motivate the restriction of just healthcare to NSF concerns the principle of equality of opportunity (and here we arrive, notice, back at the question of pattern). That is to say, it might be suggested that a commitment to equality of opportunity, which for the moment we shall accept at face value, necessitates adherence to NSF. The thought is that we have a reason of justice to bring everyone to that level of NSF as a means of promoting equality of opportunity (to fulfill life plans). The NSF is thus operating as a useful benchmark for promoting equality of opportunity. So the fact that NSF does not capture things we commonly do want to associate with healthcare might not matter (or matter less) since it is the case that it performs an important role in securing equality of opportunity, or so the thought goes. But it is actually rather easy to see that equality of opportunity does not at all justify a reliance on NSF. This is so for the simple reason that any threshold would have done in that respect. The account’s commitment, recall, was to bring everyone to an equal standard of health in order to curb inequality in opportunities (to pursue life plans). But that commitment could have been met with an adherence to a 90 percent of NSF (for everyone), or 110 percent, or 20 percent, or indeed 0 (letting everyone die). None of these is perhaps very attractive or practical. But that’s beside the point. As a matter of principle, equality of opportunity would have been achieved to exactly the same extent by providing partial health, no healthcare at all, or by commitment to (equal) enhancement (that is, beyond 100%) of NSF. Equality of opportunity, in short, does not entail confining healthcare to NSF; nor, we saw earlier, does that confinement conform to our deep-​seated intuitions, upon reflection, of what just healthcare ought to do.

21.3.2 The Pattern of Justice in Healthcare:  Fair Equality of Opportunity? There does not seem to be a good reason to think that justice in healthcare ought to be confined to treatment rather than also enhancement. But equally, as we just saw, there are problems with Rawls’s choice of the pattern of justice in health. A commitment to

466   Shlomi Segall equality of opportunity is compatible with ensuring that everyone is equally sick and miserable (see also Kamm 2001). And there are other, more particularly fairness-​based objections to grounding healthcare in Rawls’s FEOP. I can think of three such objections. The first, if successful, shows that Daniels’s account does not justify universal healthcare; the second, that it does not justify in kind healthcare; and the third, that it may not justify healthcare at all. Here is the first objection. Recall that one problem with placing healthcare under the auspices of the difference principle entailed the problem of discrimination against the rich. Now, it is not at all clear that by shifting to the FEOP Daniels manages to overcome that problem. Promoting overall equality of opportunity in individuals’ ability to pursue their life plans could actually be assisted better by differential healthcare, one that advantages the worse-​off members of society (say, the poor). In other words, far from justifying universal healthcare, Daniels’s account may end up turning the rich away from emergency rooms! Here is the second problem, the one involving the provision of in-​kind healthcare services. The underlying rationale of Daniels’s account, we said, is equality of opportunity to fulfill life plans. Admittedly, it is generally the case that good health helps us pursue our respective life plans. But surely this is not always the case. It might be the case that for some people, substituting in-​kind medical treatment for its cash equivalent would better facilitate their particular life plan. A  paraplegic concert violinist’s life plans might be better served by a Stradivarius than by an expensive wheelchair (Dworkin 2000: 61–​2). It follows that Daniels’s account not only does not justify universal healthcare; it does not even justify in-​kind provision of healthcare (see also Jacobs 2004: 200). Third, and worse still, Daniels’s FEOP-​ based account cannot justify publicly provided healthcare at all. This last criticism is the upshot of an extensive empirical work in epidemiology of the past several decades. This literature has revealed that the “social determinants of health”—​factors such as income, housing, stress-​free job, and so forth—​may account for up to 80 percent of inequalities in life expectancy between individuals. In other words, healthcare interventions, it turns out, account for only as little as one-fifth of the extent of inequality in life expectancy (Wilkinson and Marmot 2003). The ethical implications of these findings are profound, and not least for Daniels’s account. His account, we saw, aspires to boost the health of the worse-​off, in the name of curbing inequality in opportunity to pursue life plans. But if the “social determinants of health” literature is correct, then we would do well to divert resources away from the factors accounting for only 20 percent of health inequality, and onto those accounting for 80 percent of it. In short, to curb inequalities in health we should invest in the social determinants of health. In fact, we would do well, following the logic of Daniels’s account, to close down hospitals altogether (Sreenivasan 2007)! If what we ought to care about is narrowing health inequalities (in the name of equal opportunity to fulfill life plans), then that objective can be more efficiently pursued by investing all of our current healthcare expenditure in non-​medical determinants of health.

Health   467

21.4  Alternatives to Rawlsian Justice in Health: Sufficientarian Accounts Even though it remains the dominant theory, Daniels’s account of justice in health contends now with several critical alternatives. Two such alternatives might be worthy of our attention. One is a cluster of theories that all recommend a sufficientarian pattern of distribution, and the other consists of accounts whose main trait is their sensitivity to patients’ responsibility for their health, known as “luck egalitarianism.” The sufficientarian family, as indicated, comprises several theories. These include rights-​based theories (Buchanan 1984,  2009), capability-​based theories (Prah Ruger 2010, Venkatapuram 2011), “democratic egalitarianism” (Anderson 1999), and communitarianism (Walzer 1983: Ch. 3). I am less concerned here with the nuances between these than in the common ground they share in opposition to Daniels (and, we shall later see, to luck egalitarianism). A good place to start is one of the earliest responses to Daniels’s account, namely that put forward by Allen Buchanan (Buchanan 1984). To Daniels’s Rawlsian equality of opportunity account, Buchanan responded with one centered on the right to healthcare. The right to healthcare is nowadays treated as a topic worthy of its own consideration (see Wolff 2012), distinct from justice in health, so here we shall confine ourselves merely with the implications of such an account for justice, narrowly understood. The basic idea is that rather than focusing on comparative holdings of healthcare—​that is, how much healthcare people are getting relative to one another—​what we ought to care about is that individuals (or citizens) be guaranteed the amount of healthcare consistent with their civic rights. It is easy to see why this right-​to-​health account is associated with communitarianism or an ideal of “democratic equality.” Healthcare is understood here in terms of a civic right. Rather than measuring how much healthcare (not to mention health) individuals possess relative to others, we ought to specify a “decent minimum of healthcare” (Buchanan’s term). Many readers will correctly associate this suggestion with the ideal of sufficientarianism. What matters, in Harry Frankfurt’s memorable words, is not that people should be equal to each other, but rather that they will have enough (Frankfurt 1987). The common element, then, in all these accounts of justice in health is their recommendation of a sufficientarian pattern of distribution (Powers and Faden 2006). Among these accounts, though, is one variant that need not be grounded in a communitarian idea of civic rights. I refer to capability theories that draw on Amartya Sen’s and Martha Nussbaum’s work on development. These offer their own accounts of justice in health as typically focusing on the central capabilities of life and bodily integrity. These accounts are typically more global in their reach (see the question of scope in section 21.7), a feature that distinguishes them from civic accounts (“democratic equality”). Still, they do share with the latter the sufficientarian feature (Nussbaum 2006: 295, for example). These capability theories typically specify some benchmarks for meeting development goals by specifying what it takes to ensure the key capabilities with regard

468   Shlomi Segall to health (Prah Ruger 2010; Venkatapuram 2011). (This is not to deny that in principle, one could hold to an account that combines the currency of capabilities with a pattern other than sufficiency. It is just to say that capability theorists are generally, as it happens, sufficientarian.) Sufficientarianism in general and the idea of a decent minimum of healthcare in particular have much appeal. For one thing, they capture the following intuition. We think of doctors and hospitals as engaged in meeting medical needs; we don’t want to think of them as engaged in redistribution (say, of life chances). Sufficientarian accounts also lead to a clearer and more definite policy prescription. As philosophers, it tells us, we ought to find what amount of healthcare a minimally decent life would require, and then devise healthcare systems to meet that threshold. A second advantage of this account is that it avoids the problem of leveling down that has been mentioned earlier. Recall that Daniels’s FEOP-​based account was vulnerable to the suggestion that ensuring that everyone has an equally miserable unhealthy existence would, embarrassingly, be compatible with equality of opportunity. The ideal of a decent minimum of health, being non-​comparative, obviously avoids such a problem. A third feature, which is usually considered an advantage, is that this account is responsibility-​insensitive. Everyone is entitled to a basic minimum of healthcare, regardless of how they have conducted themselves beforehand. This makes the sufficientarian account immune to the so-​called “abandonment objection” (more on which in section 21.6.1). The sufficientarian approach, we can see, provides its own distinct answer to the focal question of pattern. Justice in healthcare, according to this account, must meet individuals’ basic needs. Or, if you take the civic/​communitarian version of the view, it is the care that is needed to allow them full membership in their political community (Marshall 1950; Walzer 1983: Ch. 3). Regarding the currency question, notice that unlike Daniels’s account, here there isn’t necessarily restriction of care to treatment as opposed to enhancement. (Although I cannot recall this being explicitly discussed by sufficientarians themselves.) This is another advantage of the theory: if it so happens that certain enhancements have become a pre-​requisite for full membership, sufficientarians may well include those in the healthcare package. If certain physical features, while not deviating from normal species functioning, hinder one’s “ability to appear in public without shame,” then surgically altering them ought potentially to be included in the decent minimum of care. That is an advantage of the theory. Still, sufficientarian accounts may conflict with other intuitions we have. One weakness is that sufficientarianism allows (once again) for a two-​tier healthcare system. The view says that above a certain threshold, inequalities, in this case in health or access to healthcare, do not matter. Some may accept this as unproblematic. But for many others, this is unacceptable. Even those sufficientarians (such as Roger Crisp) who are untroubled by inequalities in income, especially when they occur above a certain threshold (Crisp 2003), may balk at inequalities in access to healthcare. One way for the sufficientarian to respond to this problem is by simply raising the threshold. If “decent minimum of health” is set high enough, then, it may indeed not bother us that some people use their wealth to purchase plastic surgery, shiatsu, etc. (Debra Satz offers a

Health   469 similar sufficientarian account with regard to education (Satz 2007).) This is plausible enough. But notice that in that case sufficientarianism comes dangerously close to egalitarianism. In attempting to meet the two-​tier objection, sufficientarianism may lose its distinctness. A different problem for sufficientarianism in health is the one known as “threshold fetishism.” An account that assigns value to raising individuals to the level of decent minimum of health or healthcare, may encounter a dilemma when, inevitably, faced with scarcity. Such a version of sufficientarianism (sometimes termed “head-​count” sufficientarianism)2 may prioritize helping Jones, who is close to meeting the standard of decent healthcare, over and above helping Smith, who is much worse off, health-​wise, than Jones, and would therefore require much more help in order to reach the threshold. Suppose we could help Smith, and even significantly so, but cannot raise him up to the required threshold. Sufficientarianism, at least of this “head-​counting” version, would prefer the better-​off Jones, and to many this seems rather counter-​intuitive (Arneson 2000a: 188; Arneson 2000b: 56; Roemer 2004: 279).

21.5  Alternatives to Rawlsian Justice in Health: Luck Egalitarianism Chronologically speaking, sufficientarian accounts were the first alternatives to Rawlsian justice in healthcare. A  third theory that emerged in recent years is luck egalitarianism. Most readers of this volume will already be familiar with that term (see Arneson (2018): Ch. 2 in this volume). Very briefly, though, the core idea is that inequalities are unjust when and only when they are the product of bad brute luck, that is, disadvantages that the individual could not have (reasonably) avoided. Luck egalitarian justice in health and healthcare would, then, say that disadvantages in health (say, unequal life expectancy) are unjust whenever the disadvantaged could not have reasonably avoided them. Luck egalitarian justice in health has many advantages. To begin with, it has no need for “normal species functioning” and its problematic implications. Luck egalitarian justice in health decrees that disadvantages in health—​any disadvantage—​is unjust (if un-​chosen). In the example earlier of the two short boys, it would hold that both of them—​the one whose disadvantage (short stature) is the result of pathology (deficient growth hormone) and the one whose is not (it is the result, recall, of genetics)—​ are disadvantaged through no fault of their own. Luck egalitarianism thus decrees 2  “Head-​count” sufficientarianism attaches value only to raising individuals above the threshold. A less demanding variant of sufficientarianism will see value also in benefits to individuals who are below the threshold, whether or not this ends up raising them above it.

470   Shlomi Segall that both boys deserve the same treatment, which strikes us as the correct judgment. The same is true for other non-​pathological conditions such as menstruation pains, labor pains, and pregnancy. Assuming that these are conditions that could not have been reasonably avoided, luck egalitarian health would treat them as disadvantages, and as such as conditions that merit treatment. Whether a medical disadvantage is a matter of treatment or enhancement is therefore irrelevant from the luck egalitarian perspective. Effectively then, rather than speaking of healthcare as a means to equality of opportunity to pursue life plans, as Daniels does, luck egalitarians speak, simply, of equality of opportunity for health. For, in seeking to neutralize all non-​voluntary obstacles to good health (or more accurately, those leading to inequalities in health), luck egalitarian justice in health effectively achieves equal opportunity to be healthy. Not only is this formula much simpler than Daniels’s, it also, more importantly, solves the three problems resulting from Daniels’s particular reliance on Rawls’s fair equality of opportunity principle. The first problem, recall, was that his account ends up discriminating against the rich. Consider how luck egalitarianism handles the same issue. If, and insofar as, we focus on justice in health (and that is, to be sure, a big ‘if ’), luck egalitarian justice would seek to promote equal opportunity to be healthy among all concerned, the rich and the poor alike. Unlike Daniels’s, then, the account can justify universal healthcare. The second problem, recall, was that in his reliance on individuals’ particular life plans, Daniels’s account could not defend the provision of in-​kind health care. Luck egalitarian justice in health, in contrast, does away with this reliance on life plans. In pursuing real (some say radical) equality of opportunity for health, it simply does not register the fact that a patient would prefer to swap her medical treatment for a Stradivarius, say. On the luck egalitarian reading, society’s duty to the paraplegic (in that example) originates in that person suffering bad brute luck. It is the neutralization of that bad luck that we owe her (say, in the form of an expensive wheelchair), not its cash equivalent. Luck egalitarian justice in health also avoids the third problem for Daniels’s account, namely its inability to justify, ultimately, any public provision of healthcare, due to the “social determinants of health” objection. In forswearing the distinction between treatment and enhancement, and in abandoning a biological account of health, luck egalitarian avoids this problem as well. It seeks to ensure that individuals have an equal opportunity to be healthy, whatever it might take to achieve this (be it medical care, or simply better housing, income, and so forth).

21.6  Objections to Luck Egalitarian Justice in Health There are many objections to luck egalitarian justice in health. Some target the ‘luckist’ element of luck egalitarianism, while others target its egalitarian element.

Health   471

21.6.1 Objections to Luckism in Health In the first category, the main objection, naturally, is the harshness objection: Smith is a reckless driver; Jones is his innocent passenger. They are injured to the same degree in an accident that is clearly Smith’s fault. We can save them both, but we can only afford the payment of one ambulance service. Luck egalitarians would typically recommend saving them both, but crucially that it is fair to shift at least some of the medical bill to reckless Smith in this case. Critics, especially democratic egalitarians find this excessively harsh (Anderson 1999). It is often taken for granted that the harshness objection, if effective, sinks luck egalitarianism while lifting “democratic equality.” But it is not clear how true this is. Consider the following three variations on the above example. Suppose that the reckless driver is Bill Gates, and the innocent passenger is the slightly less wealthy Warren Buffet. Democratic egalitarians often say that they are not bothered by inequalities (say, in income) above a certain threshold of civic capabilities. It follows, in this case, that democratic egalitarians would also recommend making Gates pay for his medical bill. In this case, at least, luck egalitarianism and democratic equality are equally harsh. Consider a second variant. Here the reckless driver is a convicted criminal (say, on leave from prison). Democratic egalitarians often say that we need not be committed to securing the democratic functionings of such individuals. (Anderson 1999: 319 n76). Once again, then, democratic egalitarians recommend the same treatment as luck egalitarians. They might plausibly claim that this is not unjust. But that of course, is precisely the view taken by luck egalitarians as well. Whatever is the case, the democratic egalitarian recommendations are no less harsh. Consider a third and final variation. Suppose the innocent passenger is a citizen (of the country in question), while the reckless driver is a tourist or a mere resident. Democratic egalitarians identify no duties of distributive justice toward the latter. We owe him medical assistance as part of some basic humanitarian duty, but no egalitarian duties of justice arise. It is therefore permissible to roll on to him the cost of his treatment. Once again, we can see, democratic egalitarians end up recommending the exact same differential treatment as luck egalitarians do. Of course, democratic egalitarians may want to maintain that shifting the cost of care is the correct verdict in the case of non-​citizens and not in the case of reckless citizens, which may well be true. Still, the important point is that doing so does not seem any less harsh. If it is wrong for luck egalitarians to recommend such a course of action, pace democratic egalitarians, then this cannot be simply because it is harsh to make individuals shoulder the financial burden of their (avoidable) medical care. Once again, the so-​called abandonment objection either afflicts democratic equality as well, or is not an objection to begin with. Sufficientarian accounts are therefore not as immune to the harshness objection (the main objection to the “luckist” element of luck egalitarianism) as may initially seem. And, I want to quickly add now, luck egalitarianism is not as vulnerable to that objection. In responding to the harshness objection, luck egalitarians may make the

472   Shlomi Segall following three points. First, it is not at all clear that luck egalitarian justice in health requires the abandonment of reckless patients, as opposed to merely permitting it. Second, luck egalitarians need not permit abandoning reckless patients, as opposed to permitting shifting on to them the costs of their treatment. And third, luck egalitarian justice in health need not say that making reckless patients pay for their care is desirable (or even permissible), all things considered. They only need to hold that it is not unfair to do so.

21.6.2 Objections to Egalitarianism in Health The most prominent objections to luck egalitarianism in health focus, on its responsibility-​sensitivity. But critics may equally object simply to the value that luck egalitarians see in equal health outcomes (among those who have conducted themselves responsibly). And it is this egalitarian element that may draw their ire. For some critics it is evident that equality in health is both unachievable and undesirable (Hausman, Asada, and Hedemann 2002; Hausman 2006, 2013). The claim about equality in health being unachievable is in fact puzzling. For it is always possible, we already saw, to level down health. Killing everybody in a given set will have easily achieved equality of health. Whether or not this is desirable is of course another matter, but the point at least is that equality in health cannot be objected to merely on grounds of feasibility. Then again, critics (Hausman 2013) might be making a slightly different point. They may inquire whether equalities in health that are potentially desirable are achievable. That is a perfectly good question to pose. The puzzle, though, is what precisely ought to follow from the discovery that such a (desirable) equality is unachievable. For, wouldn’t aspiring to get as close to that equality as possible (also) be a worthwhile pursuit? In other words, if equality (at a high level) of health is desirable (something we shall investigate in a moment), then it is not clear why reducing health inequalities would not also be desirable, even while falling short of perfect equality. In short, it does not seem that the alleged infeasibility of equality in health should be any hindrance to pursuing it nevertheless. So, is equality in health desirable? This is often taken for granted, but some have provided good initial reasons to think it is not (Hausman 2006). Our concern for equality in health, these critics argue, is in fact underpinned by a concern for something else altogether. That this is so is evident, they say, by the fact that we are not moved by all inequalities in health. Consider this. We find it unacceptable that African-​ Americans have a shorter life expectancy than Americans of European ancestry. And we are appalled to find that residents of poor neighborhoods of Glasgow can have a life expectancy that is more than ten years shorter than residents of more affluent parts of that city. But, on the other hand, we don’t normally mind the inequality in life expectancy between men and women, say. We object to the former type of inequalities, so the claim predictably goes, because they testify to some underlying social injustice

Health   473 that drives them, something that is not the case with regard to the second type (the inequality between the sexes). It would follow that inequalities in health are not in themselves unjust. They are simply a useful and glaring litmus test for some underlying social injustice, or so the claim goes. However, it is possible to offer a different explanation for those intuitions. The reason we don’t mind all health inequalities is not that we are solely interested in social ills that underpin them. Rather, it is that we are interested in overall inequalities, that is, inequalities in welfare, say, and not all health inequalities contribute to that. Consider this. It is true that we are normally untroubled by the inequality in life expectancy between men and women. But would the same be true had the inequality gone the other way? That is, would we not care if it were women who had the shorter life expectancy? Now, one reason why we would feel that way is, again, the suspicion that some discrimination in the allocation of resources is taking place (a real suspicion, incidentally, in places like India and China [see Sen 2013]). But it is also possible that we would have been worried about that inequality even if there was no such worry. It might be, then, that the reason we do not mind some inequalities in health is simply that we mind inequality in currencies that are more comprehensive, say welfare or opportunity for welfare (which women have less of compared to men). To see this more clearly, think of the following. We know that the life expectancy of many people who are disabled is considerably shorter than that of those who are not. This is not directly the result of their disability. It is rather due to the fact that many health conditions (heart, kidneys) are for one reason or another aggravated when suffered by someone who is wheelchair-​bound, say. This seems like a case of inequality in life expectancy that is troubling, even though it is not rooted in any particular social ill (no more than men’s shorter life expectancy is, that is). If you share that intuition, then this would suggest that we care about health inequalities when they exacerbate inequalities in overall welfare, as opposed to those health inequalities as mere indications of some independent social injustice. Recall Sen’s point raised at the very beginning of this chapter, namely that health underlies welfare. If it turns out that health is a pretty good proxy for welfare, then equality in health would be a good proxy for equality of welfare. And that would, very plausibly, make it worthy of egalitarian concern.

21.7  The Scope of Justice in Health The debate between Daniels and his critics, we have seen so far, concerns primarily the first two key questions of currency and pattern. I want to devote the last section of this chapter to the third remaining issue, namely that of scope. What is the appropriate scope of justice in health? That question can be broken into two dimensions, space and time.

474   Shlomi Segall

21.7.1 Global Health Global health has received much attention in the past thirty years or so, and there is by now a rich epidemiological literature on global health disparities. Correspondingly, there is also now a growing body of philosophical investigation into the normative aspects of global health (Lenard and Straehle 2012; Benatar and Brock 2011; Taket 2012). It asks: space-​wise, what is the scope of justice in health? Does it obtain only domestically or also globally? Are there substantial differences, that is, between our duties in terms of medical care domestically, and the duties owed to individuals who are not fellow citizens of ours? The chief moral question about global justice in health concerns the huge disparities in health around the globe. If it is a worry for justice in health that African-​Americans live on average four years less than white Americans,3 then it surely ought to be a worry that people in Malawi can expect to lead lives that are only about half (!) as long as those living in Japan (forty-four years, as opposed to eighty-six).4 Global disparities in health highlight the potential injustice of global inequality, but it is not obvious that they raise a concern that in principle is distinct from our more general preoccupation with global justice. Global health may raise, however, some more practical (and in that sense distinct) questions of justice. One set of practical questions concerns the phenomenon of migration of health workers (doctors and nurses) from the developing to the developed world. It was, for example, estimated that at one point half of the nursing workforce trained in Ghana found its way to the UK. Developing countries, it turns out, are not only economically poorer (and sicker), but also vulnerable to losing whatever skilled workforce they have trained domestically, often at considerable expense. Wealthy (and healthy) nations reap the benefits of a highly needed workforce trained and skilled by somebody else, without offering compensation. This exacerbates health inequalities between countries, and presents a unique challenge for global justice in health. The issue has generated interesting proposals for a practical solution (Eyal and Hurst 2008). The other issue of practical global justice in health concerns medical research. Viewed on a global scale, medical research bears certain features that make it ethically quite problematic. First, pharmaceutical companies have an obvious incentive to cater for patients (whether individual or institutional) with deep pockets. This means that the overwhelming majority of medical research is conducted on ailments that afflict the rich. In contrast, malaria, TB, and some other infectious diseases do not attract nearly as much research. Second, even when research is conducted on ailments that afflict (also) people in the developing world, the drugs that are developed are protected by the global patent regime, thereby making them extremely expensive and largely out of reach of those who need them the most. And third and finally, it is often residents of the

3 

http://​www.infoplease.com/​ipa/​A0005148.html/​ (accessed 20 November 2017). WHO Health Report 2011 < http://​www.who.int/​whosis/​whostat/​2011/​en/​index.html >. * figures are for life expectancy at birth for 2009. 4 

Health   475 developing world whose bodies are used in the clinical trials for these very drugs. Big Pharma companies naturally find it cheaper and easier to conduct their clinical trials in developing countries rather than at home. The burden of medical research is, then, borne disproportionately by the world’s poor. (This may look like a simple case of exploitation. But we may note that things are complicated somewhat by the fact that the experiment offers superior medical care (for the duration of the clinical trial) than would otherwise be available to the participants (Benatar 2001).) Medical research, then, is carried out on human subjects who will, in all likelihood, never benefit from the drugs it may develop. Some political philosophers have inferred that the global patent system on drugs ought to be abolished. The challenge, of course, is how to do away with patents while preserving incentives for pharmaceutical companies to conduct research, and moreover in a way that would benefit also the developing world. There are, by now, some interesting proposals as to how to do precisely that (Pogge 2005; Banarejee, Hollis, and Pogge 2010). One proposal for a solution involves replacing the current patent regime with a global fund that would reimburse medical researchers for drugs developed in proportion to the number of healthy life-years saved (or added). This would have far-​reaching implications. Drugs for infectious diseases represent a big bang for the buck, so to speak. They are cheap to develop and administer, on the one hand, and could save millions of lives on the other. A global fund that would reimburse according to the number of life-​years saved rather than according to the purchasing-​power of the patient, would maintain the incentives for companies to conduct research, while radically shifting the direction and potential beneficiaries of current medical research.

21.7.2 Health and Time Global health adds to the complexity of justice in health. But if that isn’t complex enough, things are made even more complex by the other dimension, that of time. Assume for the sake of argument that we aim at equality in health. The question then arises: what temporal unit, exactly, should this equality refer to? Ought we to aim at equality of lifetime health expectancy, or rather at equality in health at any given moment (what is often called “time-​slice egalitarianism”), or both? Dennis McKerlie, for example, has offered the following argument in favor of (also) time-​slice egalitarianism in general, and in health in particular. He asks us to imagine a feudal society in which peasants and lords swap places every ten years, say (McKerlie 1989: 479; see also McKerlie 2013: Ch. 4; Temkin 1993: 235–​7). Lifetime equality obtains, but the outcome nevertheless seems unappealing. McKerlie and Temkin conclude that what is missing is equality also in concurrent time-​slices. But consider a slightly different case. Think of a philosophy department in which every member goes on sabbatical every seventh year, but not everyone, obviously, goes off simultaneously. Time-​slice inequality (in terms of teaching load) prevails, but there seems nothing remotely bad about this. Or consider a couple where one of them takes on more child-​caring duties in the first year of the child’s life, whereas the other

476   Shlomi Segall partner does the same in the second year (Kappel 1997: 206–​8). Again, there is lifetime (or overall) equality consistent with time-​slice inequality, and yet there seems nothing bad about this state of affairs. All of this may suggest that McKerlie’s and Temkin’s “changing-​places egalitarianism” examples prey on special features of deprivation or on the fact that lives of different directions (good first, bad later) are not really of equal value. In the absence of these special features, time-slice inequality seems to raise no moral concern. The time dimension has important implications for just health policy. A time-​slice view prioritizes individuals according to how badly off, health-​wise, they are at any given moment. This would lead to piling medical resources on the old. A lifetime view, on the other hand, would try and balance the allocation of resources so that each individual is guaranteed access to equal health (not necessarily healthcare) throughout their lives. But the value of (lifetime) equality is not the only one, it is sometimes suggested, that can underwrite rationing of resources spent on the old. Both Norman Daniels and Ronald Dworkin have provided powerful arguments for the same conclusion, but based on the notion of prudential choice behind a veil of ignorance, rather than merely that of equality (Daniels 1988; Dworkin 2000:  Ch. 8). The premise is that it is prudent, on the one hand, for the young to help fund a healthcare system that mainly helps others (namely, the old), because they will (hopefully) one day be old themselves. But also, and perhaps more importantly, it is fair for the old not to expect to have all their medical needs met. This is so because it is prudent for rational agents, gazing ahead at the beginning of their lives, to reserve some resources aside for accidents and illnesses that may befall them, especially while they are young. The latter point is important. Suppose you may have recourse to a life-​saving organ transplant only once in your life. And suppose you must decide in advance at which stage of your life you would like to receive it, if in need. Would you prefer to receive it at the age of twenty-five, forty-five, or sixty-five? Most people, upon reflection, would opt for forty-five. We think that while dying at any age, let alone at these premature points of life, is tragic; dying at forty-five maximizes that tragedy, so to speak. And the reason has to do, once again, with Daniels’s key concept of life plans and life expectations. At twenty-five we have only recently formed a coherent life plan, and, more importantly, have only begun investing in it. We may have a first degree; we may have dated several potential life partners. We may have formed expectations, in other words, but done relatively little so far to invest in them. At sixty-five we have already reaped, to a large part, much of the fruits of our life plan (fulfillment from work and family, say). At forty-five we are at the height of our investment in our life plan: we have studied, made substantial strides in terms of career, married, started a family. To end one’s life at the point where much of the investment has been made, so many expectations formed, and achieving one’s life goals has only just begun, is truly tragic. This ‘prudential life-​plan’ account also successfully explains why health systems should avoid piling all their resources, and especially their life-​saving resources, on the old.

Health   477

Acknowledgments I am grateful to Nir Eyal and Serena Olsaretti for very useful feedback.

References Anderson, E. S. (1999). “What is the Point of Equality?” Ethics 109: 287–​337. Arneson, R. J. (2000a). “Why Justice Requires Transfers to Offset Income and Welfare Inequalities.” Social Philosophy and Policy 19: 172–​200. Arneson, R. J. (2000b). “Perfectionism and Politics.” Ethics 111: 37–​63. Arneson, R. (2018). “Dworkin and Luck Egalitarianism: a Comparison,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 41–64. Arrow, K. J. (1973). “Some Ordinalist Utilitarian Notes on Rawls’s Theory of Justice.” Journal of Philosophy 70: 245–​63. Banerjee, A., Hollis, A., and Pogge, T. W. (2010). “The Health Impact Fund:  Incentives for Improving Access to Medicines.” The Lancet 375: 166–​9. Benatar, S. (2001). “Justice and Medical Research: A Global Perspective.” Bioethics 15: 333–​40. Benatar, S. and Brock, G. (2011). Global Health and Global Health Ethics. Cambridge: Cambridge University Press. Buchanan, A. (1984). “The Right to a Decent Minimum of Health Care.” Philosophy and Public Affairs 13: 55–​78. Buchanan, A. (2009). Justice and Health Care: Selected Essays. New York: Oxford University Press. Crisp, R. (2003). “Equality, Priority, and Compassion.” Ethics 113: 745–​63. Daniels, N. (1985). Just Health Care. Cambridge: Cambridge University Press. Daniels, N. (1988). Am I My Parents’ Keeper? An Essay on Justice between the Young and the Old. New York: Oxford University Press. Daniels, N. (2008). Just Health:  Meeting Health Needs Fairly. Cambridge:  Cambridge University Press. Dworkin, R. (2000). Sovereign Virtue:  The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Eyal N. and Hurst, S. A. (2008). “Physician Brain Drain: Can Nothing Be Done?” Public Health Ethics 1: 180–​92. Frankfurt, H. (1987). “Equality as a Moral Ideal.” Ethics 98: 21–​43. Hausman, D. M. (2006). “What’s Wrong with Health Inequalities?” The Journal of Political Philosophy 15: 46–​66. Hausman, D. M. (2013). “Egalitarian Critiques of Health Inequalities,” in N. Eyal, S. A. Hurst, O. F. Norheim, and D. Wikler (eds) Inequalities in Health: Concepts, Measures, and Ethics. New York: Oxford University Press, pp. 95–​112. Hausman, D. M., Asada, Y., and Hedemann, T. (2002). “Health Inequalities and Why they Matter.” Health Care Analysis 10: 177–​91. Jacobs, L. A. (2004). Pursuing Equal Opportunities:  The Theory and Practice of Egalitarian Justice. Cambridge: Cambridge University Press. Kamm, F. M. (2001). “Health and Equality of Opportunity.” American Journal of Bioethics 1: 17–​19. Kappel, K. (1997). “Equality, Priority, and Time.” Utilitas 9: 203–​25.

478   Shlomi Segall Lenard, P. T. and Straehle, C. (2012). Health Inequalities and Global Justice. Edinburgh: Edinburgh University Press. Marshall, T. H. (1950). Citizenship and Social Class. Cambridge: Cambridge University Press. McKerlie, D. (1989). “Equality and Time.” Ethics 99: 475–​91. McKerlie, D. (2013). Justice between the Young and the Old. New York: Oxford University Press. Nussbaum, M. C. (2006). Frontiers of Justice:  Disability, Nationality, and Species Membership. Cambridge, MA: Harvard University Press. Pogge, T. W. (2005). “Human Rights and Global Health: A Research Program.” Metaphilosophy 36: 182–​209. Powers, M. and Faden, R. (2006). Social Justice: The Moral Foundations of Public Health and Health Policy. New York: Oxford University Press. Prah Rugger, J. (2010). Health and Social Justice. Oxford: Oxford University Press. Rawls, J. (1971). A Theory of Justice. Oxford: Oxford University Press. Roemer, J. E. (2004). “Eclectic Distributional Ethics.” Politics, Philosophy, & Economics 3: 267–​81. Satz, D. (2007). “Equality, Adequacy, and Education for Citizenship.” Ethics 117: pp. 623–​48. Segall, S. (2010). Health, Luck, and Justice. Princeton, NJ: Princeton University Press. Sen, A. (1980). “Equality of What,” in Sterling M. McMurrin (ed.) The Tanner Lectures on Human Values. Cambridge: Cambridge University Press, pp. 197–​220. Sen, A. (2013). “India’s Women:  The Mixed Truth.” The New  York Review of Books 60, 10 October. Sreenivasan, G. (2007). “Health Care and Equality of Opportunity.” Hastings Center Report 37: 31–​41. Taket, A. (2012). Health Equity, Social Justice, and Human Rights. Abingdon: Routledge. Temkin, L. S. (1993). Inequality. Oxford: Oxford University Press. Venkatapuram, S. (2011). Health Justice:  An Argument from the Capabilities Approach. Cambridge: Polity Press. Walzer, M. (1983). Spheres of Justice: A Defense of Pluralism and Equality. Oxford: Blackwell. Wilkinson, R. G. and Marmot, M. G. (2003). Social Determinants of Health: The Solid Facts. Geneva: World Health Organization. Wolff, J. (2012). The Human Right to Health. New York: Norton.

Chapter 22

Discrim inat i on Kasper Lippert-​R asmussen

22.1 Introduction The idea that discrimination is morally wrong because it is unjust and unjust because it violates distributive justice is far from outlandish.1 However, while paradigm cases of discrimination do involve seriously unjust distribution, in principle morally wrongful discrimination can occur without distributive injustice (at least, on many accounts of what that involves). Attractive as the idea might be, we cannot explain the moral wrongness of discrimination simply in terms of just distribution. Section 22.2 defines discrimination. Section 22.3 argues that a just overall pattern of distribution can co-​exist with unjust discrimination. In view of this, the next four sections explore attempts to tie discrimination and (other non-​pattern-​based forms of) distributive injustice together. Section 22.4 contends that overall distribution is only one among several sites of distributive justice. Section 22.5 explores a deontic account of distributive justice according to which it requires people to be treated with equal respect and concern. It is submitted that this account does not condemn all instances of unjust discrimination. Section 22.6 looks at a distribuendum of distributive justice—​deliberative freedoms—​which is tailored to account for the injustice of discrimination. It argues that while deliberative freedoms are important, the injustice of discrimination cannot be accounted for simply in terms of them. Section 22.7 focuses on indirect discrimination. It rebuts the suggestion that, unlike direct forms, indirect discrimination always involves distributive injustice. The limited success of these four proposals motivates the critical assessment, presented in section 22.8, of the suggestion that the 1 

Perhaps the objection to discrimination that springs to mind first is that it violates the requirement of formal justice that like cases be treated alike and unlike cases unalike. However, some forms of discrimination, such as indirect discrimination and statistical discrimination, need not violate this requirement. It may also be suggested that formal justice can be permissibly infringed when this promotes substantive justice.

480   Kasper Lippert-Rasmussen focus of standard theories of distributive justice is misplaced, and that the injustice of discrimination derives from the ideal that citizens should relate to one another as equals. Section 22.9 elaborates the main conclusion of this chapter, to wit, that the relation between discrimination and distributive justice is considerably more complicated than appearances suggest. In particular, given a plausible account of the currency of distributive justice, morally wrongful acts of discrimination might result in just distributions and need not even clash with the ideal of distributive justice in other ways, such as by manifesting the discriminator’s vice of being unjust.

22.2  What is Discrimination? To discriminate in the generic sense is to treat people differently.2 However, treating people differently is not morally wrong as such:  thus it is not morally problematic for an aid agency to feed the starving and ignore the well fed. Yet, if something is discriminatory—​in the sense people have in mind when they object to discrimination—​ that is at least a prima facie reason for taking it to be morally wrong, and this shows that generic discrimination is not what interests us here. What additional conditions must generic discrimination satisfy to be discriminatory in the more specific sense? A natural suggestion is that discrimination is unjust differential treatment. However, some cases of differential treatment that constitute discrimination are not unjust. Thus, if young people are given priority over old people when scarce organs are allocated among needy patients, this is age discrimination, but ceteris paribus a just distribution of organs favours those who have only enjoyed short lives over those who have enjoyed longer ones. Again, there are non-​discriminatory cases of unjust differential treatment. When King David intervened to ensure Uriah’s death on the battlefield, he treated him differently from his other soldiers in a way that was unjust, but not discriminatory. Another suggestion is that discrimination is arbitrary differential treatment, or differential treatment on irrelevant grounds. This invites the question:  what is the criterion of relevance? Suppose, on the one hand, that a relevant ground for the differential treatment of two applicants for a job as a salesperson is that one applicant will generate greater profits for the company than the other. Often people have this purely economic criterion of relevance in mind when they object to racism or sexism in hiring. Given the criterion, an employer who appoints male applicants on the ground that customers tend to be sexist with the result that men outperform women in terms of revenue generated 2 

In Lippert-​Rasmussen (2013) I distinguish between various concepts of discrimination and note that different concepts might be useful in different contexts. However, in this chapter I want to employ the particular and—​for reasons given in my book—​particularly useful concept of discrimination expounded in this section.

Discrimination   481 does not discriminate against female applicants. Yet this appears to be a textbook case of discrimination. Suppose, on the other hand, that with the previous objection in mind we say that ‘irrelevant grounds’ refers to grounds that should not motivate a hiring decision, morally speaking. Here the focus on irrelevant grounds loses much of its initial appeal. We are being offered a moralized definition of discrimination, one implying that for something to be discrimination it has to have a certain moral property. This is problematic because when we move beyond paradigmatic cases of discrimination it is often controversial whether the differential treatment is unjust or morally impermissible. On a moralized definition, then, it will often be controversial whether something is discriminatory. Yet controversial cases, like gender discrimination in motor insurance premiums, are often discussed as cases where it is unclear whether the fact that they involve discrimination makes them unjust or wrong. The moralized definition is also too inclusive, as it gathers differential, non-​discriminatory treatment, such as nepotistic hiring policies too, although few would think of these as giving rise to a complaint about discrimination. Finally, on the plausible view of just distribution of scarce organs indicated above, the prioritization of younger over older patients does not amount to age discrimination. Yet it is often described this way. All of this adds up to the conclusion that a moralized definition of discrimination does not fit well with the way we talk about discrimination. A third suggestion—​the one I shall employ in what follows—​is that discrimination is differential treatment on the basis of membership of socially salient groups. A socially salient group is one in which membership significantly affects social interactions across a wide range of social relations. On this account, men and women and, in many societies, blacks and whites form socially salient groups, but a particular family and the group of people in whose surname the letter P appears twice do not. It is not a necessary condition for a group to be socially salient that (most of) its members think of themselves as forming a group, or that (most of) the members of other socially salient groups think of them in this way. However, I suspect that groups that are socially salient in my sense tend to have these further features as well. Social salience is a matter of degree. In between the examples offered at the two extreme ends of the spectrum are groups such as smokers. Disadvantageous, differential treatment of smokers is not a paradigm case of discrimination, but as smoking becomes increasingly associated with accumulated disadvantages and stigma, it becomes more like sex or race discrimination. In actual fact there are many controversial, or borderline, cases of discrimination. The case of obese people is one such. This means that the problem of degree is not confined to the social salience account of discrimination. Moreover, complaints about discrimination are almost always made by, or on behalf of, socially salient groups. Finally, the kind of harms that are distinctive of discrimination—​ resulting from the accumulated effects of many individual acts of disadvantageous, differential treatment and involving stigmatization—​are tied to individuals being divided into socially salient groups and then treated on that basis. This provides some motivation for focusing on this third kind of discrimination, which I shall label group

482   Kasper Lippert-Rasmussen discrimination. My question then becomes: how does group discrimination relate to distributive justice?

22.3  Discrimination and the justice of the overall distribution In the archetypal cases of group discrimination—​for example, sex discrimination against women and racial discrimination against African-​Americans prior to the civil rights movement—​distributive justice is obviously violated. This is pretty much true whatever theory of distributive justice we look at. If, say, that justice requires equal opportunities to use resources, these instances of discrimination involved distributive injustice, since the resource opportunities of female and African-​American individuals were much worse than those of their male and European-​American competitors (which is not to say anything about the situation now). The same claim holds if instead, distributive justice is held to require the worst-​off people to be as well off as possible, since many of the worst-​off individuals in the cases at hand were female and/​or African-​American and would have been better off in the absence of discrimination against them. Indeed this fact was an important motivating factor behind the civil rights movement. This suggests (not: entails): The closeness thesis:  (i) The existence of discrimination within a certain society entails that the demands of distributive justice are not met within that society, and (ii) it is (i) that makes discrimination morally wrong.3 Unfortunately, the closeness thesis is false.4 Consider the first component of it. If distributive justice requires a certain distribution of a particular good, like resources or opportunities for welfare, discrimination can co-​exist with the satisfaction of the demands of distributive justice. Indeed, it might be that ceteris paribus, in the absence of discrimination, the pattern of distribution required by distributive justice would not be brought into being. Consider racial discrimination against a privileged racial minority, for example Chinese people in Malaysia. This might reduce overall inequality of opportunity. Empirically, discrimination tends to target members of worse-​off groups, so the present case is atypical. More generally, if distributive justice is captured by what Nozick (1974: 155, 156) calls ‘end-​state’ principles (e.g. the difference principle) or ‘patterned’

3 

Shlomi Segall (2013) believes that discrimination is wrong in a distinctive way only if it violates equality of opportunity for welfare. Accordingly, he appears to be committed to something like the closeness thesis. 4  As I emphasize later, this is not to say that the closeness thesis is completely misguided—​that discrimination (let alone wrongful discrimination) almost never involves distributive injustice.

Discrimination   483 principles (e.g. to each according to his or her need), it will be possible that distributive justice is accompanied by wrongful discrimination. However, the fact that cases like that of the Chinese in Malaysia exist, or could exist, suffices to refute the first component of the closeness thesis. More obviously, it will be possible for distributive injustice to arise in the absence of discrimination—​for example when, through no one’s agency, an unjustly unequal distribution arises through some people’s bad, brute luck.5 In the light of this, I will explore the possibility in the following sections that the requirements of distributive justice are not exhausted by end-​ state or patterned principles. Before doing so, however, I should comment on the second component of the closeness thesis, that is, the claim that what makes discrimination morally wrong is that it clashes with distributive justice. On an influential view, injustice is one factor among others—​total welfare, for instance, being another—​that determines the wrongness of an act, or policy, or practice, and distributive injustice is one factor among others—​retributive and restorative justice are others—​determining the injustice of discrimination. If this view is correct, discrimination might be morally wrong and unjust for reasons that are independent of distributive justice. Hence, even if the second component in the closeness thesis is false, it is possible to denounce discrimination morally. For instance, one might think that discrimination, or discrimination in many of its forms: (i) violates a moral requirement of respect which is a demand of justice (Alexander 1992); (ii) involves treating the discriminatee with ‘unwarranted animus or prejudice’ (Arneson 2006: 779); (iii) is objectively demeaning in a way that makes discrimination morally wrong (Hellman 2008); or (iv) is harmful (Lippert-​Rasmussen, 2013).

22.4  The Site of Distributive Justice So far I have assumed that the site of distributive justice, by which I mean the thing that realizes distributive justice or injustice, is the overall distribution of goods within society. However, other sites can be brought into the discussion. A person, their character, and their actions can be (un)just—​perhaps in virtue of their intention to bring about outcomes that involve distributive injustice—​irrespective of how this person affects, or fails to affect, the overall distribution through their actions.6 A sexist employer who deliberately flouts ‘the equal pay for equal work’ principle, and indeed seeks to prevent its implementation, manifests the vice of distributive injustice even if, as a result of counterbalancing forces, men and women ultimately end up being equally well off.

5  This holds true even if we think there are different spheres of justice (Walzer 1983) each with its own distributive principles provided all of these are ‘end-​state’ or ‘patterned’ principles: for example, through sheer brute luck, and despite discriminatory efforts to the contrary, medical care ends up being distributed according to need, offices according to merits, and so forth. 6  A similar claim is true of super-​individual agents, for example companies and states. Hume (1790) famously discussed justice as a personal virtue.

484   Kasper Lippert-Rasmussen Another important site of distributive justice is society, where this is understood to encompass common patterns of behaviour as well as the norms and expectations associated with these patterns. A society so construed is different from a state, which is a formally organized social entity. Again, it might be true that a certain society is unjust even if the overall distributive pattern realized by the people who form this society is perfectly just. In his critique of incentive-​based justifications of unequal overall distributions, G.  A. Cohen (2008:  128) notes that a society consisting of individuals motivated by a Protestant work ethic and indifferent to the difference principle, but which happens be one in which the worst-​off are as well off as possible realizes a just overall pattern of distribution even if it, the society, is unjust. Similarly, a sexist society in which norms and patterns of behaviour are discriminatory against women, but where a feminist state successfully neutralizes the effects of these, might realize a just overall pattern of distribution. Finally, a state might be distributively unjust independently of whether the relevant overall distribution is just. If, for some odd reason, a sexist dictatorship were to seek to implement a distribution favouring men over women, but incompetently ended up with an equal distribution across genders, an unjust state would co-​exist with a just distribution.7 The existence of different sites of distributive justice means that one can ask which of these is the most important. One view here is the ecumenical one that all are important. An alternative view is that one site is more important than the others, for example because this is the site to which distributive justice centrally applies. Ronald Dworkin subscribed to the latter view. For him, distributive justice was really a matter of how the state is organized (Dworkin 1985: 205; 2000: 1–​4, 11–​14).8 G. A. Cohen took the ecumenical view. Indeed, Cohen (2008: 123–​4) thought that, owing to certain unavoidable practical limitations, such as the limited capacity of the state to make talented people work for the benefit of the worst-​off and to obtain reliable information about the extent to which people are burdened by their work, distributive justice in one site (e.g. overall distribution) might only be realized if it also obtains in other sites: society’s embodiment of an egalitarian ethos, for example. We need not enter this debate. The important point is that if the wrongness of discrimination is explained in terms of distributive injustice, it is implausible to adopt a state-​centred view of the site of the distributive injustice involved in discrimination. A sexist employer who pays male employees better wages than female 7 

The global structure is a super-​state site of distributive justice. It is relevant to discrimination (as can be seen when we consider the fact that the large role played by armed conflict in the international order might be indirectly discriminating against women); but it is rarely discussed in connection to discrimination and distributive justice. 8  In his exploration of distributive justice, Dworkin (2000: 11) sought to discover which conception of equality states an attractive ‘political ideal . . . if any does’. His exploration was based on the premise that no ‘government is legitimate that does not show equal concern for the fate of all citizens over whom it claims dominion and from whom it claims allegiance’ (Dworkin 2000: 1), and on the further premise that the question of what ‘equal concern’ requires is a matter of what the state should aim at (Dworkin 2000: 2–​3).

Discrimination   485 employees simply on account of gender acts in a way that clashes with distributive just­ ice, irrespective of the impact of his policy on overall distribution. Accordingly, the case of discrimination can be employed to support the general view that the requirements of distributive justice are not exhausted by demands pertaining to overall distribution or the state.

22.5  Deontic Distributive Justice This section has two aims. The first is to illustrate the core claim, just made, that the requirements of distributive justice are not exhausted by demands made at the level of overall distribution. The second is to show that deontic theories of distributive justice cannot save the closeness thesis. For the sake of simplicity I shall work with an egalitarian account of distributive justice, but the conclusion I  reach generalizes to non-​ egalitarian accounts of distributive justice. Derek Parfit distinguishes between two egalitarian accounts of distributive justic­e. According to telic egalitarianism: ‘It is in itself bad if some people are worse off than others’ (Parfit 1998:  4).9 Deontic egalitarianism denies this claim (Parfit 1998:  6). Unequal distributions are generally unjust, but what deontic egalitarians object to is not the inequality itself: ‘On the Deontic View, injustice is a special kind of badness, one that necessarily involves wrongdoing. What is unjust, and therefore bad, is not strictly the state of affairs, but the way in which it was produced’ (Parfit 1998: 7). This characterization calls for a way of distinguishing egalitarian from non-​egalitarian objections to the ways in which distributive outcomes are produced. For instance, Nozick is not an egalitarian, yet he too objects to certain ways in which an unequal distribution is produced (e.g. through involuntary enslavement). Presumably, deontic egalitarians differ from non-​egalitarians, like Nozick, who endorse a non-​egalitarian deontic theory of distributive justice, in the account they give of what makes the genesis of an outcome unjust. Parfit does not tell us what such an account looks like, but presumably, unlike Nozick’s account, it will be intolerant of unequal distributions.10 One important deontic egalitarian position is Dworkin’s. According to Dworkin (1985: 205), justice requires the state to treat its citizens with equal concern and equal respect,11 and this requirement implies that only distributions that come close to realizing 9  A complicating factor—​one that I ignore here but address in Lippert-​Rasmussen (2006)—​is that Parfit qualifies his statement of telic egalitarianism with ‘through no fault or choice of theirs’ (Parfit 1998: 3 n5). The qualification implies that telic egalitarians are concerned with how outcomes are produced. 10  At one point Parfit describes deontic egalitarianism as the view that ‘though we should sometimes aim for equality [of distribution], that is not because we would thereby make the outcome better . . . When we ought to aim for equality, that is always for some other moral reason’ (Parfit 1998: 6). 11  For Dworkin, the recipients of distributive justice are citizens. We can set this aside here. A cosmopolitan version of Dworkin’s theory of distributive justice can be constructed.

486   Kasper Lippert-Rasmussen equality of resources are just.12 Plausibly, however, the relevant requirement goes beyond the realization of a near-​equal distribution and might explain the wrongness of discrimination even where that discrimination does not result in an overall unequal distribution. If, say, the state refuses to hire women, it fails to treat all citizens with equal concern and respect, even if the fact that the state discriminates in hiring does not result in distributive inequality. In my view, even if Dworkin’s requirement of equal concern and respect explains the wrongness of certain forms of discrimination not involving significantly unequal distributions, many forms of discrimination do not violate the requirement of equal concern and respect. For this reason I suggest that Dworkin’s theory of justice fails to save the closeness thesis. Let me explain. First, Dworkin’s requirement of distributive justice applies to the state only. However, as we noted in section 22.4, it seems citizens act distributively unjustly, inter alia by discriminating, by failing to treat one another with equal concern and respect. Dworkin might respond that his theory is compatible with the view that a just state encourages citizens not to treat each other with unequal respect and concern; he might add that citizens have a duty to support endeavours to this end made by the state. This response, however, seems unsatisfactory for a reason pointed out by Cohen (2008: 127). Compare a state that treats its citizens with equal concern and respect, and successfully enjoins its citizens to do so as well, with one that differs from the first in the single respect that its efforts to encourage citizens to treat one another with equal concern and respect fail through no fault on part of the state. If we adopt the Dworkinian response referred to above, implausibly, we cannot describe this difference by saying that the former society is more just than the latter. Ex hypothesi, both states have done all they can to ensure that citizens treat one another with equal concern and respect. To set aside Cohen’s objection, imagine a version of Dworkin’s theory which implies that it is the duty of the state, as well as that of individuals, to treat people with equal concern and respect. Presumably, this requirement is best understood contextually so as to allow for legitimate partiality towards family members, and the like. Even so, it is unclear that all forms of discrimination violate the requirement. I have already mentioned the case of age discrimination in the distribution of scarce organs for transplantation. Arguably, to treat young and old patients alike in the competition for scarce organs would not be to treat young patients with the same degree of concern and respect as the older ones, and yet some think that giving priority to younger patients amounts to (wrongful) age discrimination.13 Again, consider statistical discrimination—​that is, differential treatment based on information about how membership of a certain socially salient group is statistically correlated with other features, such as certain kinds of criminal behaviour. While statistical discrimination often takes forms that render it unjust, it 12  Dworkin (2000: 81, 104) allows deviations from equality in cases involving severely handicapped people and ‘movie stars and captains of industry and first basemen’. 13  Recall that the closeness thesis is not restricted to wrongful discrimination.

Discrimination   487 is unclear that it necessarily violates the requirement to treat people with equal concern and respect. Certainly, Dworkin thinks it does not: he believes that giving favourable treatment to African-​American applicants to medical school on the basis of statistical confirmation that an African-​American with a medical degree is more likely to function ‘as an effective doctor, in view of society’s present needs for medical services’ (Dworkin, 1985:  299) than a European-​ American given present underservicing of African-​ Americans is just. The main conclusion of this section is this: deontic accounts of the requirements of distributive justice can explain how some cases of discrimination violate the requirements of distributive justice even if, in those cases, it cannot be said that there is an unjust overall distribution. But this does not save the closeness thesis. There are cases of discrimination in which the requirements of deontic distributive justice are not infringed—​at least, not if those requirements are cast by deontic egalitarian justice.14

22.6  Deliberative Freedoms and the Currency of Distributive Justice Another way of potentially saving the closeness thesis is by adopting an account of the good that distributive justice allocates which, unlike the more traditional distribuenda like resources, primary goods, and welfare, is tailored to manage cases of discrimination. Sophia Moreau offers this kind of an account. She suggests that discrimination violates distributive justice because a discriminator reduces the deliberative freedom of the discriminatee and thereby ‘injures’ her ‘interests’ (Moreau 2010: 147). Deliberative freedoms are ‘freedoms to have our decisions about how to live insulated from the effects of normatively extraneous features of us, such as our skin color or gender’. The normatively extraneous features are ‘traits whose costs we ought not to have to factor into our decisions in’ contexts where we enjoy deliberative freedom (Moreau 2010: 155). Anti-​discrimination law safeguards our deliberative freedoms: . . . by preventing our employers, service providers, landlords, and others from acting in ways that deny us opportunities because of [normatively extraneous] traits, so that

14  It might be said that a restricted version of the closeness thesis applying only to unjust discrimination can be supported by Dworkin’s requirement. Owing to the vagueness of the requirement to treat people with ‘equal concern and respect’, this suggestion is hard to assess (see some of the examples given in section 22.8). In any case, if the requirement of equal concern and respect is understood in such a way that agents respect it if, and only when, they treat others in a way that is deontically just, the claim—​made in the narrower version of the closeness thesis—​that unjust discrimination entails that the demands of distributive justice are not met is reduced to a tautology (given the Dworkinian account of the basic demand of distributive justice and its proposed interpretation).

488   Kasper Lippert-Rasmussen when we deliberate about such things as where to work and where to live, we do not have to think about these traits as costs. (Moreau 2010: 147)

Deliberative freedom involves a subjective as well as an objective dimension: . . . in order for me to have a particular deliberative freedom with respect to a certain decision, it has to be true not just that I believe I can make that decision without having to worry about pressures from a certain extraneous trait, but that I really am free from those pressures. (Moreau 2010: 149)

Generally, it is morally important that we are able ‘to deliberate about and decide how to live in a way that is insulated from’ certain ‘pressures stemming from extraneous traits of ours’ (Moreau 2010: 147). Indeed, one of the core injustices in paradigm cases of discrimination is that a job applicant is obliged to worry that her application will be turned down on account of her race or gender. Moreau herself endorses a non-​relational theory of deliberative freedoms according to which people are entitled to a certain set of deliberative freedoms regardless of how much deliberative freedom others enjoy. However, other theories of distributive justice might treat deliberative freedoms as the distribuendum. For instance, one could take the egalitarian view that distributive justice requires everyone to have equal amounts of deliberative freedom. Despite these attractive features of Moreau’s account, any account that is narrowly focused on deliberative freedoms is flawed, and ultimately the loss the deprivation of deliberative freedoms involves should be subsumed under a more general category of distribuendum. The leading objection to Moreau’s account is that there are cases of wrongful discrimination not involving any restriction of deliberative freedom.15 Accordingly, if the distribuendum of distributive justice is constituted by deliberative freedoms, what makes discrimination wrongful or unjust cannot be that it clashes with distributive justice. The notion of ‘costs’ is ambiguous between costs incurred through choosing one option rather than another—​opportunity costs—​and costs incurred whatever one chooses. It is the former that are relevant to deliberative freedom, but pace Moreau’s account, the latter are not irrelevant to the moral wrongness of discrimination. Suppose everyone faces the same opportunity sets, with the important qualification that the expected value of any available opportunity is 10 per cent higher for a man than it is 15 

Another problem is set by cases where differential treatment on the basis of traits such as race and gender is not morally wrong. These are not counterexamples to Moreau’s account: she acknowledges that we only have a right to deliberative freedom in certain contexts, not in others. But at this point we do want to know how we can tell contexts that involve a right to deliberative freedoms from those that do not. Moreau’s account offers little guidance at this point beyond the suggestion that any sorting criterion will reflect the need to balance the value of deliberative freedoms against ‘other important values’ (Moreau 2010: 163).

Discrimination   489 for a woman. However, there is a sense in which women ‘do not have to think about [their sex as a cost]’. It is not the case that if they make one choice rather than another they will be penalized because of gender: for example, they will avoid informal social sanctions if, and only if, they choose to stay home rather than to pursue a career. Whatever choice they make, they are penalized on that ground. Ironically, they can therefore decide in ‘a way that is insulated from the pressures or burdens caused by certain extraneous traits’ (Moreau 2010: 149). Not only might it be true that agents believe that they ‘can make that decision without having to worry about pressures from a certain extraneous trait’, but it is also true that, objectively speaking, they are ‘free from those pressures’, since objectively they can do nothing to avoid them (Moreau 2010: 149). Subjectively as well as objectively, the pressures are deliberatively irrelevant. Hence, the disadvantage unjustly endured by women here simply consists in their set of opportunities being worse, not in the sets of deliberative freedom being worse. This means deprivation of deliberative freedoms cannot be what makes discrimination wrong or unjust. So even if deliberative freedoms are the relevant distribuendum, the closeness thesis is not vindicated.

22.7  Does Indirect Discrimination Always Violate Distributive Justice? As was argued in section 22.3, it is quite easy to see that direct discrimination need not result in outcomes that violate distributive injustice. However, the situation seems to be different with indirect discrimination, which arguably involves distributive injustice by necessity. This is so, for instance, according to John Gardner (1996: 360), for whom the legal duty not to indirectly discriminate ‘is essentially a duty of distributive, not corrective, justice’. Accordingly, it might be suggested that a version of the closeness thesis in which it is restricted to indirect discrimination is true. To assess this proposal we need a working definition of indirect discrimination. Here I will rely on work by Andrew Altman. Altman says direct discrimination exists where disadvantageous ‘acts or policies’ are aimed at certain socially salient groups. By contrast, there is indirect discrimination where acts or policies are ‘not aimed—​explicitly or surreptitiously, consciously or unconsciously—​at persons for being members of a certain social[ly salient] group’, but where these acts or policies ‘have the effect of disproportionately disadvantaging the members of a particular [socially salient] group’ (Altman 2011: 8). To illustrate this definition, consider the famous Supreme Court ruling, Griggs v.  Duke Power, which historically introduced the legal notion of indirect discrimination. According to Duke Power’s contested rules of promotion, employees seeking advancement were required to establish that they had a high school degree or pass a high school equivalence test. While the company had a past history of direct racial discrimination, the plaintiff conceded that the challenged promotion regime did not reflect direct discrimination today. However, while the company had no discriminatory

490   Kasper Lippert-Rasmussen intent in imposing the high school equivalence test, this test would in effect prolong the period of time in which African-​Americans were seriously under-​represented at senior management level. It would do so because African-​Americans were seriously under-​ represented among those who passed the test, and because, in virtue of historical direct discrimination, no African-​Americans presently occupied senior management posts. The Supreme Court ruled that the company policy ‘though fair in form’ was ‘discriminatory in operation’ and thus illegal.16 The reason one might think that indirect discrimination necessarily involves distributive injustice becomes apparent once we focus on the ‘disproportionate disadvantage’ component of Altman’s definition. For what is it for a disadvantage to be disproportionate? This question divides into two sub-​questions. What are the two relata that are being said to be disproportionate? Under what conditions is the relation between these two relata disproportionate? Starting with the first question, there are at least two possible answers, each of which requires careful specification. The first is as follows. The two relata are the benefits that the allegedly discriminating agent gains or produces for others, on the one hand, and the disadvantage to members of the group that is said to be the object of indirect discrimination, on the other. The second answer is this. The two relata are, first, the position of the group that is said to be the object of indirect discrimination and, second, the position of other groups. By way of illustration, suppose that a certain test used to determine admissions at a university is ‘not aimed’ at excluding women but results in a disproportionately low number of women being admitted. On the second view of the relevant relata, if the percentage of male admissions is very high and the percentage of female admissions is very low, the policy is indirectly discriminatory. (If, however, the gap between these two numbers is small, then, unless we take a severe view of proportionality, the policy is not indirectly discriminatory.) This cannot be claimed on the first view. Here, even if the gap between the percentage of male and female applicants admitted is high, the relevant policy might not be indirectly discriminatory. There are two reasons why. First, the admission policy might be very beneficial to society at large; its abandonment might result in, say, newly graduated doctors being much less better qualified. Second, the negative relatum could be interpreted comparatively across groups as well as comparatively across outcomes. With a comparison across groups, a policy will involve a ‘disproportionate disadvantage’ if it makes the putative object of indirect discrimination worse off than other groups (e.g. when fewer women than men are admitted). With a comparison across outcomes, however, the key comparison is between the level of benefits the putative object of indirect discrimination will enjoy given the relevant policy and the level of benefits it will enjoy under the relevantly alternative policy—​for example, we would need to ask whether fewer women are admitted under the present test than under some alternative test. The test used might affect the total number of applicants, and because of this it could turn

16 

Griggs v. Duke Power Co., 401 U.S. 424 (1971).

Discrimination   491 out that a test resulting in a higher percentage of women being admitted actually grants places to fewer women overall because it reduces the total number of applicants. It may be that, by and large, if a policy involves a ‘disproportionate disadvantage’ on one of these interpretations it will do so on the latter as well. However, the two interpretations diverge when groups are better off, absolutely speaking, if there are inequalities between them. Here a policy might be indirectly discriminatory on the first interpretation but not the second.17 When people complain of indirect discrimination they rarely have the first of the two interpretations of the two relata in mind. Most accept that if a certain policy has, in reality, very beneficial effects it is not indirectly discriminatory. Consider the visual tests used when pilots are being hired on commercial airlines. In principle, such tests might not aim to exclude sight-​impaired people even if, in effect, very few (non-​clairvoyant) such people are able to pass them. But given that the hiring of sight-​impaired pilots might have very bad consequences, none will complain that the tests are indirectly discriminatory. This suggests we should focus on the second interpretation. As indicated, this interpretation needs to be specified in relation both to the reading of ‘disproportionate advantage’ that is comparative across groups and to the reading that is comparative across outcomes. Complaints about indirect discrimination rarely involve the latter interpretation. This is because, in the absence of a palpable benefit to society as such, the people who raise them tend to infer straightforwardly from group inequality (e.g. most of applicants admitted are men) to indirect discrimination. Given this, let us focus on the specification involving comparison across groups. And let us add to this specification the view that the group inequality that indirect discrimination involves is disproportionate when it involves distributive injustice. On this view, it may seem that, necessarily, indirect discrimination invites a complaint about distributive justice. For if a policy involves indirect discrimination it now follows that it generates inequality between groups of a kind involving distributive injustice. While this implication is significant, it holds only for a certain range of theories of distributive justice—​notably, relational theories of distributive justice. Moreover, even if we adopt a theory of this sort, not all cases of indirect discrimination will involve distributive injustice. Let me try to make good these claims. I define a relational theory of distributive justice as one that holds the following: A distribution is just only if people’s distributive positions are suitably related to one another. Many theories are relational. Egalitarianism, according to which distributive just­ ice requires that no one is worse off than others (or significantly worse off than

17 

Admittedly, interpretations of ‘disproportionate disadvantage’ other than the two proposed here are available. For simplicity’s sake I ignore these.

492   Kasper Lippert-Rasmussen others) through no fault or choice of their own, is one. Theories of comparative desert on which a distribution is just only if the members of every pair of equally deserving people are equally well off are also in this category. Other theories of distributive just­ ice are non-​relational, however. The leximin principle of distributive justice on which distributive justice requires the worst-​off people to be as well off as they can possibly be is non-​relational. So is prioritarianism, since for prioritarians, distributive justice requires the morally weighted sum of benefits to be maximized, and benefits count for more, morally speaking, the worse off the recipients are on some absolute scale. So is sufficientarianism, according to which distributive justice requires that everyone have enough on some absolute scale. Obviously, this is not the place even to try to adjudicate between these theories of distributive justice. We can note, however, that on the non-​relational side of this divide indirect discrimination need not involve distributive injustice. The fact that a given policy grants benefits to a certain group that are disproportionately large relative to the much smaller benefits it brings to the group being subjected to indirect discrimination does not entail that the policy prevents the worst-​off people from being as well off as possible. Nor does it entail that the sum of morally weighted benefits is not maximized. Nor does it entail that someone does not have enough. Hence, if indirect discrimination necessarily involves distributive injustice, this must be because the distributive justice is relational. Relational theories face an important objection, however:  the levelling down objection. This objection is often formulated as one that targets egalitarian theories of just­ice (Parfit 1998: 17–​19). This is quite understandable, because egalitarianism is the most prominent relational theory of distributive justice. However, the objection in fact undermines a broad range of theories, including all relational theories. If distributive justice requires people’s distributive positions to be suitably related to one another, it must be more just to move from a situation in which people’s relative positions are not suitably related to one another to one in which they are, even if everyone is worse off, absolutely and in every respect, under the latter distribution, than they were initially. However, if one distribution is better than another for no one in any respect, it cannot be better in any respect, including in terms of distributive justice. Hence, relational theories of justice are false. That, in essence, is the levelling down objection to relational theories. It does not decisively refute relational theories of distributive justice (Temkin 1993: 245–​82; Persson 2008). But it certainly poses a formidable challenge. I shall not discuss the difficult issues that arise here. My point is simply that any theory of distributive justice permitting one to claim that indirect discrimination necessarily involves distributive injustice will be saddled with the levelling down objection, and that, accordingly, anyone who thinks that indirect discrimination necessarily involves distributive injustice must fend off the levelling down objection! Let us set aside the levelling down objection and consider whether even relational theories of distributive justice are compatible with the notion that some instances of indirect discrimination do not involve distributive injustice. There are at least four reasons

Discrimination   493 to think they are. First, if we adopt a strict view of proportionality in relation to indirect discrimination on which any inequality is disproportionate, weak relational views of distributive justice holding that only significant inequalities involve distributive injustice will imply that some forms of indirect discrimination do not involve distributive injustice. Second, the benefits invoked to assess whether a given practice is indirectly discriminatory may differ from those that matter according to the relevant relational theory of distributive justice. The relevant theory may require people to be equally well off in their opportunities for welfare (Arneson 1989). However, complaints about indirect discrimination invariably involve a much narrower dimension of benefit, such as access to university courses or (as in the Griggs v. Duke Power ruling) to senior management jobs. Given this, a policy that is indirectly discriminatory may even prevent a distributive injustice! Third, while indirect discrimination is tied to inequality across groups—​no one has ever complained about being subjected to indirect discrimination though not qua member of a certain salient group—​it is arguable that distributive justice is not concerned with the distribution across groups as such: inequalities across groups matter from the point of view of distributive justice only to the extent that they are tied to inequalities among individuals. Admittedly, if there is inequality between two groups, there has to be inequality between the individuals involved. However, it is a mistake to think that the elimination of inequality between two groups necessarily reduces inequality among individuals. Suppose a policy that is indirectly discriminatory against women could be replaced with another policy which, on average, would leave men and women equally well off. Suppose also that the alternative policy will generate new, huge intra-​group inequalities: the best-​off women will be much better off than the worst-​off women; likewise the men. In this kind of case, on an egalitarian view, the policy that is indirectly discriminatory is preferable to the one that is not. Finally, there are several plausible views about how to determine whether a certain group is disadvantaged relative to another (Lippert-​Rasmussen 2013: 63; Doyle 2007). A natural thought is that group averages should be compared. However, in some kinds of case this measure is inappropriate. Suppose that the policy in place leaves men better off on average than women, and that an alternative policy would eliminate this difference. Suppose also that under the first policy a handful of men are very much better off than they would be under the second, but that it is also true that under the second almost all men are significantly better off than they would be under the first policy.18 Because almost all men benefit from the second policy, it seems absurd to say that the first gives men an advantage over women. Let me summarize this section. Indirect discrimination need not involve a problem of distributive justice if we assume that the correct theory of distributive justice is

18 

The point still goes through if we also assume that almost all women are better off under the unequal distribution.

494   Kasper Lippert-Rasmussen non-​relational. Even if we drop this assumption there are several reasons why an indirectly discriminatory policy need not involve distributive injustice, so we cannot save the closeness thesis by restricting its scope to indirect discrimination. While this implication is theoretically important, it is practically less important. In particular, it does not demonstrate that paradigmatic cases of indirect discrimination do not involve distributive injustice. As indicated previously, they do so, more or less, whichever of the plausible theories of distributive justice we subscribe to.

22.8  Egalitarian Social Relations and Discrimination From the conclusions just set out it is tempting to infer that standard theories of distributive justice are not well placed to account for the wrongness of discrimination. To the extent that one is wedded to the idea that discrimination is morally wrong at least in part because it is unjust, one will presumably surmise from this that standard theories of distributive justice do not adequately specify the demands of justice. The standard theories need to be supplemented, if not replaced, to accommodate the injustice of discrimination. This thought parallels one that emerges in the work of various critics of luck egalitarianism, including Elizabeth Anderson (1999) and Samuel Scheffler (2003).19 In their view, what is really important from the point of view of egalitarian justice is that social relations are egalitarian in character. It is simply a mistake to think justice is a matter of how people’s distributive positions per se—​whether these are cast in terms of resources, welfare, or something else—​compare. And this mistake distorts egalitarianism, and by implication other relational theories of justice. This critique is interesting here because, intuitively, paradigm cases of discrimination are incompatible with the kind of social relations embodying equality of moral standing that matter to social relations egalitarians. Jim Crow-​style racial discrimination clearly upheld hierarchical social relations, as did sex discrimination of the sort that deprived women of the right to vote and access to higher education. The obvious question, then, is whether the view of justice championed by Anderson and Scheffler—​social relations egalitarianism—​is well suited to account for the injustice of discrimination. The crucial issue is whether there are forms of wrongful discrimination that would be tolerated within the social relations egalitarian framework. If not, and if it is also correct that standard theories of distributive justice cannot account adequately for wrongful discrimination, this supports the view that these need to be supplemented with something like social relations egalitarianism. If there are such wrongful forms of discrimination, the case of discrimination does not show that social relations egalitarianism is 19 

For critical assessments of social relations egalitarianism, see Lippert-​Rasmussen (2015a–​b; 2018).

Discrimination   495 strengthened—​at any rate, as long as we go along with the idea that discrimination is an infringement of distributive justice. In tackling this issue, it is important to specify what the relevant sort of equal social relations are. In Elizabeth Anderson’s view, the ‘proper negative aim of egalitarian just­ ice is to end oppression, which by definition is socially imposed’ (Anderson 1999: 289), while the proper positive aim ‘is . . . to create a community in which people stand in relations of equality to others’ (Anderson 1999: 288–​9). Such a community is one that ‘integrates principles of distribution with the expressive demands of equal respect. This version of social relations egalitarianism—​democratic equality—​guarantees all law-​ abiding citizens effective access to the social conditions of their freedom (Anderson 1999: 289). Democratic equality is concerned with the distribution of individual goods only insofar as an unequal distribution might cause inequality in social relations—​for example, the bowing and scraping of the poor before the rich. Negatively, one can characterize a society embodying democratic equality as one in which there are no ‘relations between superior and inferior persons’ (Anderson 1999:  312). A  society of this kind manifests the ‘equal moral worth of persons’, where this involves, negatively, a denial of ‘distinctions of moral worth based on birth or social identity’ and, positively, a state of affairs in which ‘all competent adults are equally moral agents: everyone equally has the power to develop and exercise moral responsibility, to cooperate with others according to principles of justice, to shape and fulfill a conception of their good’ (Anderson 1999: 312). These ideals are naturally interpreted in a way that rules out certain forms of direct discrimination. However, it is far from clear that all forms of discrimination would clash with democratic equality. Take statistical discrimination. Undoubtedly, its systematic targeting of badly-​off minorities may carry the message that their members are second-​ class citizens, so it might be condemned by the social relations egalitarian. However, cases can be imagined in which this line of thought falters. We can, for example, suppose that in each and every socially salient group there is a category of crime within which members are statistically over-​represented, and that all groups are, as a result, targeted by statistical discrimination to the same extent. In this kind of case statistical discrimination by the police may not symbolize inferiority. Moreover, even in real-​life circumstances not all kinds of statistical discrimination threaten our social standing as equals. Statistical discrimination by tax units, targeting the rich, and the kind of statistical discrimination involved in preventive police work focusing on violent crime by young males, hardly undermine equal social relations. Such discrimination is essentially connected neither with hostility to, nor beliefs about the inferior status of, those targeted. Those tasked with the implementation of a policy of statistical discrimination may even be more favorably inclined towards those they are subjecting to statistical discrimination than they are to others. They may think of the former as having a superior status. This notwithstanding, some will insist that these forms of discrimination are wrongful because they fail to treat people as individuals. Similar claims can be made about indirect discrimination when it results in unjust but moderate inequality in resources of a kind compatible with citizens relating to one

496   Kasper Lippert-Rasmussen another as equals. It seems, then, that social relations egalitarianism may not offer an account of the injustice of discrimination that is superior to that offered by the standard theories.

22.9 Conclusion This chapter has asked what discrimination is, and has then explored the relationship between discrimination and distributive justice, focusing especially on what I have called the closeness thesis. It has argued that the wrongness of discrimination is not closely tied to distributive injustice, since: (1) a just overall distribution can co-​ exist with wrongful and unjust discrimination; (2) indirect discrimination need not always be unjust, or indeed otherwise wrongful; and (3) direct discrimination does not necessarily violate deontic distributive theories of justice. Of course, in making these points I do not mean to belittle the distributive injustices, or other forms of moral wrong, involved in paradigm cases of discrimination. As I have noted, discrimination may be morally wrong, exclusively, or primarily, for reasons other than its violation of distributive justice. One obvious such reason is that it very often harms people, especially discriminatees. Also, it should be kept in mind that overall distribution is not the only site of distributive justice. Finally, we sometimes use a moralized concept of discrimination implying that, for an act or practice to amount to discrimination, it must be unjust or morally impermissible.20 The closeness thesis is formulated for non-​moralized discrimination, and if we were to substitute a moralized notion for the notion of group discrimination, we might discover that (3) is false.21 However, as I have noted in passing (see note 14), this is due first and foremost to the vagueness of the demand that we treat people with equal concern and respect; on some interpretations of this demand the denial of (3) becomes a tautology.

Acknowledgements I thank David Axelsen, Morten Brænder, Rasmus Sommer Hansen, Søren Flinch Midtgaard, Jens Damgaard Thaysen, Miklós Zala, and, especially, Serena Olsaretti for helpful criticisms of an earlier version of this chapter.

20  I take it that an act might be unjust and yet morally permissible all things considered, so the disjunction signals two importantly different moralized notions of discrimination. 21  The substitution of a moralized notion of discrimination for my notion of group discrimination turns (2) into a contradiction.

Discrimination   497

References Alexander, L. (1992). ‘What makes Wrongful Discrimination Wrong?’  University of Pennsylvania Law Review 141(1): 149–​219. Altman, A. (2011) ‘Discrimination’, in E. N. Zalta (ed.) The Stanford Encyclopedia of Philosophy. http://​plato.stanford.edu/​archives/​spr2011/​entries/​discrimination/​ (accessed 20 November 2017). Anderson, E. S. (1999). ‘What is the Point of Equality?’ Ethics 109(2): 287–​337. Arneson, R. (1989). ‘Equality and Equal Opportunity for Welfare’. Philosophical Studies 56(1): 77–​93. Arneson, R. (2006). ‘What is Wrongful Discrimination?’ San Diego Law Review 43(4): 775–​807. Cohen, G. A. (2008). Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Doyle, O. (2007). “Direct Discrimination, Indirect Discrimination and Autonomy.” Oxford Journal of Legal Studies 27(3): 537–​53. Dworkin, R. (1985). A Matter of Principle. Oxford: Clarendon Press. Dworkin, R. (2000). Sovereign Virtue. Cambridge, MA: Harvard University Press. Gardner, J. (1996) ‘Discrimination As Injustice’. Oxford Legal Studies 16(3): 353–​67. Hellman, D. (2008). When is Discrimination Wrong? Cambridge, MA:  Harvard University Press. Hume, D. (1790) A Treatise of Human Nature. Book III. London: Collins/​Fontana. Lippert-​ Rasmussen, K. (2006). ‘Telic versus Deontic Egalitarianism’, in N. Holtug and K. Lippert-​Rasmussen (eds) Egalitarianism: New Essays on the Nature and Value of Equality. Oxford: Oxford University Press, pp. 101–​24. Lippert-​Rasmussen, K. (2013). Born Free and Equal? A Philosophical Inquiry into the Nature of Discrimination. New York: Oxford University Press. Lippert-​Rasmussen, K. (2015a). Luck Egalitarianism. London: Bloomsbury Publishing. Lippert-​Rasmussen, K. (2015b). ‘Luck Egalitarians vs. Relational Egalitarians: On the Prospects of a Pluralist Account of Egalitarian Justice’. Canadian Journal of Philosophy 45(2): 220–​41. Lippert-​ Rasmussen, K. (2018). Relational Egalitarianism:  Living as Equals. Cambridge: Cambridge University Press. Moreau, S. (2010). ‘What is Discrimination?’ Philosophy & Public Affairs 38(2):143–​79. Nozick, R. (1974). Anarchy, State, and Utopia. Oxford: Blackwell. Parfit, D. (1998). ‘Equality and Priority’, in A. Mason (ed.) Ideals of Equality. Oxford: Blackwell Publishers, pp. 1–​20. Persson, I. (2008). ‘Why Levelling Down could be Worse for Prioritarianism than for Egalitarianism’. Ethical Theory and Moral Practice 11(3): 295–​303. Scheffler, S. (2003). ‘What is Egalitarianism?’ Philosophy and Public Affairs 31(1): 5–​39. Segall, S. (2013). Equality and Opportunity. Oxford: Oxford University Press. Temkin, L. (1993). Inequality. Oxford: Clarendon Press. Walzer, M. (1983). Spheres of Justice. New York: Basic Books.

Chapter 23

Race Bernard R. Boxill

As my chapter’s title suggests, my project is to assess the considerations for supposing that racial differences tend to lead to violation of distributive justice. Undoubtedly the races tend to treat each other badly. In practically every multiracial society that has ever existed on earth the distribution of goods and burdens among the races is not a just distribution. Of course, every society is to some extent unjust. But multiracial societies always seem more unjust. It does not follow that such societies are necessarily unjust. Tendencies to injustice may be controlled. Natural inequalities in talent may tend to encourage injustice, but Plato argued that they could be the basis of a just society (Plato 1992). Similarly John Rawls’s famous principles of justice, the fair equality of opportunity principle, the difference principle, and the principle of redress are designed to counteract the tendency of natural inequalities to encourage distributive injustice (Rawls, 1971). Presumably then, with appropriate precautions, multiracial societies may be reasonably just, even if racial differences encourage injustice. But I cannot base my discussion of race and justice on the theories of Plato and Rawls. These theories are designed to resist or offset the tendency of natural inequalities in talent to cause injustices. Consequently, they must be irrelevant to my discussion because the races are not unequally or even differently talented. Whites did not enslave blacks because whites and blacks are naturally unequally or differently talented. Of course the belief that the races are naturally unequally or differently talented has helped to cause distributive injustice. But it does not follow that the means used to prevent inequalities of talent from causing distributive injustice will be the same as, or even resemble, the means used to prevent racial differences from causing distributive injustice. I will also avoid discussing the theories of justice of Rawls and Plato because they are ideal theories. Such theories focus on describing societies in which no one behaves unjustly. The idea is that the ideals described give us something to aim at. It is a bad idea. It can give us no suggestions about how to get to the ideal because the ideal is so far away. I’ll proceed on the assumption that our efforts are better spent on understanding and remedying the injustices going on under our noses than in speculating about how to reach distant ideals.

Race   499 Let us begin with the concept of race, since the idea that racial differences tend to cause injustice can make sense only if we know what race is.

23.1 Race Today many people in America will tell you that the races are social constructions. They mean in part that there are no important biological differences between the races despite the reliably inherited and often highly visible differences between them like differences in skin color. Of course some members of a race are not visibly different from the members of another race: some black people have white rather than black skins and are otherwise indistinguishable from white people. Nevertheless as race has been socially constructed these people are black because among their ancestors is someone, male or female, with a black skin. In such cases especially, one might wonder why people are usually so strongly disposed to distinguish the races. It is not because white-​skinned black people are culturally unlike the white-​skinned white people. Culture has nothing to do with the social construction of race; culturally similar people are often members of different races, and culturally dissimilar people are often members of the same race. The reason why people are so strongly disposed to distinguish the races has mostly to do with the alleged character and talent of the different races, rather than with culture. Black people, even the white-​skinned ones are widely believed to be stupid, lazy, undisciplined, violent, rambunctious, and oversexed. These widespread beliefs are the work of racial ideologies. A racial ideology is a vast set of strongly connected claims about the members of the different races. Multiracial societies are necessarily always suffused with racial ideologies. This is because races, as social constructions, are created by racial ideologies. The claims in racial ideologies are usually, but not always, obviously derogatory. Consequently, racial ideologies are usually naturally associated with attitudes that correspond with the derogatory claims that compose them. For example, the racial ideology that creates blacks is associated with and generates contempt for them. Often the attitude thus generated generates or reinforces further derogatory claims that expand and reinforce the ideology. The claims or attitudes that help in the creation of race differ for the different races. In the United States blacks are a race partly because they are believed to be innately generally unintelligent and lazy, while Asians and Jews are races partly because they are believed to be innately highly intelligent and motivated. The claims or attitudes in question also often change, though usually while maintaining their general character. Sometimes however they change drastically, while nevertheless retaining the resilience to recover their former venom. The stereotype of Jews as dishonest, stingy, cunning, and vengeful, is fading; one can only hope for good. The discovery that races are social constructs is recent. In the nineteenth century and most of the twentieth century, races were assumed to be “natural kinds.” That is, it was widely and falsely believed that there were deep, innate, and important

500   Bernard R. Boxill biological differences between the races. It was also widely and falsely believed that these differences revealed themselves both in the visible physical differences between the races and in the equally striking intellectual and moral differences between the races that many people claimed to see. Thomas Jefferson gave a list of these differences in his Notes on the State of Virginia. Blacks were unintelligent and unimaginative compared to whites, as well as oversexed and disposed to seek immediate gratification; and although they appeared to be courageous, this was probably because they lacked foresight. The only genuine virtues that Jefferson allowed blacks to possess were integrity, benevolence, gratitude, and fidelity, though it would be easy to argue that he believed that these too were the result of their stupidity. Finally, with an affectation of diffidence, Jefferson opined that the various differences between the races that he had listed were formed by “nature” not experience, and even suggested that the races might be different species (Jefferson 1982 [1787]: 138–​ 43). Many of these claims or speculations were probably in circulation before Jefferson, but most of them became standard after his book was published and read widely. The idea that races are social constructions presents us with a puzzle. It requires that we believe that for centuries vast numbers of people, whole civilizations, including learned and intelligent men like Jefferson, took outrageously false views about other people for granted, even when those people lived right next to them. Feminists will remind us that the phenomenon is not unique. Men have always lived next to women and yet continue to have outrageously false views about them too. Part of the explanation of both phenomena is that the false claims about black people and women are ideological claims. Such claims are often difficult to even doubt because they are tightly connected to a vast network of mutually connected and supporting views that include the background contexts in which theories and claims are judged to be true or false. As a result, doubting them is like calling a whole world into question. Ideological claims are also hard to rebut because as parts of an ideology they are always highly advantageous to the many people in the society that the ideology has helped to put into positions of power, and naturally these people use their influence to suppress or undermine any challenges to them. For example, since it was highly advantageous for whites to have black slaves, whites tended to justify black slavery. Further, nature has often cooperated with ideologies by providing them with the basis for advancing specious teleological arguments in their favor. For example, male supremacists have always argued that providence showed its intention that women be domestics by making women the sex that bears children. Finally, ideological claims are hard to rebut because, being so widely believed, they tend to cause the people they are about to acquire the characteristics they claim that nature gave those people. As a result, turning to the facts to rebut ideological claims often seems to confirm the claims. Rousseau had escaped that snare when he mocked Aristotle’s claim that “some are born for slavery” with the observation that Aristotle “mistook the effect for the cause”(Rousseau 1978: 48). These considerations may help explain why people find it difficult to escape from the clutches of an ideology they are already enmeshed in. Similar considerations

Race   501 also suggest how people can become enmeshed in ideologies that are already well established. But how do ideologies come to exist and to be established? People cannot become enmeshed in an ideology before it exists. Aristotle could not have mistaken the effect for the cause before the cause existed. Jefferson could not have made his false claims about blacks because he was enmeshed in the ideology of race before that ideology existed. Assuming that ideologies have not always been around, and do not start to exist already fully formed, we must suppose that they begin with a few false claims that then multiply and develop into the vast network of false claims that is an ideology. But why do people believe these false claims before they become part of an ideology and acquire the backing that an ideology provides them? Eric Williams summarized a well-​known answer to this question in his book Capitalism and Slavery: “Slavery was not born of racism: rather, racism was the consequence of slavery” (Williams 1994: 19–​20). He meant that the white plantation owners enslaved blacks because they wanted to maximize their profits and blacks were the best and cheapest workers available. And then trying to justify their injustice they invented and persuaded themselves and others to believe the false claims that eventually developed into the ideology of race. They might have failed in these endeavors if they had been less powerful and influential, or if most other whites did not also profit from slavery. But they were powerful and influential and most other whites did also profit from slavery. Consequently, most white people were readily persuaded to believe the false claims the plantation owners had invented, the claims spread, stuck, grew more elaborate, and when slavery became profitable enough to fund specious scientific research to support them, developed into an ideology. Williams’s account of the rise of racial ideology and the construction of black-​ skinned people as a race makes no appeal to the visible and often striking physical differences between the races, especially the difference in skin color. This “color blind” account of the rise of racial ideology suggests that had white people been better and cheaper workers than black people they would have been enslaved instead of black people and an ideology would have arisen constructing them instead of black-​ skinned people into a race. The second of the preceding two suggestions seems implausible. Perhaps the plantation owners would have enslaved white people if white people had been the cheapest and best workers available; white people have often enslaved other white people. Indeed those siding with the Williams-inspired account of the rise of racial ideology have often claimed that plantation owners treated their white indentured servants as badly as they treated their black slaves. But it does not follow that they could have developed an ideology that made their white slaves into a race. This is because the ideology of race requires that people believe that the races are made by nature rather than invented by themselves. That belief is of course false, but it cannot be altogether implausible. The belief that the socially constructed races are made by nature has the following specious plausibility: the members of a race do usually share some specific easily noticed physical characteristic that is created by nature. The most influential early definition of race as a creation of nature, Immanuel Kant’s for example, made that fact central to the division of humanity into different races (Kant

502   Bernard R. Boxill 2007: 153). By establishing that nature made the races innately different from each other in some obvious and undeniable ways, it eased the social construction of race which proceeded by extending the list of innate characteristics to include peculiar moral and mental characteristics. Thus Kant claimed that the races also differed naturally in their moral and mental characteristics (Kant 2007: 930). This helps explain why even the socially constructed races are normally easily identified by some physical naturally inherited characteristics. If these considerations are sound, plantation owners might have enslaved white people had they been the best and cheapest workers available, but they would never have been able to create an ideology that made these people into a race. This is important because it is the alleged natural and inherited characteristics of a race supposedly persisting indefinitely over the generations that help make the idea of race so dangerous. The black slaves of America could not be simply emancipated and then expected to progress like anyone else because their black skins had helped to make them into a race. But, as Jefferson pointed out, the emancipated white slaves of ancient Rome simply merged into the general population (Jefferson 1982 [1787]: 143). I contend that this is because they had not been constructed into a race. Why did these white slaves not become a race? Aristotle had attempted to make white slaves into a race with his theory that at least some of them were “natural” slaves. This theory had nearly all the earmarks of the more recent theories of black inferiority in the eighteenth and nineteenth centuries that make black slaves into a socially constructed race. But history tells us that it never caught on in ancient Athens or ancient Rome, and that the white slaves of those cities never became such a race. I suggest that this was because the white slaves in question were not a natural race in the sense already explained. Centuries later the Spanish revived Aristotle’s theories and used them to make the Native Americans they enslaved into a socially constructed race, but those Americans were of course a natural race. Another problem with the “color-blind” theory of racial construction, is that it cannot explain why the members of all the recognized socially constructed races usually have highly visible natural and inherited characteristics like skin color that distinguish them from the members of other races. Indeed, these socially constructed races are routinely described as the white, black, yellow, and brown races. This would not be possible if skin color had nothing to do with the creation of these groups as races, or at least it would be possible only if the properties that supposedly were the basis of their becoming a race were highly correlated with their skin color. But in that case their skin color would be reliable evidence of those properties, and consequently of deep and important differences between the races—​contradicting the major premise of social constructivism. It may be objected that this contradiction depends on the assumption that the properties in question are fixed by nature, rather than being the result of human beings adapting to their environment. For example, it could be argued that blacks were the best workers because they lived in environments that compelled them to be good workers to avoid extinction. But this argument assumes that white people inhabit the same environments and that the adaptations in question persist even when they move to different environments. These considerations are not meant to suggest that races are

Race   503 not socially constructed. They are meant to suggest that physical differences between the natural races can play an important part in the generation of the socially constructed races, even if no one took them to be indicative of the deep and important differences between the natural races. Martin Delany’s theory of how skin color played an important role in black enslavement further supports suggestions that the “color-blind” theory of racial construction is mistaken. Delany dismissed the idea that the African was enslaved “on account of hatred to his color,” and denied that anyone had ever “dreamed of ” the “absurd idea of the natural inferiority of the African” before that idea was “adduced by the slave-​holders, and their abettors in justification of their policy.” But he argued the African’s color facilitated his enslavement by whites because it made him “foreign to the sympathies” of whites and consequently elicited “less interest” from them in his “favor” when they enslaved him (Delany 2004: 48–​51). His point can be elucidated by Hume’s account of sympathy. Hume believed that sympathy, the transfusion of passions among people, depended on the “great resemblance among all human creatures” that “nature has preserv’d” (Hume 2000 [ 1739–40]: 207). Presumably then, and Hume makes the point explicitly, the greater the resemblance among people, the greater the sympathy, or transfusion of passions among them. Consequently, since nature has preserved a greater resemblance among the members of a natural race than among the members of different natural races, it follows that sympathy is more facilitated among members of a race than among the members of different races. Consequently, whites are likely to feel the misery of white slaves more keenly than they feel the misery of black slaves, and since everyone prefers to feel less misery than more, will other things equal, prefer to enslave blacks rather than to enslave whites. Jefferson’s comment that the black skin on his slaves’ faces was an “immoveable veil of black” covering its possessor’s emotions (Jefferson 1982 [1787]: 138) suggests another role that skin color could have played in the enslavement of blacks. The comment suggests that for Jefferson meeting a black person was like meeting a masked person. A masked person tends to be threatening because his mask prevents others from using his face to help make educated guesses of what he is thinking and preparing to do, and consequently deprives them of the security they normally have among unmasked people. Of course, Jefferson was deeply prejudiced when he made the comment, but it helps us to see how dramatic differences in skin color could exacerbate the suspicion that strangers naturally feel when they meet each other. It would deprive them of access to the natural and universal language of facial expressions that often ease the mutual suspicions of strangers who lack a common learned language. Certainly, the resulting heightened suspicion between such strangers would reduce sympathy between them and facilitate the enslavement of one by the other, even if neither took the other’s skin color as evidence of any deep difference between them. If these considerations are sound, greed and pride motivated the enslavement of blacks, and differences in skin color facilitated that prolonged tragedy and the social construction of the black race. A  different facilitator of racial construction, though not of slavery, should also be mentioned. Jefferson can again serve to introduce it.

504   Bernard R. Boxill Although he endorsed the widely held claim that blacks were intellectually inferior to whites, he scoffed at the idea that it justified enslaving blacks with the comment that the idea implied that Sir Isaac Newton could justifiably enslave the rest of us (Jefferson 2006 [1809]: 205). Nevertheless, Jefferson’s endorsement of the claim that blacks were intellectually inferior to whites had a political purpose, which was to justify expelling blacks from the country. And many whites who shared his dislike of slavery nevertheless shared his dream of an all-​white America. Lincoln, we should remember, shared that dream and actually took steps to realize it. Its source was not greed, for the dreamers understood that the enslavement of blacks which they condemned was profitable. Its source was the Enlightenment idea that nature intended and morality demanded the maximal eventual development of our talents. Since it was widely believed that our capacities for developing our talents were naturally inherited and unequally distributed among the races, it followed, given apparently reasonable assumptions, that nature discouraged and obligation forbade race mixing. Accordingly, Jefferson championed the deportation of blacks from America, and Immanuel Kant, who was particularly insistent that nature intended the maximal development of our talents, argued that it had determined that dark-​skinned people from hot countries could never adapt to life in the cold countries where white people lived (Kant 2007: 209). In sum, the beliefs that together construct the black race have many sources and serve different purposes. Greed and the love of domination are important sources: they helped to motivate the enslavement of black people, but differences in skin color facilitated that tragedy as well as the ensuing social construction of the black race, which then helped prolong the tragedy. Further, as I have just suggested, another source of the social construction of the black race was yet another ideology, namely, the ideology of the Enlightenment which complicated the social construction of race with distinctive beliefs justifying the forcible separation and segregation of the races.

23.2  The Problem Given the beliefs that constitute the construction of the races, it is clear that multiracial societies tend to be peculiarly prone to violations of distributive justice. Indeed, the claim to that effect seems almost a tautology. People with the beliefs that construct the races will almost certainly tend to treat those they perceive to be of different races than themselves with less consideration that justice requires. Since the troublesome beliefs are false, the remedy might seem to lie in better education. But though improved education has an assured place in the fight against injustice, there is room for an additional approach. Racism is a complex matter and understanding it requires study and close attention. But the false beliefs that constitute the social construction of the races are not the result of subtle mistakes that can be corrected only by even more subtle reasoning. Most of these beliefs would not have been countenanced in the first place were it not for the material inequality of the races when they first met, and for their unfamiliarity

Race   505 with each others’ appearances. Of course, once the ideology supporting these beliefs had established itself, the beliefs maintained their hold on people’s minds even after the races became apparently familiar with each other. The case of Jefferson amply demonstrates that claim. But in such cases the continuing inequality of the races, both in terms of their authority and power, tends to keep the races strangers to each other, though even in such cases we have it on the authority of W. E. B. DuBois that there is nevertheless some appreciable improvement in mutual understanding (DuBois 1999 [1903]: 107). Sometimes DuBois was too optimistic. For example, he once wrote as if he thought that a dramatic widening of sympathy had occurred in the nineteenth century. In his famous words: The nineteenth century was the first century of human sympathy—​the age when half wonderingly we began to descry in others that transfigured spark of divinity that we call Myself; when clodhoppers and peasants, and tramps and thieves, and millionaires and—​sometimes—​Negroes, became throbbing souls. (DuBois 1999 [1903]: 136)

Probably he was moved to write these lines because in the nineteenth century Americans abolished slavery and the British emancipated the slaves in the Caribbean. And his provocative claim made some years later, “All Art is propaganda,” suggests that he gave credit for the alleged widening of sympathy in the nineteenth century to the great writers of that century like Dickens, Hugo, and Stowe (DuBois 1995:  514). Unfortunately, there are good reasons to believe that whites abolished slavery in America because it was in their interests to do so, not as legend has it because Stowe’s Uncle Tom’s Cabin widened their sympathy so as to include blacks. And though the British historian W. E. H. Lecky seems to have claimed very influentially that the British emancipation of slaves in the Caribbean was morally motivated, his claim is in fact still the subject of considerable debate. Eric Williams’s Capitalism and Slavery was the first to challenge it, starting what is now referred to among economic historians as the “Williams debate;” and though establishment economists seem recently to have concluded against Williams, others contend that these economists might have failed to consult Williams’s Oxford doctoral dissertation that contains a better presentation of the evidence for his challenge. And indeed those economists cite evidence suggesting that Lecky’s famous claim might not even have been about the British abolition of slavery in the Caribbean (Williams 2014: xi–​xxiv). In section 23.3 I will defend affirmative action as a justifiable means of both educating us about racism, and of overcoming the unfamiliarity of the races with each other, and loosening the hold on them of the false beliefs that constitute race. I  will focus mostly on showing that affirmative action is consistent with the strictest justice if only backward-looking considerations are not impatiently dismissed, though I will spend some time too on its educational possibilities.

506   Bernard R. Boxill

23.3 Affirmative Action At first, the term “affirmative action” was used to refer to special efforts to ensure that applicants for positions at government contracting agencies be considered without regard to their race, creed, color, or national origin. This could require that the contracting agencies pay special attention to the race, creed, color, and national origin of applicants simply to be able to deliberately offset their own tendencies to give unfair preference to applicants of certain races, creeds, colors, and national origins. Today, however, the term is used more broadly to refer to policies that not only try to offset tendencies that give unfair preference to applicants of certain races, creeds, colors, but actually give preference to members of various groups, usually African Americans, women, and Hispanics, in the competitions for jobs and promotions and for places in universities and colleges. Many consider such policies to be obviously unjust because they violate the principle of equal opportunity, which is a basic principle of distributive justice. Since that principle requires that places and positions in a society be awarded to those best qualified for them, and affirmative action requires that we give preference to blacks (or women), it requires that we violate the principle of equal opportunity, and consequently seems unjust. One of the considerations favoring the principle of equal opportunity as a basic principle of distributive justice is that filling positions with the best qualified will better satisfy our needs and desires than filling positions with the less qualified. This consideration suggests that the principle of equal opportunity could be made even more attractive as a principle of justice if it was modified to become the principle of fair equality of opportunity. According to that principle everyone must be educated so that the equally talented have the same chances to become qualified for desirable positions. Obviously this modification of the equal opportunity principle would increase the number and improve the qualifications of people filling most positions in the society, including those that we consider crucial. A problem with this line of argument for the equal opportunity principle lies in the ambiguity in the meaning of “qualification.” If states establish medical schools in order to train doctors to serve their people, but the school graduates doctors who go into practice only in the big cities, it seems that the school may consider the likelihood of practice in a rural area to be qualification for admission and preferentially admit candidates who show reliable evidence of that likelihood. Similarly being black may be a qualification for admission to a medical school if white doctors seem to avoid treating black patients. These examples have breathed life into the case for affirmative action because they show that it can be designed so as to be entirely consistent with the principle of fair equality of opportunity, and consequently cannot be ruled out on the ground that it necessarily violates that principle. Two new arguments have come to the forefront. The first is that affirmation action will help end racial segregation, which is the main cause of

Race   507 racial inequality in America today. The second is that affirmative argument will improve the education of white students in universities and colleges. These two arguments are of the general sort widely referred to as forward-looking arguments. The so-​called backward-looking argument that affirmative action can be just compensation for the damages of past injustice was seriously discussed in the early debates over affirmative action, but critics raised so many objections to it that it is now generally dismissed. One of the most effective objections against it is that it is “backward looking.” For example, Elizabeth Anderson complains that it is “backward looking; it aims to restore justice by undoing the wrongs of the past,” adding that “In focusing on past discrimination” it suggests that “discrimination is only in the past, and that current group disadvantage consists only of the inherited effects of past events.” She concludes that an adequate defense of affirmative action “would not locate the rationale for affirmative action in fast receding events of the past, but in present—​and future oriented concerns”(Anderson 2010:  135, 140, 141). These complaints are mistaken. Compensatory affirmative action does not aim to “restore” justice by undoing the wrongs of the past. It aims to prevent the present and future injustice of refusing to make compensation owed for past injustice. Obviously refusing or failing to make compensation that one owes is an injustice and a present one. Thus the rationale for compensatory affirmative action is not in “fast receding events of the past” but is the “present” concern to prevent the present and future injustice of refusing to make compensation when it is owed and due. Further, the claim that compensation is owed for the damages caused by past injustices is perfectly consistent with current group disadvantage being caused partly by present discrimination. That claim does not deny that forward-looking arguments for affirmative action can also be made. Accordingly I will swim against the current and defend the compensatory argument for affirmative action, at least when its beneficiaries are supposed to be black people. Indeed, I will go further and also argue that neither of the forward-looking arguments mentioned can stand as independent arguments for affirmative action. Their chief value is the help they provide in justifying the backward-looking compensatory argument for affirmative action. The plainest principle in the theory of compensatory justice is that when someone acts unjustly and his injustice harms others, he must compensate them. If it is not true then justice cannot be important or ever secure. If I steal a bicycle I must return it to its owner and my returning it, restitution, is part of my compensating him. Of course most applications of the simple principle are more complicated. For example, restitution may be impossible if I lost the bicycle that I stole. We need a definition of compensation to cover such cases. Robert Nozick’s definition will do: Something fully compensates a person for a loss if and only if it makes him no worse off than he otherwise would have been” (Nozick 1974: 57). On that account, a person can be compensated for a loss even when the loss did not result from an injustice, and even when he has no right to demand it. Further, on the same account, anyone can compensate a person for a loss even when it did result from an injustice. However, when

508   Bernard R. Boxill a person suffers a loss that did result from an injustice he has a right to demand compensation from the wrongdoer. As John Locke suggested, this right follows from his right to property, and consequently from his right to life. A person having the right to demand compensation for a loss from a wrongdoer who caused that loss is, however, consistent with other people compensating him, or even with other people being obligated to compensate him, especially if the wrongdoer lacks the means to compensate him or cannot be compelled to do so. For example, Locke showed that the beneficiaries of an injustice that caused losses may be required to bear the costs of compensating those who bear the losses, even if the beneficiaries in question are entirely innocent of the injustice that caused the losses (Locke 1988: 390–​1.) This conception of compensation allows that compensation for the damages caused by an injustice may disrupt what was considered to be a just pattern of distribution before the injustice. If equality in holdings is supposed to be a just pattern of distribution, that pattern will be disrupted if I must use my holding to compensate others for unjustly destroying their holdings. The argument that affirmative action can be justified as compensation can now be quickly summarized:  white America has for centuries unjustly harmed its black inhabitants and citizens; first it enslaved them, then it segregated them, discriminated against them, intimidated them, and cheated them. It built institutions that saved whites from the trouble of actually having to do these things, and that effortlessly assigned blacks to disadvantaged positions in society. If any blacks escaped the onslaught they had to live in fear and apprehension that their luck would not last, given that they satisfied the one sufficient condition for being battered by the more powerful whites. Finally, white America created, sustained, and inflamed the ideology of race that has stifled blacks’ aspirations, undermined their confidence in their abilities, and insulted them. Affirmative action would give blacks some compensation for these unjustly caused losses. Consequently, affirmative action is justified. The critics’ main objections to this argument can also be as quickly summarized. I list the three most common: First affirmative action compensates very few, since its main beneficiaries are people who are too young to have suffered from the onslaughts I have listed. Second affirmative action gives preference to blacks over whites who are too young to have engaged in the racially motivated injustices listed earlier. Consequently even if its beneficiaries deserve compensation, it makes innocents bear the costs of compensating them. Third, the main beneficiaries of affirmative action are the most qualified blacks; consequently, it compensates blacks who have suffered the smallest losses; but justice surely demands that those who suffer the greatest losses from injustice must be compensated before those who suffer smaller losses. For these three reasons compensatory justice cannot justify affirmative action. The first objection fails because black youngsters appearing on the market for the first time have been harmed by the race-​based injustice that harmed their parents. A favorite argument declares that compensating people for the harms they suffer because of harms caused to their parents and grandparents by past injustices is logically impossible. The crucial premise in that argument is that compensating such people would

Race   509 require making them no worse off than they would have been had the injustice not been committed, but that this is impossible because had the injustice not been committed the people’s parents and grandparents, and consequently the people themselves would not have been conceived. However, as I have shown elsewhere, it can be parried by limiting claims for compensation to harms caused after the conception of each generation of blacks after the injustice in question (Boxill 2003). The second objection also fails because whites who are too young to have engaged in harm, causing racial injustice, are nevertheless its beneficiaries, and consequently can be required to bear the cost of compensating those it has harmed. Anderson is disinclined to press this argument because she thinks that it is “divisive,” and the objection it tries to meet can be met more effectively because “there is no injustice in sharing the costs of widespread injustice” (Anderson 2010: 139, 140). That claim may be true in certain circumstances, such as when those who committed the injustices that caused the damages are dead or not known or cannot bear the costs of the compensation, or when their beneficiaries are not known or cannot bear the costs of the compensation. Otherwise the claim is plainly false. Is there no injustice when poor innocents are forced to share the costs of cleaning up the widespread damages caused by a careless or malevolent billionaire? Doesn’t the public rightly fulminate at the injustice of requiring it to pay for repairing the costs of industrial pollution when the polluters are known and capable of paying? And, in the case at hand, the wrongdoers, or at least their beneficiaries, are well known and capable of paying. If Anderson does not like my argument that whites are the innocent beneficiaries of earlier injustices against blacks, her own forceful argument that segregation is an unjust, opportunity-​hoarding institution makes whites the beneficiaries of injustice too—​and also probably guilty. The third of the objections, that affirmative action compensates the most qualified, and consequently those who have suffered the smallest losses is the most serious, though it is not decisive. It can be met by implementing additional compensating devices besides affirmative action that target the less qualified and more harmed. Or perhaps the most qualified have become the most qualified through their own efforts, although they suffered losses equal to, or greater than, the losses suffered by the less qualified. Consider the second of the forward-looking arguments mentioned earlier. Its basic claim is that white students in elite universities are likely to come from the same social and economic backgrounds, and likely, consequently, to share the same basic assumptions about society and the world. These similarities may lead them to miss the truth, or to fail to appreciate it properly. This is a loss to them even if their assumptions are true, since they would better understand and appreciate the truth of their assumptions if they were challenged. And if their shared assumptions are false, their loss would be even greater. In such circumstances preferentially admitting black students could improve the education it offered its white students even if the preferentially admitted students had lower test scores and grades than the other students. This argument may be challenged on the ground that the best qualified students usually make the best educational environments for each other. Being equally

510   Bernard R. Boxill knowledgeable they will challenge each other and force everyone to do his or her best. The weaker students that affirmative action will bring into the university will undermine this useful competition and not be taken seriously enough by the other students to usefully challenge their basic assumptions. Universities could surmount that difficulty by recruiting highly qualified foreign students. That strategy might improve the education of white students in elite universities—at least diversity improves education—but it would not get black students into the university or, as far as I can see, do anything for race relations in America. Another strategy would be to preferentially admit only the best black students who could presumably interact on equal terms with their white peers. But this strategy is likely to end up getting few if any black students into the university because the university would probably try to improve the epistemic diversity of its student body by preferentially admitting the best white lower-​class students instead of the best black students. In other words, the case for epistemic diversity does not provide a case for preferentially admitting black students. It only provides a case for admitting students with different backgrounds from the ones its students already have. A case for preferentially admitting black students must make their race a necessary part of the argument for admitting them preferentially. Here is one. Black people do not differ from white people merely in having different background assumptions: they also differ from white people in having background assumptions that involve a much better understanding of racism than white students. White people generally do not know much about racism. They do not have to. They are its beneficiaries but effortlessly so, since their institutions do a lot of the dirty work for them. Black people know more about racism than white people because they are its victims, and the powerful motive of self-​preservation compels them to try to understand it and the people through whom it is so dangerous. DuBois’s comment, “Of them I am singularly clairvoyant. I see in and through them” was meant to make the point (DuBois 1975: 29). And his further comment that black people are “gifted with second sight in this American world” indicates that he thought that they generally shared his clairvoyance (DuBois 1999 [1903]: 10). These comments may claim too much, but they have some plausibility, given that understanding in any area is won by study and attention, and black people have powerful motives for studying and paying attention to racism. If so, DuBois’s comments provide the basis for a forward-looking argument for the preferential admission of the best qualified black students into elite universities. The close livelong study of racism, aided by the reflections of earlier generations of black people all similarly preoccupied with the same study, will give these students an understanding of racism that helps resolve the serious problems that it poses. This argument is both backward-looking and forward-looking. Backward-looking because the beneficiaries of the beneficiaries of the affirmative action it proposes deserve compensation for the damages that they and their ancestors have suffered because of racial discrimination; and forward-looking since they are assumed to be people who will help solve the problem of racism. The above two-pronged argument has advantages over Anderson’s forward-looking argument for affirmative action noted earlier, namely, that it will help to end racial

Race   511 segregation which is the main cause of racial inequality in America today. Her argument is that the beneficiaries of affirmative action will be “those individuals best placed to act as agents of racial integration,” and consequently of the end of racial inequality (Anderson 2010: 148). But ending racial inequality is not the same thing as ending racism. Racial inequality is not a necessary condition for racism. Whites in America may still hate blacks after racial inequality has ended. They may hate them even more especially if racial inequality is ended because of affirmative action, which many whites already believe to be unfair. When these considerations are placed next to my earlier argument that the social construction of the black race in America has to a considerable extent been inspired by the dream of an all-​white America, the case for the superiority of my argument should be clear.

References Anderson, E.(2010). The Imperative of Integration. Princeton, NJ: Princeton University Press. Boxill, B. R. (2003). “A Lockean Argument for Black Reparations.” The Journal of Ethics 7(1): 63–​91. Delany, M. R. (2004 [1852]). The Condition, Elevation, Emigration and Destiny of the Colored People of the United States, and Official Report of the Niger Valley Exploring Party, ed. Toyin Falola. New York: Humanity Books. DuBois, W. E. B. (1975). Darkwate. New York: Kraus-​Thomson. DuBois, W. E. B. (1995). WEB DuBois: A Reader, ed. David Levering Lewis. New York: Henry Holt and Company. DuBois, W. E. B. (1999 [1903]). The Souls of Black Folk, ed. Henry Louis Gates Jr and Terri Hume Oliver. New York: Norton. Hume, D. (2000 [1739–​40]). A Treatise of Human Nature, ed. David Fate Norton and Mary J. Norton. Oxford: Oxford University Press. Jefferson, T. (1982 [1787]). Notes on the State of Virginia, ed. William Peden. Chapel Hill, NC: University of North Carolina Press. Jefferson, T. (2006 [1809]). The Essential Jefferson, ed. Jean M. Yarborough. Indianapolis, IN: Hackett. Kant, Immanuel (2007). Anthropology, History, and Education, ed. Gunter Zoller and Robert B. Louden. Cambridge: Cambridge University Press. Locke, J. (1988). Two Treatises of Government, ed. Peter Laslett. Cambridge:  Cambridge University Press. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. Plato (1992). The Republic, trans. G. M.  A. Grube, rev. C. D.  C. Reeve. Indianapolis, IN: Hackett. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rousseau, J. J. (1978). On The Social Contract with Geneva Manuscript and Political Economy, ed. Roger D. Masters. New York: St Martin’s Press. Williams, E. (1994  [1944]). Capitalism and Slavery, ed. Colin A. Palmer. Chapel Hill, NC: University of North Carolina Press. Williams, E. (2014). The Economic Aspect of the Abolition of the West Indians Slave Trade and Slavery, intr. William Darity. Lanham, MD: Rowman and Littlefield.

512   Bernard R. Boxill

Further Reading Alcoff, I. M. (2006). Visible Identities: Race, Gender, and the Self. New York: Oxford University Press. Balfour, L. (2003). “Unreconstructed Democracy:  W.  E.  B. DuBois and the Case for Reparations.” American Political Science Review 97(1): 33–​44. Cohen, A. I. (2009). “Compensation for Historic Injustices: Completing the Boxill and Sher Argument.” Philosophy and Public Affairs 37(1): 81–​102. Darby, D. (2009). Race, Rights, and Recognition. Cambridge: Cambridge University Press. Roberts, R. C. (2003). “The Morality of a Moral Statute of Limitations on Injustice.” Journal of Ethics 7: 115–​38. Shelby, T. (2004). “Race and Social Justice: Rawlsian Considerations.” Fordham Law Review 72: 1697–​7 14. Zack, N. (1993). Race and Mixed Race. Philadelphia, PA: Temple University Press.

Chapter 24

Work Paul Gomberg

Work raises issues of justice: unequal pay for the same work, discrimination in hiring and promotions, insufficient pay to allow workers to live decently, physical injuries at work, and harms caused by work activity itself to the worker’s health, abilities, and sense of self-​worth. While all these issues are important, this chapter focuses on the last. Some workers are benefited and others harmed by their work. Workplace harms are unjust, and perfectionist solutions to them are superior to Rawlsian proposals.

24.1 Preliminaries 24.1.1 What is Work? Normally, humans work: adult humans spend much of their lives working to serve either the larger community or their families; this is expected of them, making it “normal.” A norm is a social expectation of conduct (Gomberg 1997, 2002), and working is normative in all human societies. Philosophical writing on social justice has been more concerned with distribution of things (such as food and housing) and services (such as healthcare) than with how work is distributed. Yet work may seem to be organized unjustly:  many cannot find work (are unemployed); many workers’ tasks are mind-​ numbing and subordinate them to a supervisor; the work of others, however, is enjoyable and self-​directed, winning them social esteem and material security. However, the process by which some end up with good work and others with bad work or no work can seem unfair. I paraphrased being unable to find work as being unemployed, but work and employment are not the same. In some societies work typically takes the form of employment, where some are paid to work for others who issue them instructions; in other societies people work but never for pay.

514   Paul Gomberg Work differs from other human activities in having a purpose beyond the activity itself. Some human activities have no further purpose beyond the activity, for example solving puzzles or playing basketball for fun. To say we do it for pleasure or amusement is not to give a purpose beyond the activity.1 Work aims to produce, change, or maintain something: to produce a coil of steel, repair a copier, or lubricate a motor. The product or state that is the goal is contingently related to the activity itself: no steel may result; the copier may still not work; the motor may deteriorate. The goal or purpose is separable from the activity narrowly defined, but work must have a goal or purpose. While a goal or purpose is necessary to make an activity work, it is not sufficient. I may eat vegetables and whole grains to improve my health, but that is not my work. Human beings separate production from consumption; their activities of obtaining, transporting, processing, and cooking foods become work, and work becomes a characteristic human activity. Ordinarily, eating is not work. However, if, on an infomercial, I am paid to eat stuff and say “Delicious!”, then my eating is part of the social organization of work, and its purpose, to sell the stuff, is separable from the eating. If I tell stories to entertain others, this may or may not be work; if my story-​telling is a socially organized responsibility to entertain others, it is my work. Work that we enjoy is like play in that the activity itself is enjoyable. These comments do not give a fully adequate characterization of what distinguishes work from non-​work, but they are a start. A full account may be difficult, in part because goal-​directed activities which are avocational or aimed at self-​improvement may not be considered work by some, but may be considered to be “my work” by the person who does them. Paradigmatically, work is a socially organized contribution to a larger group. The work of a society produces goods or provides services people need—​food, shelter, clothing, child care—​and, in some societies, creates other goods and services such as healthcare and automobile repair. Work includes cultural activities when these are someone’s organized contribution to society: then playing sports and singing can be work. Human beings work, and human societies organize the work of creating the things we need—​ including music, science, and philosophy. In gatherer–​hunter societies typically younger men hunt, and women, youths, and older men gather or trap game. People contribute as they are able to tasks of obtaining food and other goods (Lee 1979). In stratified societies, where elite classes can exclude others from control of productive resources (especially land, but also water, draft animals, tools, and other essential means of life) (Fried 1967), divisions of labor become more elaborate. Some organize and control production: they issue instructions to others (slaves, bondservants, wage-​workers). The rise of modern capitalism alienated workers from land and changed their labor. English enclosures of the seventeenth through the early nineteenth centuries abrogated traditional use-​rights to land and forced many to 1 

Aristotle thought that pleasures were proper to the activities they complete; see, for example, Nicomachean Ethics, Book X, 1175a20–​b3. Activities done solely for the pleasures proper to them are done for their own sake.

Work   515 work for wages (Polanyi 1944; Neeson 1993); high unemployment limited wages of the employed. Detailed division of labor in manufacture and industry made workers’ tasks monotonous and mind-​numbing (Smith 1776), harms beyond the older one of labor that punishes the body or may kill the worker.

24.1.2  Harm at Work, Equal Opportunity, and Freedom of Occupation In The Market Experience Robert Lane writes, “Those whose jobs offer self-​direction, substantive complexity, and challenge, variety, little supervision, and intrinsic satisfaction of excellence or self-​determination may be called members of the privileged class” among workers. They “do freely and (usually) with social approval and support what they want to do and usually do best” (1991: 302). Professionals are at the top of this privileged class. Most workers are not so fortunate. Their work is closely supervised, sometimes under the direct eye of another person, sometimes under the control of a machine or assembly line which makes apparent any failure to keep up. Many have little control over their labor tasks. When I was a manual postal clerk in the 1970s I was often assigned to a particular task (sorting letters or larger flat envelopes, boxing letters that had been machine-​sorted) only to be told by a supervisor to go to another operation (sorting parcels, sorting sacks of mail coming down a slide). There was little point in thinking about what I was doing or attempting to organize my time, as I could abruptly be told to do something else. Other labor tasks may be interesting and challenging at first, while being mastered, but then descend into dull routine; I enjoyed learning to use a number keypad to audit military money order reports but soon found myself bored by the repetition of a simple, already mastered task. Much labor is repetitive, commonly leading to injuries. Workers are bored and frustrated, and may be very sore or in pain. There is an immense literature on these effects (Terkel 1974; Doppelt 1981; Ezorsky 1987; Sessions and Wortman 1992; Striffler 2005). More importantly, work changes us. Complex work develops our abilities, making us capable of even more complex tasks. Self-​direction (making decisions about how work is to be done) makes us generally more confident and capable in planning and executing a plan. Melvin Kohn and Carmi Schooler write, “Jobs that facilitate occupational self-​ direction increase men’s [the study’s subjects were men] ideational flexibility and promote a self-​directed orientation to self and to society; jobs that limit occupational self-​direction decrease men’s ideational flexibility and promote a conformist orientation to self and society” (1983: 152). Moreover, apart from self-​direction, the substantive complexity of work also affects us: Kohn and Schooler write, “substantive complexity actually does have a causal impact on one pivotal aspect of psychological functioning, ideational flexibility” (124). Put crudely, complex work that we control tends to make us smarter and more autonomous; simple, closely supervised work tends to make us stupid,

516   Paul Gomberg servile, and diffident about our abilities to make and execute plans.2 We regard bodily injuries at work as harms. Functional impairments caused by work are also harms. Because some work benefits while other work harms the worker, it seems that people should at least have equal opportunity for desirable work, and freedom to choose an occupation they prefer. Equal opportunity is usually understood as fair competition for positions more desirable, either extrinsically (pay) or intrinsically (the work itself). Discrimination by race, age, or gender violates equal opportunity to compete; the probability of attaining a position is affected by irrelevant considerations. Freedom of occupation concerns how laws or government directives can affect the availability of work: licensing requirements for healthcare professionals or requirements of service as a condition of being offered training can be regarded as limits on freedom of occupation. For those holding a negative view of freedom (Berlin 1969), freedom of occupation raises issues of liberty not raised by equal opportunity. For others, however, freedom of occupation and equality of opportunity serve the same fundamental interests. Rawlsians have argued that our work affects our highest-​ order interests in developing our two moral powers—​a sense of justice and a conception of what is good—​and our higher-​order interest in advancing our good. I have argued (Gomberg 2007) that it is good to develop our abilities, to contribute developed abilities to a human community through work, and to earn esteem for those contributions. Lucas Stanczyk, in a discussion of freedom of occupation (a chapter of From Each: A Theory of Productive Justice, unpublished at this time) appeals to our interest in the growth of intellectual, social, and moral capacities. On the views of Rawlsians, myself, and Stanczyk, we should evaluate social norms and practices by whether they advance or impede fundamental goods or interests. The three views share a conception of the citizen or human person as having important interests in their own intellectual and moral development. Accepted norms and practices may affect both equal opportunity and freedom of occupation, making unlikely or even impossible the realization of these interests. For example, concerning equal opportunity, I argue (2007: Ch. 3) that in all human societies children are socialized for the opportunities that will be available to them as adults in numbers that approximate the availability of those positions—​call this the Socialization Principle. Still, where desirable positions are limited, there are more aspirants than such positions; so access to them is decided by competition. Socialization reduces the number of competitors and tilts the results toward some and away from others, making equal opportunity impossible. Losers in the competition for the best jobs may have work which further stunts their abilities and is stigmatized. Thus, many are deprived of important goods—​developing complex abilities, contributing developed abilities to society, and earning esteem for their contributions—​and of a fair chance to obtain them. 2  Kohn and Schooler emphasize that the effects are reciprocal; those who are more self-​directed and ideationally flexible tend to have work that is more self-​directed and complex. They could make inferences about the effects of work through a ten-​year longitudinal study. The effects are tendential, like the effect of smoking on lung cancer.

Work   517 In a related vein, Stanczyk argues that the fundamental interest freedom of occupation should advance—​to have work which contributes to our moral and intellectual development—​is frustrated for millions who work on difficult, monotonous jobs picking crops, processing or butchering food, assembling electronics, etc. where market competition makes work fast-​paced, repetitive, and destructive of body and spirit; these workers are typically deprived of meaningful freedom to choose an occupation among a range of diverse opportunities. In another work Stanczyk (2012) starts from the premise that justice requires that all be provided at least minimal healthcare; he argues that providing that care may impair freedom of occupation. If people are unwilling to do the work necessary to provide healthcare to poor or rural populations, then states may justly compel workers to provide these services, for example by making it a condition of medical licensing that professionals serve underserved populations or by subsidizing medical training in return for service. In affluent countries these workers still retain freedom of choice among diverse occupations which contribute to their moral and intellectual development. In Ghana, nurses are prohibited from emigrating to provide nursing services elsewhere until they have provided five years of service. Stanczyk argues, however, that because alternatives for Ghanaian nurses are much more limited, their freedom of occupation is abridged, but not unjustly.

24.1.3 How the Organization of Work can Create Injustice If opportunities to engage in challenging work that develops our abilities is skewed toward some (white males from economically and educationally advantaged families) and away from others (in the United States, women and especially black and some immigrant workers, disadvantaged white people), then there seems to be social injustice. Let us say that social practices and institutions create injustice when (1) they harm some people but not others (or some much more than others); (2)  it cannot be adequately justified that those but not the others should be harmed (or experience greater harm); and (3) alternative practices and institutions are possible where those unequal harms (and comparable harms) are absent. Practices and institutions that harm people unnecessarily are bad; those that harm people unequally (without sufficient justification of the inequality) and unnecessarily are unjust. These words are not intended as a full account of social injustice but as an initially plausible account which will allow us to discuss injustice at work.3 A person’s probability of having a better or worse work-​life is influenced by the class and educational background of her family, its status as immigrant or not, its national 3 

I use this definition of social injustice to advance the argument. Ultimately my defense of justice at work is that it enables us to flourish together; appeals to an intuitive idea of justice play no role.

518   Paul Gomberg origin, and its social “race” (in the United States greatly influenced by race) as well as by gender.4 Opportunity is very unequal. In the United States students from families in the lowest twenty-five percent and the lowest fifty percent in income made up only three percent and ten percent respectively of the student population of elite colleges (Carnevale and Rose 2003). Elite colleges provide access to better jobs and careers, and these lead to longer lives, less morbidity, and greater happiness (Marmot 2004; Layard 2005). Poorer life prospects created by class and race are a social injustice—​provided that greater equality and better prospects are possible. People’s lives can be worse in at least two ways. People can be harmed by lack of access to goods and services such as housing or healthcare. But, as we have seen, people’s work or lack of work can also harm them: greater probability of suicide or divorce as a result of unemployment; loss of mental ability due to monotonous work; lower life expectancy caused by subordination to others at work (Marmot 2004). Our concern in this chapter is injustice related to the work process and its organization. In one sense, injustice in the organization of work is distributive injustice, in another sense not. When a phenomenon can be measured and compared to its context, we can speak of its distribution: of stars in a patch of sky, of weeds in the lawn, or of complex work among workers5—​this last an issue of distributive justice in this broad sense. There is, however, a narrower sense of “distributive justice” in which distribution refers to the social distribution of good things and services and of income and wealth that can be exchanged for these. Those good things and services are produced by people’s labor. Contributive (or productive) justice concerns how labor itself (and its absence) can benefit or harm workers, an issue of distributive justice in the broad, but not the narrow, sense of that phrase. The distinction between distribution (in the narrow sense) and contribution is important. Suppose the following are true: (1) what makes life good are the good things that labor produces; (2) there are inevitably fewer of these good things than people will be induced to want—​that is, they will be in limited supply; (3) because (1) and (2) are true, there will be competition to see who gets goods in limited supply; (4) any competition for goods in limited supply will unfairly favor some over others (Gomberg 2007:  Ch.  3). Under these assumptions distributive justice is impossible. If, on the contrary, we assume that what makes life good is what we do, particularly how we develop our abilities and contribute them through labor to the benefit of a human community with which we identify, then it becomes possible for all to attain the good simultaneously, provided that labor is shared (in ways I  will sketch). Under these

4  I mention gender separately because it plays a slightly different role. It does organize inequality within the working class but does not generally ground as much social segregation as the other categories; for example, women are only slightly underrepresented at elite colleges. 5  Samuel Arnold (2012: 106–​7 and n41) uses a measure of the skill required of the occupants of different occupations in how they interact with data, people, and things—​essentially a measure of how much complex intelligence people must exercise in their work.

Work   519 assumptions the most important goods are not in limited supply and justice is possible (Gomberg 2016). The rest of this chapter discusses whether harms resulting from the way work is organized constitute injustice and, if so, why. I  consider in detail two approaches to work and justice, Rawlsian political liberalism and my egalitarian perfectionism, showing how each would argue for an organization of work that did not harm workers. Rawls thought that a political conception of justice, grounded in shared ideas that were part of democratic culture, would avoid excessive political division and the coercion that would be needed to control a divided society: it would enable an overlapping consensus between various comprehensive religious and philosophical views. However, I argue, a Rawlsian egalitarianism that would avoid injustice in the organization of work is less likely to attract consensus than egalitarian perfectionism. To establish social injustice I must do more than argue some are harmed more than others without sufficient justification.6 That much seems clear already, but will be explained more carefully in relation to Rawlsian views of harm and my own. We must also consider how it is possible to have a society where such harms do not occur. In section 24.2 we explore Rawlsian approaches to this problem and in section 24.3 a perfectionist approach grounded in the view that work is a good. Section 24.4 argues that Rawlsian political liberalism would lead to a less stable society than would egalitarian perfectionism.

24.2  Rawlsian Approaches to Justice at Work John Rawls wrote little about work and did not fully explain his ideas. Hence in looking at “Rawlsian approaches” we must mine Rawls’s theory to show how his views about work can be incorporated into his theory of justice. The most explicit comment on the quality of work is in A Theory of Justice in §79 “The Idea of a Social Union.” In a well-​ ordered society: . . . the worst aspects [of the division of labor] can be surmounted: no one need be servilely dependent on others and made to choose between monotonous and routine occupations which are deadening to human thought and sensibility. Each can be offered a variety of tasks so that the different elements of his nature find suitable expression. . . . The division of labor is overcome not by each becoming complete in

6 

Aristotle and IQ theorists of the twentieth century believed that workers doing menial work were suited by nature for those tasks; the belief is unjustified (Gomberg 2007: Ch. 10; 2010: 13–​18).

520   Paul Gomberg himself, but by willing and meaningful work within a social union of social unions in which all can freely participate as they so incline. (1999a: 464)

In a section of Theory, addressing justice between generations and what justice requires us to save for future generations, Rawls writes that “we are not bound to go on maximizing [what we save for future generations] indefinitely” (1999a: 257). Then he writes: It is a mistake to believe that a just and good society must wait upon a high material standard of life. What men want is meaningful work in free association with others, these associations regulating their relations to one another within a framework of just basic institutions. To achieve this state of things great wealth is not necessary. In fact, beyond some point it is more likely to be a positive hindrance, a meaningless distraction at best if not a temptation to indulgence and emptiness. (Rawls 1999a: 257–​8)

Here Rawls cites the passage about meaningful work in a social union of social unions. When we have reached minimal abundance, it is better, he believes, to organize work so that it allows for exercise of decision-​making and intelligence than to create greater wealth. However, Rawls does not support this belief within the framework of his overall theory. I shall try to do this. In Rawls’s later thought a theory of justice must elaborate the principles of justice and institutional arrangements implied in the modern conception of democratic citizenship or, if that conception is insufficiently determinate, implied in the values which underlie our idea of democratic citizenship (1999b: 306). The democratic citizen has two highest order interests—​to develop and live by a conception of justice and to develop and revise a conception of the good—​and a higher order interest to advance their conception of the good (1999b: 365–​7; cf. 2006: 74). Primary goods are what the democratic citizen needs in order to develop and exercise the two moral powers. The principles of justice insure to citizens these primary goods. We will consider, first, what primary goods might be affected by conditions at work and why attaining these goods requires that work be complex and self-​directed and, second, how justice at work might be required by principles of Rawlsian justice. Rawls presents five primary goods as normally necessary for developing our two moral powers and pursuing our conception of the good: (1) basic liberties of thought and conscience and the political liberties; (2) freedom of movement and choice of occupation among diverse opportunities; (3) powers and prerogatives of offices and positions of responsibility; (4)  income and wealth; and (5)  the social bases of self-​respect. He briefly explains why each good is needed to allow people to develop their moral powers and achieve the good as they conceive it (2006: 308–​9; cf. 1999b: 366). The basic liberties and income and wealth seem not to address the harms of mindless labor and subordination at work. That leaves three other candidates. Choice of occupation among diverse opportunities “allow[s]‌the pursuit of diverse final ends and give effect to a decision to

Work   521 revise and change them” (2006: 308). These Rawlsian justifications of freedom of occupation seem relevant to ways in which a free choice of an occupation—​or a change of occupation—​may allow the worker more effectively to realize their conception of the good (by being a musician if making music is part of their conception of the good) but not to monotony and servility. However, if we are to provide to the worker the social bases of self-​respect and the powers and prerogatives of positions of responsibility, it may be necessary to eliminate servility and mindlessness at work. The absence of work might undermine a robust sense of our own worth or weaken abilities needed to develop and pursue a plan of life (2006:  xlvii; cf. 1999c:  50). Workplace norms governing the chain of command could also be argued to have negative consequences for a positive sense of self (Doppelt 1981). Rawls writes that depriving some of equal liberties would undermine citizens’ self-​respect, “publicly establishing their inferiority” and hence damaging self-​esteem (1999a: 477). Likewise subordination to a boss—​when work occupies so many waking hours—​seems to undermine workers’ sense of their own worth, diminishing the value of workers’ ideas and making their will largely impotent; norms of workplace subordination also “publicly establish our inferiority” and can be “humiliating and destructive of self-​esteem.” Continued monotonous labor can undermine workers’ confidence in their abilities, as Arthur Kornhauser’s interviews of autoworkers in the 1950s showed (1965), and can harm their abilities, as Kohn and Schooler (1983) show. Such monotony would seem to damage workers’ confidence in the worth of their life plans, confidence needed “to pursue their conception of the good with zest and to delight in its fulfillment” (Rawls 1971: 178; cf. 1999a: 155). The detailed division of labor—​work narrowly focused on a single task—​can harm other abilities relevant to the highest order interests of a democratic citizen. The Aristotelian Principle holds that we enjoy the exercise of developed abilities and that the exercise of more complex abilities is more enjoyable than of simple ones (within the limits of what we can do without excessive stress) (Rawls, 1999a: 274). Rawls writes that “in the design of social institutions a large place must be made for [the Aristotelian principle], otherwise human beings will find their culture and form of life dull and empty. Their vitality and zest will fail as their life becomes a dull routine” (377). These remarks remind us of Rawls’s belief that we should strive to eliminate “monotonous and routine occupations” and servile dependency at work (464). Rawls links self-​respect to the Aristotelian principle and our need to exercise complex abilities: When activities fail to satisfy the Aristotelian Principle, they are likely to seem dull and flat, and to give us no feeling of competence or a sense that they are worth doing. A person tends to be more confident of his value when his abilities are both fully realized and organized in ways of suitable complexity and refinement. (Rawls 1999a: 386–​7)

Since we spend so much time working, we can attribute to Rawls an incipient version of the conclusion that Kohn and Schooler reached from their research that dull routine labor tends to undermine self-​confidence generally for both women and men

522   Paul Gomberg (1983:  20–​47; cf. Kornhauser 1965). Hence it could damage our ability to form and pursue a plan of life with confidence: these social conditions tend to undermine workers’ self-​respect. What Rawls calls “powers and prerogatives of offices and positions of responsibility” (2006: 308) is the other primary good apparently relevant to the social organization of work. Rawls includes this primary good because these powers “give scope to various social and self-​governing capacities of the self.” Samuel Arnold (2012) develops the following comparison: just as the basic liberties are the social basis of the internal good of self-​respect so, for Rawls, positions of authority and responsibility are the social basis of the internal goods of social agility and self-​government and control.7 When our work requires us to interact skillfully with others and judiciously control our conduct, we develop the internal resources of managing complex social interactions and governing our conduct. (Teachers who strive to get the best out of their students are familiar with the constant learning process and the difficulty of acquiring both abilities.) Arnold argues that while offices and positions of authority and responsibility cultivate social and self-​ governing abilities, work that is substantively complex causes us to develop greater intelligence and skill (as Kohn and Schooler’s research showed). Arnold defines intelligence as “ability to reason, to plan, to solve problems, to think abstractly, to comprehend ideas, and to learn” (101). While most positions of responsibility and authority (what Rawls seems to have in mind) require complex mastery, some work is substantively complex without involving authority over others or requiring social skill. Arnold’s argument complements earlier arguments that conditions at work should support workers’ self-​respect. Together they show how conditions at work can either foster or undermine a variety of internal resources: a positive sense of ourselves and confidence in that positive assessment, social skill, ability to govern our conduct, and intelligence and virtuosity. These internal resources are needed for us intelligently to form and revise our conception of the good, to pursue that conception with confidence in its value, and to realize that conception of the good. These ideas are deeply Rawlsian, as Rawls writes, “Rational autonomy  . . .  rests on intellectual and moral powers. It is shown in [citizens’] capacity to form, to revise, and to pursue a conception of the good, and to deliberate in accordance with it” (2006: 72). Just as the basic liberties provide social support for these “intellectual and moral powers,” so do powers and prerogatives of offices and positions of authority, responsibility, and complexity. Rawls argues that, as citizens, we need these abilities. So justice requires conditions of work that foster, rather than undermine, these internal resources. In this section I set for myself two tasks: first, to show why Rawls’s primary goods require that our work be complex and self-​directed and, second, to show how the 7 

Arnold has an excellent discussion of the difference between the internal good of self-​respect and its social basis; norms of justice can try to provide the social basis of self-​respect, not self-​respect. His discussion of self-​respect adheres to Rawls’s official position that its social bases are the basic liberties; but see Rawls (1999a: 478–​9) for the suggestion that the difference principle may apply to the social bases of self-​respect.

Work   523 requirement that work be meaningful in these ways would be incorporated into the principles of justice. What I  have written completes the first task. As for the second: Rawls nowhere indicates that freedom from subordination at work or from excessive monotony of work should be among the basic liberties.8 The principle of fair equality of opportunity seems to address equal and fair chances or prospects to obtain good work; probabilities can, in theory at least, be equal but very low for all because such work is scarce.9 Our concern is that anyone who experiences excessive subordination or monotony at work will be harmed. Therefore, as Arnold argues, the difference principle is the relevant principle to guarantee meaningful work.10 The difference principle says that inequalities are allowable provided they improve the situation of the least-​well-​off group.11 The idea is familiar that income and wealth inequalities can create an incentive to the talented to develop and exercise their talents, hence increasing productivity and goods available, hence conceivably increasing income and wealth for those with the least. Can a parallel argument be made with respect to powers of office, that inequalities in these lead to an increase in authority and power for those with the least? It seems not: centralization of control of production (greater inequality of powers between workers and their bosses) has led to a decrease in authority, power, and complexity of tasks for those with least (compared with earlier production in small shops). If the least-​well-​off group benefits, it must be because centralization of powers of office raises productive efficiency so much that, taking all the primary goods into account, the increase in income and wealth for the worst-​off more than compensates for loss of influence and increased monotony at work. There are two problems with this idea. First, evidence that inequalities in the powers of positions enhance productivity is debatable at best. Arnold (2012: 109–​14) argues for some skepticism. Citing Murphy (1993: 43–​5), he points out that the rational organization of labor tasks does not require that workers be excluded from the planning process or confined to a single task. Moreover, owners may centralize power for reasons other than efficiency. Second, as Rawls suggests (1999a:  257–​8 and quoted earlier), greater control over one’s work benefits the worker more than additional income and wealth.

8  But O’Neill (2008) argues that Rawls’s justifications of the basic liberties can be understood as applying with comparable force to control over workplace decision-making. 9   But not in reality, as I argue in 2007: Ch. 3. Freeman (2007a: 135) suggests that fair equality of opportunity might require wide dispersal of powers and prerogatives, but does not develop this suggestion. 10  The point that the difference principle applies not only to differences in income and wealth but also to differences in positions of authority and responsibility is stressed by Freeman (2007a: 113–​15, 134; 2007b: 107–​8) and, citing Freeman, by O’Neill (2008: 48–​51; 2009: 385). Arnold develops it best. 11  The difference principle applies only to primary goods that can vary. In Theory Rawls says those are only powers of office and income and wealth (1999a: 80); the others are equal for all. Was he considering the social bases of self-​respect? If these are allowed to include work conditions, then these could be unequal, and the difference principle could apply to them. My explanation of work and self-​respect develops Freeman and O’Neill’s position; relevant citations are in n. 9, 10, and 11; for Rawls’s official position, see n. 8.

524   Paul Gomberg Considering Rawls’s index of primary goods, the least-​well-​off group benefits from wide dispersal of powers of positions. There are, then, two ways that Rawls’s theory can be developed to include justice at work. In order to develop social and intellectual skills needed by the democratic citizen, workers must have authority over their own work (there must be wide dispersal of powers of positions), and work must be substantively complex; the same conclusion could come from considering the social bases of self-​respect, if they are allowed to be unequal, which they would be if they include powers of positions at work. The difference principle requires maximizing the index of primary goods, including powers of positions, for those with least; so, on either way of developing Rawls’s view, work would need to be re-​organized to give workers more control over work and greater job complexity.12 Rawls’s preferred institutional realization of the principles of justice was what he called “property-​owning democracy”—​private ownership of productive capital in a competitive market economy surrounded by taxation policies, particularly limits on bequests, to prevent concentrations of wealth. Much literature on property-​owning democracy (Freeman 2007a, 2007b; O’Neill, 2008, 2009; Hsieh, 2005, 2008a, 2008b, 2009, and 2012; O’Neill and Williamson 2012) argues that wide dispersal of wealth would empower workers. Some of this literature is not well tied in with discussions of primary goods and how these advance democratic citizens’ interests; Freeman’s suggestions (2007a: 133–​6) tie things together but are undeveloped. Rawls, when explaining the difference principle, focuses on money and even assumes very unequal distribution of powers of office (2001:  126). The present discussion somewhat mechanically considers the primary goods and their basis in the interests of a democratic citizen, slighting issues of stability and institutional realization of justice. A more integrated development of Rawls’s theory is needed.

24.3  Aristotelian/​M arxist Perfectionism and Justice at Work Rawls came to believe that A Theory of Justice had offered a comprehensive moral doctrine to defend justice as fairness, but a society founded on a comprehensive doctrine can only be sustained over time by the “oppressive use of state power” (2006: 37). He thought this a fatal objection to A Theory of Justice; a conception of justice must, he thought, derive from our idea of democratic citizenship or the values that underlie that idea. He thought that social justice could not be defended based on a view of human nature. I disagree with Rawls. A contributively just society can be defined as one where 12 

This is a left interpretation of Rawls and is hard to reconcile with much that he writes. For discussion, see Gomberg (2010: 22–​4).

Work   525 labor is shared and each has both opportunity and duty to contribute simple labor and to develop and contribute more complex abilities. I have argued (2007) that contributive justice creates social conditions that allow everyone to achieve a constellation of important goods: to develop abilities, to contribute developed abilities through work, and to earn esteem for those contributions. Under these conditions it is possible for all of us to have good lives, according to a shared understanding of what makes life good. In this section I present this alternative road to justice at work. In section 24.4 I defend its superiority to Rawlsian views. Social labor (work serving others’ needs) is both a natural and a normative good. The assertion that social labor is a natural good does not imply that it is a good for everyone. Compare:  sexual activity and intimate attachments are natural goods but not good for everyone (not everyone is capable of them or wishes to be sexually active or have attachments). Some things make life go better, others worse. Having friends makes our lives better. So does work serving others. We are the least individually self-​sufficient and most social of primates. Work is part of our humanity; through work we care for and are cared for by others (cf. Kittay 1999). Work is also a normative good. Human social cooperation is possible because we are bound by norms, shared expectations of conduct enforced by esteem (Gomberg 1997). We naturally develop a sense of reciprocity and fairness (Tomasello 2009). Because caring for one another’s needs is central to human life, it is normative (expected) that we work to provide for others. When we reach maturity we expect that our labor will contribute to the good of others just as others’ labor has contributed to our good. The connection between work and being a respected adult is universal in human societies, including those of gatherer–​hunters (Leacock and Lee 1982: 1–​20). Hence, work is simultaneously a natural and a normative good. This conclusion may seem paradoxical. If work is good for us, why do we need a norm that we should work? But norms requiring us to do things which benefit us are often needed. Consider care for children. Those who have experienced the joys (and trials) of being parents can testify that it can be a great good. Still, often parents have to do things that they don’t want to do. Something can be both a great good and an important duty. There is no paradox here. The conclusion that social labor is a good is compatible with recognizing that it is a harm when it injures our bodies, stunts our minds, or undermines self-​worth. Applied to work, the Socialization Principle (see section 24.1.2) implies that, where labor is divided—​a few doing engaging, complex, self-​directed labor (Lane’s “privileged class” of workers) and many others labor that is mindless, monotonous, and under others’ control—​some children will be socialized with aspirations and abilities appropriate for complex, self-​directed work, while others will be socialized with skills and aspirations for necessary but simple labor (Gomberg 2007: Ch. 3). This labor harms them because it rewards servility and does not sufficiently engage human abilities. Under this division of labor it is impossible for all to have good lives. The solution is to share labor. Consistently with the Socialization Principle, it is possible for all simultaneously to have complex, engaging work as part of a generally good life if the residue of necessary simple

526   Paul Gomberg labor is shared. Sharing simple labor makes it possible for each person to develop, contribute, and earn esteem for contributing more complex skills and abilities, provided complex labor too is shared as well among all with the relevant mastery. For example, in a hospital, doctors would help clean up; housekeeping staff would have opportunity and encouragement to develop more complex skills and to contribute mastered abilities (Gomberg 2007: Ch. 7). This argument assumes a constellation of good things: to develop our abilities, to contribute developed abilities to a social group with which we identify, and to earn esteem for these contributions.13 These goods importantly benefit those who attain them. This is a very weak perfectionism (Chan 2000; Wall and Klosko 2003). It does not imply what Rawls came to call a “comprehensive doctrine,” but only a limited view of what makes a human life go well.14 When labor is shared, esteem for our work is in unlimited supply (in hierarchical societies it is in limited supply). Each can be esteemed for contribution of simple labor; sharing labor removes the negative stigma (this work too can be enjoyable, and no one’s life is consumed by simple labor). Each has opportunity to master complex abilities and to earn esteem for their contribution. Each can earn esteem for helping others to acquire new abilities. Each can earn esteem for contributions as a philosopher, architect, landscape designer, or physician. Each can earn esteem for displaying goodwill by contributing in all these ways and thus enabling simultaneous flourishing. This does not mean equal esteem for all, but why be obsessed with quantitative measures (exact equality)?15 Esteem is specifically for what one does. Consider, as examples, the contributions to social philosophy of Rawls’s A Theory of Justice and Eva Feder Kittay’s Love’s Labor; if forced to rank them, we would say that Rawls’s contribution is greater. But so what? Kittay points out that dependency is the normal condition for all, at least as children, a point we could generalize, since we all depend on social labor. A theory of justice must start with the fact of human dependency on others. Kittay’s contribution can be esteemed for what it is; ranking her contribution in comparison to Rawls’s is odious (Gomberg 2016). The flourishing of each can contribute to the flourishing of all. Those with complex mastery in areas other than philosophy can enrich my life, filling it with music, crafts, and visual art, with beautiful buildings, landscapes, and neighborhoods, developing our understanding of nature, or solving environmental problems. Other social philosophers can enrich my life by participating in a community of mutual criticism and support. As

13 

Rawls (1999a: 73) endorses the second and probably the other two as well; on his later view, however, these are not intrinsic goods but implied by democratic citizenship. He is not consistent about labor being a good (Gomberg 2010: 22–​4). 14  See Kraut (2007) for an excellent defense of what he calls “developmentalism,” a better name for the view. I have stuck with the more traditional—​and somewhat pejorative—​“perfectionism.” 15  In a competition, precise quantitative measures and ranking may be necessary because small differences can decide who wins.

Work   527 simple and complex labor are shared, the community of social philosophers (and other communities) could greatly expand. Relationships are transformed. The social psychology of vanity, jealousy of others’ accomplishments, and, on a group level, racism, arises from competition for limited positions of prestige (Gomberg 2007: Ch. 8). When labor is shared, esteem earned from contribution of complex abilities is not scarce, and others are no longer competitors for it. While our abilities and contributions would be different, social competition for limited positions of prestige would disappear. Opportunities to contribute complex mastery could be greatly expanded as complex abilities could be broken down in more fine-​grained ways (Gomberg 2007: Ch. 7). We attain the goods of the constellation by helping others to attain those same goods. In a society where labor is shared people would be bound by norms of contributive justice: each has an opportunity (and a duty—​to cover the days one may not be feeling so great) to contribute both simple and complex abilities to society; each has opportunity (and a duty) to develop more complex abilities, not only as a child, but throughout their life; each has opportunity and a duty to help decide how to organize social cooperation. In a contributively just society, where our labor makes it possible for each to grow, learn, and develop abilities, contributing is a natural good for which we seek opportunity. Moreover, it is a normative good, a duty arising from norms of contributive justic­e. Where there is a norm (a shared expectation of conduct enforced by esteem) that each should contribute routine labor, develop their abilities, and contribute more complex labor, esteem is accorded to those who contribute. While we naturally think well of those who contribute to our good, norms of contributive justice make this connection between contribution and esteem necessary (Gomberg 2007: Ch. 13). Anthropological study of egalitarian groups shows how such social organization works (Gomberg 2007: Ch. 9); norms of contributive justice would make it possible in a modern complex society. Doing for others is not the only good, but it is a central social good; it creates a context for the development and expansion of human creativity. Because we are flourishing together, norms of contributive justice are “robust in the face of reflection” (Gomberg 2007: 159). How would contributive justice be realized institutionally? It may seem that norms of contributive justice are compatible with market economies. I am skeptical (Gomberg 2007: Chs 12 and 13; 2016). In order to contribute our labor, we need nutrition, shelter, and other material goods. The norm that each should have opportunity to contribute requires that each receive these goods. Distribution of material needs serves and advances contributive justice. We tend to assume that people should be paid to work; however, there are two reasons to think that paying people to work would undermine a contributively just society. First, in that society we would naturally seek opportunity to contribute our abilities; enjoyment of work is reinforced by our recognizing, on reflection, that such contribution helps to make life good. Paying people to work conveys the opposite message, that labor is a disutility which requires compensation (as Adam Smith thought). I speculate that this message influences people now; moreover, there is

528   Paul Gomberg evidence that paying people to do things they enjoy can detract from their intrinsic enjoyment of the activity (Deci et al. 1999). Second, paying people to work would weaken the motivating force of norms of contributive justice. Under those norms each receives what she needs because receiving these things enables her to contribute. Each has a duty to contribute so as to enable others’ contributions; group-​centered norms are woven into daily life. If receiving what we need depends on our working, our motivational focus narrows: we think of labor as a means of providing for self and family, rather than for the needs of others and as a duty of justice. It seems narrower motives could overwhelm and weaken motives of contributive justice. The argument of these last two paragraphs is that paying people to work would undermine motives of contributing as a good and a duty. Institutionally, there is reason to believe a society of contributive justice is a moneyless communist society (Gomberg 2016). My defense of contributive justice should be understood as follows: the description of this section is a speculation, imperfectly supported by evidence, that all would be better off in a contributively just society. This is a consequentialist defense of a society based on a contributive conception of justice.

24.4  Political Liberalism or Egalitarian Perfectionism? In this section I compare Rawlsian political liberalism with contributive justice. I argue, first, that the two approaches are both radical enough that they would raise similar transition problems from current injustice to a socially just society and, second, that contributive justice based on a perfectionist view of human good would be more stable than Rawlsian political liberalism—​exactly the opposite of what Rawls thought. For both the left-​Rawlsian view developed in section 24.2 and contributive justice there is a problem of transition from current injustice to a society conceived as socially just. Transition to a society of contributive justice is, I acknowledge, likely to be violent. Does left-​Rawlsianism have an advantage in avoiding a violent transition? I doubt it. When Rawls’s primary goods are “thickened” so that they imply that subordination to a boss must end and workers be given scope to exercise social judgment and other complex abilities at work, the changes required would encounter resistance that could be overcome only by considerable use of force. Contributive justice has an advantage and a disadvantage in making this transition. The advantage is in the simplicity of its argument compared with that of left-​ Rawlsianism. The transition to a just society as conceived by each view depends on persuading many that changes need to be made. The left-​Rawlsian argues something like this:  our conception of democratic citizenship implies that citizens are free and equal; this freedom and equality requires the capacity to develop the two moral powers

Work   529 and the means to advance our conception of the good; the development of the moral powers requires internal capacities of developing and revising a plan of life, of forming and being disciplined by concerns of justice, and of self-​respect; developing these internal capacities requires ending subordination to a boss at work and work that gives scope to the exercise of complex abilities. Compare this left-​Rawlsian argument with that for contributive justice: we do better when we develop our abilities, practice them in our work, and earn esteem for our social contributions; when labor is divided with most doing simple work, most children will be socialized with abilities and aspirations for simple labor; such socialization harms children just as simple labor harms workers; it is possible for all to develop and contribute more complex abilities only if labor is shared. Each claim is broadly empirical—​ granted that each is also speculative and backed only imperfectly by evidence. The social philosopher just assembles them into an argument. If further evidence and discussion supports these theses, they are fully capable of persuading others. The argument for contributive justice is simpler, more accessible, and more persuasive; the argument for left-​ Rawlsianism is based on a debatable interpretation of the implications of democratic citizenship or of the values that ground it. Contributive justice also has a disadvantage. In section 24.3 I argued that, institutionally, it requires that work not be paid, a very radical change. The argument depended in part on the claim that paying people obscures the fact that labor is a good. But since people now are paid and since much work is oppressive of and harmful to the worker, it is harder to persuade people that, ideally, work is an important good. Even with that qualification, however, I believe that, on balance, because of the simplicity of the argument for contributive justice, it is easier to persuade people of that conception than of the left-​Rawlsian view. Hence the transition to a society of contributive justice is more likely. More importantly, contributive justice, once in place, would be stable. In a contributively just society each could see that its norms enable all to flourish simultaneously, and state power would not be needed to sustain them. In contrast, there would likely be a continual need for the “oppressive use of state power” to maintain the norms of left-​Rawlsianism when these norms are combined with a market-​based money economy. I have argued that paying workers would undermine norms of contributing as a duty of justice. Here I generalize that argument: not just pay for work but a market exchange economy cannot be combined with egalitarian norms of justice. Market interactions in a stable economy encourage anti-​egalitarian ways of thinking. Specifically, market practices encourage bargaining to advance self-​interest.16 They imply norms:  fraud, theft, dishonest bargaining, and reneging on agreements are wrong; we have a right to property justly received through exchange or gift and are free to exchange or give

16 

Joseph Carens (1981) has argued that this need not be so, that markets can be combined with moral incentives for labor. For my reply, see Gomberg (2007: 146–​7).

530   Paul Gomberg it as we wish. These norms are the kernel of an idea of social justice. Libertarian social philosophy develops and rationalizes this sense of justice. Just as paying people to work undermines working from a sense of duty to others (motives of self-​and-​family interest would swamp motives of justice), so market transactions in general encourage a narrowly self-​interested approach to all our everyday economic activities. We think of these activities from the individual point of view, not from a general social point of view. To be stable, a conception of justice must make sense to people in everyday life. Markets entrench ways of thinking that are inimical to egalitarian notions of justice, whether left-​Rawlsian or contributive. A Rawlsian property-​owning democracy (or even market socialism) would retain a money economy and market relationships which ground individualist, libertarian interpretations of democratic citizenship; left-​Rawlsianism tries to marry those economic institutions with radically egalitarian policies—​such as prohibiting bequests—​which challenge the norms of property implied in market exchange. That wedding will not take place. Because market economies encourage a narrow focus on self-​interest and reward practices that advance these in either wage-​bargaining or consumption, people in stable market societies tend to reject egalitarian norms that are based on viewing issues of social justice from a social rather than an individual perspective. To maintain egalitarian norms in the face of their seeming so unnatural to people in everyday life would require continual interference by the state power, and such interference would be experienced as oppressive. To the extent a market society was stable and people therefore rationalized market exchanges (Gomberg 2007: 144–​5 for an explanation of rationalization), imposing egalitarian norms of justice would tend to destabilize it. Socialist revolutions of the twentieth century popularized more egalitarian ideas and tended to destabilize capitalist societies; however, with the collapse of the Soviet Union and the blatantly aggressive capitalism of “communist” China, the natural tendency of stable market economic relations to encourage anti-​egalitarian ways of thinking about justice is manifesting itself in Anglo North America, Europe, and elsewhere. Either we keep markets and their associated ideas of justice, rights, and ownership (and a conception of democratic citizenship that will be interpreted in light of these); or we look at the obvious conditions under which human life goes well and change our social and economic institutions so as to make a good life possible for all (Gomberg 2007: Chs 12 and 13; 2016). Contributive justice requires only that we do our part in a system where we all flourish together.

Acknowledgments I thank Serena Olsaretti for exhaustive, detailed, and helpful comments on three drafts, and Andrea Veltman for useful criticism of the penultimate draft.

Work   531

References Arnold, S. (2012). “The Difference Principle at Work.” Journal of Political Philosophy 20(1): 94–​118. Berlin, I. (1969). “Two Concepts of Liberty,” in I. Berlin (ed.) Four Essays on Liberty. Oxford: Oxford University Press. Carens, J. H. (1981). Equality, Moral Incentives, and the Market. Chicago, IL:  University of Chicago Press. Carnevale, A. P. and Rose, S. J. (2003). Socioeconomic Status, Race/​Ethnicity, and Selective College Admissions. New York: The Century Foundation. Chan, J. (2000). “Legitimacy, Unanimity, and Perfectionism.” Philosophy and Public Affairs 29(1): 5–​42. Deci, E. L., Koestner, R., and Ryan, R. M. (1999). “A Meta-​Analytic Review of Experiments Examining the Effects of Extrinsic Rewards on Intrinsic Motivation.” Psychological Bulletin 125(6): 627–​68. Doppelt, G. (1981). “Rawls’ System of Justice: A Critique from the Left.” Noûs 15(3): 259–​307. Ezorsky, G. (ed.) (1987). Moral Rights in the Workplace. Albany, NY:  State University of New York Press. Freeman, S. (2007a). Rawls. London: Routledge. Freeman, S. (2007b). Justice and the Social Contract. Oxford: Oxford University Press. Fried, M. H. (1967). The Evolution of Political Society:  An Essay in Political Anthropology. New York: McGraw-​Hill. Gomberg, P. (1997). “How Morality Works and Why it Fails: On Political Philosophy and Moral Consensus.” Journal of Social Philosophy 28(3): 43–​70. Gomberg, P. (2002). “The Fallacy of Philanthropy.” Canadian Journal of Philosophy 32(1): 29–​66. Gomberg, P. (2007). How to Make Opportunity Equal:  Race and Contributive Justice. Oxford: Blackwell Publishing. Gomberg, P. (2010). “Dilemmas of Rawlsian Opportunity.” Canadian Journal of Philosophy 40(1): 1–​24. Gomberg, P. (2016). “Why Distributive Justice is Impossible but Contributive Justice would Work.” Science and Society 80(1): 31–​55. Hsieh, N. (2005). “Rawlsian Justice and Workplace Republicanism.” Social Theory and Practice 31(1): 115–​42. Hsieh, N. (2008a). “Survey Article:  Justice in Production.” Journal of Political Philosophy 16(1): 72–​100. Hsieh, N. (2008b). “Workplace Democracy, Workplace Republicanism, and Economic Democracy.” Revue de Philosophie Économique 9(1): 57–​78. Hsieh, N. (2009). “Justice at Work:  Arguing for Property-​Owning Democracy,” Journal of Social Philosophy 40(3): 397–​411. Hsieh, N. (2012). “Work, Ownership, and Productive Enfranchisement,” in M. O’Neill and T. Williamson (eds) Property-​Owning Democracy: Rawls and Beyond. Malden, MA: Wiley Blackwell, pp. 147–​62. Kittay, E. F. (1999). Love’s Labor:  Essays on Women, Equality, and Dependency. New  York: Routledge. Kohn, M. L. and Schooler, C. (1983). Work and Personality: An Inquiry into the Impact of Social Stratification. Norwood, NJ: Ablex Publishing.

532   Paul Gomberg Kornhauser, A. (1965). Mental Health of the Industrial Worker: A Detroit Study. New York: John Wiley and Sons. Kraut, R. (2007). What is Good and Why. Cambridge, MA: Harvard University Press. Lane, R. E. (1991). The Market Experience. Cambridge: Cambridge University Press. Layard, R. (2005). Happiness: Lessons form a New Science. New York: The Penguin Press. Leacock, E. and Lee, R. (eds) (1982). Politics and History in Band Societies. Cambridge: Cambridge University Press. Lee, R. (1979). The !Kung San:  Men, Women, and Work in a Foraging Society. Cambridge: Cambridge University Press. Marmot, M. (2004). The Status Syndrome:  How Social Standing Affects our Health and Longevity. New York: Times Books. Moriarty, J. (2009). “Rawls, Self-​Respect, and the Opportunity for Meaningful Work.” Social Theory and Practice 35(3): 441–​59. Murphy, J. B. (1993). The Moral Economy of Labor. New Haven, CN: Yale University Press. Neeson, J. M. (1993). Commoners: Common Right, Enclosure and Social Change in England, 1700–1820. Cambridge: Cambridge University Press. O’Neill, M. (2008). “Three Rawlsian Routes towards Economic Democracy.” Revue de Philosophie Économique 9(1): 29–​55. O’Neill, M. (2009). “Liberty, Equality and Property-​Owning Democracy.” Journal of Social Philosophy 40(3): 379–​96. O’Neill, M. and Williamson, T. (eds) (2012). Property-​Owning Democracy: Rawls and Beyond. Oxford: Wiley-​Blackwell. Polanyi, K. (1944). The Great Transformation. Boston, MA: Beacon Press. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (1999a). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press. Rawls, J. (1999b). Collected Papers, ed., Samuel Freeman. Cambridge, MA: Harvard University Press. Rawls, J. (1999c). The Law of Peoples. Cambridge, MA: Harvard University Press. Rawls, J. (2001). Justice as Fairness: A Restatement, ed. Erin Kelly. Cambridge, MA: Harvard University Press. Rawls, J. (2006). Political Liberalism, expanded edn. New York: Columbia University Press. Sessions, R. and Wortman, J. (eds) (1992). Working in America: A Humanities Reader. Notre Dame: University of Notre Dame Press. Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations, ed. E. Cannan. Chicago, IL: University of Chicago Press, 1976. Stanczyk, L. (2012). “Productive Justice.” Philosophy and Public Affairs 40(2): 144–​64. Striffler, S. (2005). Chicken:  The Dangerous Transformation of America’s Favorite Food. New Haven, CN: Yale University Press. Terkel, S. (1974). Working. New York: Ballantine. Tomasello, M. (2009). Why We Cooperate. Cambridge, MA: MIT Press. Wall, S. and Klosko, G. (eds) (2003). Perfectionism and Neutrality: Essays in Liberal Theory. Lanham, MD: Rowman and Littlefield Publishers.

Chapter 25

Expl oitat i on Benjamin Ferguson and Hillel Steiner

25.1 Introduction Broadly speaking, the term ‘exploitation’ can be applied to a number of different kinds of interaction. Mineral deposits, mistakes, workers, and wombs are all said to be exploited. One sense of exploitation, often applied to the extraction of resources, or capitalizing on an opponent’s mistaken move in a board game, connotes mere use or advantage taking. Exploitation in this sense need not imply any form of distributive injustice, or indeed, any wrongdoing at all. However, when we claim that a group of workers is exploited, we use a second sense of the word. Exploitation in this normative sense involves taking unfair advantage and is a form of distributive injustice. In what follows we shall focus solely on exploitation in the normative sense. When Alice steals Bob’s wallet, she takes it without his consent and makes him worse off than he was. Conversely, when Bob buys an apple from Alice’s fruit stand, the transaction is consensual and mutually beneficial. Transactions like the apple purchase, in which all parties are better off and no parties worse off, are strictly Pareto-improving transactions. Some accounts of exploitation include non-​consensual and/​or harmful transactions such as theft, while others limit their scope to consensual and strictly Pareto-improving exchanges. At least regarding normative analysis, it matters little whether we extend the term beyond consensual and mutually advantageous cases, for it is easier to say what has gone wrong when consent is lacking and harm has been done. Explaining the injustice of transactions that are both Pareto improving and consensual, on the other hand, is more difficult. In fact, some philosophers and economists have claimed that the reason it is difficult to explain the injustice of such transactions is that they are simply not impermissible. They argue, (1) if a transaction is Pareto improving and consensual, then it is not impermissible.

534    Benjamin Ferguson and Hillel Steiner After all, they reason, if everyone involved is better off transacting and everyone freely agrees to transact, what grounds can there be for condemning the transaction? If a Pareto-improving transaction does not occur, then both parties will be worse off than they otherwise would be. While (1) has some prima facie plausibility, it is difficult to endorse in the face of some particular transactions that many people feel are impermissible. Sweatshop labour, for example, is often condemned for being exploitative. Although some sweatshop labour may be harmful to the labourer (and so violate the Pareto condition) or forced (and so be nonconsensual), many sweatshop contracts are both Pareto improving and consensual. Yet it still seems there is something morally amiss with these transactions. Sweatshop labour appears to provide an example of a transaction that is both Pareto improving and consensual, but nevertheless impermissible. That is, (2) There exist some transactions that are Pareto improving, consensual, and impermissible. Clearly (2) conflicts with (1). One reason for defending (1) is welfarist. After all, as we have noted, if the transaction does not occur, both parties will be worse off. But what reasons might we have for accepting (2)? What else might be required in order for a transaction to be permissible? A common response from those who offer accounts of exploitation is that a transaction may also be impermissible if it is unfair.1 Pareto-​superiority and consent are not sufficient for the permissibility of a transaction: transactions must also be fair in order to be permissible. While this response sounds reasonable, in order to justify an acceptance of (2) and a rejection of (1) we require an account of what it means for a transaction to be unfair. This is precisely what accounts of exploitation attempt to provide.

25.2  Two Kinds of Fairness If exploitation is a matter of taking unfair advantage, then Alice exploits Bob only if the benefit she receives from transacting is greater than it ought to be and the benefit Bob receives is less than it ought to be. Though the transaction is mutually beneficial,

1 

Different accounts of exploitation offer different accounts of why the transaction is unfair. According to the classical Marxist account of exploitation, the transaction is unfair if the socially necessary labour time embodied in the goods being transferred is unequal. According to Steiner’s account, an unfair distribution of advantage leads to an unfair distribution of gains from transaction. According to Alan Wertheimer’s account (1996), transactions are unfair when their terms differ from those that would be offered in hypothetical competitive markets. We discuss these and other accounts of fairness in the sections that follow.

Exploitation   535 the distribution of these benefits is unfair. What each transactor’s gain ‘ought to be’ is determined by a particular criterion of fairness. Different accounts of exploitation utilize different fairness criteria to ground exploitation claims. We can call the feature of the transaction that allows Alice to extract greater benefit from Bob than she would otherwise have extracted ‘Bob’s disadvantage’. So, suppose Bob has forgotten to eat breakfast and he finds himself in front of Alice’s fruit stand, desperate for an apple. Seeing that Bob’s demand for an apple is extraordinary, Alice doubles her price. In this case, Bob’s hunger is his disadvantage and this disadvantage allows Alice to gain more than she would have from Bob had he not been so hungry. And, of course, in this scenario Bob gains less than he would have because his hunger means he must pay more for the apple. Whether this transaction is an exploitation will depend on whether the difference in the terms of the transaction caused by Bob’s hunger is unfair, according to some criterion of fairness. Existing accounts of the unfairness involved in exploitative transactions take two general forms: they may be historical or ahistorical. According to historical accounts of unfairness, in order to determine whether the difference in the distribution of the benefits is unfair we need to know what causes the disadvantage. We can partition the historical causes of Bob’s disadvantage into three categories. Either Bob was disadvantaged by a person, or he was not. If he was not, then we may say his disadvantage is ‘natural disadvantage’. For example, Bob may be naturally disadvantaged because, unforeseeably, a tree fell on his home during a storm. On the other hand, if Bob was disadvantaged by a person, then either it was another person, or it was himself. We may call these forms of disadvantage ‘self-​caused’ and ‘other-​caused’, respectively. So, Bob’s disadvantage is other-​caused if Alice sets his home on fire and it is self-​caused if Bob burns his own home. These forms of disadvantage are shown in Fig. 25.1.

Person

Self

Natural

Other

fig. 25.1  Forms of disadvantage

Most—​though not all—​historical accounts draw on an approach to distributive justice that has become known as ‘luck egalitarianism’. For present purposes, we may consider luck egalitarianism in its sloganistic form, according to which ‘it is bad when persons are worse off than others through no fault of their own’. Luck egalitarianism implies that both other-​caused disadvantage and natural disadvantage are bad from the point of view of justice. Thus, historical accounts of fairness based on luck egalitarian approaches to distributive justice imply that if Bob’s disadvantage is either other-​caused disadvantage or natural disadvantage then it is, broadly speaking, unfair when this disadvantage allows Alice to gain more—​and Bob less—​than either would have gained

536    Benjamin Ferguson and Hillel Steiner absent the disadvantage. So, if Alice gains more from Bob because Carol has stolen Bob’s breakfast, then Alice’s higher price is unfair because Bob’s disadvantage was other-​ caused. The same holds if Alice’s gain is brought about by Bob’s natural disadvantage; perhaps his breakfast blew away in the wind. However, luck egalitarian approaches to distributive justice also imply that if Alice is able to charge Bob more for the apple because he forgot to eat breakfast, then her higher price is not unfair. This is because Bob’s disadvantage was self-​caused. To reiterate, for historical accounts of fairness, the source of disadvantage is relevant for determining the fairness of the transaction. Often, but not always, historical accounts rely on broadly luck egalitarian theories of distributive justice. According to ahistorical accounts of unfairness, the cause of the disadvantage is either always irrelevant (which is a strong form of ahistoricism), or irrelevant in certain contexts (which is a weaker form) for determining whether the thereby modified terms of a transaction are unfair. One defender of the strong form of ahistoricism is Robert Goodin, who claims that exploitation consists in a failure to protect the vulnerable (i.e. disadvantaged) in our interactions with them ‘regardless of the particular source of their vulnerability’ (Goodin 1987). Alternatively, Ruth Sample (2003)  defends a weaker version of ahistoricism, claiming that we exploit others when we fail to consider whether their basic needs are met when we transact with them. For Sample, this obligation is not predicated on the reasons these persons’ basic needs are unmet and therefore, is ahistorical. However, unlike Goodin, her focus is limited to disadvantage that falls below a certain threshold. When persons’ basic needs are unmet, regardless of the reason they are unmet, we ought to constrain our advantage over them according to Sample. Thus, her criterion of fairness implies a commitment to a weak form of ahistoricism. According to ahistorical accounts of fairness, the unfairness of exploitation is not—​or not only—​located in prior distributive injustice. Rather, it is grounded in the (non-​historical) properties of the present transaction.

25.2.1 Two Kinds of Advantage Taking Because their focus is on the features of the present transaction, ahistorical accounts often place an emphasis on the attitude, intentions, or motive of the exploiter. Transactions ordinarily described as exploitations often seem to involve a blatant disregard for the exploited’s situation or a disrespect for their person. This advantage-​taking feature should be distinguished from the maldistribution—​or unfairness—​involved in exploitation. Approaches to exploitation differ not only according to whether they employ historical or ahistorical criteria of fairness, but also according to their interpretations of ‘advantage taking’. One interpretation of what it means to take advantage, endorsed by Jon Elster, is rather minimalist. Elster claims we take advantage of another whenever we ‘[add] an option to his opportunity set that is better both for him and for [us] than the option preferred by him prior to the expansion’ (Elster 1982). On this account, Alice can be said to ‘take advantage’

Exploitation   537 of Bob (and vice versa) simply by engaging in a mutually beneficial transaction with him.2 Accounts that employ Elster’s minimal sense of advantage taking will characterize exploitation simply as an unfair, but Pareto-improving transaction. What makes the transaction unfair will, of course, depend upon the notion of fairness the account endorses. On the other end of the spectrum lie accounts that claim advantage taking involves engaging in such transactions with attitudes, intentions, or motivations that are morally amiss. For example, it may be claimed that in order for Alice to exploit Bob the transaction in which they engage must be not only unfair, but she must be aware that it is so and yet fail to constrain her advantage over Bob.3 These two distinctions—​whether exploitation’s unfairness is to be understood ahistorically or historically (Table 25. 1) and whether the presence of inappropriate attitudes is necessary for exploitation—​can, in theory, be variously combined. However, in practice, accounts of exploitation that employ ahistorical criteria of fairness generally also place a strong emphasis on the attitudes of the exploiter towards the exploited, and accounts that employ historical criteria of fairness tend to favour Elster’s account of advantage taking. Table 25.1 Four approaches to exploitation   Attitudes unnecessary

Attitudes necessary

Historical fairness

Steiner, Roemer

Ferguson

Ahistorical fairness

Marx

Goodin, Sample

In the sections that follow, the distinction between historical and ahistorical accounts of fairness takes centre stage. Section 25.2 focuses primarily on Robert Goodin’s (1987) and Ruth Sample’s (2003) ahistorical approaches to exploitation. Both accounts also place a strong emphasis on the attitudes involved in advantage taking. The section will also briefly consider the classical Marxist account of exploitation. Though Marx is not concerned with the attitudes of exploiters, his account of exploitation is based upon an account of fair transaction that can be construed as ahistorical. Section 25.3 addresses two historical approaches proposed by John Roemer (1982a, 1982b) and Hillel Steiner (1984, 1987, 2010, 2013). Each of these approaches implicitly endorses Elster’s minimalist account of advantage taking. Section 25.3 concludes with a brief outline of a historical approach proposed by Benjamin Ferguson (2013) that also includes an attitudinal approach to advantage taking. Section 25.6 revisits the tension between the claim in (1) that Pareto improvement and consent are sufficient for the permissibility of a transaction, and the claim in (2) that some transactions that are both Pareto improving and consensual are, nevertheless, impermissible. 2  Note that advantage taking in Elster’s sense is not morally amiss. Rather, advantage taking in this sense becomes exploitative when the Pareto improving transaction is also unfair. 3  See Ferguson (2013) for an outline and defence of an account that takes this approach.

538    Benjamin Ferguson and Hillel Steiner

25.3  Ahistorical Accounts of Fairness As we noted in the introduction, exploitative transactions are unfair transactions, and this unfairness may, broadly speaking, be understood in two ways. Transactions can be unfair in a historical sense when Alice does better than she would have—​and Bob worse than he would have—​because of a prior unjust distribution of advantage. Alternatively, transactions can be unfair in an ahistorical sense when they diverge from a criterion of fairness that does not make reference to the justice of Alice and Bob’s distributive shares prior to entering the transaction. Rather, the transaction is judged fair or unfair according to its own (non-​historical) properties or to the non-​ historical properties of the transacting parties situations. Let’s revisit Alice and Bob’s apple transaction. Suppose that the transaction is unfair. This means a fair transaction is one in which Bob pays less than he does in the actual transaction. There are many ahistorical criteria of fairness that we might use to justify the claim that the terms of the actual transaction are unfair to Bob. For example, we might believe that a fair transaction is one in which both transactors receive equal utility gain from transacting. Alternatively, Marx argued that Alice exploits Bob when the amount of labour that went into making the apple is less than the amount of labour it took Bob to earn the wages he used to purchase it. Equal utility gain and equality of embodied labour time are both ahistorical fairness criteria because each refers to the value of the goods exchanged in the present transaction. In fact, they are very similar ahistorical criteria since each argues that the exchange is fair only when it is equal. They differ according to the currency they use as the equalisandum. Alternative ahistorical criteria of fairness do not require equality in the exchange. Rather, they imply that a transaction is fair when it surpasses a particular threshold. According to these criteria, Alice’s transaction with Bob is unfair when—​on Sample’s account—​she fails to take Bob’s basic needs into account, or—​as Goodin argues—​when she fails to act in a way that protects Bob from his vulnerability to her in the transaction. In what follows we first consider Goodin’s strong form of ahistoricism, followed by Sample’s more constrained approach.

25.3.1 Goodin’s Account According to Goodin’s account, exploitation in general involves ‘playing for advantage in situations where it is inappropriate to do so . . . [it] consists in playing games of strategy in circumstances which render them somehow inappropriate’ (Goodin 1987). He identifies four kinds of situation where playing for advantage is inappropriate: when playing against those who have ‘renounced playing for advantage themselves’; or ‘who are no match for you in games of advantage’; or ‘who are unfit or otherwise unable to play in games of advantage at all’; or ‘when your relative advantage derives from others’ grave

Exploitation   539 misfortunes’ (Goodin 1987). All four situations are, he claims, ‘manifestations of one particular kind of wrong’, namely, a failure to honour the moral norm of ‘protecting the vulnerable . . . regardless of the particular source of their vulnerability’ (Goodin 1987, emphasis added). Thus, for Goodin exploitation occurs when we use others’ vulnerabilities to inappropriately secure some form of benefit in games of strategy, irrespective of the vulnerabilities’ source. Goodin’s claim that any form of vulnerability can generate exploitation places his account clearly in the ahistorical camp. Goodin’s understanding of vulnerability is central to his account of exploitation. He claims vulnerability is both relational—​‘a full specification will tell us who is vulnerable to whom with respect to what’—​and relative—‘A is more vulnerable to B (1) the more control B has over outcomes that affect A’s interests and (2) the more heavily A’s interests are at stake in the outcomes that B controls’ (Goodin 1986). That is, A is vulnerable to B with respect to x when B has the ability to control x and x has some impact on A. While this analysis is promising as an account of vulnerability, it threatens to make the conception of exploitation that is based upon it too wide in scope. The claim that we ought to constrain our advantage when we interact with vulnerable others is convincing in many cases. Surely Alice is wrong to use her advantage to extract labour from Bob for very low wages in a sweatshop. But is she also obliged to constrain her advantage over Bob if he is a robber baron who, after making his millions, has ‘renounced playing for advantage’?4 What if her advantage over Bob derives from a grave misfortune that Bob himself is responsible for bringing about? If Bob is vulnerable to Alice just in case Alice can control some outcome or action that impacts Bob, it is unconvincing to claim that all cases in which Alice fails to make allowance for this vulnerability amount to exploitation. After all, we are willing to spend more money on food when we are hungry, more willing to purchase products if we have seen them advertised, and more desperate for a new job if we are currently unemployed. In the broadest sense, each of these situations creates vulnerability, yet it is not clear that it is inappropriate to take advantage of all of these vulnerabilities. Indeed, all Pareto-improving transactions involve vulnerability in Goodin’s sense. If Alice owns an apple that Bob wants, then regardless of the reason Bob wants the apple, it is true that he is vulnerable to her with respect to the apple since she has the ability to control the apple, and whether or not Bob’s obtaining the apple has some impact on him. Yet, far from being impermissible, in most cases, taking advantage of Pareto-improving transactions should be encouraged. Of course, some of these transactions may be unfair, but this unfairness must consist in something other than taking advantage of vulnerability in Goodin’s sense. It seems that Goodin fails to explain what is special about those transactions we ordinarily call exploitative. Nevertheless, Goodin is correct that vulnerability or disadvantage is necessary for exploitation. If Bob were not in some way vulnerable or disadvantaged vis-​à-​vis Alice, then Alice could not gain more and he less than either ought to gain from the transaction. However, as the previous paragraph makes clear, some form of vulnerability is present 4 

Thanks to Mike Otsuka for drawing our attention to this point.

540    Benjamin Ferguson and Hillel Steiner in nearly every transaction, yet not every transaction is an exploitation. Thus, though Goodin identifies a necessary condition for exploitation, it is not sufficient. Perhaps, in light of this potential overextension, we should search for a notion of vulnerability that is more constrained in scope. What we are after is a restricted sense of vulnerability of which it is more intuitively ‘unfair’ to take advantage. For such an approach, we turn to Ruth Sample’s account of exploitation.

25.3.2 Sample’s Account Sample argues that exploitation in general involves ‘interacting with another being for the sake of advantage in a way that degrades or fails to respect the inherent value in that being’ (Sample 2003). She claims these failures of respect fall into three domains: we can fail to show proper respect for other persons by neglecting what is necessary for the person’s basic needs, by taking advantage of past injustices, or by engaging in a transaction involving ‘an aspect of that person’s being that ought not to be commodified’ (Sample 2003). According to Sample’s first domain of disrespect, we ought to constrain our advantage over others when those with whom we transact are so impoverished that their basic needs are unmet. We treat others unfairly when we fail to respond appropriately to their impoverishment, and this is so regardless of the reason these needs are unmet. Since the historical reasons for others’ impoverishment play no role in generating this obligation, this domain employs an ahistorical notion of fairness. However, unlike Goodin’s unconstrained concept of vulnerability, Sample’s appeal to a threshold level of well-​being is more limited in scope. That we ought to constrain our advantage in the face of severe poverty does not necessarily imply that we ought to constrain our advantage when we transact with those who are less poor, but nevertheless vulnerable in Goodin’s wider sense. Sample also argues that we ought to constrain our advantage even when interacting with the vulnerable who are not impoverished. However, this obligation is limited to situations in which their vulnerability is the result of past injustice. Sample’s second domain of disrespect employs a historical notion of fairness. Whether we ought to constrain our advantage when interacting with those whose basic needs are met, but are nevertheless vulnerable, does depend upon the reason for their vulnerability. These domains provide two distinct reasons for Alice to constrain her advantage. For example, suppose a very hungry Bob approaches Alice’s stand to purchase an apple. If Bob is starving and his basic needs are unmet, then according to Sample’s first domain, Alice ought to constrain her advantage over Bob—​perhaps by offering him the price she would extend to a non-​hungry customer.5 On the other hand, if Bob’s basic needs are met (but he is still quite hungry), then the reason he is hungry becomes important.

5 

What exactly showing proper respect for Bob requires, in terms of transaction prices, is left unspecified in Sample’s account.

Exploitation   541 Though Sample does not endorse any particular theory of distributive injustice, we can consider the example in light of a luck egalitarian account. In this case, if Bob is hungry because Carol stole his lunch, then Alice ought to constrain her advantage, but if Bob is hungry because he forgot to eat breakfast, then Alice is not obliged to constrain her advantage. Of course, these two domains may also be combined: Bob’s basic needs may be unmet because of a past injustice. Taking advantage of persons in these situations seems, on Sample’s account, to be particularly unjust. The final domain of disrespect in Sample’s account is disrespect that issues from trading in non-​commodifiable goods, such as human organs or sexual intercourse. While it is true that such transactions are occasionally referred to as exploitative, it is less clear that these kinds of transaction are exploitations in the sense we have been considering. Without denying that they may be degrading or disrespectful, it does not seem that transactions falling only within this domain are cases of unfair advantage taking. Indeed, Sample acknowledges that when the other forms of disrespect are not present, but a person violates the non-​commodification condition they are ‘not exploited . . . however, [they] may be degraded if the appropriate case against commodification can be made’ (Sample 2003). For this reason, in the remaining discussion we will restrict our attention to Sample’s first two forms of disrespect. There are many points in favour of Sample’s account. It captures the intuition that we ought to constrain our advantage in light of others’ vulnerabilities, but it avoids the overextension of Goodin’s account by placing a restriction on this obligation when the vulnerabilities are less severe, that is, in cases where our interactors’ basic needs are met. Nevertheless, there is one concern that arises not only in the case of Sample’s account, but also in Goodin’s account. The concern is that inasmuch as these ahistorical (or partly ahistorical) accounts imply unconditional duties on the part of the would-​be exploiter not to take advantage of his or her superior position in certain situations, these duties are themselves exploitable.

25.3.3 The Moral Hazard Objection Suppose Bob is a wealthy banker who takes a calculated gamble. If the gamble goes well, Bob will gain a great deal, but if it does not, his basic needs will go unmet. According to Sample’s first form of disrespect, if Bob loses the gamble and his basic needs are unmet, then Alice is obliged to constrain her advantage when transacting with him. Failure to do so amounts to exploitation according to Sample’s conception. But suppose Alice’s obligation is known to Bob prior to his decision to take the gamble. In particular, he knows that if the gamble goes well, the large gain is his to keep, but if it goes poorly, his loss will be subsidized by Alice, since she is obliged to constrain her advantage over him.6 The

6 

Alice’s constraint of advantage can be characterized as a subsidy since that proportion of the gains from transacting she refrains from claiming, accrues to Bob.

542    Benjamin Ferguson and Hillel Steiner obligation for Alice to constrain her advantage in the event of Bob’s loss creates a moral hazard that Bob can use to reduce the cost of losing the gamble. In effect, Alice’s obligatory subsidy is a form of insurance for Bob’s gamble that he does not pay for. Alice’s obligation is tantamount to a moral vulnerability that, if enforced, would also be an empirical vulnerability. Taking advantage of the vulnerability created by the moral hazard is unfair and—​since exploitation involves taking unfair advantage—​it seems appropriate to say that Bob exploits Alice when he does so.7 According to the basic needs domain, Alice exploits Bob if she fails to constrain her advantage while transacting when his basic needs are unmet. Yet, according to the above argument, when Bob takes advantage of the moral hazard by claiming this advantage from Alice, he exploits her. This situation creates a conflict between two mutually exclusive and exhaustive acts.8 A desideratum of any normative theory is that it provides noncontradictory guides to action. Yet here the theory implies that the subsidy is both obligatory and impermissible. It cannot be the case that both conflicting obligations are valid: either Alice is not obliged to constrain her advantage or Bob is not obliged to constrain his advantage (which is bestowed upon him by her having such an obligation).9 The problem with Sample’s claim that Alice must constrain her advantage over Bob—​ even when Bob is responsible for his own vulnerability—​is that it is insensitive to personal responsibility. The reason Bob is vulnerable is relevant to determining whether Alice is obliged to constrain her advantage. Requiring Alice to constrain herself when Bob is responsible for his own vulnerability makes Alice vulnerable to being exploited by Bob. A solution to this tension is to deny that Alice exploits Bob when she fails to constrain the advantage she gains from Bob’s self-​caused vulnerability. Indeed, concerns about this form of exploitation provide part of the motivation for luck egalitarians’ omission of self-​caused disadvantage from the domain of injustice. Of course, this luck egalitarian claim is controversial. Critics of the approach have advanced the ‘unacceptable outcomes objection’, which claims ‘the luck egalitarian embrace of responsibility implausibly justifies leaving people without assistance in thoroughly bad situations if they are responsible for bringing them about’ (Stemplowska 2009).10 Though we believe the objection is misplaced, Goodin disagrees, arguing:

7 

Note that Alice’s obligation to Bob is a conditional obligation. Alice is not obliged to transact with Bob, but if she does, then she is obliged to constrain her advantage over him. However, provided the transaction is (strictly) Pareto improving, Alice will prefer to transact with Bob. So, though Alice is not morally obliged to transact, she is prudentially advised to do so, and if she does, then she must constrain her advantage over Bob. It is this conditional obligation that suffers from the moral hazard objection. 8  The two possible acts are Alice’s provision of the subsidy and Bob’s refusal of the subsidy when the two transact. If Alice provides the subsidy, then Bob has not refrained from accepting it; if Bob refrains from accepting the subsidy, then Alice has not provided it. The two acts are mutually exclusive. Since Alice’s failure to provide the subsidy (¬φ) is equivalent to Bob’s refusal to accept the subsidy (¬¬φ, i.e., φ) the two acts are exhaustive. 9  The moral hazard problem is explored in more detail in Ferguson (2016). 10  It should be noted that Stemplowska favours the luck egalitarian account and argues against the objection.

Exploitation   543 . . . [o]‌nce all their opportunities for self-​help have passed . . . the situation is beyond their control. Others, however, may still be able to act so as to avert harm to them. To suggest that those others should (or even that they may) stand idly by and watch people reap the bitter fruits of their own improvidence is surely absurd. (Goodin 1986)

It is important to note though that we can both accept the unacceptable outcomes objection—​agreeing with Goodin that we are obliged to aid those whose disadvantage is self-​caused—​and reject the claim that Alice exploits Bob if she fails to constrain her advantage when his vulnerability is self-​caused. We may claim that Alice is obliged to constrain her advantage despite the fact that when she subsidizes Bob she allows him to treat her unfairly. We do not need to abandon the intuition that what Bob does is unfair in order to claim that Alice is nevertheless obliged to provide aid. Rather, it is possible to claim that a concern for Bob’s welfare, now that his basic needs are unmet, overrides a concern for fairness. What we claim here is that it is incorrect to use the concept of exploitation to explain the obligation Alice has towards Bob because, as the above example shows, exploitation is a responsibility-​sensitive concept. If these arguments are correct, then Goodin’s and Sample’s omission of personal responsibility in their accounts is misplaced. Whether Alice exploits Bob will depend, importantly, on the historical reasons Bob is disadvantaged. In addition to the moral hazard objection a second, practical, concern arises for both Goodin’s and Sample’s accounts. Though Goodin tells us we ought to protect all those who are vulnerable and Sample claims that exploitation involves a failure to respect those with whom we interact, neither provides a concrete account of how the obligations to protect or respect others should be operationalized. Recall Alice’s sale of an apple to Bob. Let us suppose that the sale price of $2 for the apple is unfair because it violates Sample’s second domain of disrespect. By selling Bob the apple for $2, Alice takes advantage of Bob’s vulnerability that was caused by past injustice. Though Sample’s domain explains the wrongfulness of Alice’s action, it does not specify what price would ensure Alice’s interaction with Bob was respectful. Should she sell the apple for $1? For $1.50? Of course, the ‘fair’ price will likely depend upon some contextual details that have not been specified in this case, but a complete account of exploitation should at least provide some general principles that can be applied to specific contexts. This concern is not an objection to either of the two accounts—​rather, it represents an important detail that must be specified if the accounts are to properly guide our decisions in transactions.

25.3.4 Marx’s Account One account that does provide a concrete guide to when a transaction is fair is the classical Marxist account of exploitation. According to Marx, workers are exploited when they work more hours than are needed to produce the goods they consume. So, if Bob works for Alice for ten hours, but the wages he is paid by Alice only allow him to

544    Benjamin Ferguson and Hillel Steiner purchase a bundle of goods that comprise eight hours of work, Bob is exploited by Alice. Thus, according to Marx a transaction is non-​exploitative just in case the labour time embodied in the goods exchanged is equal. Since Marx’s concern was primarily with employment contracts, the scope of his account is somewhat narrower than the accounts we have considered so far. Nevertheless, the idea that people should receive remuneration that is equivalent to the amount of work they do is normatively compelling. Marx’s account of fairness, like those of Goodin and Sample, can be seen as an ahistorical account. The transaction is unfair when the labour time embodied in the exchanged items is unequal regardless of the historical reasons for this inequality. However, Marx’s account does not suffer from the same moral hazard problem that plagues Goodin’s and Sample’s accounts because the amount of embodied labour Alice is obliged to transfer to Bob depends on the amount of labour Bob provides for Alice. The reliance on equality of labour exchanged ensures that there is no distributive unfairness in the transfer between Alice and Bob. Unfortunately, the account suffers from other problems. In particular, the labour theory of value that underpins the equal exchange notion of fairness in Marx’s account has long been criticized for being difficult to operationalize. Determining whether the amount of embodied labour transferred between Bob and Alice is equal requires a measure of labour. However, a convincing measure of embodied labour has been notoriously difficult to come by. Clearly a fair exchange is not one in which the ‘raw’ amount of time laboured is equal, for some workers may be more talented than others. In order to incorporate differences in labour, Marx argues that the equalisandum is not mere ‘raw’ labour, but ‘socially necessary labour’. Yet this move simply passes the definitional buck to ‘socially necessary labour’. We now need a criterion for distinguishing socially necessary labour from unnecessary labour. This lack of a convincing definition of socially necessary labour, along with some other well-​ known problems for Marx’s labour theory of value—​its inability to reliably explain the relationship between labour value and prices and the arbitrariness of labour as the value numéraire—​have led most economists to reject Marx’s labour theory of value.11 Two critical responses to the classical Marxist account of exploitation have emerged in light of these difficulties with the labour theory of value. The first is espoused by Robert Nozick, who concluded that ‘with the crumbling of the labor theory of value, the underpinning of [the Marxist] theory of exploitation dissolves’ (Nozick 1974). For Nozick, the labour theory of value was essential to the Marxist conception of exploitation. The second critical response originates with G. A. Cohen. Cohen argues that though the labour theory of value is untenable, . . . what really animates Marxists, whatever they may claim . . . is not that the capitalist gets some of the value the worker produces [as the labour-​theory implies], but that 11  For a discussion of the arbitrariness of labour as the value numéraire see Roemer (1982a). The transformation problem is discussed in Samuelson (1971) and an overview of the debate between Samuelson and Marxist economists is provided in Bronfenbrenner (1973).

Exploitation   545 he gets some of the value of what the worker produces . . . [workers] do not create value, but they create what has value. (Cohen 1979)

Nozick noted that the particular form of exploitation found in the classical account is formulated in terms of the labour theory of value, and that this theory of value is false. But Cohen points out that amended versions may succeed where Marx’s original account fails. Cohen’s observation heralded a research programme that sought to provide alternative accounts of equal exchange based on equalisanda other than equality of embodied labour time, many of which also employed historical accounts of fairness. In section 25.4 we consider two such accounts, John Roemer’s equality-​based account of exploitation and Hillel Steiner’s rights-​based account.

25.4  Historical Accounts of Fairness The two accounts of exploitation that we focus on in this section—​John Roemer’s property relations account and Hillel Steiner’s rights-​based account—​draw upon Marx’s idea that exploitation involves unequal exchange, but they diverge from the classical Marxist account by offering different accounts of what must be equalized in order for a transaction to be fair. Unlike Sample and Goodin, these two accounts employ historical approaches to fairness. Both accounts also adopt Elster’s minimal notion of advantage taking, though, as we will see at the end of the section, this adoption poses problems for both accounts.

25.4.1 Roemer’s Account Roemer begins his property relations account by noting that we can think of a group as being exploited if it has some hypothetical ‘conditionally feasible alternative under which its members would be better off ’, which he captures in the following two conditions (Roemer 1982b). A coalition S in a larger society N is exploited if and only if: (R1)  Th  ere is an alternative, which we may conceive of as hypothetically feasible, in which S would be better off than in its present situation. (R2)  Under this alternative, the complement to S, the coalition N − S = St, would be worse off than at present. (Roemer 1982a) Various forms of exploitation can be specified in Roemer’s model by altering the withdrawal conditions that specify the hypothetical situations. Roemer focuses on three withdrawal conditions that, he claims, capture feudal, capitalist, and socialist

546    Benjamin Ferguson and Hillel Steiner exploitation. Under specific withdrawal conditions coalitions are exploited if (R1) by withdrawing from society and creating their own economy they can achieve utility levels that dominate the utility frontier available to the coalition when it remains in the original economy, and (R2) the coalition’s complement fares worse in this respect if the coalition withdraws (Roemer 1982a). Roemer outlines the following three withdrawal conditions: Feudal exploitation.The coalition can take with it its own endowments of both alienable (transferable, nonhuman property) and inalienable (i.e. skills and talents) resources. Capitalist exploitation. The coalition can take with it its per capita share of society’s alienable property. Socialist exploitation. The coalition can take with it its per capita share of all endowments, alienable and inalienable. (Roemer 1982a) For example, taking sweatshop labourers as coalition S, we can say that they are capitalistically exploited according to Roemer’s model if and only if the utility they could achieve by withdrawing from society and creating their own economy with their per capita share of assets, is greater than the utility they gain by working in sweatshops and the utility of others in society (St)—​including the sweatshop owners—​is less after the labourers withdraw. An attractive feature of Roemer’s account is that it separates the structure of an exploitative transaction (conditions R1 and R2) from the ethical content that allows us to condemn the transaction (the withdrawal conditions). The normative withdrawal conditions of exploitation are specified by various approaches to distributive justice, while the structural conditions specify the way in which these distributive elements are manifest in an exchange. Roemer’s framework allows us to agree about whether coalitions are capitalistically or feudally exploited even if we disagree about whether the withdrawal conditions that characterize these forms of exploitation are normatively compelling. However, this does not mean Roemer’s account entirely escapes criticism. Though few philosophers have expressed doubts about the necessity of Roemer’s conditions, many argue that R1 and R2 are not sufficient to establish the existence of an exploitative transaction. While exploitation does require Alice to gain and Bob to lose, this gain and loss must come about in the right way. One concern is that R1 and R2 do not ensure that an exploiter’s gain comes at the expense of the exploited. Imagine two islands societies, S and St, each unknown to the other. Suppose S has $5 of assets and St has $15. If the assets of the islands were combined ($5 + $15 = $20) and the inhabitants of each were allowed to withdraw with their per capital share of assets ($20.2 = $10), (R1) and (R2) would imply that St exploits S, since S would be better off with $10 and St would be worse off with $10. Yet, it does not seem appropriate to say that S exploits St. This is because the wealth of St does not depend on—​come at the expense of—​S action. Roemer recognizes this problem and

Exploitation   547 adds a further condition: ‘capitalist exploitation can be said to exist only if one coalition is gaining at the expense of another’ (Roemer 1982c), that is, ‘if the would-​be exploited withdrew with their own endowments, the complementary coalition would be worse off ’ (Sensat 1984). Though this condition brings Roemer’s account closer to capturing the relation of exploitation between the two groups, Jon Elster argues that the account is still insufficient. Elster asks us to consider two groups that have ideological interactions, but no economic interaction. One group, S, is hedonistic and prefers leisure to work. The other, St, has a puritanical work ethic and disdains the hedonism of S. This disdain causes St to work harder than they otherwise would, leading to a greater productivity for St. The hedonists in S, on the other hand, find the puritanical work ethic contemptible and this attitude encourages their hedonism. Consequently, the presence of St negatively affects the productivity of S. If St did not exist, then S would be better off. If S did not exist, then St would be worse off. Conditions (R1) and (R2) are satisfied. But it seems absurd to claim that St exploits S precisely because St does not appropriately cause S to be worse off (Elster 1982). Roemer acknowledges that certain ‘bizarre’ counterexamples may be produced unless the following ‘dominance’ condition is added: (R3)  St is in a relationship of dominance to S. However, as Kymlicka, Bertram, and even Roemer himself point out, the addition of R3 is ad hoc. Roemer acknowledges that ‘since dominance is undefined, and is as elusive a concept as exploitation, the addition of (R3) is ad hoc’ (Roemer 1982b). Even if an analysis of dominance were provided, the condition would remain ad hoc—​at least for Roemer’s purposes—​for two reasons. First, as Kymlicka points out, appeals to dominance or power are ‘disconnected from the “ethical imperative” [Roemer] identifies as the basis of exploitation theory’, namely, distributive injustice (Kymlicka 1990). Second, as Bertram notes, . . . the different withdrawal conditions were supposed to bring to light the normative differences which exist between Marxists and neo-​classicals. The invocation of a further condition is not going to satisfy someone who considers that the forced transfer of income through taxation to finance welfare programmes is exploitative. (Bertram 1988)

So, although the first problem of dependence can be handled by requiring that one coalition gain at the expense of the other, the problem that a lack of dominance presents for Roemer’s account is more substantial. But, we hasten to point out that it is not fatal. The presence of R3 is indeed ad hoc if our goal is, like Roemer’s, to provide a purely distributive account of exploitation. However, it may turn out that exploitation is not, in fact, a purely distributive concept. An adequate conception of exploitation may need to move beyond Elster’s minimalist account of advantage taking to include a richer analysis that makes room for concepts like power, domination, and intention. A modified

548    Benjamin Ferguson and Hillel Steiner version of Roemer’s account that offers an analysis of what it means for St to be in a relationship of dominance to S may successfully narrow the scope of Roemer’s initial two conditions. Such an account would move from the top-​left quadrant of the grid we presented in section 25.2.1 to the top-​right quadrant. Finally, we would like to mention a pragmatic reason for moving away from Roemer’s account. Roemer’s work on exploitation includes not only the property relations account we have outlined here, but also an ‘equal exchange’ account, which, we have suppressed for reasons of space.12 Together both of Roemer’s accounts have generated a large number of interesting results for Marxist theory. However, Roemer’s focus on hypothetical economies and the exploitation of coalitions S and St makes it difficult to determine whether a single transaction between Alice and Bob is exploitative. Roemerian exploitation is an aggregate phenomenon that is not easily applied to individual transactions. If we wish to explain how Marx’s concern with class struggle can be analysed in terms of exploitation and inequality without relying upon the labour theory of value, then Roemer’s account is extremely useful. However, if our aim is to explain how individual transactions can be Pareto improving, consensual, but unfair, another account may be more useful. With this in mind we turn now to Hillel Steiner’s rights-​based account of exploitation.

25.4.2 Steiner’s Account Steiner’s guiding aim is to show how exploitation can be understood as a form, specifically, of injustice (Steiner 1984, 1987, 1994, 2010, 2013). Despite its longstanding association with the idea of injustice, he observes, the currently numerous theories of distributive justice are all devoid of any account of exploitation. To remedy that lacuna and exhibit that connection, Steiner begins by focusing on the paradigmatic form of injustice—​violations of moral rights. He asks us to imagine an auction where Bob is selling some objects or services, Xs, and Alice enters the winning bid of $1.50.13 Under what conditions is it plausible to regard those Xs as having been undersold? That is, under what conditions can we say that an X’s $1.50 price is unjustly low and should have been higher? A common response, especially prior to the advent of neoclassical economics in the 1870s, has been to claim that an X is worth more than $1.50 because the cost of producing it was greater than $1.50. Although this objective conception of value is endorsed by both classical Marxist and recent cost-​of-​production (Reiff 2013) accounts, Steiner shares modern mainstream economics’ rejection of this response, inasmuch as it relies upon the pre-​neoclassical 12 

A detailed treatment of Roemer’s unequal exchange account can be found in Roemer (1982a). Roemer also provides a very useful (and less formal) summary of this work in Roemer (1982b). Elster offers a critique of Roemer’s property relations approach and endorses the unequal exchange account in Elster (1982). 13  Auctions are markets writ small.

Exploitation   549 view that things exchanged have a value which is not determined by their sellers’ and buyers’ preferences. That is, he accepts persons’ actual preference orderings as the determinants of prices. Those preferences are taken as given: they are not to be gainsaid by reliance upon some notion of value that is independent of them. And, moreover, they cannot be gainsaid by any liberal theory of justice that aspires to be neutral as between different conceptions of ‘the good life’. This does not, however, oblige us to accept an X’s $1.50 price as just. It does not do so because, under different circumstances, it might have sold for a higher price, say, $2.00. So the question is: What are those counterfactual circumstances, and how does their absence make the $1.50 price unjust? One such circumstance, we might be tempted to think, is that others—​someone else or Alice herself—​have preferences that place a higher value than $1.50 on an X. But this counterfactualized circumstance is insufficient to prove that $1.50 is an unjust price. Justice does not require anyone to value an X at all, let alone at more than $1.50. As previously indicated, we are to take persons’ actual preferences as given. Suppose there is someone, Carol, who does value an X at more than $1.50. Why, then, did she not outbid Alice? One reason might be that she couldn’t afford to do so. But, again, although this explains why Alice won, it doesn’t suffice to render her $1.50 price unjust. What we need to know is why Carol couldn’t afford a higher bid. Why was her wealth endowment insufficient for that? Suppose her endowment was insufficient because she had previously lost some of it in a risky investment venture. That, however, would not sustain a claim that the $1.50 price was unjust. It was Carol’s actual preference to undertake that risk and, hence, the potential loss it entailed. She committed no injustice in doing so and her incurring that loss was not unjust. But suppose, instead, that the reason why Carol’s endowment was insufficient is that, on her way to the auction, she was robbed. That is, had she not been robbed—​had she not been unjustly deprived of some of what she had a moral right to possess—​her outbidding Alice would have occurred and Bob’s Xs would have sold for, say, $2.00 each. It is by virtue of that prior injustice—​that unjust reduction of Carol’s endowment—​that Alice’s $1.50 bid was successful and thereby deprived Bob of the additional $.50 per X that he would have acquired in the absence of that injustice. Bob, in short, is exploited to the tune of $.50 per X. This, then, represents the core of a conception of exploitation that can be clearly and directly construed as it commonly is construed: namely, as a form of injustice. An exploitation is a transaction in which the exploitee (Bob) gains less—​and the exploiter (Alice) more—​than they would each have gained had a prior injustice not occurred. What can count as such a prior injustice will, of course, depend on some wider theory of distributive justice. Each individual’s entitlements—​what they have just rights to possess and what they are justly permitted to do with those possessions—​will vary according to which particular account of distributive justice (Rawlsian, Dworkinian, left-​libertarian, Nozickian, Rothbardian, etc.) one endorses. We can amplify this conception of exploitation by indicating a further range of variables that might count as instances of that prior injustice.

550    Benjamin Ferguson and Hillel Steiner So, for instance, an alternative reason why Carol did not outbid Alice might be that she was forcibly prevented from attending the auction. If, according to some theory of distributive justice, Carol’s moral rights include various claims, liberties, powers, and immunities that entitle her to participate in the auction, her forcible prevention from doing so amounts to an injustice. Similarly, if her failure to attend the auction was due to someone depriving her of a notice of the auction’s occurrence—​a notice which she was morally entitled to receive—​that, too, could amount to an injustice. More complexly, if someone imposed a surcharge on the selling price of an X, that could amount to an injustice. Suppose there was an imposed surcharge of 10 per cent. Then, although Carol was prepared to pay up to $2.00 for each of Bob’s Xs, she was unwilling to pay that extra 10 per cent, that is, $2.20. Alice’s winning bid of $1.50 incurs the same surcharge, and she has to pay $1.65. In this case, Bob gets only $1.50 for each of his Xs; the surcharger gets $.15; and Alice gets Bob’s Xs, whereas, in the absence of that surcharge, she would have had to bid more than $1.50 to acquire them. What is happening here is that the surcharger is, in effect, exercising partial ownership over the wealth endowments of the bidders.14 Whether we should consider that exercise to be an injustice—​that is, whether the bidders are justly entitled to full ownership of their endowments—​depends, of course, on the particular theory of distributive justice we endorse. Yet another form of prior injustice, sufficient to entail that the $1.50 price is exploitative, consists in what may previously have happened to the exploitee, Bob, himself. So far, the prior injustices we’ve considered have all been ones committed against a third party, Carol. It is the violation of Carol’s rights—​with regard to her wealth endowment and her auction participation—​that has supplied the condition necessary for Bob’s exploitation by Alice. But, equally, it can be a prior violation of the moral rights of Bob himself that supplies that condition. Thus, suppose that, prior to the auction, Bob himself was robbed. Before that robbery, Bob’s reservation price for his Xs was, say, $1.80 each: he would not have sold the Xs for less than $1.80 each.15 The effect of the robbery, in diminishing his endowment, is to lower his budget line to the point where his reservation price is only, say, $1.20. It is thus by virtue of that robbery that Alice is able, exploitatively, to purchase his Xs for as little as $1.50 each. Steiner’s conception of exploitation as injustice thereby analyses it as consisting in a pair of sequenced events: a prior injustice which consists in the violation of a relevant

14 

On the composite nature of ownership, and its consequent interpersonal divisibility, see Honoré (1961). 15  Cf. Steedman (1987): ‘The simplest example of a reservation price is that price below which an owner will refuse to sell a particular object in an auction. Since the owner could always, in principle, enforce such a price by outbidding everyone else, this leads immediately to the more general concept of a reservation price as that price at which the owner of a fixed stock will choose to retain some given amount from that stock, rather than supply more.’

Exploitation   551 person’s rights—​the exploitee’s or some third party’s—​followed by the exploitative transaction itself.16

25.4.3 Advantage Taking Both Steiner’s account and Roemer’s accounts of exploitation are based upon historical accounts of fairness: for both, a prior injustice is necessary for a subsequent exploitation. The two accounts also share a certain amount of flexibility with respect to the theories of distributive justice that underpin this prior injustice. Just as Roemer’s account is compatible with a variety of withdrawal conditions, so too does Steiner’s account allow for a variety of theories of distributive justice. Indeed, if the theory of distributive justice that is employed in Steiner’s account is the same as that which gives content to the withdrawal conditions in Roemer’s account, then the scope of the resulting conceptions of exploitation will be very similar. Suppose that the withdrawal conditions for Roemer’s account are resource egalitarian: agents may withdraw with their per capita share of alienable endowments. Then a coalition S is exploited if, after withdrawing, Roemer’s conditions R1 and R2 are satisfied. That is, if the members of the coalition had their per capita share of alienable assets they would be better off and their compliment, St, would be worse off. Similarly, if a resource egalitarian theory of distributive justice is combined with Steiner’s account, then Alice exploits Bob in a transaction just in case Bob gains less, and Alice more, than they each would have gained because of a prior unequal distribution of alienable resources. The scope of Steiner’s and Roemer’s conceptions differ (rather trivially) because Roemer’s conditions focus on coalitions, while Steiner’s are expressed in terms of individuals and (more significantly) because of the evaluations of whether the person or coalition is better or worse off. Roemer’s evaluation of whether a coalition is better (worse) off depends on whether, in the post-​withdrawal economy, there is a utility frontier available to the coalition that dominates (is dominated by) the frontier available prior to withdrawal. Reaching this frontier may require a number of transactions between the agents in the coalition. On the other hand, on Steiner’s account, a person is better (worse) off if they gain more (less) utility from an individual transaction, rather than a series of transactions. Roemer’s account focuses on how a coalition fares in the long run, making it more appropriate for questions related to the state of particular socioeconomic classes while Steiner’s focus on how agents fare in particular transactions makes his account more appropriate for evaluating the state of individuals. So, while the extensions of the two historical accounts are not identical, they are very similar. This similarity suggests a possible problem for Steiner’s account. If the extension 16  Since some of those prior injustices constitute unjust reductions in persons’ endowments, it follows that an exploitative transaction can also be the result of prior exploitative transactions. If Bob receives less than he is justly entitled to for his Xs, the bid that he can enter at a subsequent auction, for Doug’s Ys, will be less than it would have been had he (Bob) not been exploited in his sale of the Xs.

552    Benjamin Ferguson and Hillel Steiner of Roemer’s account, sans a dominance condition, is too wide, then it seems the same could be true of Steiner’s approach. And indeed, such a criticism has been raised by Steven Walt, who argues, . . . if a title to objects resulting from a rights violation is invalid and exploitation involves a rights violation, then it follows that most if not all unequal bilateral exchanges are cases of exploitation. But this is implausible. The presence of exploitation is unquestionable. Its omnipresence is questionable. (Walt 1984)

The problem Walt identifies is common to many historical accounts of justice. A past rights violation may infect all ‘downstream’ transactions, making the holdings resulting from all subsequent transactions unjust. Since there is good reason to believe that there exists at least one rights violation in the history of transactions leading up to Alice’s sale of an apple to Bob, there is good reason, on Steiner’s account, to believe not only that this transaction is exploitative, but indeed, that most transactions are exploitative. In this case, the extension of Steinerian exploitation exceeds our ordinary understanding of exploitation. One reason we may think the scope of exploitation characterized by Steiner’s and Roemer’s accounts outstrips our ordinary understanding of exploitation is that, though the accounts capture an unfair—​or unjust—transaction, the conditions that the accounts provide do not ensure that this unfairness comes about ‘in the right way’; that is, in a way that justifies our calling the transactions exploitative. Both accounts seem to allow for a kind of ‘accidental exploitation’. However, we may think that exploitation should be tied more tightly to the attitudes or intentions of the exploiter. What exercises us about sweatshops, for example, is not merely that some individuals happen to gain unfairly at the expense of others, but rather that they seek out vulnerable individuals from whom they can be certain to extract large profits. Though Roemer and Steiner both provide compelling analyses of the unfairness involved in exploitation, their approaches do not seem to account for the sense in which exploiters take advantage of their victims. In fact, in this sense it seems that the ahistorical accounts fare much better. Though Goodin’s claim that any form of vulnerability could generate obligations to constrain our advantage was too broad, his remark that exploitation involves playing for advantage in an inappropriate way seems to be on the right track. One approach, suggested by Ferguson (2013), is that a conception of exploitation that aligns more closely with our ordinary understanding of the concept can be provided by adding an ‘awareness’ condition to historical approaches like Steiner’s and Roemer’s.17 On this approach, exploitation differs from mere unfair transaction because exploitation involves cases in which Alice both gains unfairly at Bob’s expense and has a true

17 

Indeed, the awareness condition can be seen as an operationalization of Roemer’s dominance condition.

Exploitation   553 belief—​is ‘aware’—​that she does so. When Alice sees that Bob is unjustly disadvantaged and yet, in light of this awareness fails to constrain her advantage over him, not only does she gain unfairly from Bob, but she actively takes advantage of him. Advantage taking in this sense, then, is stronger than Elster’s minimalist claim that we take advantage of others when we engage in a mutually beneficial transaction with them. Not only does the ‘awareness condition’ limit the scope of these theories, but Ferguson’s account can be seen as a hybrid account that incorporates concerns from both ahistorical and historical approaches. It combines the thicker notion of advantage taking implied by ahistorical accounts’ concern for the attitudes of the exploiters towards those they exploit with the concern for personal responsibility incorporated in the historical accounts’ approach to fairness. As a hybrid account, however, Ferguson’s approach has its drawbacks. In particular, the satisfaction of the awareness condition depends on persons’ beliefs; thus, on Ferguson’s account, whether an act is exploitative depends on whether it has a certain intensional description. However, as Steiner has argued, . . . whether an act is just doesn’t depend on . . . its having a certain intensional description. It depends, rather, on its having a certain extensional description: that is, on whether it’s compatible with a certain distribution of freedom, with a certain set of rights. (Steiner 1994)

In this sense, the link between exploitation and injustice is weakened. On Ferguson’s account exploitation is not a specific form of injustice—​rather, it becomes a wrongful, or impermissible response to prior injustice.

25.5 Conclusion As was noted in the Introduction, some philosophers and economists believe that: (1) If a transaction is Pareto improving and consensual, then it is not impermissible. Contrarily, it is widely claimed that: (2) There exist some transactions that are Pareto improving, consensual, and impermissible. All the theories of exploitation examined here support (2) and reject (1): that is, all exploitative transactions are morally impermissible. Whether we should regard legal interference with such transactions as morally warranted will thus depend on (i) how closely our conception of exploitation is associated with that specific form of moral

554    Benjamin Ferguson and Hillel Steiner impermissibility that we call injustice; and (ii) whether we believe that injustices are the only form of morally impermissible acts that warrant legal interference.

References Bertram, C. (1988). ‘A Critique of John Roemer’s General Theory of Exploitation’. Political Studies 36: 123–​30. Bronfenbrenner, M. (1973). ‘Samuelson, Marx, and Their Latest Critics’. Journal of Economic Literature 11: 58–​63. Cohen, G. (1979). ‘The Labour Theory of Value and the Concept of Exploitation’. Philosophy and Public Affairs 8: 338–​60. Eatwell, J., Newman, P., and Milgate, M. (eds) (1987). The New Palgrave:  A Dictionary of Economics. London & New York: Macmillan. Elster, J. (1982). ‘Roemer Versus Roemer: A Comment on New Directions in the Marxian Theory of Exploitation and Class’. Politics and Society 11: 363–​73. Ferguson, B. (2013). The Paradox of Exploitation: A New Solution. Ph.D. thesis, London: LSE. Ferguson, B. (2016). ‘Exploitation and Disadvantage’. Economics and Philosophy 32(3): 485–​509. Goodin, R. (1986). Protecting the Vulnerable. Chicago, IL: University of Chicago Press. Goodin, R. (1987). ‘Exploiting a Situation and Exploiting a Person’, in Andrew Reeve (ed.) Modern Theories of Exploitation. London: Sage, pp. 166–​200. Guest, A. (1961). Oxford Essays in Jurisprudence, First Series. Oxford: Oxford University Press. Honoré, A. (1961). ‘Ownership’, in A. Guest (ed.) Oxford Essays in Jurisprudence, first series. Oxford: Oxford University Press, pp. 107–​47. Kymlicka, W. (1990). Contemporary Political Philosophy: An Introduction. Oxford: Clarendon Press. Nozick, R. (1974). Anarchy, State, and Utopia. Oxford: Blackwell. Reeve, A. (ed.) (1987). Modern Theories of Exploitation. London: Sage. Reiff, M. (2013). Exploitation and Economic Justice in the Liberal Capitalist State. Oxford: Oxford University Press. Roemer, J. (1982a). A General Theory of Exploitation and Class. Cambridge, MA:  Harvard University Press. Roemer, J. (1982b). ‘New Directions in the Marxian Theory of Exploitation and Class’. Politics and Society 11: 253–​87. Roemer, J. (1982c). ‘Property Relations vs. Surplus Value in Marxian Exploitation’. Philosophy & Public Affairs 11: 281–​313. Sample, R. (2003). Exploitation: What It is and Why It’s Wrong. Boulder, CO: Rowman and Littlefield. Samuelson, P. (1971). ‘Understanding the Marxian Notion of Exploitation:  A Summary of the So-​Called Transformation Problem between Marxian Values and Competitive Prices’. Journal of Economic Literature 9: 399–​431. Sensat, J. (1984). ‘Exploitation’. Nous 18: 21–​38. Steedman, I. (1987). ‘Reservation Price and Reservation Demand’, in John Eatwell et al. (eds) The New Palgrave:  A Dictionary of Economics, vol. 4. London & New  York:  Macmillan, pp. 158–​9. Steiner, H. (1984). ‘A Liberal Theory of Exploitation’. Ethics 94: 225–​41.

Exploitation   555 Steiner, H. (1987). ‘Exploitation:  A Liberal Theory Amended, Defended and Extended’, in A. Reeve (ed.) Modern Theories of Exploitation. London: Sage, pp. 132–​48. Steiner, H. (1994). An Essay on Rights. Oxford: Basil Blackwell. Steiner, H. (2010). ‘Exploitation Takes Time’, in J.  Vint et  al. (eds) Economic Theory and Economic Thought: Essays in Honour of Ian Steedman. London: Routledge, pp. 20–​9. Steiner, H. (2013). ‘Liberalism, Neutrality and Exploitation’. Politics, Philosophy & Economics 12: 335–​44. Stemplowska, Z. (2009). ‘Making Justice Sensitive to Responsibility’. Political Studies 57: 237–​59. Vint, J., Metcalfe, S., Kurz, H. D., Salvadori, N., and Samuelson, P. (eds) (2010). Economic Theory and Economic Thought: Essays in Honour of Ian Steedman. London: Routledge. Walt, S. (1984). ‘Comment on Steiner’s Liberal Theory of Exploitation’. Ethics 94: 242–​7. Wertheimer, A. (1996). Exploitation. Princeton, NJ: Princeton University Press.

Chapter 26

Pu blic G o od s David Miller and Isaac Taylor

26.1 Introduction Public goods, as commonly defined by economists, are goods that are non-​excludable (if they are provided, they must be provided to everyone in the relevant population) and non-​rival (one person’s enjoyment of the good does not reduce the value of the good to others).1 A textbook example of a public good is national defence. Once institutions of national defence have been created, they necessarily protect everyone in a particular country from invasion, and one person’s benefitting from being so protected does not reduce the quantity or quality of protection others receive. When goods are supplied publicly, however, they are vulnerable to the free-​rider problem. Since individuals will receive the benefits whether or not they contribute to their production, each person has an incentive not to contribute, and to ‘free-​ride’ on the efforts of others. The more people act on this incentive, the less of the good will be supplied. Even those willing to contribute their fair share of the cost will lack assurance that others will do their part. Everyone may be a net loser as a result. For this reason, even classical liberals justified the state in using its coercive powers to supply certain public goods at the taxpayers’ expense (Mill 1965: 968–​9; Smith 1976: 723). Stepping outside of the textbooks, however, we see that in the real world few of the goods that are classified as public are strictly either non-​rival or non-​excludable. For example, crowding effects may mean that the value of a good diminishes the more people have access to it (think of open roads). And very few goods are inherently non-​ excludable, even if it is infeasible or impractically costly to exclude people from them (de Jasay 1989: 61–​3). Even the good of clean air, for instance, might become excludable as technology advances: it could be pumped into bio-​domes that were accessible 1  Although this is the definition most often given by economists, and is the one that we will use here, in other contexts public goods are defined by other conditions. For an outline of the different ways of defining public goods, see Cullity (1995, 3–​4, 32–​4).

Public Goods   557 only to those who are willing to pay a fee. Moreover goods that remain public, in the sense that everyone is free to take advantage of them, may nonetheless be ‘privatized’ by identifying those who actually make use of them and charging a fee: this happened historically in the case of lighthouses, for instance (Coase 1974). In this way the free-​rider problem can be overcome, and the good in question might be produced by an entrepreneur rather than the state. Given that few goods are strictly public in the textbook sense, it becomes a matter of political decision as to which goods should be provided as public goods, open to all without charge. Various reasons may be given for supplying goods in this way. We may, for example, lack the technology to exclude people without incurring unreasonable cost (as in the case of clean air). Or people may be regarded as having rights of access to certain goods, such as works of art and other cultural artefacts, as part of their cultural heritage, and so museums and art galleries are provided free of entry charges. A further reason might have an egalitarian flavour. Although we could leave individuals to purchase protection from physical attack from private firms, for example, and let those willing to pay more (or, more likely, those able to pay more) buy better protection, we have a norm that tells us that every citizen should receive equal levels of physical protection, regardless of their willingness or ability to pay. We therefore provide a public police force to protect each individual, and cover the costs through general taxation. The non-​rivalry condition can also be approximated, by supplying enough police so that crowding issues do not arise. Similar concerns may encourage us to supply a number of environmental goods publicly (Light 2000: 222–​4), although efficiency and technological constraints may also be at work here. Our discussion here will apply to goods that are public in this wider sense as well as to goods that are strictly public, by virtue of non-​excludability and non-​rivalry, since we believe that similar issues of distributive justice arise in both cases.2 How should we judge the supply of public goods by the state? Economists usually apply an efficiency standard: does the overall benefit created by providing a good outweigh the cost to the taxpayer in diminished consumption of private goods (Samuelson 1954)? Political philosophers, meanwhile, tend to focus on issues of legitimacy and political obligation. Is it legitimate for the state to use its coercive power to tax citizens to pay for public goods (Gaus 2011: 529–​38; Quong 2011: 88–​91)? Are citizens under an (enforceable) obligation to contribute (Simmons 1979: Ch. 5; Klosko 1992)? Our focus here, however, will be on an issue that has received far less attention, the issue of distributive justice. Assuming that in general the state can legitimately supply public goods—​both goods that are strictly public, and goods provided without charge as public goods, as indicated—​how should we decide which goods are to be supplied and how the ensuing costs should be allocated? These costs are usually collected through

2  For some purposes it may be important to separate the two cases. Serena Olsaretti, for example, applies the term ‘public goods’ only to goods that given current technology meet the two conditions of non-​excludability and non-​rivalry and describes goods that are made public, even though they could in principle be supplied as private goods, as ‘socialized goods’ (Olsaretti 2013).

558    David Miller and Isaac Taylor taxation,3 but different ways of taxing individuals could be envisaged: a flat rate tax and a progressive tax are two options. On the benefit side of the equation, as well, the distribution of gains will depend upon which public goods are supplied. If the state funds national parks, for example, it will disproportionately benefit nature-​lovers, while public roads are likely to benefit those who own a car more than those who do not. Both the range of goods provided and the funding mechanism used to support them therefore raise important questions of distributive justice, and in this chapter we discuss how one might begin to answer these questions.

26.2  Existing Liberal Theory and Public Goods Leading liberal theories of distributive justice have thus far failed to address the public goods problem adequately. Consider first John Rawls’s theory. In A Theory of Justice, Rawls understands that achieving justice requires the supply of a number of public goods. Some public goods—​he mentions national defence—​are so important that the state must provide them in order to avoid the free-​rider problem (1971: 267). Also mentioned are cases where some agents, left to their own devices, will impose harms on others through negative externalities (1971: 278). If a factory pollutes a river that a nearby village relies on for drinking water, for example, the state will need to step in and impose environmental regulations to ensure that the public good of clean drinking water is available. Although Rawls is certainly correct to assume that justice requires the supply of a number of important public goods (just shares in private goods could not be maintained without an effective system of law and order, for instance), he says little that would help us to determine exactly which goods fall into this category. So, although we could stop the unjust harm of the pollution in the river in the case above by imposing environmental regulations, an alternative would be to create property rights in the river, perhaps awarding them to the villagers (Coase 1960). This would prevent the factory polluting without the villagers’ say-​so, but would also allow the villagers to sell the right to pollute their river, if they thought that they could get drinking water from elsewhere. Rawls provides no grounds for deciding which of these two solutions is to be preferred. In consequence, rather few public goods are likely to be justified in this way. Elsewhere, Rawls acknowledges that governments typically supply a number of public goods aside from those strictly required by justice. Public parks and free museums, for example, are hardly needed to uphold just shares or prevent harm, yet we often see governments supporting their provision. Rawls argues, however, that for this to 3 

Though not always. Conscription, for example, may be used to pick up some of the costs of national defence.

Public Goods   559 be legitimate, the good in question must pass a ‘unanimity test’, originally devised by the economist Knut Wicksell (Rawls 1971: 282–​3; Wicksell 1958). This requires that the supply of such a public good, along with the tax schedule that will be used to fund it, must be unanimously agreed to, or at least nearly so. In his later book Political Liberalism, Rawls relaxes this condition, and allows that additional public goods may be supplied by the state if approved by a majority vote (1993: 246). In both cases, Rawls appears to think that justice has nothing to say about the supply of these goods, and as long as the supply schedule passes the relevant test (either unanimity or a majority vote), no injustice is done to citizens. But surely issues of justice do arise here. For we would certainly want to judge the manner in which the costs of providing these public goods has been distributed by a standard of social justice, even after it has been voted through or has received nearunanimous approval. What if, for example, a majority voted on a distribution of costs that adversely affected the holdings of the least well-​off members of society (Miller 1999a: 156)? Since Rawls’s difference principle, which governs distributive justice for certain private goods, requires that socio-​economic inequalities should always work to the advantage of a representative member of the least well-​off group, supplying public goods according to a schedule which worsened the position of this group must be unjust. For another liberal theory of justice that pays insufficient attention to public goods, we can look at Ronald Dworkin’s principle of ‘equality of resources’. It is clear from the outset that Dworkin is not overtly interested in public goods, since he understands equality of resources as ‘a matter of equality in whatever resources are owned privately by individuals’ (2000: 65). Nonetheless, Dworkin does acknowledge, at a number of points, the need to address the distributive issues that public goods raise. For one thing, he notices that negative externalities brought about by the way individuals choose to use their fairly acquired resources may undermine equality of resources, and to correct for this justice may require the supply of what are in effect public goods, for the value of the goods for which individuals bid in the auction that Dworkin uses to model equality of resources may in part be public in character. The example he gives is of an otherwise pristine Georgian square in which one prospective property owner intends to build a house in the form of a glass box (2000: 156). If he succeeds, he will destroy the public good of an architecturally harmonious square, which forms at least part of the value of the other properties to their owners. If equality of resources is to prevail, this public good must be protected (by zoning regulations or tort law) so that bidders in the auction can more accurately anticipate the outcome of their bids (and thus make the best use of their bidding tokens).4 So, for Dworkin, at least some public goods must be protected to ensure a just distribution in private goods. But what of additional public goods, those not required to 4  See Dworkin (2000: 156–​7). To show why planning regulation is better than a free-​for-​all in such cases, Dworkin imagines a ‘pre-​auction auction’ in which people bid under full knowledge of others’ motives, intentions, future transactions, etc., and there are no transaction costs to deter people from making combined bids.

560    David Miller and Isaac Taylor uphold justice? Dworkin addresses the issue of individuals who prefer that some of their society’s resources should be used to produce public goods from which they derive no personal benefit. He gives the example of Charles, an environmentalist, who argues for preserving the habitat of the snail darter (a small fish) at the expense of building a new dam, because of the intrinsic value he attributes to different species’ continued existence (2000: 26). Given Charles’s beliefs, the survival of the snail darter is a public good (it is non-​excludable and non-​rival); so, on the other side of the argument, is the building of the dam for those who favour cheap electricity. Dworkin’s response is that such conflicts of preference should be decided procedurally, for example by taking a majority vote on the building of the dam. Like Rawls, then, Dworkin ignores issues of distributive justice that arise when public goods of this kind are at stake. In certain other cases, however, Dworkin appears to take a different view, and to be willing to extend special treatment to some preferences for public goods. In his earlier paper ‘Liberalism’, he discusses cases where a failure to provide public goods will mean that the availability of certain ways of life will be irrevocably lost. If housing developments mean that there is little countryside left, those whose way of life requires large tracts of undeveloped land will be disadvantaged. Here, liberal neutrality itself can justify the state in providing the public goods that are essential to keep open the possibility of pursuing the threatened way of life, according to Dworkin (1986: 202). It is unclear, though, on what grounds such claims can be made. As has been pointed out elsewhere (de-​Shalit 1995: 126; Miller 1999a: 158–​9), a liberal egalitarian society cannot expect to impact equally on all ways of life; it makes some of them, such as that of a reclusive hermit or a religious fanatic, hard to pursue. What makes the claim of someone whose way of life involved roaming the wild countryside, and who therefore has preferences for certain environmental public goods, any different? As our last case study of the role given to public goods within liberal theories of distributive justice, we may consider Michael Walzer’s approach in Spheres of Justice (1983: Ch. 3). Walzer discusses public goods under the heading of ‘communal provision’, and he begins by noting that all political societies have devoted some part of their resources to providing collectively for the needs of their members. But he also notes that the content of this provision has varied considerably: ‘the ancient Athenians, for example, provided public baths and gymnasiums for the citizens but never provided anything remotely resembling unemployment insurance or social security’ (1983: 67). Walzer’s view is that the choices made about public good provision always reflect culturally specific understandings of what a ‘common life’ requires, and that these choices can only be made by democratic deliberation. So he is reluctant even to identify a set of basic needs that all societies must provide for as a matter of justice: ‘the idea of need and the commitment to communal provision do not by themselves yield any clear determination of priorities or degrees’ (1983:  66). Like Rawls and Dworkin, therefore, Walzer concludes that justice in the provision of public goods is largely a matter of the procedures used to decide which goods will be supplied and by what means, rather than a matter of the substance of what is provided.

Public Goods   561 A very different perspective is provided by libertarian theorists, who are sceptical about the possibility of any public goods being provided by the state as a matter of justic­e. These philosophers claim that political association should as far as possible be based on the voluntary consent of individuals, and they doubt that such individuals will agree to be taxed to provide public goods from which others will also benefit. By withholding consent, each can hope to free-​ride on the consensual contributions of others—​though if all think like this, no public goods will actually be provided. Contemporary libertarian thinking on public goods undoubtedly takes as its starting point Robert Nozick’s discussion of the principle of fairness. Nozick asks us to imagine a group of 364 neighbours each of whom voluntarily spends one day a year operating a public address system broadcasting across their neighbourhood. They read philosophy, play music, and tell funny stories. The turn of the 365th neighbour to operate the PA system then arrives. This neighbour has genuinely benefitted from the PA system on the other 364 days of the year, but, according to Nozick, this does not show that he has any obligations to pick up his fair share of the costs now, by sacrificing a day to operate the system. Nor, says Nozick, can others coerce him to do so (Nozick 1974: 93–​5). Some libertarians would distinguish between justice and fairness (de Jasay 1989:  212–​14), restricting justice to enforceable obligations, and say that although it would be unfair of this individual not to pick up the costs (Nozick is ambiguous here), it would nonetheless be unjust to force him to do so. For Nozick and other libertarians, then, to be concerned with distributive justice in the supply of public goods is wrongheaded. If the 364 neighbours voluntarily decided to produce a good from which the 365th neighbour cannot be excluded, it cannot be unjust that he receives the same benefit without paying the same cost. If justice is based on consent, whatever distribution of the costs and benefits of an activity that individuals voluntarily engage in (or refrain from) must be the just distribution. If we cannot require that individuals pay towards the cost of public goods, even when they benefit from them, it is likely that a large number of public goods which we typically see modern states supplying would be lost. While Nozick argues that a just state can use its coercive power to ensure the supply of personal protection, the libertarian approach may lead to other public goods such as public transportation, clean air, public parks, and publicly funded education being not supplied at all. The undesirable consequences of this view have not been lost on all libertarians. David Schmidtz (1991), for one, attempts to reconcile the provision of a wide range of public goods with the libertarian insistence on voluntarism by introducing the device of assurance contracts. He envisages private entrepreneurs approaching the group of individuals who would benefit from the provision of a particular good and asking them to contribute towards its cost. If everyone pays the required amount, the public good is supplied by the company using the funds collected. If anyone refuses to pay, everyone else gets their money back. This supposedly overcomes the free-​rider problem since, first, each individual knows that refusal to pay will mean that the good definitely will not be supplied, and, second, each has assurance that everyone else will contribute (and so will not fear being exploited).

562    David Miller and Isaac Taylor There are a number of questions and problems for Schmidtz’s approach (Miller 1991). But setting these aside, note that Schmidtz has not provided us with a proposal for how to distribute the costs and benefits of public goods, any more than has Nozick. The private firms would presumably only attempt to supply public goods that benefitted everyone in the relevant community (since this is the only way they would be able to secure the unanimous consent necessary for their venture to be a success). And the distribution of the costs would be whatever induces all individuals to consent. But this is consistent with massive inequalities in benefit. While everyone must gain on balance if they are going to sign up, it may be profitable for the entrepreneurs to supply public goods that benefit a majority considerably, but make signing up barely worth the while of the minority. For many people, this seems unfair, and the unfairness would also amount to injustice.

26.3  Toward a Liberal Theory of Public Goods Our discussion of existing liberal theory suggests that public goods may enter into theories of distributive justice at two points. First, certain public goods may be required by justice, either because (like law and order) they are required for social cooperation to take place at all or because they are needed to ensure that the value of fair shares is not undermined (like zoning restrictions for Dworkin). Second, Rawls, Dworkin, and Walzer all accept that, once basic justice is achieved in this way, the state may supply additional public goods, either if the distribution of costs can be agreed upon at least nearly unanimously (in early Rawls) or if a decision is made democratically to supply them (in later Rawls, Dworkin, and Walzer).5 We will call the first category of public goods—​those required by justice—‘essential public goods’, and the second category—​those not required by justice—​‘discretionary public goods’. While the liberal theories outlined above implicitly acknowledge this division, they do not offer a precise specification of which public goods are required by justice, and which therefore are to be counted as essential. Nor do they specify how the costs and benefits of discretionary public good supply should be distributed, relying instead on democratic procedures or the unanimity condition to yield an answer. It will be the task of the rest of this chapter to address these issues. Before proceeding, though, we should note a challenge which says that public goods do not raise any issues of distributive justice in the first place. This challenge can take a number of forms. At its simplest, it points out that public goods are non-​excludable, and so we need not worry about distributive justice when they are supplied, since equal access is assured. Strictly speaking, nothing is being distributed; we simply need to 5 

For a slightly different, tripartite, distinction, see Miller (1999a: 159–​60).

Public Goods   563 supply the good and everyone will benefit. If a national park does not charge an entrance fee, for example, everyone will have access to it once it is provided, and no issues of distributive justice arise. There are two mistakes in this line of reasoning. First, it ignores the fact that public goods often impose costs. Maintaining a national park, for example, requires staff and resources to prevent activities like illegal hunting, tree-​felling, littering, and so on. These costs must be picked up somehow. One way would be to charge a sufficiently high entry fee, of course, but this would turn the good into a private good, and we may have reason to avoid doing this (efficiency, technology, or morality may rule it out). If the costs are picked up by taxation, instead, we need to come to a decision about how to distribute the tax burden. Should we use a flat-​rate tax, a progressive tax on income, or a tax on consumption, for example? This decision must undoubtedly be informed by considerations of distributive justice. The second mistake the challenge makes is to suppose that equal access to a good like a national park is all we require in order to justify it. But note that different individuals benefit by different extents from such public goods. Nature-​lovers will generally visit the park more often than others, and those with a car will be able to visit more often than those without, even if all these groups formally have equal access. Inequality of benefit appears immediately to raise questions of distributive justice. Our challenger may, however, respond by claiming that theories of distributive just­ ice should concern themselves, not with the distribution of benefit or welfare directly, but rather with the distribution of the means to benefit or welfare. Just as we should not worry that two individuals with the same income end up with different levels of welfare because one has expensive tastes for champagne and caviar and the other has more modest tastes for beer and hamburgers, neither should we worry that two individuals receive different levels of welfare as a result of having equal access to a national park. By creating the park, it is assumed, we add the same resource to everyone’s bundle. But this claim overlooks the obvious point that the value of the resource differs greatly from person to person, and we cannot ignore this when deciding what public goods to provide. If free art galleries were supported instead of national parks, for instance, some individuals would enjoy significant gains and others would lose. Considerations of distributive justice must apply when such decisions are made (Miller 2004: 129–​130). Another reason for thinking that (at least some) public goods do not raise issues of distributive justice may be advanced by environmental philosophers. They claim that public goods involving certain natural features—​a preserved forest, the continued existence of a species, a place of unique biological diversity—​have an intrinsic value, irrespective of the way human beings value them, and they should be protected simply in virtue of this. Of course, humans may also attach value to these things. They may enjoy visiting these places, or they may gain satisfaction simply from knowing they exist. But the environmentalists’ point is that these goods have an independent value, which would persist even if all the human beings on earth were to die out (e.g. Sylvan 2003). And if this is correct, public policy should not concern itself with the way in which different individuals may derive benefits from these goods; the non-​anthropocentric value of nature alone gives us a reason to promote them.

564    David Miller and Isaac Taylor In response, note that at least half of the question of distributive justice remains for the environmentalists: assuming there are costs involved in preserving the natural world, these could be allocated in different ways.6 But, aside from this, the green conception of value is deeply controversial, and so may not be the appropriate basis for public policy. And, even if we could come to some agreement about non-​anthropocentric values, this does not necessarily have any implications for requirements of justice: acknowledging something as a non-​anthropocentric value does not necessarily mean that we have a (justice-​based) reason to promote or preserve it (O’Neill 1992).

26.4  Essential Public Goods We now turn to a discussion of what distributive justice requires in the supply of public goods, starting with what we called ‘essential public goods’, whose provision is itself a matter of justice. Consequently, the distribution of both costs and benefits will be determined by the general theory of justice that we favour. Consider, for example, how a Rawlsian might think about a public good such as national defence. Having first determined what level of defence is necessary to safeguard individual liberties and other primary goods, he would then distribute the costs of paying for defence through the tax system in whatever way the difference principle recommends—​the aim being to ensure that the economic position of the worst-​off group in society, net of tax, is as high as possible (presumably this would point towards a progressive tax schedule). For this approach to work, we need to know which public goods should be counted as essential. As we have already suggested, justice will require the supply of those public goods that are necessary for any social co-​operation at all, such as the institutions that provide basic security. Public goods may also be required to correct for externalities that would otherwise interfere with individuals’ justly acquired rights, for example a system of tort law. But many theories of justice also hold that there should be a social minimum that no individuals should fall below. The minimum is usually cashed out in terms of basic needs that must be met for an individual to live a minimally decent life (in their society) (Rawls 1993: 7; Miller 1999b: 210). If this is correct, and we have obligations to meet (at least some of) the basic needs of fellow citizens, then it seems that justice may require supply of a number of further public goods to serve this function. These public goods would also count as essential. But note that there may sometimes be alternative ways of meeting individuals’ basic needs: the supply of public goods may be one option, but private resource transfers may be another. Take the basic need for health. This might be met by creating the public good of a largely disease-​free environment through public health measures, alongside a 6  Casal (1999) examines a way environmentalists might go about allocating these costs, albeit based on the benefit to others of environmental goods rather than the requirement of others to support the intrinsic value of the environment.

Public Goods   565 publicly funded healthcare service. Alternatively, however, the same need could be met by allowing a market in private healthcare services to develop, with transfer payments to those who cannot otherwise afford to buy the services. When both options are available, which should we choose? One suggestion is that public goods should be preferred when they are the only or best way of meeting (at least one individual’s) basic needs (Miller 2004: 136–​7). Another is to supply public goods when they are the only feasible way of meeting (at least one individual’s) basic needs (Taylor 2013: 126). What is to be said for each proposal? The former formulation has the implication that if a small cost saving arises from meeting individuals’ basic needs through public goods, those public goods are required by justic­e, since that would be the ‘best’ way of meeting individuals’ basic needs. But it may be objected that a presumption in favour of individual liberty means that as far as possible we should leave it up to individuals how they organize themselves to provide for basic needs. For example, just because digging private wells would cost slightly more than creating a national company to deliver drinking water, it does not appear to follow that the nationalized solution is the one required by justice (cf. Simmons 2001:  34). Although each person may be worse off in terms of resources when they need to dig their own wells, if everyone prefers this solution, what injustice is done? And even if not everyone prefers digging wells, why should the presumption always be in favour of supplying the public good? These considerations may suggest that public goods should be required by justice when the stricter condition, that they are the only feasible way in which individuals’ basic needs can be met, obtains. Against this, though, if society has a justice-​based obligation to ensure that the basic needs of its citizens are met, it may be thought that it should do so without imposing unnecessary costs on them while doing so. While individuals may be quite capable of digging wells to meet their basic needs for water, it may be quite costly for some of them to do this, perhaps significantly more so than paying a national water company. If we think that basic needs should be met in the most efficient way possible, then perhaps this will lead us to accept the more permissive condition, that public goods should be used for this purpose when they are the only or best way in which basic needs can be met.7 Without coming to any firm conclusion regarding which suggestion is more plausible, we now turn instead to a rather different approach suggested by Stuart White. According to White, public goods are required by justice (and are therefore ‘essential’, in our terms) if they would be bought by individuals in a hypothetical market where each was ignorant 7  An additional problem for both suggestions is that the demand they impose to meet individual needs might involve excessive costs. For example, if the only way of curing a rare and debilitating disease was to carry out expensive medical research, these formulations imply that we are to divert all resources not already committed to meeting basic needs to this purpose. It is often thought, though, that we are only required to meet others’ basic needs up to a particular level (Dworkin 2000: 309–​10). It seems, then, that we should add a further condition to require that basic needs should be met by public goods only when this can be done at a reasonable cost. In this vein, the UK’s NHS will only fund medical research if it is likely to create a sufficient number of additional quality-adjusted life years (‘QUALYs’), that is, additional years of life spent in good health, per pound spent on research.

566    David Miller and Isaac Taylor of their own particular need for those goods (White 2003:  100–​6). White’s proposal builds on the device of hypothetical insurance markets found in Ronald Dworkin’s political theory (Dworkin 2000: 73–​83). The idea is as follows. Citizens deciding whether or not to supply a particular public good—​street lighting of a certain intensity, say—​ may each have different preferences for it. Some may want a lot of street lighting: perhaps they live in an area with high crime rates. But they also want to know how much street lighting justice requires them to provide. They therefore imagine themselves in an insurance market, where they know the population-​wide risk of crime, but not their own exposure to risk, and consider how much street lighting they would buy if they had the average earning potential in society, and costs were distributed equally. The level of street lighting the average citizen would choose, says White, is the level justice requires (2003: 105). Unlike the approaches discussed earlier, White does not aim to reduce the requirements of justice in public good supply to upholding just distributions of private goods and meeting individuals’ basic needs. His approach might count a public good as essential even if it does not serve either of these functions. We can imagine, for example, the average member of society bidding for a level of street lighting well above what would be necessary to meet their, or anyone else’s, basic need for protection from crime. But what would make street lighting at this intensity a requirement of justic­e? What could be said to somebody who refused to pay for it on the grounds that he preferred to see the stars at night? In this case, it seems, relying on a hypothetical choice to guide public good provision would not yield only essential goods, and the approach might therefore be thought to be deficient.

26.5  Discretionary Public Goods In section 26.4 we suggested that states are required as a matter of justice to provide public goods when these are essential to protect basic rights, either directly (as in the case of national defence), or to prevent individuals from infringing others’ rights (as in the case of tort law), or finally because there are needs that are best catered for through public provision (as in the case of a public health service). A state that only did this, however, would look very different from most democratic states today, which provide a wide range of cultural, recreational, and environmental amenities that are available to all, either free or at subsidized prices, and that therefore count as ‘discretionary’ public goods. The supply of such goods, we have suggested, is not itself a matter of justice. But justice enters the picture as soon as we have to decide which goods are to be provided and how the costs of providing them are to be allocated. Unlike in the case of essential public goods, we cannot simply distribute the costs so that whatever prior principle of distributive justice we endorse—​the difference principle, say—​is served by that distribution. Instead, we should assume that privately held goods are already distributed as justice requires, and ask what expenditures on public

Public Goods   567 goods will maintain this distribution. The rationale for this is given by Rawls, who writes ‘[t]‌here is no more justification for using the state apparatus to compel citizens to pay for unwanted benefits that others desire than there is to force them to reimburse others for their private expenses’ (1971: 283). Since supplying discretionary public goods is not within the remit of justice, we should not attempt to justify their supply by their contribution to justice. Instead, we should investigate how discretionary public goods should be supplied in a way that leaves the justice of the existing distribution intact; discretionary public good supply should be ‘justice-​preserving’ rather than ‘justice enhancing’ (Miller 2004: 131n4). Whereas Rawls, as noted, thought that justice could be preserved if public good supply was agreed to unanimously, or as he later suggested, through a majority vote, we have argued that there should be an independent standard of justice by which we judge the supply. If a majority voted in favour of supplying a single public good which benefitted them, but not the minority, while requiring everyone to contribute to the cost, we would surely want to judge the resulting distribution unjust. In this section, three principles purporting to furnish us with such a standard are discussed.

26.5.1 The Benefit Principle Liam Murphy and Thomas Nagel offer the following principle. The cost to each individual of supplying discretionary goods should be proportional to the benefit each receives from such a package (Murphy and Nagel 2002:  85). Thus, if one individual benefits twice as much from the overall supply of discretionary goods, she should pay twice as much. They call this the ‘benefit principle’. To see how this principle would work in practice, imagine a society made up of two individuals: Adam and Betty. The government supplies two discretionary goods: publicly funded television and free art galleries. Adam, a couch potato, benefits enormously from public television; he receives forty units of benefit from these (but no benefit from art galleries). Betty, an art-​lover who enjoys occasionally visiting art galleries, receives twenty units of benefit from these (but no benefit from public television, which she never watches). Suppose that the cost of supplying each public good is fifteen units, measured in the same metric as the benefit each receives.8 So, the cost of supplying both public goods together is thirty units. How should we distribute this cost between Adam 8  A worry here may be that we cannot possibly measure benefit and cost using the same metric; while an individual’s benefit from public goods is in the form of subjective utility or preference-​satisfaction, the cost each pays is usually monetary (in the form of taxation). However, economists often use a device known as the ‘contingent valuation method’ to convert the subjective benefit of each individual into a monetary amount. Surveys are taken asking individuals ‘How much would you be willing to pay for public good x to (continue to) be supplied?’ or ‘How much would it take to compensate you for public good x to be taken away?’ The answers given are thought to represent the benefit the respondent receives as a monetary value. This method has well-​known practical problems, but we assume that, at least in theory, a monetary value reflecting benefit can be revealed in this way.

568    David Miller and Isaac Taylor and Betty? The benefit principle, as outlined above, says that Adam, who benefits twice as much from these discretionary goods as Betty, should pay twice as much; thus Adam pays twenty units tax while Betty pays only ten units. The resulting distribution can be summed up in Table 26.1. This principle faces a number of potential objections. For one thing, it appears to treat society like a company rather than a community, which some may find objectionable (Miller 2004: 140). But more troublingly, it does not provide criteria for determining whether more (or different) discretionary public goods should be provided in order to ensure a just distribution of the costs and benefits of the overall package. It simply takes the discretionary public goods that are currently supplied as given. While in theory a society might supply all the public goods that anyone would find valuable, in practice it will be a matter of choosing between different possible combinations of goods. Murphy and Nagel do, at one point, claim that public provision should be preferred to market provision whenever ‘the market will in certain respects make everyone worse off than they could be’ (2002: 86, emphasis added). But there do seem to be a number of cases where we would wish to supply more (or different) public goods that are of benefit to some but not others. For example, suppose that in our example, the government could also provide subsidized opera at a cost of twenty units. This benefits Betty by twenty units, but it does not benefit Adam at all. The benefit principle would then yield the distribution of costs shown in Table 26.2. Moving from the position indicated in Table 26.1 to that in Table 26.2 does not make everyone in society better off. In fact, Adam loses overall: his net benefit is reduced from twenty to fifteen. The benefit principle cannot, therefore, endorse such a move, but neither can it oppose it. In both tables, costs are strictly proportional to benefits. However, some might say that Table 26.2 is intuitively fairer than Table 26.1, since net benefit from

Table 26.1 Arts and television Adam

Betty

Benefit

40

20

Cost

20

10

Net benefit

20

10

Table 26.2 Arts, television, and opera Adam

Betty

Benefit

40

40

Cost

25

25

Net benefit

15

15

Public Goods   569 the overall scheme of public good supply has been equalized. At the very least, we want a distributive principle to be able to decide between the two possibilities.

26.5.2 The Equal Benefit Principle David Miller provides a principle that overcomes this problem. His ‘equal benefit principle’ requires that each individual’s net benefit from supply of discretionary goods should be equal (2004: 142–​3). The rationale for this is that the state’s ability to provide public goods efficiently, making everyone better off, can be seen as a windfall gain for the citizens, and equality is the default principle for distributing such gains. Thus, in the two-​person example above, this principle requires that we move from Table 26.1 to Table 26.2, since this equalizes net benefit. The principle, then, is more determinate than Murphy and Nagel’s proportional benefit principle. There are, however, objections to Miller’s principle too. One is that it may be impossible to create a package of public goods that yields roughly equal benefit to everyone. Suppose that within the society there is a small group of people to whom no discretionary goods are of any value at all: all of their welfare derives from the possession of private goods. In this case the only way to implement the equal benefit principle would be for the state to provide no discretionary public goods to anyone. This would be unfortunate, given the benefits that would have to be forgone. Moreover, it is not clear that justice must prohibit the supply of public goods only to some, provided the non-​ beneficiaries are made no worse off as a result (which might be achieved if they can be identified and exempted from tax). To insist otherwise would fall foul of the famous ‘levelling down’ objection to egalitarian principles (Parfit 1995). A second problem with the equal benefit principle arises once we realize that the costs of providing different public goods will vary. The principle permits those with expensive tastes for more costly public goods to be subsidized by those with more modest tastes. So, in our two-​person example, imagine that Betty’s tastes change in such a way that for her to benefit to the same extent as previously, funding for art has to be doubled from fifteen to thirty (she now only enjoys looking at the very finest works of art, say). In this case, the equal benefit principle recommends loading the costs of this increase on to both Adam and Betty. For each to have an equal net benefit, the distribution of costs and benefits shown in Table 26.3 is now required: Table 26.3 Betty’s expensive taste for art Adam

Betty

Benefit

40

40

Cost

32.5

32.5

7.5

7.5

Net benefit

570    David Miller and Isaac Taylor In this scenario, the cost of Betty’s taste for expensive art is partially offloaded onto Adam, who has more modest tastes for cheap television. In other words, the overall net benefit which Adam gets out of the public good regime as a whole has been reduced (from 15 to 7.5) through no fault of his own.9 Is this fair? Now, the fact that someone needs more resources to satisfy their expensive tastes does not, in the view of some philosophers, mean that transferring the additional resources to them must always be problematic from the point of view of distributive justice. G. A. Cohen, for example, argued that only expensive tastes that were voluntarily developed should be exempt; expensive tastes that were unavoidable may rightly call for more resources to satisfy them (Cohen 1989: 922–​3). But drawing this distinction will be of little help here. It is likely that expensive tastes for at least some public goods will be voluntarily developed (as Betty’s taste for art might be). In these cases, it seems that the equal benefit principle would recommend objectionable transfers. Miller is aware of this issue, but holds that even in these cases such transfers may not be objectionable. In a society which is genuinely communitarian, he holds, ‘more people will be willing to allow public good production to be cross-​subsidised, subject to the rider that . . . each person should make an effort to demand less expensive rather than more expensive goods’ (2004: 145). For Miller, the problem of expensive tastes may not count against his principle, since if society is a genuine community, people may be willing to subsidize expensive tastes in this way. In this vein, Cohen has pointed out that libraries do not charge their borrowers according to the costs of the books they borrow.10 They charge the same fee, or nothing at all, to those who prefer to read expensive art books and those whose taste is for cheap novels. Nor is there any reason to regard this as unfair (Cohen 2004: 12). The question this raises is whether the equal benefit principle is suitable as a guide to public good provision under all circumstances. What if a society’s ethos is less communitarian, with the result that its members are less willing to cross-​subsidize those with more expensive tastes for public goods? A possible response here is that public good supply will in any case be radically reduced in such a society. Miller had previously argued that the level of trust necessary for public good schemes to be successfully upheld would be lacking in societies whose members were not united by a common national identity (1995: 91). Thus, the equal benefit principle will provide guidance in most cases where justice is at stake. Nonetheless, we ought to be able to say something about what justice requires in public goods provision where communal motivations are lacking. Consider, for example, the supply of global public goods when the contributors

9 

Also note that Betty’s tastes were relatively expensive to begin with. She required two public goods (art galleries and opera) at a cost of thirty-​five to benefit by forty, while Adam only needed one (television) at a cost of fifteen. 10  A public library is a public good for all those eligible to use and borrow from it (e.g. all residents in a particular city). It is not strictly non-​rival since the very book I want to read may have been taken out by someone else, but for the most part my use of the library does not diminish the opportunities of others to enjoy its amenities.

Public Goods   571 come from a number of distinct national communities. Trust, fostered by a common national identity, is certainly not the main motivating factor inducing the supply of these goods. Instead, production usually comes about for reasons of self-​interest (Barrett 2007). For example, countries with more resources and more to gain may take it upon themselves to supply a global public good unilaterally. This may be seen as unfair: just because one country is keen enough to provide a good that also benefits other countries by itself, it does not follow that the remaining countries do not act unfairly in free-​riding on the first country’s efforts. Yet if we want to establish what a fair contribution would be, we cannot rely on the reasoning given here to support the equal benefit principle, since subsidizing expensive tastes in the way that the principle may require would be objectionable in this context. Ought we therefore to look for a different approach?

26.5.3 The Hypothetical Insurance Market We have already discussed Stuart White’s use of an insurance model to determine which public goods are required by justice. Our worry was that it appeared capable of licensing more than the provision of goods that we would normally consider essential. But perhaps the same model can help to explain what distributive justice requires with respect to the costs and benefits of discretionary public goods, whose supply is not itself a matter of justice. White’s model, to recall, invites us to consider what, on average, the members of a society would choose to do by way of purchasing public goods through taxation if they were ignorant of their particular needs but knew about the general incidence of needs in their society. But now we are no longer considering public goods that aim to meet basic needs, but rather public goods that aim to meet simply the preferences of individuals. The natural extension, then, is to suggest that public goods should be supplied if they would be bought by the average member of society if ignorant of their personal preferences. What schedule of public goods, and associated costs, would be chosen? The choosers would be guided by considerations of expected utility. They would know the costs of providing each public good being contemplated, and they would also know how likely it was that they would turn out to have a preference (and how strong the preference would be) for that particular good. Maximizing expected utility would require them to choose public goods that were sufficiently cheap and widely preferred so as to make it more likely that they would benefit, on balance, from having the good supplied. The problem here, however, is that the choosers, under conditions of ignorance, would have no reason to connect benefitting from a good to contributing to its costs of production. Since they are only concerned about average levels of benefit, they would not be worried if some people had to make large contributions towards goods from which they benefitted little or not at all. White does not give any precise answer as to how costs are to be distributed, but it is clear that he takes it that this should be

572    David Miller and Isaac Taylor independent of benefit received. On pragmatic grounds, he argues that, since we cannot know how much each individual benefits from public goods, we cannot charge them on the basis of their use (2003: 104–​5). But while the data that would tell us how much each person in society benefits individually from particular public goods is indeed likely to be impossible to collect, this does not mean that we should not take benefit into account when choosing a distributive principle for costs. We may not be able to implement the resulting principle perfectly, but it would at least provide guidance. Here, for example, are some plausible conjectures about the distribution of benefits between different groups: the rich benefit more than the poor from public art galleries; residents of a city benefit more than residents of the surrounding countryside from a public park within the city; car-​owners benefit more than non-​car owners from motorways. We can make these judgements without knowing exactly how much each member of these groups benefits. In all of these cases, it is relatively easy to find ways of linking together benefit and cost: in the first case, we could pay for art galleries from the proceeds of progressive taxation; in the second case, we could pay for public parks predominantly through a city-​wide tax rather than a county-​or state-​level tax; and in the third case we could pay for motorways predominantly through a road tax paid by car users, rather than more general taxes on income or consumption. The fact that costs and benefits cannot be perfectly aligned does not mean that we should abandon the principle entirely when thinking about what distributive justice requires in this area. Separating costs from benefits in the way White proposes also leads to some further counter-​intuitive implications. Reverting to our original Adam/​Betty example, imagine that they are joined by a third individual, Chris, who does not benefit from any discretionary public goods (he only cares about consuming private goods). According to the hypothetical insurance market approach, however, he might still be required to contribute to the costs of both art galleries and television. The average individual in the hypothetical insurance market does not know whether she will have the preferences of Adam, Betty, or Chris, but since supplying both art galleries and public television as public goods maximizes her expected benefit, she would opt for both of these.11 If all three individuals had the same income, the just distribution of costs and benefits on this approach would be as shown in Table 26.4. In this case, Chris is being forced to pay for a package of public goods even though he does not receive any benefit from it. As we suggested earlier, it might be reasonable to require him to pay for essential public goods, even if he turns out not to benefit from them, since he is under an obligation to contribute towards his fellow-​citizens’ basic needs being met and to help uphold justice. But when the goods being supplied are discretionary, this sort of argument seems implausible. Why should he contribute to public goods that are not necessary to secure justice and that do not provide for basic needs? By wholly severing the amount one is required to contribute towards 11  The expected benefit of supplying both would be 10; of supplying TV only, 8.3; of supplying art galleries only, 1.7; and of supplying nothing, 0.

Public Goods   573 Table 26.4  Chris joins society Adam

Betty

Chris

Benefit

40

20

0

Cost

10

10

10

Net benefit

30

10

–10

public goods from the benefit one receives, the hypothetical insurance market is an implausible mechanism for determining a just distribution of costs and benefits of discretionary goods.

26.6 Conclusion In this chapter, we have tried to accomplish three things. First, we have argued that the compulsory provision of public goods poses serious problems of distributive just­ ice that the existing literature fails to address. Liberal egalitarian theories such as those of Rawls, Dworkin, and Walzer all resort to procedural solutions to the problem, for example simply allowing democratic majorities to vote for the public goods that they prefer. Meanwhile libertarian theories that set the bar higher by requiring unanimous consent have difficulty in explaining how any public goods can legitimately be supplied at all. Second, we have proposed that any plausible application of principles of distributive justice to public goods needs to begin by drawing a distinction between essential and discretionary goods. The former are those goods that justice itself requires us to provide, and that citizens can therefore be required to contribute to as a matter of justice. Yet there are different ways of cashing out ‘essential’ here, and we have not defended a particular account. Further work is needed to decide whether, for example, a public good can be interpreted as essential when it provides the only feasible way of meeting individuals’ basic needs, or when it is the most efficient way. Third, we have examined several principles that have been proposed to govern the supply of discretionary goods. Although these goods are not demanded by justice itself, if they are to be supplied the resulting distribution of benefits and costs brings principles of distributive justice into play. We considered three possibilities:  that costs should be made proportional to benefits received, that levels of benefit net of costs should be equalized, and that the overall cost-​benefit package should be such as would on average be chosen in a hypothetical insurance market. Each of these principles has strengths, but also some counter-​intuitive implications in particular cases. Once again the topic is one that stands badly in need of further research.

574    David Miller and Isaac Taylor

Acknowledgements We would like to thank Serena Olsaretti for very helpful written comments on this chapter.

References Barrett, S. (2007). Why Cooperate? The Incentive to Supply Global Public Goods. Oxford: Oxford University Press. Casal, P. (1999). ‘Environmentalism, Procreation, and the Principle of Fairness’. Public Affairs Quarterly 13(4): 363–​76. Coase, R. H. (1960). ‘The Problem of Social Cost’. The Journal of Law and Economics 3(1): 1–​44. Coase, R. H. (1974). ‘The Lighthouse in Economics’. The Journal of Law and Economics 17(2): 357–​76. Cohen, G. A. (1989). ‘On the Currency of Egalitarian Justice’. Ethics 99(4): 906–​44. Cohen, G. A. (2004). ‘Expensive Taste Rides Again’, in J. Burley (ed.) Dworkin and His Critics: With Replies by Dworkin. Oxford: Basil Blackwell, pp. 3–​29. Cullity, G. (1995). ‘Moral Free Riding’. Philosophy & Public Affairs 24(1): 3–​34. de Jasay, A. (1989). Social Contract, Free Ride: A Study of the Public Goods Problem. Oxford: Oxford University Press. De-​Shalit, A. (1995). Why Posterity Matters. London: Routledge. Dworkin, R. (1986). A Matter of Principle. Oxford: Clarendon Press. Dworkin, R. (2000). Sovereign Virtue:  The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Gaus, G. (2011). The Order of Public Reason: A Theory of Freedom and Morality in a Diverse and Bounded World. New York: Cambridge University Press. Klosko, G. (1992). The Principle of Fairness and Political Obligation. Lanham, MD: Rowman & Littlefield. Light, A. (2000). ‘Public Goods, Future Generations and Environmental Protection’, in A. Anton, M. Fisk, and N. Holmstrom (eds) Not for Sale: In Defense of Public Goods. Boulder, CO: Westview Press, pp. 209–​25. Mill, J. S. (1965). ‘Principles of Political Economy with Some of their Applications to Social Philosophy’, in J. M. Robson (ed.) Collected Works of John Stuart Mill, Vols II–​III, Toronto: University of Toronto Press. Miller, D. (1991). ‘Public Goods without the State’. Critical Review 7(4): 505–​23. Miller, D. (1995). On Nationality. Oxford: Oxford University Press. Miller, D. (1999a). ‘Social Justice and Environmental Goods’, in A. Dobson (ed.) Fairness and Futurity:  Essays in Environmental Sustainability and Social Justice. Oxford:  Oxford University Press, pp. 151–​72. Miller, D. (1999b). Principles of Social Justice. Cambridge, MA: Harvard University Press. Miller, D. (2004). ‘Justice, Democracy, and Public Goods’, in K. Dowding, R. E. Goodin, and C. Pateman (eds) Justice and Democracy: Essays for Brian Barry. Cambridge: Cambridge University Press, pp. 127–​49. Murphy, L. and Nagel, T. (2002). The Myth of Ownership: Taxes and Justice. Oxford: Oxford University Press. Nozick, R. (1974). Anarchy, State, and Utopia. Oxford: Basil Blackwell. Olsaretti, S. (2013). ‘Children as Public Goods?’ Philosophy & Public Affairs 41(3): 226–​58.

Public Goods   575 O’Neill, J. (1992). ‘The Varieties of Intrinsic Value’. The Monist 75(2): 119–​37. Parfit, D. (1995). ‘Equality or Priority?’ The Lindley Lectures. Lawrence, Kansas: University of Kansas. Quong, J. (2011). Liberalism without Perfection. Oxford: Oxford University Press. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (1993). Political Liberalism. New York: Columbia University Press. Samuelson, P. A. (1954). ‘The Pure Theory of Public Expenditure’. The Review of Economics and Statistics 36(4): 387–​9. Schmidtz, D. (1991). The Limits of Government:  An Essay on the Public Goods Argument. Boulder, CO: Westview. Smith, A. (1976). An Enquiry into the Nature and Causes of The Wealth of Nations, ed. R. H. Campbell, A. S. Skinner and W. B. Todd. Oxford: Clarendon Press. Simmons, A. J. (1979). Moral Principles and Political Obligations. Princeton, NJ:  Princeton University Press. Simmons, A. J. (2001). Justification and Legitimacy:  Essays on Rights and Obligations. Cambridge: Cambridge University Press. Sylvan, R. (2003). ‘Is There a Need for a New, an Environmental, Ethic?’,  in A. Light and H. Rolston (eds) Environmental Ethics: An Anthology III. Malden, MA: Blackwell, pp. 47–​52. Taylor, I. (2013). ‘Practices, Institutions, and Global Public Good Regimes’. Raisons Politiques 51: 121–​35. Walzer, M. (1983). Spheres of Justice:  A Defence of Pluralism and Equality. Oxford:  Martin Robertson. White, S. (2003). The Civic Minimum: On the Rights and Obligations of Economic Citizenship. Oxford: Oxford University Press. Wicksell, K. (1958). ‘A New Principle of Just Taxation’, in R. A. Musgrave and A. T. Peacock (eds) Classics in the Theory of Public Finance. London: MacMillan, pp. 72–​118.

Chapter 27

Cultural and Re l i g i ou s M inori t i e s Peter Jones

Very many societies nowadays are religiously diverse. Some are conspicuously multifaith and also include many people who profess no faith. Others have a clearly predominant faith but contain minorities who subscribe to other faiths or none, and the predominant faith may itself be divided by differences of doctrine or tradition. The same pattern of diversity characterizes the cultural make-​up of most modern societies. Indeed, differences are often simultaneously cultural and religious. These differences readily spawn questions of distributive justice. How, for example, should religious freedom, or the freedom to live according to a culture, be distributed amongst a population? Should a society be scrupulously egalitarian in its treatment of religions and cultures or can it justly privilege one over others? What sort of distribution is genuinely equal? Should it ignore differences in the demands made by people’s cultural or religious allegiances, or should it be sensitive and responsive to them? If a group has fewer social and economic opportunities than others because of its religious or cultural commitments, is that a price it should pay for those commitments or an injustice that calls for remedy? A number of things complicate the answers to these questions. One is that religions and cultures do not come in neatly symmetrical forms. They each have features peculiar to themselves which can make it difficult to know what it is to treat their adherents fairly in relation to others. Another is that religious or cultural practices that harm or violate the rights of others have no claim to just accommodation; but whether a practice is indeed harmful or rights-​violating can be controversial. A third is that an intrinsically unobjectionable religious or cultural practice can conflict with a law or public policy that has a genuinely non-​discriminatory purpose; we have then to decide which should override the other. A fourth is that justice need not be the only value at stake. A society may, for example, have reason to fear that accommodating religious and cultural diversity beyond a certain point will threaten its cohesion and stability; we may therefore find ourselves having to take account of considerations other than justice.

Cultural and Religious Minorities    577

27.1  Religion and Culture If we are to secure justice amongst religious and cultural groups, we have to identify which groups count as religious or cultural. The First Amendment of the US Constitution provides that Congress shall make no law prohibiting the free exercise of religion; hence the scope of ‘religion’ is crucial to what the Amendment protects (Greenawalt 2006: 124–​56). In the great majority of cases, groups claiming to be religious locate themselves within one of the world’s major faiths that are, by common consent, religions. But there can be difficult cases, and the US Supreme Court has sometimes been willing to construe religion in a way that includes nontheistic conscientious belief.1 European jurisdictions have also grappled with the issue of religion’s definition (Ahdar and Leigh 2013:  139–​55), though they sometimes avoid it by following the European Convention on Human Rights (ECHR) in protecting ‘religion or belief ’. If we adopt that formula, it need not matter that a body of belief sits on the boundary between religion and other forms of belief since, even if it is not religious, it can qualify for protection as a ‘belief ’. However, that formula simply relocates the issue of definition since we have then to decide what constitutes ‘belief ’ within the meaning of the formula. ‘Culture’ is an even more inclusive and open-​ended term. In one usage, people’s culture encompasses everything that belongs to the milieu in which they exist and that contributes to their conception of the world and their form of life. That usage has merit, but it is too inclusive for our purposes. Our concern is with the cultural differences associated with ‘multiculturalism’. Will Kymlicka has usefully distinguished between two sorts of multicultural society: multination and polyethnic (1995: 10–​33). Kymlicka identifies a nation as ‘a historical community, more or less institutionally complete, occupying a given territory or homeland, sharing a distinct language and culture’ (1995: 11). Belgium and Switzerland are multination states. So are societies that include indigenous peoples, such as Australia and New Zealand. Multination societies have incorporated national minorities through colonization or conquest or more or less consensual acts of confederation. Polyethnic societies, by contrast, have become culturally diverse through immigration. The US has long been polyethnic. Many European societies, such as the UK, France, Germany, and the Netherlands, have become significantly polyethnic more recently. Canada and Australia are both polyethnic and multination societies. In this chapter, we shall consider only polyethnic multiculturalism. The ‘cultural’ in ‘cultural minority’ will therefore describe the different sorts of belief, custom, and way of life that migrant groups have brought with them from elsewhere and bequeathed to subsequent generations. The cultures within which people actually conduct their lives are, of course, much more complex and intercultural than that, since people of migrant 1 

United States v. Seeger, 380 U.S. 163 (1965); Welsh v. United States, 398 U.S. 333 (1970).

578   Peter Jones heritage both influence and are influenced by the societies to which they migrate. In addition, everyone is now prey to influences stemming from all parts of the globe. But, as long as significant cultural differences persist, so will the issue of how a society should respond to them. Religious and cultural differences are often heavily overlapping, especially in polyethnic societies. For many groups, religion is central to their culture, but their culture has also shaped the tenets, institutions, and practices of their religion. Indeed, it frequently makes no sense to separate the two. On the other hand, religious and cultural differences are not co-​extensive. The same religion can straddle different cultures, as do Christianity, Islam, and Buddhism for example, and religious differences can exist within a single culture. Equally, cultural differences typically include differences that are non-​religious, such as language and literature, or religious only in part, such as diet, dress, and social mores. But, even though culture and religion are not wholly identical in range, the issues to which they give rise in polyethnic societies are often the same or similar and, in what follows, they will often, though not always, be considered together.

27.2  Religious and Cultural Freedom It is now widely accepted that a society should accord all of its citizens equal freedom to pursue whatever religious faith they embrace, including the freedom to pursue none. If its government promoted the faith of some and prohibited that of others, it would act unjustly. Indeed, we might suppose that there is no need to single out religious liberty for special attention; we might conceive it as merely one aspect of the equal freedom people should have to conduct their lives as they see fit. That general freedom will also entitle them to conduct their lives in accordance with their culture, either alone or in association with others. Consider, for example, how religious liberty figures in Rawls’s theory of justice (1999, 2005). His theory is intended to govern a society’s basic structure. Its two principles provide that citizens should enjoy equal basic freedoms and that resources should be distributed according to the difference principle. Citizens should then be at liberty to use their freedoms and resources to pursue whatever conception of the good they wish, religious or non-​religious. Rawls does not single out cultural differences for special mention, but the basic structure within which those are manifested and pursued might also be governed by his two principles. A number of factors complicate this simple template for equal religious and cultural liberty. One is that people should not be free to engage in just any religious or cultural practice merely because it is religious or cultural. The principle of equal freedom itself entails that some may not use their freedom to infringe the freedom of others; they may not, for example, impede the practice of religions they deem false. In addition, a society can justifiably forbid practices that are wrongful or harmful in other ways. This, however, can be more difficult territory. Different religions or cultures frequently incorporate

Cultural and Religious Minorities    579 different beliefs about what constitutes proper and improper conduct. Those different beliefs constitute part of the very diversity for which a multifaith or multicultural society is supposed to provide. So should not those differences evoke mutual toleration? Toleration does indeed belong to the realm of conflicting beliefs and practices. We tolerate only that to which we object; we do not tolerate that to which we take no exception. Different religions are conflicting religions, which is why mutual toleration has been crucial to achieving peaceful co-​existence in religiously diverse societies. Mutual toleration may also be needed and appropriate for some non-​religious cultural differences, although people may find large swathes of one another’s cultural differences entirely unobjectionable or positively admirable. Consider, for example, the literatures, arts, and cuisines of different cultures. However, toleration becomes of doubtful merit when it permits the harming or wronging of third parties. Then the ‘costs’ of toleration are borne by someone other than the tolerators. Some classic examples of cultural practices, which Western societies generally deem intolerably harmful, are honour killing, forced marriage, and female clitoridectomy (or ‘female genital mutilation’). But the issue of toleration is not entirely absent from third-party cases. Consider male circumcision. Some deem this intolerable because harmful; others believe it harmless; still others hold that, while it may inflict limited harm, its practice within Judaism and Islam should be tolerated, given its significance for those faiths. A similar trio of positions exists on ritual slaughter. There is no simple way of reaching consensus on issues of this sort, though we should neither exaggerate the number of issues that actually attract dissensus, nor overlook the dissensus that frequently exists within cultural or religious groups. The debate is about what should be restricted, rather than whether anything should be restricted. Few of us are willing to stomach unlimited toleration or to accept that distributive justice requires it. A further complication is the wide range of matters that need to be addressed when a society moves from being largely mono-​faith and monocultural to being multifaith and multicultural, especially in a relatively short period of time. For example, if chaplaincies have been publicly funded for hospitals and prisons, the range of religions for which they provide will need to be extended to reflect the changed make-​up of the society. Efforts may be needed to ensure that new minorities are adequately represented in the society’s governmental institutions and in its public services such as the police. Educational curricula may also need adjustment. If the society has publicly funded faith schools, equitable treatment will require it to make similar provision for faiths newly present in the society. Healthcare may need delivery with a new sensitivity to cultural differences, and so on. These are not always readily describable as matters of religious or cultural ‘liberty’, but they belong in the same domain insofar as they concern the scope people enjoy to function in a society as citizens with specific religious or cultural allegiances. The measures through which a society should provide for the diverse cultural and religious make-​up of its population are often described as policies of ‘accommodation’. If a multi-​ faith or multicultural society is to treat its minorities justly, it must accommodate, rather than remain blind to, their differences.

580   Peter Jones On the other hand, how much and what sort of accommodation there should be is often highly controversial. How far, for example, should school curricula adapt to new minorities, and how far should new minorities adapt to a society’s established pattern of education? The very existence of publicly funded faith schools is controversial but so has been the funding of particular faith schools. In Britain, state-​funding for Muslim schools was long ​delayed, partly because of concerns about the quality of female education they would deliver, and partly because of fears that they would exacerbate social divisions in ways that other faith schools generally do not (with the chastening exception of Northern Ireland). Translation services may seem to be instruments of accommodation, but may be curtailed in the belief that cajoling minorities into learning the society’s majority language is good for both them and the society. For the most part, policies of accommodation do not accord privileged or ‘morethan-equal’ treatment to minorities. Indeed, taken in the round, they often provide less-than-equal treatment in the belief that it is not unfair that, in some degree and in some respects, the society’s majority culture should remain its prevailing culture. For example, in polyethnic societies it is generally accepted that the society’s only public language should be its majority language, at least for most purposes; or, if a society is both multinational and polyethnic, its national or indigenous minorities may be accorded language rights that its migrant minorities are not. Sometimes accommodation does, however, take the form of special treatment and it is that form that has attracted the most debate within political philosophy.

27.3  The Issue of Exemption The most conspicuous instances of this form of accommodation are exemptions. These allow some citizens, for reasons of religion or culture, not to comply with laws or rules with which others have to conform. In many societies, motorcyclists and construction workers are legally required to wear safety helmets. Those laws create problems for devout Sikhs who believe they are religiously obliged not to substitute other headgear for their turbans; turban-​wearing Sikhs are often therefore exempted from safety helmet legislation. Their faith also requires them to carry a kirpan (a ceremonial dagger or sword) and, for that reason, Sikhs are also often exempt from laws prohibiting the carrying of knives in public. Orthodox Jews and some Muslims believe animals slaughtered for meat must be killed without first being stunned, which conflicts with the animal welfare legislation of many societies. Some countries, such as the UK and Germany, exempt Jews and Muslims from that legislation so that they can practise ritual slaughter; others, such as France, Denmark, and Norway, do not. In the US, subscribers to a Native American religion, whose ceremonies involve the smoking of peyote (an intoxicating drug derived from a cactus plant) are exempt from laws prohibiting the non-​medical use of drugs. The Amish are exempt from mandatory high school education laws and

Cultural and Religious Minorities    581 from a host of other public regulations that are at odds with their religiously based way of life. Over 2,000 US statutes are reckoned to include these sorts of religious exemption (Ryan 1992: 1445). Sometimes the beneficiaries of exemptions are institutions rather than individuals (Shorten 2015). In the UK and the US, organized religions (churches, mosques, synagogues, etc.) are exempt from laws forbidding discrimination in employment on grounds of gender and sexual orientation. If they were not, some, such as the Roman Catholic Church, would face a conflict between their doctrinal and legal obligations. Organized religions also frequently benefit from tax exemptions. Exemptions may be granted by legislatures, but they may also fall within the competence of judiciaries. In the US, courts will exempt a claimant from a law or rule that they judge, all things considered, infringes his constitutional right to free exercise. In European jurisdictions, courts can take similar action if they judge a law or rule to infringe a claimant’s human right to manifest his religion or belief (ECHR Article 9). Given the special nature of exemptions, some argue that they are best decided upon by democratically accountable legislatures (Barry 2001:  321). US Supreme Court judges have sometimes taken that view (McConnell 2000: 2). Others suggest that judiciaries are better equipped to do the job, since exemptions are often claimed by relatively powerless minorities who are more likely to receive a fair hearing from a court (Greene 2009: 1001–​2). The laws or rules from which people seek exemption are not usually deliberate efforts to deprive a group of its religious or cultural freedom. If they were, they could be readily identified as unjust, unless the freedoms they removed were sufficiently morally objectionable to justify prohibition. They are typically laws that have a bona fide non-​discriminatory purpose, such as reducing the risk of injuries or mitigating animal suffering or curbing drug abuse. It just so happens that, in pursuit of that purpose, the law intersects with a group’s cultural or religious practice and renders it either impermissible or unusually costly. The clash is most acute when, in the absence of an exemption, people would be unable to comply with norms of their religion or culture. If, for example, Orthodox Jews were required to attend court on a Saturday, and the Catholic Church were required to make women eligible for the priesthood, they would be compelled to flout the requirements of their faith. Often, however, a law, rather than rendering compliance with a group’s norms impossible, exacts a price for its continued compliance. In the absence of an exemption, the devout Sikh would have to give up motorcycling and construction work, and the Orthodox Jew would have to give up eating meat. That sort of cost is normally said to ‘burden’ a person’s religious liberty. How are we to evaluate exemptions? Do they privilege religious and cultural practices in ways that are unjust? Or do they secure distributive justice by removing inequalities in the way different groups are affected by their society’s public and private arrangements? Before turning to those questions, we need to be aware of two distinct sorts of distribuend that may be an exemption’s concern.

582   Peter Jones

27.4  Two Sorts of Distribuend The distribuend that relates most immediately to people defined by religion or culture is their freedom to pursue their religion or culture. Sometimes, however, a different distribuend can be at issue—​people’s opportunities to do things other than practise their religion or culture, but which might be impaired because of their religion or culture. When, for example, people suffer religious discrimination in employment, the primary injustice they suffer is the loss of an employment opportunity rather than a loss of religious freedom. In the US and most European countries, the distinction between these two sorts of distribuend is marked by the difference between two bodies of law. In the US, people’s freedom of religion is protected by their constitutional right to free exercise, and in European jurisdictions by their human right to manifest their religion or belief. The discrimination laws of those countries have a different purpose: they aim to prevent people from being unfairly denied opportunities because of their religious or cultural identity,2 in matters such as education and employment. When we turn to legal exemptions, which of these distribuends is the dominant concern is sometimes clear and sometimes not. When organized religions are exempted from discrimination laws relating to gender and sexual orientation, the dominant concern is clearly their religious freedom. By contrast, the exemption of Sikhs from motorcycle helmet legislation matters more obviously for their ability to ride a motorcycle than for their freedom to practise their religion. But, because the loss of a non-​religious good can be both a bad in itself and a ‘burden’ upon the loser’s religious freedom, it is often quite impossible to discern which distribuend drives the case for an exemption. It could, of course, be both. In what follows, I shall consider exemptions on the assumption that they aim primarily to preserve religious or cultural freedom and to avoid burdening that freedom. That, however, is merely a working assumption; I do not mean to exclude the possibility that the other sort of distribuend may sometimes assume primacy. If it does, that will affect the kind of argument that justifies (or fails to justify) the exemption.

27.5  Exemptions and Equality Even though exemptions provide for people to be treated differently, they are most commonly defended as instruments of equality. That defence frequently points to the

2  Unlike religion, ‘culture’ is commonly not a named characteristic protected by discrimination law, unsurprisingly given its indefinite scope. However, it can receive a degree of legal protection insofar as it overlaps with religion or belief or falls within the domain of race or ethnicity.

Cultural and Religious Minorities    583 different conditions of majorities and minorities. Typically, a society’s arrangements will have been constructed to suit its historical majority. For example, Christianity has shaped the structure of the working week and the timing of public holidays in most Western societies. Those arrangements are out of kilter with the practices and festivals of other faiths, and exemptions provide a way of reducing their disadvantageous impact upon minorities (Kymlicka 1995: 108–​15; Nussbaum 2008). Mitigating inequalities amongst religions does not, however, deal with another possible inequality: differences in the way that public policy impacts upon religious and non-​religious forms of life. Christopher Eisgruber and Lawrence Sager (2007) argue that a just society will secure equal liberty for the pursuit of spiritual and secular commitments and projects. Legislators and policy-​makers should treat those committed to religious forms of life with neither favour nor disfavour but with ‘equal regard’. But how does equal regard justify exemptions for the religious? Eisgruber and Sager’s answer is that spiritual interests are unusually vulnerable to hostile or neglectful treatment by governments. Exemptions remove unfair disadvantage and restore equal liberty to the religious. In another egalitarian defence, David Miller (2002) argues that equality of opportunity can require more than equal access to the opportunities that are already present in a society. It requires a match between opportunity sets and the aims and aspirations present amongst a population. If different people wish to engage in different sports, such as football, swimming, and hockey, equality of opportunity is not satisfied by providing football facilities only, even if those facilities are equally available to all. Different sports need to be provided for roughly in proportion to the demand for them. Similarly, if through immigration a society comes to include religious or cultural groups whose aims and ambitions are new to it, equalizing opportunities requires the accommodation of those aims and ambitions, which may require exemptions. Jonathan Quong also defends exemptions as instruments of equal opportunity. The just Rawlsian society, he argues, would afford all of its citizens ‘a reasonable or adequate chance to pursue their conception of the good’ (Quong 2006:  58). Rules that deny minorities the same opportunities as the majority to combine their cultural or religious commitments with basic civic opportunities, such as employment and education, violate fair equality of opportunity. Consider the following case (Quong 2006: 62–​6). Jeff, a liberal Protestant, and Jonah, an Orthodox Jew, both wish to become police officers. The police force operates a rota that requires all officers to work for some weekends each month. That rota deprives Jonah of the opportunity to join the police, since, as an Orthodox Jew, he is unable to work on Saturdays. It also means that Jeff and Jonah have unequal opportunity sets: Jeff can combine his religious beliefs with service as a police officer, while Jonah cannot. In these circumstances, justice, Quong argues, requires the police to exempt Jonah from Saturday working. That exemption will not privilege Jonah, since he gains no advantage not already enjoyed by Jeff and other police officers. If Jonah’s faith were accommodated in a privileged fashion (e.g. he was allowed to work for fewer hours than others but for the same salary), fair equality of opportunity would veto that accommodation.

584   Peter Jones One question that may affect our thinking on exemptions and equality is whether we should hold people responsible for their religious beliefs or cultural commitments. That issue is at the core of a dispute between Bhikhu Parekh and Brian Barry. Parekh is another theorist who holds that equal opportunity requires a society to accommodate the diverse religious and cultural commitments of its population so that no group is better or worse off than others (2006: 261). Opportunity, he argues, is a ‘subject-​dependent concept’ in that a facility, a resource, or a course of action is only a mute and passive possibility and not an opportunity for an individual if she lacks the capacity, the cultural disposition, or the necessary cultural knowledge to take advantage of it’ (2006: 241). An opportunity that people can take up only by jettisoning their religion or culture is really no opportunity at all: A Sikh is in principle free to send his son to a school that bans turbans, but for all practical purposes it is closed to him. The same is true when an orthodox Jew is required to give up his yarmulke, or the Muslim woman to wear a skirt, or a vegetarian Hindu to eat beef as a precondition for certain kinds of jobs.

Barry (2001: 32–​8), by contrast, holds that a person’s opportunity set is an entirely objective matter. If people are subject to the same rules, they possess the same opportunity sets. Different beliefs and preferences may make some less inclined than others to take up an opportunity, but they still possess that opportunity. Without an exemption, the devout Sikh has the opportunity to ride a motorcycle; he will simply opt not to take it up. Likewise, in the absence of ritual slaughter, the Orthodox Jew has the same opportunity as others to eat meat even though he will choose not to. On Barry’s view, we can reasonably require people to assume responsibility for the costs of complying with their religious or cultural commitments. Laws may burden people unequally but Barry denies that they must be unfair in doing so. A law prohibiting rape is felt more keenly by potential rapists; banning smoking in confined spaces inconveniences smokers more than others; prohibiting drink-​driving frustrates the toper but not the teetotaller. None of those laws is unfair because it ‘burdens’ some more than others. Similarly, a law whose impact is felt more keenly by those who possess particular religious or cultural commitments is not unfair merely for that reason (Barry 2001: 34–​5, 40–​1). Neither Parekh nor Barry claims to be a luck egalitarian but the dispute between them clearly resonates with that position.3 We are more likely to agree with Barry if we conceive people’s religious or cultural commitments as self-​assumed rather than as consequences of circumstance and fate. That issue takes us into difficult territory and Miller and Quong would question our need to enter it. Their arguments seek to establish fair background conditions for the pursuit of religious or cultural commitments whatever their genesis. But there is another dimension to Barry’s argument that they, and 3 

Barry expressly denies that his position is based on the chance/​choice distinction; Barry 2002: 215–​ 19, responding to Mendus 2002.

Cultural and Religious Minorities    585 others, can less easily avoid: why should it matter more or differently if a law frustrates a religious or cultural commitment rather than an ordinary preference? Consider a modified version of Quong’s example. Suppose that Jeff too does not want to work on Saturdays but his reason, unlike Jonah’s, is not religious. He wants to devote his Saturdays to his family or to playing football for his local team. Can Jeff then claim, like Jonah, that, if he has to work on Saturdays, his opportunity-​set will be inferior to those of others so that he too should receive an exemption? If not, we have to find reason why Jonah’s religious commitment should matter for fair equality of opportunity while Jeff ’s family or sporting commitment should not.

27.6  Privileging and Balancing We might, of course, deny there is any such reason (Bedi 2007; Cornelissen 2012). Or we might follow Eisgruber and Sager (2007: 78–​120) in attempting to justify exemptions as instruments that give no special status to religion but merely protect it from intended or unintended discrimination. However, it is doubtful whether mere non-​discrimination can justify all the exemptions they believe to be justified (Koppelman 2006:  574–​83; Laborde 2014; White 2012). Of those who hold that religion should have a privileged status, some find the justification in religion’s being a distinct and ultimate good; that provides reason enough for its special treatment (Garvey 1996; Koppelman 2006:  McConnell 2000). That claim is, however, deeply controversial, especially as a claim that the state should endorse (Gedicks 1998; Rawls 2005). It is also extravagant in deeming any religion to be good merely in virtue of being a religion and without reference to its particular content. A more broadly based case for privileging religious and cultural claims appeals to the idea of identity. Ideas of identity and its recognition have assumed growing significance in the modern world (Appiah 2005; Parekh 2008; Patten 2000), and exemptions are sometimes conceived as instruments for according recognition to religious and cultural identities (Galeotti 2002; Quong 2002, 2006; Seglow 2003). However, those particular identities do not always matter more to people than identities of other sorts, even though some may think they should, and it seems odd to make identity-​recognition the primary purpose of exemptions. There is also an incongruity in holding that what really matters about a religious faith is the identity it bestows upon those who adhere to it (Jones 1999). The most promising sort of justification for religious or cultural exemptions focuses on the norm-​governed nature of the conduct for which they provide. This justification is often formulated in the language of conscience: we have reason, albeit prima facie, neither to compel people to act contrary to conscience nor to burden their consciences, even inadvertently. If we use that language, we do best to understand ‘conscience’ generously so that it encompasses all religious or moral motivations, rather than only those

586   Peter Jones that conform to a peculiarly Protestant notion of conscience. Conscience will normally encompass both religious and non-​religious convictions, but that can count in its favour. Why should conscience matter? For Martha Nussbaum (2008: 164–​74), ‘conscience’ describes the faculty with which we search for ultimate meaning, and that faculty needs a protected space within which to function. However, that makes conscience seem more epistemic than normative in significance and a ‘search for meaning’ does not describe how many adherents relate to their faith. Other commentators stress the social reality of non-​legal norms. Rather than being merely individual convictions, cultural norms are always, and religious norms are commonly, rooted in communities. Waldron (2002:  24) emphasizes the way in which community-​based norms are socially enforced as part of an established way of life and Miller (2002: 53) stresses the social costs that people can incur if they defy their community’s norms. Social sanctions may indeed attach to norms, but that is not invariably true, and it seems at best an ancillary consideration:  we should take people’s conscientious convictions seriously even if they are not subject to community coercion. Conscience more obviously matters because it matters to people that they should do right and not do wrong. Subscribing to a religious faith typically includes being subject to a set of prescriptions. As Brian Leiter (2013: 34, 92–​133) observes, it is the ‘categoricity’ of religious commands that most obviously grounds their claim to accommodation, although categoricity can yield an equivalent claim for conscientious non-​religious convictions. For those who believe themselves subject to categorical prescriptions, non-​ compliance is not an option. They may remain capable of defying and be legally free to defy those prescriptions, but religiously or morally they are not. In that sense, the Sikh is not free to cut his hair, the Muslim to ignore the Islamic obligations of prayer, the Sabbatarian to work on the day of rest, and the non-​religious doctor to perform an abortion he believes wrong. Conscientious commitments generate reasons for conduct that are different in kind from those manifesting tastes and preferences, and the experience of being made to do wrong or burdened for doing right is quite different from that of having a preference disappointed (Jones 2015a). That is how Jonah’s objection to Saturday working differs from Jeff ’s. That may satisfy us that conscientious convictions have, prima facie, a claim to accommodation. Or we may prefer to make conscientiousness part of a larger story about the demands of human integrity and well-​being (Bou-​Habib 2006), or the social bases of self-​respect (Shorten 2010), or the respect to which people are entitled as persons (Leiter 2013: 69–​91). Argument of this sort may seem less well suited to claims of culture, but in polyethnic societies it is the prescriptive elements of culture, rather than culture indifferently, that are taken seriously as candidates for exemption. That is why the cultural practices that receive exemption are so frequently religious practices as well. Unsurprisingly, not everyone is persuaded by the claims of conscience. Some object that subsuming the claims of religion within those of conscience must result in religion’s receiving less than its due (Koppelman 2009; Trigg 2012), and others that privileging claims of conscience over claims of other sorts is arbitrary and unfair (Arneson 2010).

Cultural and Religious Minorities    587 Even if there is reason to treat claims of religion or culture (or some of them) differently from claims of other sorts, the case for a particular exemption has still to be made. That is because it competes with other legitimate concerns, most obviously the purpose of the law from which exemption is sought. That purpose may be sufficiently pressing to override the case for exemption and, even if it is not, that case can remain heavily contingent upon circumstances. If ritual slaughter were discovered to inflict greater animal suffering than currently thought, the case for its exemption could easily disappear. So too could the case for permitting the carrying of kirpans if Sikhs began putting them to violent use. Since 9/​11, through no fault of their own but reasonably nonetheless, Sikhs have generally been required to surrender their kirpans during airflights. Another factor is the extent to which an exemption will relieve some of a burden only to shift it onto others; if, for example, we exempt the conscientious objector from military conscription, someone else has to take his place (Leiter 2013: 99). US federal law now provides that government may substantially burden the free exercise of religion only in furtherance of ‘a compelling governmental interest’,4 but the circumstances that can bear on exemptions are too varied for any such simple formula to be of much use. Weighing is also necessary on the other side of the equation. The degree to which a law adversely affects a group’s religious or cultural life obviously matters. So does the centrality or obligatoriness of the affected practice for a faith or culture. Allowing legislators and judges to inquire into such matters, especially when they are religious, is thought improper by some (Trigg 2012: 116–​17), but, if a group wants an exemption, it cannot reasonably expect them to remain unexamined (Greene 2009: 998–​1,000). If assessing the case for an exemption requires this sort of complex balancing of competing considerations, is it really a matter of distributive justice?

27.7  Accommodation and Non-​C omparative Justice The idea of distributive justice that is most commonly invoked in this context is comparative in nature. Whether the good at stake is distributed justly amongst a population turns upon how much each person holds relative to others. Sometimes, however, justice can take a non-​comparative form. If, for example, a person possesses a right to religious freedom but is denied that right by his government, that denial is unjust and 4 

Religious Freedom Restoration Act 1993 (RFRA). The test of ‘compelling state interest’ was especially associated with the Supreme Court’s judgement in Sherbert, and its sidelining many years later with the Court’s judgement in Smith; Sherbert v. Verner, 374 U.S. 398 (1963); Employment Division v. Smith, 494 U.S. 872 (1990). Following Smith, Congress reinstated the test through RFRA, though the Supreme Court subsequently ruled RFRA unconstitutional for states and localities, though not for federal law. For the history and critical discussions of the test, see Eisburger and Sager 2007: 240–​76, and Hamilton 2005: 203–​37.

588   Peter Jones non-​comparatively so. We do not need to know how the government treats others before we can know that it has treated the person unjustly. (On non-​comparative justice, see Feinberg 1974; Hoffman 1993; Olsaretti 2004: 33–​7.) Distributive and non-​comparative justice are not mutually exclusive. A  person or group can simultaneously enjoy both or neither. If, for example, a government denies some individuals a religious freedom that it grants to others, those individuals may suffer non-​comparative injustice, in being denied a freedom to which they are entitled, and distributive injustice, in being allowed less freedom than others. Nevertheless, the distinction between the two is significant. In particular, what the proper make-​up of a good that is to be distributed amongst a population is, is quite different from the question of how it should be distributed amongst them. When the religious or cultural demands of some clash with the public rules or private arrangements of others, the relevant issue may be, not how one person’s freedom compares with another’s, but what precisely should be the scope of the freedom to which a person (any person) is entitled. Consider, for example, the kind of decision that a court has to make when it faces a claim that someone’s human right to freedom of religion, as laid down in the ECHR, has been violated. That could be the claim of a Christian Sabbatarian whose employer requires him to work on Sundays;5 or a Muslim pupil whose school will not allow her to wear a jilbab rather than the shalwar kameez approved by the school’s uniform policy;6 or a Hindu whose local authority will not allow his body to be cremated by open pyre when he dies.7 In judging such cases, the court has to take account not only of the first clause of Article 9 of the ECHR, which asserts everyone’s freedom ‘to manifest his religion or belief, in worship, teaching, practice and observance’, but also of its second clause, which subjects that freedom to ‘such limitations as are prescribed by law and are necessary in a democratic society in the interests of public safety, for the protection of public order, health or morals, or for the protection of the rights and freedom of others’. It has to decide upon the scope of the religious freedom to which any human being has a right in all of the circumstances that bear on the case. Making that decision is an exercise in mapping non-​comparative justice. The freedom at issue here is the object of a human right and is therefore a freedom to which all are equally entitled. To that extent, the issue of non-​comparative justice is set within a context of equal distribution. But that notion of equality cannot determine the precise scope of the religious freedom to which a person (any person) is entitled. Similarly, when US courts determine the scope of the ‘free exercise’ to which people are entitled, their decision has to be one of non-​comparative justice, at least in the first instance. Once a number of such decisions are in place, a court may decide a case by comparing it with others and trying to ensure that the parties in that case are treated similarly to those in previous cases (cf. Eisgruber and Sager 2007:  78–​120). But the 5 

Copsey v. WWB Devon Clays Ltd [2005] EWCA Civ 932. Begum v. Denbigh High School [2006] UKHL 15. 7  Davender Kumar Ghai v. Newcastle City Council [2009] EWHC 978 (Admin); [2010] EWCA Civ 59. 6 

Cultural and Religious Minorities    589 court’s decision will then be parasitic upon other decisions that were non-​comparative in nature and will take for granted that those other decisions were non-​comparatively just. The same logic applies to the efforts of law-​makers when they set about deciding the rightful domain of religious liberty. If we drill down still further, we may begin to question how much the idea of justice, even in its non-​comparative form, contributes to determining the proper scope of religious freedom. As we have seen, a court in determining whether an individual’s human right to religious freedom has been infringed has to take account of the considerations listed in the second clause of Article 9 of the ECHR relating to various public goods and the rights and freedom of others. In human rights jurisprudence, a claimant’s interest in religious freedom vies with those competing considerations, but his human right does not. Rather, he has a human right only to so much religious freedom as the court deems justified, taking into account all of the relevant considerations. So the precise details of the religious freedom to which a claimant has a human right are determined by the balance of considerations that the court assesses. The human right itself and the demands of non-​comparative justice are outcomes rather than contributors to that balancing process. The same logic applies when legislators balance competing considerations in deciding whether a particular exemption would be justified all things considered. Of course, we can say that the balancing process should be conducted fairly: it should be conducted impartially and should give due weight to all of the interests at stake. A person might therefore intelligibly complain that a court or legislature has denied him the religious freedom to which he is entitled because it has weighed his interests too lightly. But for both decision-​makers and complainants the critical matter is the balance of interests that determines the proper scope of the right. These remarks are not intended to imply that ideas of justice, either distributive or non-​comparative, are inconsequential for religious or cultural freedom. As we have seen, there are many issues relating to those freedoms for which principles of justice are straightforwardly determinative. But, in the case of exemptions, those principles frequently present us with a question rather than an answer.

27.8  Accommodation and Discrimination Law We saw earlier that efforts to accommodate religious and cultural commitments can aim to protect two different sorts of good: either (i) the enjoyment of religious or cultural freedom; or (ii) the opportunity to enjoy goods other than religion or culture. I now want to consider how justice relates to the second sort of good. A legal exemption may aim to secure a good of that sort, and some of the arguments for exemption cited previously, particularly those of Quong and Parekh, could be mobilized in its defence. Here,

590   Peter Jones however, I want to focus more narrowly on the goods typically protected by discrimination law: employment and access to goods and services. The basic demand of justice in this area is straightforward. If we are committed to treating people equally, their opportunities should be unprejudiced by their religious or cultural affiliations. It would be unjust, for example, if employment as a doctor or an engineer were conditional upon a person’s faith or if publicly provided healthcare were made differently available to citizens of different cultures. Restricting an opportunity to a specific religion or culture is, of course, sometimes justified. It is not unjust that only Catholics should be eligible for the Catholic priesthood, and it may not be unjust that recruitment for social work targeted at people of a particular culture should be limited to people of that culture. But, for the general run of jobs and services, people’s religion or culture will be of no relevance and should not prejudice their access to them. In this area, the form of justice pre-​eminently at issue is distributive justice. In section 27.7, we saw how we might map out the religious freedom to which a person (any person) is entitled, and go on to characterize a denial of that freedom as non-​comparatively unjust. The ideas of non-​discrimination and equal opportunity do not similarly imply a specific set of non-​religious or non-​cultural opportunities that should be available to all, and deprivation of which would be non-​comparatively unjust. Rather they hold that the employment and service opportunities present in a society, whatever they happen to be, should be available on fair terms to all, irrespective of religious or cultural allegiance. Even so, discrimination can be non-​comparatively as well as distributively unjust. Consider a simple case of direct discrimination. An employer, A, refuses to employ an applicant, B, because B is a Muslim. A’s refusal is an act of religious discrimination because she treats B less favourably than she treats or would treat those of other faiths or none. The phrase ‘less favourably’ may suggest that the injustice B suffers is necessarily comparative, but, treating B ‘less favourably’ can be the property that identifies A’s act as an act of discrimination rather than the feature that describes its wrongness. The injustice that B suffers consists in A’s denying him employment because of his religious affiliation, and we can identify A’s refusing to employ B for that reason as unjust without reference to how A treats others. Even in this case, however, we can also identify A’s discriminatory act as distributively unjust. A treats B unjustly, not only in denying him an opportunity, but also in denying him equal opportunity, without good reason. Distributive justice most frequently provides the impetus for anti-​discrimination measures. Some groups, such as racial or religious minorities, are more likely than others to suffer discrimination, and the main concern of legislators in prohibiting discrimination has been to safeguard the equality of opportunity of those groups. The distinction between direct and indirect discrimination is crucial for the issue of accommodation. Law prohibiting direct discrimination does not require employers and providers to engage in accommodation, but law prohibiting indirect discrimination does. Indirect discrimination arises when an employer or provider has a rule or practice whose effect, though not its intention, is to disadvantage one group relative to others. If, for example, a factory-​owner requires her employees not to have facial hair,

Cultural and Religious Minorities    591 she discriminates indirectly against male Sikhs whose faith requires them not to cut their hair. The rule effectively bars Sikhs from employment at the factory, and indirect discrimination law requires the employer to exempt Sikhs from it. If, however, she can show that her rule is justified in spite of its discriminatory effect, she will not normally be held guilty of indirect discrimination as a legal offence. If, for example, she can persuade a court that her production process requires strict standards of hygiene that are incompatible with facial hair, she will not be required to accommodate unshaven Sikhs.8 Unlike direct discrimination, indirect discrimination is not usually deliberate. (Arguably, if it were deliberate, it would be direct discrimination.) It arises as the side-​ effect of a rule or practice that has been put in place for a bona fide non-​discriminatory reason. That gives it a different ‘moral feel’ from direct discrimination, but it can still be unjust. If A’s rule or practice deprives B of an employment opportunity without good reason, B suffers an injustice, even though the injustice reflects no malice on A’s part. And, as with direct discrimination, the injustice B suffers can be both non-​comparative and distributive. In other respects, however, indirect discrimination differs significantly from direct, and raises issues closely akin to those of legal exemption. First, indirect discrimination law requires employers and providers to accommodate people’s racial and religious characteristics insofar as they can do so without significantly frustrating their own endeavours. It may require employers and providers to set aside the indirectly discriminatory practice or to exempt members of the disadvantaged group from it. There is an element of privileging here in that typically only a limited number of characteristics, usually including race and religion, are protected in that way. Jonah’s wish not to work on Saturdays for religious reasons would be protected, but Jeff ’s identical wish, grounded in family or footballing reasons, would not. Second, indirect discrimination law, like legal exemption, involves a trade-​off between competing interests. At some point, an employer’s or provider’s obligation to accommodate gives way to the justified pursuit of her own purpose. Where that point should be set is open to argument. In the US the legal test is ‘business necessity’, and in the UK it is ‘proportionate means of achieving a legitimate aim’. The cross-​over point can also vary in form. It can be either a fixed threshold or a more flexible balancing of the interests at stake in each case.The duty of accommodation arising from law on religious and racial discrimination differs from legal exemption in falling, not on the state or society at large, but on particular members of civil society: employers and providers of goods and services. These can be large impersonal organizations, but they can also be small outfits or single individuals. That too may affect our thinking on how much accommodation justice requires (Jones 2015b). Direct and indirect discrimination law are both sites of another sort of accommodation: exemption of the religious from discrimination law. For doctrinal reasons, religious organizations and individuals sometimes want to discriminate in relation to characteristics protected by discrimination law. That is uncontroversial when a church

8 

This example is based on Panesar v. The Nestle Company Ltd [1980] IRLR 64 CA.

592   Peter Jones or mosque discriminates on grounds of religion in appointing a priest or imam; it becomes more controversial if the church or mosque wants to discriminate religiously in appointing, for example, a caretaker or a gardener. But the religious sometimes want, and are permitted, to discriminate on other grounds. In the UK, for example, organized religions are allowed to discriminate in employment on grounds of gender, sexual orientation, gender reassignment, and marital status, insofar as that discrimination is required by their doctrines or to avoid conflict with the strongly held convictions of a significant number of the religion’s followers. In matters other than employment, ‘organizations relating to religion or belief ’ can take account of sexual orientation in imposing restrictions on their membership, participation in their activities, choosing to whom they provide goods and services, and in the use and disposal of their premises, subject to the same conditions concerning doctrine or strongly held convictions.9 Religious individuals are not similarly exempt. Quite how the competing claims of religious belief and sexual orientation should be reconciled has become one of the most heated issues of discrimination law (Malik 2013).

27.9  Equality, Establishment, and Recognition A final subject of distributive justice is the relative standing that different religions and cultures enjoy in a society. As well as upholding free exercise, the US Constitution provides that ‘Congress shall make no law respecting an establishment of religion’. Non-​establishment less clearly has a distributive dimension than free exercise. Its most obvious aim is to prevent the state’s becoming an instrument of religious oppression. That purpose may, of course, be an adjunct to free exercise; historically, states in establishing a religion have usually curtailed their citizens’ freedom to pursue other religions. Nowadays, however, several European societies, such as England, Scotland, and Denmark, retain established churches, and yet afford their citizens freedom to pursue other faiths on terms roughly equal with the established religion. Moreover, the zeal with which non-​establishment has been maintained in the US cannot be explained wholly by the demands of free exercise. The distributive objection to a state’s establishing a religion is that, in doing so, it violates the equal status and respect it should accord its citizens. In giving privileged status to a religion, the state gives privileged standing to those who subscribe to it. In the US, non-​establishment has been understood to require strict state neutrality on religious matters, but quite what that neutrality demands with respect to prayers and ‘moments of 9  Equality Act 2010, schedule 3, para. 29; schedule 9, paras 2 and 3; schedule 23, para. 2. See Sandberg 2011: 100–​30, for further details and discussion of religious exemptions relating to discrimination law in Britain, and Minow 2007, for an overview and discussion of these issues as they have arisen in the US.

Cultural and Religious Minorities    593 silence’ in schools, the location of nativity scenes and religious monuments, reference to God in the Pledge of Allegiance, and a host of other matters, has been fiercely contested (Greenawalt 2009; Nussbaum 2008: 224–​353). Many other broadly ‘liberal’ societies also subscribe to the general idea of state neutrality, but implement it less thoroughly than the US. Even those without established churches often have public institutions in which the legacy of Christianity, or one of its forms, remains visible. Frequently, however, those legacies are sufficiently minor and anodyne not to be felt as injustices by those who subscribe to other faiths. Some religious minorities even favour continued establishment in the belief that it provides a degree of public recognition for religion in general; disestablishment, they fear, would be less a victory for religious equality than for secularism (Modood 1994). As that view indicates, nowadays contests in this area are often between religion and secularity rather than different religions. The French tradition of laïcisme, and the controversies concerning hijabs and burkas to which it has given rise, is but one example of that contest (Laborde 2008). If we turn to the equivalent issue for culture, justice is often said to require equal recognition for cultures (e.g. Taylor 1994). As Colin Bird explains (Chapter 11, this volume), recognition is a multifaceted and controversial idea (see also McBride 2013; Thompson 2006). If we should accord equal recognition to cultures, what sort of recognition should that be? Perhaps the most obvious answer is status-​recognition: in recognizing people as persons or citizens of equal status, we should recognize their cultures as those of persons or citizens of equal status and treat them accordingly. That sort of recognition does not require us to evaluate cultures and find them to be of equal value. It is not the merit of a culture but the status of its bearers that does the work. But recognition of that sort is often deemed inadequate. According to Charles Taylor, equal recognition demands that we recognize ‘the equal value of different ways of being’ (1992: 51). It requires ‘that we all recognize the equal value of different cultures; that we not only let them survive, but acknowledge their worth’ (1994: 64, emphases in the original). ‘Value’ or ‘worth’ here equates with merit, so that we are called upon to recognize cultures as of equal merit. That is altogether more problematic. The beliefs and values of one culture often conflict with those of another, so that requiring people to recognize others’ cultures as of equal merit with their own can make little sense. (Compare the incoherence of requiring religious adherents to recognize other religions as equal in value with their own.) Taylor acknowledges that we currently lack the cross-​cultural criteria needed to make a comparative assessment of cultures, and also that there is something odd in making equal merit-​recognition a matter of moral obligation, but he still suggests that ‘equal value’ should be our working presumption (1994: 66–​73). Yet, if it really were feasible to compare and evaluate the many different cultures humanity has evolved, it seems most unlikely that we would discover all to be of equal merit (Jones 2006). Even if we limit ourselves to status recognition, it is not clear that a society always can or should give equal status to people’s cultures. Unless we adopt an unusually narrow conception of ‘culture’, comprehensive neutrality amongst cultures, unlike neutrality

594   Peter Jones amongst religions, will often not be feasible for a polyethnic society. While a society can do much to enable its population to live according to their different cultures, it cannot shuffle every aspect of culture out of the public domain and into communal ghettoes; nor can it represent cultures equally in all aspects of its public life, such as its system of law. Nor is this just a matter of practicability. If a minority of Poles migrate to Greece and a minority of Greeks migrate to Poland, we would not think it incumbent on each society to reinvent itself so that its culture becomes Polish and Greek in equal measure. Some asymmetry between the culture of the host society and the cultures of its migrant minorities is generally deemed acceptable as well as unavoidable. Kymlicka, for example, argues that, in opting to migrate, people voluntarily relinquish the rights they possessed as members of their original national culture. They do not forfeit all cultural rights, and their new society should offer them fair terms of integration, but they have no right to a separate societal culture nor to the cultural integrity and self-​government that an indigenous minority might claim (Kymlicka 1995: 94–​100, 2001: 152–​76). How to provide for all of its diverse cultural groups in a manner that is fair, whilst simultaneously integrating them into a society possessing a genuine sense of commonality, is the major challenge facing contemporary multicultural societies.10

Acknowledgements I am grateful to Ian O’Flynn and Serena Olsaretti for their helpful comments on earlier versions of this chapter.

References Ahdar, R. and Leigh, I. (2013). Religious Freedom in the Liberal State, 2nd edn. Oxford: Oxford University Press. Appiah, K. A. (2005). The Ethics of Identity. Princeton, NJ: Princeton University Press. Arneson, R. J. (2010). ‘Against Freedom of Conscience’. San Diego Law Review 47(4): 1015–​40. Barry, B. (2001). Culture and Equality. Cambridge: Polity Press. Barry, B. (2002). ‘Second Thoughts—​and Some First Thoughts Revised’, in P. Kelly (ed.) Multiculturalism Reconsidered. Cambridge: Polity Press, pp. 204–​38. Bedi, S. (2007). ‘What is so Special about Religion? The Dilemma of the Religious Exemption’. Journal of Political Philosophy 15(2): 235–​49. Bird, C. (2018). ‘The Theory and Politics of Recognition’, in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 235–55. Bou-​Habib, P. (2006). ‘A Theory of Religious Accommodation’. Journal of Applied Philosophy 23(1): 109–​26. Cornelissen, G. (2012). ‘Belief-​Based Exemptions: Are Religious Beliefs Special?’ Ratio Juris 25(1): 85–​109. 10 

For a sample of views on how that challenge should be met, see Miller 1995; Modood 2007; Parekh 2006.

Cultural and Religious Minorities    595 Eisgruber, C. L., and Sager, L. G. (2007). Religious Freedom and the Constitution. Cambridge, MA: Harvard University Press. Feinberg, J. (1974). ‘Noncomparative Justice’. Philosophical Review 83(3): 297–​338. Galeotti, A. E. (2002). Toleration as Recognition. Cambridge: Cambridge University Press. Garvey, J. H. (1996). ‘An Anti-​ Liberal Argument for Religious Freedom’. Journal of Contemporary Legal Issues 7(2): 275–​92. Gedicks, F. M. (1998). ‘An Unfirm Foundation: The Regrettable Indefensibility of Religious Exemptions’. University of Arkansas Little Rock Law Journal 20(3): 555–​74. Greenawalt, K. (2006). Religion and the Constitution, vol. I:  Free Exercise and Fairness. Princeton, NJ: Princeton University Press. Greenawalt, K. (2009). Religion and the Constitution, vol. II:  Establishment and Fairness Princeton, NJ: Princeton University Press. Greene, A. (2009). ‘Three Theories of Religious Equality . . . and of Exemptions’. Texas Law Review 87(5): 963–​1007. Hamilton, M. A. (2005). God vs. the Gavel: Religion and the Rule of Law. Cambridge: Cambridge University Press. Hoffman, J. (1993). ‘A New Theory of Comparative and Noncomparative Justice’. Philosophical Studies 70(2): 165–​83. Jones, P. (1999). ‘Beliefs and Identities’, in J. Horton and S. Mendus (eds) Toleration, Identity and Difference. London: Macmillan, pp. 65–​86. Jones, P. (2006). ‘Equality, Recognition and Difference’. Critical Review of International, Social and Political Philosophy 9(1): 23–​46. Jones, P. (2015a). ‘Liberty, Equality, and Accommodation’, in T. Modood and V. Uberoi (eds) Multiculturalism Rethought. Edinburgh: Edinburgh University Press, pp. 126–​56. Jones, P. (2015b). ‘Belief, Autonomy and Responsibility:  The Case of Indirect Religious Discrimination’, in G. B. Levey (ed.) Autonomy, Authenticity, Culture. London: Routledge, pp. 66–​86. Koppelman, A. (2006). ‘Is it Fair to Give Religion Special Treatment?’ University of Illinois Law Review 2006(3): 571–​603. Koppelman, A. (2009). ‘Conscience, Volitional Necessity, and Religious Exemptions’. Legal Theory 15(3): 215–​44. Kymlicka, W. (1995). Multicultural Citizenship: A Liberal Theory of Minority Rights. Oxford: Clarendon Press. Kymlicka, W. (2001). Politics in the Vernacular. Oxford: Oxford University Press. Laborde, C. (2008). Critical Republicanism:  The Hijab Controversy and Political Philosophy. Oxford: Oxford University Press. Laborde, C. (2014). ‘Equal Liberty, Non-​Establishment and Religious Freedom’. Legal Theory 20(1): 52–77. Leiter, B. (2013). Why Tolerate Religion? Princeton, NJ: Princeton University Press. McBride, C. (2013). Recognition. Cambridge: Polity Press. McConnell, M. W. (2000). “The Problem of Singling Out Religion.” De Paul Law Review 50(1): 1–​47. Malik, M. (2013). ‘Religion and Sexual Orientation:  Conflict or Cohesion?’,  in G. D’Costa, M. Evans, T. Modood, and J. Rivers (eds) Religion in a Liberal State. Cambridge: Cambridge University Press, pp. 67–​92. Mendus, S. (2002). ‘Choice, Chance and Multiculturalism’, in P. Kelly (ed.) Multiculturalism Reconsidered. Cambridge: Polity Press, pp. 31–​44.

596   Peter Jones Miller, D. (1995). On Nationality. Oxford: Oxford University Press. Miller, D. (2002). ‘Liberalism, Equal Opportunities and Cultural Commitments’, in P. Kelly (ed.) Multiculturalism Reconsidered. Cambridge: Polity Press, pp. 45–​61. Minow, M. (2007). ‘Should Religious Groups be Exempt from Civil Rights Laws?’,  Boston College Law Review 48(4): 781–​849. Modood, T. (1994). ‘Establishment, Multiculturalism and British Citizenship’. Political Quarterly 65(1): 53–​73. Modood, T. (2007). Multiculturalism: A Civic Idea. Cambridge: Polity Press. Nussbaum, M. (2008). Liberty of Conscience:  In Defense of America’s Tradition of Religious Liberty. New York: Basic Books. Olsaretti, S. (2004). Liberty, Desert and the Market. Cambridge: Cambridge University Press. Parekh, B. (2006). Rethinking Multiculturalism: Culture Diversity and Political Theory, 2nd edn. Basingstoke: Palgrave Macmillan. Parekh, B. (2008). A New Politics of Identity. Basingstoke: Palgrave Macmillan. Patten, A. (2000). ‘Equality of Recognition and the Liberal Theory of Citizenship’, in C.  McKinnon and I. Hampsher-​ Monk (eds.) The Demands of Citizenship. London: Continuum, pp. 193–​211. Quong, J. (2002). ‘Are Identity Claims Bad for Deliberative Democracy?’  Contemporary Political Theory 1: 307–​27. Quong, J. (2006). ‘Cultural Exemptions, Expensive Tastes, and Equal Opportunities’. Journal of Applied Philosophy 23(1): 53–​7 1. Rawls, J. (1999). A Theory of Justice, revised edn. Cambridge, MA: Belknap Press. Rawls J. (2005). Political Liberalism, expanded edn. New York: Columbia University Press. Ryan, J. E. (1992). ‘Smith and the Religious Freedom Restoration Act:  An Iconoclastic Assessment’. Virginia Law Review 78(6): 1407–​62. Sandberg, R. (2011). Law and Religion. Cambridge: Cambridge University Press. Seglow, J. (2003). ‘Theorising Recognition’, in B. Haddock and P. Sutch (eds) Multiculturalism, Identity and Rights. London: Routledge, pp. 78–​93. Shorten, A. (2010). ‘Cultural Exemptions, Equality and Basic Interests’. Ethnicities 10(1): 100–​26. Shorten, A. (2015). ‘Are There Rights to Institutional Exemptions?’ Journal of Social Philosophy 46(2): 242–​63. Taylor, C. (1992). The Ethics of Authenticity. Cambridge, MA: Harvard University Press. Taylor, C. (1994). ‘The Politics of Recognition’, in A, Gutmann (ed.) Multiculturalism. Princeton, NJ: Princeton University Press, pp. 25–​73. Thompson, S. (2006). The Political Theory of Recognition. Cambridge: Polity Press. Trigg, R. (2012). Equality, Freedom and Religion. Oxford: Oxford University Press. Waldron, J. (2002). ‘One Law for All: The Logic of Cultural Accommodation’. Washington and Lee Law Review 59(1): 3–​34. White, S. (2012). ‘Religious Exemptions: An Egalitarian Demand?’ Law and Ethics of Human Rights 6(1): 97–​118.

Chapter 28

L anguag e Alan Patten

28.1 Introduction Linguistic diversity can be found in every society in the world. This is obvious in major cosmopolitan cities, where speakers of hundreds of different languages live in close proximity. It is also apparent in countries such as Canada, Switzerland, Belgium, Spain, India, South Africa, and Nigeria, where more than a quarter of the population are members of historically rooted national minorities that speak a language different than the majority’s. In the United States, although English is predominant, about a fifth of the population report using a language other than English in their homes. Whenever more than one language is in use, people face the problem of how to communicate with one another. An important and fascinating challenge for social scientists is to understand how people work out this communication problem. What are the determinants of linguistic behavior under conditions of diversity? Why, for instance, do people acquire some languages and not others? And why do some languages rather than others come to be used in particular domains of communication? As social scientists start to get a grip on this kind of question, it is natural for a second more directly practical sort of question to arise. If the first kind of question concerns explanation, this second one is about policy. What, if anything, can be done to shape or influence the linguistic habits and practices of a particular society? What instruments do public policy actors have at their disposal for shaping linguistic practice, and how can they most effectively use these instruments to accomplish whatever language-​ related objectives they may have? These questions could, in principle, arise for a range of different actors, but they arise most obviously for the state. Many citizens in a democratic state will have preferences about societal patterns of language use and will consider using the state as a tool for advancing these preferences. So what can the state accomplish? Given what we know about the determinants of linguistic behavior, what impact can the instruments of state policy have on language practices?

598   Alan Patten As soon as one formulates these questions about policy, however, it immediately becomes apparent that there is a third kind of question that also needs to be asked: a normative question. The policy question is focused only on means. It asks about what the state can hope to achieve with the various instruments at its disposal. By their very nature, however, we do not care about means for their own sake: we care about them as instruments for advancing certain ends. So what are these ends? What ends would it be appropriate for the state to pursue, and what other sorts of standards (besides ends) ought to guide the state’s policies relating to language? And if the short answer to these questions is that, in a democracy, the state should pursue the ends and obey the constraints that its citizens authorize, then what ends and constraints would it be appropriate for citizens—​for “we the people”—​to authorize? A fully satisfying answer to the normative question is not possible without grappling with the explanatory and policy questions. “Ought” implies “can,” and in any practical context our views about what “can” be done have to be compatible with a plausible understanding of how people actually behave linguistically, of how they are likely to respond to various incentives, penalties, processes of habituation and socialization, and so on. In this sense, the three kinds of questions are interconnected and one cannot make defensible normative judgments in isolation from a consideration of explanatory processes and policy options. The point of departure for the present discussion, however, is that “can” does not imply “ought.” Even once we get clear about the set of things that can be done, a question remains about which of them should be done. It is this normative problem that is at the heart of recent work on language rights by political philosophers.1 When others address the normative problem, a number of different goals are sometimes posited as being especially relevant to evaluating language policies. Many sociolinguists assume that preserving vulnerable languages and maintaining linguistic diversity are the appropriate goals of language policy. Political scientists highlight goals such as political stability and national unity, while economists focus their attention on economic efficiency and growth, public good provision, and other economic outcomes. For political philosophers, while all of these ends are important, there is one standard that is arguably most important of all. A  society that preserved linguistic diversity, or that achieved political stability or high levels of economic growth, would not have succeeded at what is most important if did not also realize justice. From the standpoint of political philosophy, the fundamental standard for evaluating different ideas and proposals about language policy is what Rawls has called the “first virtue of social institutions”—​the achievement of justice (Rawls 1999). But what does political philosophy have to say about the demands of justice with respect to language—​about “linguistic justice” as we might call it? In the small but growing 1  Kymlicka and Patten 2003 is a collection of essays by political philosophers about the normative issues surrounding language policy and language rights. Other treatments of this topic include Green 1987; Kymlicka 1989b, 1995, 2001; Réaume 1991, 1994, 2000; Taylor 1992; Carens 2000; Van Parijs 2000, 2011; Patten 2001, 2003, 2009, 2014; De Schutter and Ypi 2012.

Language   599 literature on this topic, it is possible to find two main approaches to this question. One framework of analysis emphasizes the instrumental importance of language for distributive justice. From this perspective, although language policies and patterns of language use are not themselves a matter of distributive concern, these or other linguistic facts may be consequential for the distribution of that which does matter for justice. The other framework attaches at least some non-​instrumental importance to language. From this point of view, part of what makes a distribution just is that it appropriately attends to the interests that people have in the use, the success, and/​or the treatment of their languages. In this contribution, I’ll rather briefly discuss some issues arising in the instrumental framework and then turn to a more detailed engagement with the non-​instrumental framework. In a concluding section, I’ll reflect, again quite briefly, on how the concerns of the two frameworks might be integrated with one another. Throughout the discussion, the policy implications of the various claims about linguistic justice will be kept in view. In particular, we shall consider whether justice supports a nation-​building policy, in which the state encourages everyone in a society to converge on a single national language, or whether it lends support to a minority rights policy, in which public services are offered, and public business is conducted, in more than one of the languages spoken in the society.

28.2  The Instrumental Framework An instrumental account of linguistic justice brackets the question of whether language is itself an object of distributive concern. It posits a language-​independent conception of distributive justice, and then explores the ways in which language diversity might help or hinder the realization of distributive justice so conceived. Obviously, a satisfactory discussion of such an account would need to specify the language-​ independent conception in enough detail to allow the relevant causal mechanisms to be identified. I will not attempt to justify any particular conception here, but will assume, for the purposes of discussion, that distributive justice requires some degree of equality of opportunity and social mobility. I also suppose that distributive justice, whatever exactly it consists in, normally rests on a willingness of persons to accept certain burdens for the sake of the claims of others. These may be burdens of taxation and redistribution, or of democratic deliberation, or of some combination of these and other expectations. Many different causal mechanisms might plausibly be expected to connect linguistic diversity and distributive (in)justice. In what follows, I draw attention to a few of the most important of these mechanisms. Some arise from linguistic diversity itself; others are triggered by familiar policy responses to linguistic diversity. The first four mechanisms hinge on the further idea that, despite its differences, a linguistically diverse society may still have a dominant language. This is the language spoken by most

600   Alan Patten people and used most of the time as a lingua franca by native speakers of different languages. It is the language in which much of the economy operates, and in which other social institutions and practices are typically conducted. A linguistically diverse society with a dominant language is a society in which at least some people speak as their first language a language other than the dominant one. From an instrumental perspective, this leads to several challenges for distributive justice: (1) Exclusion through lack of proficiency in the dominant language. If some members of language minorities lack proficiency in a society’s dominant language, they will be excluded from many economic, social, and cultural opportunities. (2) Exclusion of language minorities through insistence on dominant-​language instruction. Concerned to avoid situations in which minorities lack proficiency in the dominant language, the state may insist that all schools provide instruction in the dominant language. But this may lead to its own problems if non-​native speakers of the dominant language struggle to achieve literacy and other skills in an instructional setting of this kind. Lackluster development of these important skills can also lead to exclusion from opportunities. (3) Exclusion through opportunity-​ hoarding by dominant-​ language speakers.2 Language differences may provide a salient coordination point around which a group of people can mobilize to provide for themselves a greater share of social benefits and opportunities, with a lesser share left for “out-​groups.” Since the linguistic differences that give rise to this pattern of exclusion may be fairly superficial in character, such as accent, surname, etc., a mechanism of this kind may diverge in its implications from a mechanism based on a lack of proficiency in the dominant language. Note too that language-​related opportunity-​hoarding may occur, not just between, but also within language groups. Variations in accent and dialect based on region, class, and ethnicity may align with other factors to produce objectionable inequalities. (4) Exclusion through segregation by public policy. In some situations, minority language status correlates with socioeconomic status: native speakers of a minority language tend disproportionately to be disadvantaged socioeconomically. In such a case, efforts to serve minority speakers in their own language may (perhaps unintentionally) end up reinforcing disadvantage rather than relieving it. For a variety of reasons, public education and public services that mainly cater to disadvantaged people are often associated with undesirable outcomes. Other causal pathways that generate an instrumental relationship between language and distributive justice depend less on the presence of a dominant language and more on the fact that people have an identity based on their (first) language. They identify themselves as members of a group of speakers of the language and/​or are

2 

On “opportunity hoarding,” see Anderson 2010.

Language   601 so identified by others in their society; and they value, or are assumed to value, the language and the language group. Under conditions of linguistic diversity, these actual or presumed attachments can present several different challenges for distributive justice: (5) Social solidarity weakened by linguistic diversity. Some political theorists believe that the realization of distributive justice depends on the presence of strong ties of social solidarity grounded in a widely shared sense of national identity or political community (Miller 1995, Van Parijs 2000, 2011, Kymlicka 2001). Linguistic differences might make it more difficult for the requisite sense of common nationality to form and be sustained. (6) Social solidarity weakened by attempts to impose majority language onto minorities. When language minorities feel attached to their language and language community, they may resent attempts by the state to promote a common national language through policies that subordinate or marginalize their own language. In this scenario, unequal treatment makes it difficult for minorities to join in the sense of common nationality needed to sustain distributive justice. For anyone interested in promoting distributive justice, the empirical issues raised by these different causal mechanisms are quite complicated. A major challenge is that the mechanisms seem to pull in different directions. Mechanisms (1), (4), and (5) would seem to recommend a nation-​building approach to language policy. To prevent exclusion and to support the emergence of a common national identity, policymakers should seek to promote a common national language. Combined with other policies (anti-​discrimination, anti-​racism, the promotion of inclusion as an aspect of national identity, etc.) this approach offers a recipe for engaging with mechanism (3) as well. The successful diffusion of a common national language should tend over time to lessen the differences that make language-​related opportunity-​hoarding possible and should generally diminish the salience of language as a basis for identity politics. On the other hand, nation-​building would not help with, and would in fact exacerbate, mechanisms (2) and (6). Aggressive attempts to diffuse a common national language through language immersion programs risk leaving minorities behind in their development of critical skills such as literacy, numeracy, and so on (mechanism (2)). And the nation-​building project may stir up resentment amongst minorities who perceive it as an obstacle to the survival and flourishing of their own language community (mechanism (6)). Many people will look at these different factors and conclude that, on balance, the nation-​building strategy is the best one to follow over the long run. As was just noted, mechanisms (1), (3), (4), and (5) are broadly supportive of this hypothesis. Mechanism (2) pulls in the opposite direction but its pull will often be quite weak. Even if mechanism (2) is operative, there may be little to gain from a distributive justice perspective in substituting a minority language-​based education for a majority language one. If most of the society’s opportunities are available only in a dominant-​language medium, then

602   Alan Patten minorities have little choice but to learn the dominant language, even if the learning process leaves them trailing behind in other areas. In light of mechanism (2), it might make sense to offer bilingual instruction on a transitional basis, but the medium-​term goal of such a policy is the nation-​builder’s one of convergence on a common public language. Mechanism (6) might also be rather weak in many contexts. A  nation-​ building project that is fairly successful at promoting a common national language may be resented by minorities at first. But, if subsequent generations are treated inclusively, there is reason to hope that the effects predicted by mechanism (6) will subside over time. Advocates of the nation-​building approach point out, as evidence for their approach, that many societies that do a satisfactory job of realizing distributive justice possess a common national language. Some of these societies may never have been terribly diverse linguistically in the first place (in part because of restrictive immigration policies), but this is not true in an important range of cases. France is the classic example of a society that was characterized by substantial linguistic diversity (around the time of the French Revolution) but which, over the course of the century that followed, consciously and very successfully set out to diffuse French as the common language of the country. Unlike France, the United States has never officially designated its dominant language as the national language, but it has been no less insistent at publicly privileging English over minority languages. Of course any general argument for nation-​building on instrumental grounds ought to come with significant caveats. Some language minorities are able to generate a fairly wide range of opportunities within their own communities. To the extent that this is true, mechanisms (1), (3), and (4) become less of a threat to the realization of distributive justice, while mechanism (2) may be more pronounced in its effects (greater numbers of minority-​speaking families will lack proficiency in the majority-​language and will suffer under an educational regime that insists on majority-​language instruction). More generally, the nation-​building strategies that worked well in the nineteenth and early twentieth centuries may be less successful today faced with language minorities possessing their own institutions and public culture, and having a strong sense of identity and attachment to their own language. Under conditions where a strong sense of national identity based on minority language attachment has been mobilized, mechanism (6) may be very powerful indeed. Still, with some appropriate caveats, it would not be terribly wrong to align the instrumental framework with the nation-​building approach to language policy. This makes it all the more important to explore whether the instrumental framework is the full story about the relationship between language and justice. Is language of relevance to distributive justice in an instrumental way only, or are there other respects in which language matters to justice as well? The remainder of the contribution is devoted to answering this question. We shall see that there is indeed a non-​instrumental dimension to linguistic justice, and that, once this dimension is appreciated, any preference for nation-​building needs further qualification.

Language   603

28.3  A Non-​Instrumental Framework: The Neutrality Model I begin with a simple proposal about how language might be considered an object of distributive concern in its own right. On this simple proposal, language matters non-​instrumentally because it is something that people care about for a variety of reasons that do not reduce back to an instrumental connection with distributive justice. The appropriate way for the state to accommodate this fact about people’s preferences is to adopt a stance of neutrality. When the state is neutral, it refrains from deciding which language-​ related preferences will be successful or unsuccessful, and instead offers support to all of them in an evenhanded fashion. In section 28.4, we will consider some other accounts of language’s non-​instrumental significance and ask whether they add anything essential to the neutrality model. One alternative argues that treating language as a mere preference—​something that people care about—​fails to appreciate the fact that language is a matter of “identity” for some people. A second claims that language-​related disadvantages should be of special concern in a theory of justice because they are part of a person’s unchosen circumstances. I shall argue that neither of these alternatives offers a decisive challenge to the neutrality model, although the argument from choice does leave some lingering philosophical questions, and the argument from identity may be relevant to balancing the claims of the instrumental and non-​instrumental frameworks. Many people care about their first language or languages—​that is, the language or languages that they learnt and used as young children. (Henceforth, for ease of exposition, I  shall assume that people have just one first language, although clearly some people learn and use more than one language from a very young age.) They self-​identify with the (local) community of speakers of the language. They are proud of the language and of the cultural achievements that have been expressed through it. They take pleasure in using the language and encountering others who are willing to use it. They hope that their (local) language community will survive and flourish into the indefinite future. They find that a number of their other preferences are linked with, and may be expressions of, their identification with their language community. In some contexts, they feel respected and affirmed when others address them in their language and denigrated when others use a different language. And so on. It is true that some people are scarcely aware of having these attitudes and preferences. They are comfortably cocooned in an environment in which their first language enjoys pre-​eminence and in which enjoying the benefits of the language is as automatic as enjoying the air that they breathe. The attitudes and preferences that I am attributing to people, however, are dispositional in character: persons have them if they would consciously be moved by them under conditions in which the language is not pre-​eminent but is in meaningful competition with one or more other languages in at least some significant contexts of communication. This is not to say that all people care about their first

604   Alan Patten language. Some people may be genuinely indifferent and would remain so even under the appropriate counterfactual conditions. But the attitudes and preferences I  have mentioned are instantly familiar, and I assume they are widely, if not universally shared. Now the mere fact that people care about their language is not sufficient to justify the claim that language matters non-​instrumentally to justice. People care about all sorts of things: the car they drive; the clothes they wear; the home they live in; the company they keep; and so on. Theorists of justice would not normally single out the make of the car a person drives as an object of distributive concern just because it is something that some people happen to value. So, how then does justice connect with the various preferences people have? John Rawls’s influential answer is procedural in form (Rawls 1999, 2005). Faced with a wide range of different preferences, what justice requires is a fair framework in which people can strive to satisfy their preferences. On this approach, which Rawls calls a “pure procedural” conception of justice, there is no framework-​independent notion of what constitutes a just distribution of particular objects of preference. A just distribution of such objects is simply whatever distribution emerges from people interacting within a fair framework. The theoretical problem under this approach is to specify the framework that shapes distributive outcomes such that the distribution that emerges can be considered just whatever it is. To investigate whether linguistic preferences imply any non-​instrumental justice claims, then, we need to know what procedure or framework could have this effect of conferring legitimacy on language-​related distributive outcomes such as the success or failure of particular languages and the degree to which people feel respected by the norms of language usage. Following Rawls’s terminology again, we might call the conditions that define such a procedure “fair background conditions” (Rawls 2005). Our question, then, is what fair background conditions consist in once it is observed that some people may care deeply about linguistic outcomes. What conditions could make it such that it could reasonably be said that linguistic outcomes are just (from a non-​instrumental perspective) whatever they are? The question being posed here parallels the question that might be asked when people care about particular religious outcomes. There is no particular religious outcome that is mandated by non-​ instrumental considerations of justice. Instead, justice is secured procedurally through fair background conditions, and the question of religious justice resolves itself into a question about what those conditions are. One approach to specifying fair background conditions might be characterized as minimalist. Key requirements of a minimalist approach are: –​ protection of a set of basic liberties (such as those identified by Rawls); – freedom from discrimination in the economy and civil society; – a social minimum and/​or a requirement that all persons have a decent or fair share of resources with which to pursue their ends; – various programs and policies that promote what Rawls calls “fair equality of opportunity.”

Language   605 Although each of these conditions needs further specification and refinement, a general sketch is enough for our purposes. The crucial claim is that any particular distributive outcome can be considered just, so long as it arises in a social system in which these requirements are satisfied. Thus, if some people live in one-​bedroom downtown apartments, others live in suburban bungalows, and others in spacious rural farmhouses, this distribution of living arrangements raises no justice-​based concern so long as it is generated in a context where the minimalist background conditions are realized. Likewise, there is no non-​instrumental justice concern about a particular level of success (or failure) of Welsh in Great Britain so long as the minimalist requirements are in place. So long as Welsh-​speakers enjoy freedoms of speech and association, as well as access to a fair share of resources with which to pursue privately whatever projects and ambitions they happen to have (including the promotion of a Welsh-​language community in Britain), then justice is compatible with any outcome for Welsh. The important question for our purposes is whether minimalism is enough for procedural justice. In one respect, of course, the minimalist conditions are extremely demanding. There are few societies on earth in which they are completely satisfied. And yet it is important to ask whether they would be enough if they were completely satisfied. A negative answer might suggest that justice has a quite different shape—​even in our imperfect world—​than the one implied by minimalism. Reflection about the case of religious justice suggests that minimalism is insufficient. Consider a state that officially establishes a particular religion. It is careful not to impose the religion on anybody coercively, so no basic liberty is violated. But it does extend to the religion certain privileges that are not given to other denominations, such as tax breaks, subsidies, and special access to public land and property. The minimalist conditions are satisfied but, if you think about it from the standpoint of other denominations that are struggling to maintain themselves, the outcome will hardly seem justified by the background conditions. Those belonging to a disfavored denomination will quite reasonably complain that, far from being fair, the background conditions are tilted against them. Religious justice might seem like a special case in certain respects. For example, someone might argue that all establishment—​even where it is non-​coercive—​violates a basic liberty. But the problem with minimalism is apparent in other areas too. For instance, imagine that tickets for professional football, baseball, and basketball games are exempted from taxation (on the grounds that they are “American pastimes”), but other sporting events are subjected to regular taxation. The minimalist would be satisfied, but again fans of the other sports would have good grounds to complain of unfairness in background conditions. In both of these cases, the state is siding with a particular preference or set of preferences held by some of its citizens over preferences held by others. It is extending benefits and privileges to the objects of some preferences and not to others, or it is imposing burdens and obligations on the objects of some preferences and not others. People holding disfavored preferences can reasonably feel that they are denied a fair

606   Alan Patten opportunity to realize their own conception of the good life. For this reason, they can reasonably object to the notion that the background conditions under which they are striving to realize that conception are fair. If this analysis is correct, then an alternative to the minimalist specification of fair background conditions might be formed by adding to the minimalist conditions a requirement of state neutrality. Neutrality, as I understand it, is a claim about how the state should treat the various goods and activities that its citizens value.3 The state treats different preferences neutrally when, relative to an appropriate baseline, it extends the same benefits to, and imposes the same burdens on, the goods and activities valued by those preferences. On the neutralist specification of fair background conditions, a particular distributive outcome is considered just if and only if the minimalist requirements are satisfied and the state is treating neutrally the preferences that are implicated in that outcome. It is easy to overlook the neutrality condition when thinking about justice because in so many cases it is satisfied negatively—​by the state not doing certain things. The state treats different preferences neutrally by not offering any benefits to, or imposing any burdens on, particular objects of preference, but instead by striving to establish a general framework in which citizens enjoy rights and liberties to pursue whatever preferences they have as well as a fair share of resources with which to do so. What is going wrong in the religious establishment and sporting preference cases is a departure from neutrality understood in this negative way. In contexts where it could quite reasonably do nothing, the state is actively backing particular options valued by some, but not all, of its citizens. However, this do-​nothing or privatization strategy for realizing neutrality does not make sense for every object of preference. With some goods and activities, the state cannot help but support some particular options, or, if it can help it, it would do so only at severe cost in other values. Language is one of those goods for which the privatization strategy is obviously unsatisfactory (Kymlicka 1995, Carens 2000, Patten 2001). Public institutions cannot wash their hands of any involvement with particular languages, and leave language up to the market and civil society. Since public institutions use language to communicate both internally and with the public, they must decide in which particular language or languages to do so. Because the privatization strategy is unworkable in relation to language, it is sometimes said that neutrality is impossible or incoherent in this area. According to Will Kymlicka, The idea that government could be neutral with respect to ethnic and national groups is patently false. ( . . . ) In the areas of official languages, political boundaries, and the division of powers, there is no way to avoid supporting this or that societal culture. (Kymlicka 1995: 110–​11)

3 

I develop an account of neutrality in Patten 2012 and 2014: Ch. 4.

Language   607 If linguistic neutrality is impossible or incoherent, then supplementing the minimalist conditions with a neutrality condition might not make much of a difference for the requirements of justice with regard to language. Language’s non-​instrumental contribution to justice would still be very marginal. But this argument rests on an inadequate analysis of neutrality. It is not true that a state must choose between supporting this or that language. It can support two (or more) languages by offering public services in each of those languages and by giving citizens the right to conduct public business in any of those languages. Of course equal support of this kind does not guarantee equal outcomes for the different languages concerned. But neutrality of outcome (or “neutrality of effects,” as it is sometimes known) is not in any case an attractive demand to make of public policy (Kymlicka 1989a, Arneson 2003, Patten 2012, 2014). The conception of neutrality that is relevant to fair background conditions and procedural justice is neutrality of treatment. The crucial point is that privatization is just one approach to realizing neutrality of treatment. Public institutions follow the privatization strategy by offering no special benefits to, and imposing no special burdens on, particular goods and activities that people value. An alternative strategy for realizing neutrality would involve positive provision of benefits tailored to particular goods, so long as equivalent benefits were also tailored for and extended to rival goods. Following Joseph Carens, this strategy can be termed “evenhandedness” (Carens 2000). A  municipality might practice evenhandedness in its public parks, for example, by providing equipment and facilities for different sports and recreational activities according to the distribution of preferences in the community.4 Consider two scenarios. In the first, there is only one language in which all public business is conducted—​the majority language. In the second, public business can be conducted in either the majority language or the largest minority language. Public services and public education are offered in both languages, and measures are adopted to make it possible for people to use either language while participating in public institutions (in the courts, the legislatures, the military, the bureaucracy, etc.). Intuitively, there is a major difference between these scenarios for minority speakers. If, in the first scenario, they have trouble satisfying their language-​related preferences, it would be hard to justify such an outcome on grounds of procedural justice. Given that public policy is a major determinant of linguistic success, and that public policy shows a preference for the majority over the minority language, it would be hard to tell minority speakers that the frustration of their preferences is legitimated by the fairness of the

4 

This “evenhandedness” variant of neutral treatment remains distinct from neutrality of effects. Neutrality of treatment is achieved by equalizing the public resources (the “treatment”) extended to each activity for which there is citizen demand. Neutrality of effects, by contrast, seeks to equalize along the dimension of outcome, that is, the success of the activities in question. Since some activities are relatively expensive and some forms of success have hard-​to-​satisfy pre-​conditions (e.g., large numbers of participants), neutrality of treatment and neutrality of effects are not equivalent. The state could equalize treatment and still one would not expect equal levels of success. For elaboration, see Patten 2012, 2014.

608   Alan Patten background conditions under which the outcome arises. Contrast this with the second scenario. In this situation, the support given to the largest minority language makes it less likely that the language-​related preferences of that minority will be frustrated. And, if those preferences are frustrated, a procedural justification of such an outcome starts to become more plausible. It can be argued that minority speakers have no legitimate complaint, given that they enjoy the same rights and liberties as others, and given that their language enjoys public recognition and support that is comparable to that enjoyed by majority-​speakers. The neutrality model helps to account for these judgments. Whereas in the first scenario, public institutions are clearly siding with the linguistic preferences of the majority, in the second scenario they offer neutral treatment of the two languages by adopting the evenhandedness strategy. As a consequence, the background conditions are fair and—​ so long as the minimalist conditions are also satisfied—​it is plausible to regard outcomes as just whatever they are. If the neutrality model is accepted, then the non-​instrumental contribution language makes to distributive justice is far from trivial. When people care about their first languages, it is not enough for justice merely to guarantee certain basic liberties and access to a fair share of resources. Justice also requires that public institutions be neutral towards the objects of different personal preferences. In the case of linguistic preferences, the only way to realize this neutrality is by evenhandedly extending public recognition to each of the languages that people want to use publicly. Only when minority language rights are recognized and protected can background conditions be regarded as fair and outcomes considered just whatever they are. I have been framing this argument in terms of Rawlsian ideas of pure procedural just­ ice and fair background conditions. But it is possible to arrive at the same conclusion via Ronald Dworkin’s theory of equality of resources (Dworkin 2000). Dworkin argues that, in a context where people have diverse preferences over heterogeneous resources, markets are essential for realizing distributive justice. Imagining that a society’s resources are allocated through an idealized auction in which each participant starts out with equal purchasing power, Dworkin argues that such a procedure would result in a distribution that satisfies what he calls the envy-test: no person would prefer anybody else’s bundle of resources to his own. In such a distribution, each person’s share, measured in terms of the opportunity cost it imposes on others, is equal. In developing the view, Dworkin notes an important underlying assumption. The bundles that are put up for auction must not be arbitrarily defined in ways that suit some preferences and not others. If the auctioneer were to sell off land in lots the size of football fields, for instance, then the envy test would still be satisfied, but the resulting distribution would not be an equal one. An auction “provides a more genuinely equal distribution,” Dworkin argues, “when it offers more discriminating choices and is thus more sensitive to the discrete plans and preferences people in fact have” (Dworkin 2000: 151). The auction should respect a “principle of abstraction” requiring that resources be put up for sale in the most abstract form possible, such that they are as finely tuned as possible to the various plans and preferences that people might have. If this principle is not

Language   609 satisfied, then there is no guarantee that the auction will bring to light the true opportunity costs for others of letting some particular person control a particular resource. Dworkin’s principle of abstraction is relevant to our discussion because it entails that the state’s (the “auctioneer’s”) choices have to be responsive to the actual preferences of citizens. In typical cases, this means that the state has to avoid imposing preferences held by some but not all citizens onto the operation of the market: it has to give the market the greatest possible space in which to operate. But Dworkin does recognize that, in some situations, letting actual markets determine outcomes—​“privatization” as I termed it earlier—​is undesirable. Even here, however, he thinks hypothetical markets can play a heuristic role. In an intriguing exchange with G. A. Cohen about fairness in the acquisition policies of public libraries, Dworkin suggests that the right approach would mimic market justice as far as possible (Dworkin 2004). It would be sensitive to the reading preferences of different patrons, as well as to the costs of acquiring particular books. Dworkin’s discussion here has important implications for non-​instrumental linguistic justice. As we have seen already, language policy is one of those decisions that cannot simply be left up to the market. Obviously, some choices about language can and should be left up to individuals to make for themselves, but the auctioneer (the state) must use some language or languages, and it would be incoherent to insist that this decision could be put up for auction. This leaves open the possibility of mimicking market justice, however, by shaping language policy in response to the same sorts of factors that would determine outcomes in actual markets. This means making language policy sensitive to the distribution of linguistic preferences in the community, as well as to the costs of accommodating particular languages. This Dworkinian argument for linguistic evenhandedness illuminates several nuances that were not apparent in the Rawlsian argument. One is that it is natural to think of an evenhanded language policy as involving a certain amount of pro-​rating. Small language groups do not, as a matter of non-​instrumental justice, have a claim on as much linguistic support and recognition as large groups. The argument for pro-​rating rests on two previously stated assumptions plus a third. The assumptions we have seen already are that each individual should be thought of as having equal purchasing power, and that the market outcomes that are to be mimicked in determining the demands of justice with respect to language policy are sensitive to cost. The third assumption needed to make the case for pro-​rating is that, when it comes to language support and recognition, there are increasing returns to scale. On a per capita basis, it is less costly for public institutions to support and recognize a widely spoken language than one that is spoken by only a small minority. Pro-​rating is compatible with the Rawlsian argument sketched earlier—​it is one way of understanding what exactly neutral treatment involves—​but its rationale is more fully motivated by introducing Dworkin’s argument together with the assumption of increasing returns. The second nuance relates to what economists call “market failure.” Dworkin is aware that actual markets sometimes produce sub-​optimal outcomes because of externalities. Some goods are underprovided on the market because they require some level of cooperation between a number of people. When somebody whose cooperation is needed

610   Alan Patten does not cooperate this impacts on others in ways that are not captured by market prices. (Dworkin’s example is the disruption by a single property owner of an effort by neighbors to give their immediate neighborhood a uniform architectural style). In cases of this kind, Dworkin argues for a “principle of correction,” which would impose constraints on particular choices (e.g., by zoning particular neighborhoods) in order to better track the true opportunity costs associated with market allocations. The principle of correction is potentially relevant to language policy, since linguistic outcomes are often influenced by externalities and collective action problems (Laitin 1993; Van Parijs 2011; Patten 2017). Speakers of a minority language may feel a non-​ instrumental attachment to their language but have instrumental reasons for using the majority language in various important settings. Even if they care more about the non-​instrumental reasons, they may find themselves in a collective action problem that makes it rational for them to act on the instrumental reasons. To the extent that everyone reasons this way, the destruction of the minority language may be a real danger. The Dworkinian approach explains why it may be legitimate, in some situations, for the state to take special protective measures on behalf of minority languages that would otherwise be vulnerable in a strictly evenhanded policy regime.

28.4  Alternative Accounts On the neutrality model, language contributes non-​instrumentally to justice because it is something that people care about for reasons that do not reduce to an instrumental connection between language and distributive justice. Many people care about their first language, want to have opportunities to use that language, and want to see the language and the community of speakers of that language survive and flourish into the future. From a non-​instrumental perspective, justice requires that public institutions cater to these preferences and attitudes in an evenhanded fashion. The same support and recognition ought to be given to minority languages that is given to the majority language. At the level of policy, the argument supports a fairly robust set of minority language rights. Subject to some limits imposed by numbers and costs, minority speakers can reasonably expect to be served in their own language by public institutions, and to be able to use that language when they participate in those institutions. To some observers, the conclusion of this argument—​that there are significant non-​ instrumental claims of justice at stake with language policy—​will seem correct, but the argument for it unsatisfactory. One possible reason for dissatisfaction is that the neutrality model may seem to overlook the importance of language for identity (May 2001). The language a person speaks—​especially a person’s first language—​matters to them in ways that are not captured by the more clinical idea of “preference” that was used in developing the neutrality model. When something matters to a person’s identity, it makes them psychologically vulnerable. It implicates dignity and the sense that life is worth living, and leaves a person exposed to feelings of shame and humiliation. Identity

Language   611 is also central to a person’s practical reasoning. To have an identity focused on X is, in part, for X to figure in one’s reasoning about what to value and how to act. However, it is not obvious that insisting on the identity dimension of language makes much of a difference as far as the neutrality model is concerned. It is true that, in presenting the neutrality model, I  referred to individual preferences, but this was an analytic term meant to cover a range of different pro-​attitudes (desire, attachment, commitment, etc.) including the complex set of attitudes that one has in mind in attributing a particular identity to somebody. Substituting “identity” for “preference” may end up being little more than a semantic change. The challenge may be sharpened further, however, by insisting that identity does make a substantive difference by strengthening the protections that are owed to weaker, more vulnerable languages beyond the protections offered by the neutrality model. Objects of identity, on this picture, require better-​than-​neutral treatment. But in other domains where identity is invoked, identity claims do not imply more than the fair treatment promised by neutrality. Consider an applicant to an elite university whose family has attended the university for generations. It might be devastating for this person not to be accepted. The parents might be beside themselves. The university might truly be central to the family’s identity. But it is doubtful that identity considerations should make any difference at all to admissions decisions beyond reinforcing the importance of fair admissions procedures. Or consider someone whose identity is focused on a particular religious community. It is far from clear that special measures—​beyond those called for by neutrality—​ought to be taken to protect or preserve the community in question. Even if identity is at stake, many people would judge that the state’s responsibility is to establish fair background conditions, and then to allow religious outcomes to work themselves out through the choices of individuals. We will return briefly to the theme of identity in section 28.5. First, I want to consider in more depth a different kind of concern about the neutrality model. The concern is that the neutrality model overlooks the important distinction between a person’s choices and her circumstances. By failing to conceptualize a person’s linguistic situation as part of her circumstances, the model we have been exploring misses a major reason why support for disadvantaged languages excites the concern of liberal egalitarians. Moreover, the choice–​circumstance distinction makes a difference to policy, helping to explain the sense shared by some commentators that mere neutrality is not enough. Neutrality is consistent with smaller, weaker languages being crushed in what appears to be a neutral competition. Once it is allowed that a person’s linguistic situation is a matter of her circumstances rather than her choices, justice may mandate protections for vulnerable languages beyond the evenhandedness promised by neutrality (Kymlicka 1989b; De Schutter and Ypi 2012). The distinction between a person’s choices and her circumstances is a cornerstone of the strand of liberal egalitarian theory that has come to be known as “luck egalitarianism.” The distinction is meant to capture the idea that some features of a person’s life are subject to her choice, while others are simply given by outside circumstances. In the view of luck egalitarians, disadvantages that can be traced back to conditions that are

612   Alan Patten or were subject to choice do not, as a matter of justice, ground as strong a claim for assistance or compensation. Disadvantages that cannot be avoided through appropriate choice, by contrast, should stimulate a full measure of egalitarian concern: they call for prevention, assistance, and compensation. The neutrality model does not rely on luck egalitarian assumptions. But perhaps luck egalitarianism can ground a superior conception of the non-​instrumental relation between justice and language? One feature of a person’s situation that certainly counts as part of their circumstances is their first language. Whether this or that language is the one that a person first learns and uses as a young child is not something that is ever subject to a person’s choice. It is more-​or-​less entirely a matter of the linguistic capacities and choices of their family and of any other formative institutions to which they are exposed at a young age. Some commentators have suggested that it is this fact—​that one’s first language is not a matter of choice—​that grounds a luck egalitarian analysis of linguistic justice. Nobody chose to be raised a Welsh-​speaker, and thus it is unfair for Welsh-​ speakers to have to face various disadvantages that are escaped by English-​speakers. For this argument to go anywhere, it will need to specify the relevant disadvantages more precisely. The mere fact that some feature of a person’s life is unchosen is not enough to ground a justice claim. Many given features of a person’s life are of no great importance for justice. Whether a man goes bald in his forties, or maintains a full head of hair, is largely determined by his biological make-​up and there may be little he can choose to do to prevent nature from running its course. But since differences with respect to this part of a person’s circumstances do not correspond to major disadvantages or inequalities, there is no reason to think that justice is at stake. The first question then is whether there are disadvantages that can be traced back to the fact that people have different first languages. One reason to think that there might be was encountered in our earlier discussion of the instrumental framework. If a person’s first language is also the only language they are able to speak proficiently later in life, and that language is not a medium in which many of the society’s economic and social opportunities are available, then they may find themselves at a serious disadvantage. It is true that they could have tried to learn new languages as an adult, including the society’s dominant language. If they opt not to, then this makes their dominant-​ language competences a matter of choice not circumstance. But it is hard for adults to learn new languages and it is quite possible that some will not be successful if they try. The language repertoire of people who speak a non-​dominant language as their first language, and who are unable to learn the dominant language, should be considered as part of their circumstances. Given the disadvantages to which such a repertoire leaves them exposed, it is plausible to think that they have some kind of claim for assistance or compensation. Although considerations of justice do seem to arise in this kind of situation, the analysis has not taken us beyond the instrumental framework. As we saw earlier, it is plausible to assume that access to an adequate (perhaps equal) set of social and economic opportunities is a condition of justice. And once this assumption is made, it immediately follows that anyone who cannot speak the dominant language of a society (where

Language   613 there is one) is facing a serious disadvantage. But, as with the instrumental argument considered earlier, this version of the luck egalitarian argument is broadly supportive of a nation-​building approach to language policy rather than a minority rights one. If it is true that adults sometimes have trouble mastering a new language, then the most effective response may be to do a better job of teaching the dominant language to all people when they are still children. It is possible that, in some cases, even childhood language training may be ineffective. It is notoriously hard to teach English-​speaking children a second language proficiently. But what seems quite unlikely (even if not impossible) is the conjunction of two facts: on the one hand, a language is unable to provide adequate opportunities to its speakers, and, on the other, childhood dominant-​language education for speakers of this language is likely to be ineffective. At best, there may be a few cases that are exceptions to this generalization, and in these cases some special form of minority language protection may be the best approach.5 Another version of the luck egalitarian argument locates the disadvantage faced by native speakers of some languages in a different place. It is not so much that they cannot speak the dominant language but that they care about using their own native or first language. It might seem that we have already explored this possibility. After all, this is where the neutrality model started from as well. But there are a couple of twists that a luck egalitarian can introduce to the argument that pushes it in new directions. Both twists have to do with responsibility for preferences. Liberals typically assume that individuals should be considered responsible for their preferences: the mere fact that some preference is likely to be relatively unsuccessful does not ground a justice claim. People with unsuccessful preferences might be able to argue that their preferences have been treated non-​neutrally. This was the possibility we explored in section 28.3. But if their preferences are treated neutrally and yet they are still relatively unsuccessful, then they have no further justice-​based complaint. This liberal assumption about preferences is itself sometimes given a luck egalitarian justification. People are held responsible for their preferences because they are assumed to have the opportunity, over the course of their lives, to influence what preferences they have. The first twist offered by luck egalitarians accepts the liberal assumption about preferences (if only for the sake of argument), but argues that it still leaves space for a certain kind of justice claim. Even if individuals are responsible for their preferences, they are not responsible for the fact that particular preferences are difficult or costly (or conversely easy or cheap) to satisfy (Kymlicka 1989b; Cohen 1989, 2004; De Schutter and Ypi 2012). Welsh-​speakers might be responsible for the attachment they feel to their language, but they are not responsible for the fact that only a tiny fraction of UK citizens speak Welsh and that this makes it difficult and costly for them to enjoy the flourishing of their language. However, this version of the luck egalitarian argument strikes me as unsuccessful. The same reasons that motivate the view that individuals are responsible for their

5 

For discussion of these special cases, see Patten 2014: Ch. 6.

614   Alan Patten preferences are also reasons for resisting the roundabout suggestion that individuals have a claim for assistance or compensation whenever their preferences are difficult or costly to satisfy for reasons that are outside of their control. Consider, for instance, a stock example of somebody with expensive tastes, such as Harold, who needs champagne to get the same level of satisfaction that others derive from mass-​produced beer. Almost nobody would maintain that Harold is owed compensation or assistance: his is the sort of case that drives liberals to endorse responsibility for preferences. But of course the factors that make his tastes costly to fulfill—​the difficulty of producing champagne, the fact that many people around the world desire it, etc.—​are completely outside of Harold’s control. As a general matter, if people are rightly considered responsible for their preferences, it is because they have the opportunity, over the course of their lives, to influence what preferences they have. Suppose that Harold did have this opportunity. Then that would count as a reason, not just for holding him responsible for his preferences, but for dismissing his complaint that he is unfairly burdened by the external factors that make his preferences expensive. After all, he could have avoided the burdens in question by developing different preferences. Likewise, it makes little sense to hold people responsible for their linguistic preferences, but then to turn around and say that they are none the less owed compensation or assistance for the externally determined factors that make those preferences difficult or costly to satisfy. This brings me to the second luck egalitarian twist, which questions whether it is indeed reasonable to hold people responsible for their linguistic preferences. Linguistic attachments may have been developed at a very young age and they may be very strongly held. It may simply not be true that a person had a genuine opportunity over the course of her life to replace her attachment to her language with some other set of attachments. This is a fundamental question for any account of linguistic justice, indeed for any account of distributive justice. It is hard to say anything very decisive, but there are a couple of skeptical observations about this latest version of the luck egalitarian argument that are worth noting. The first is that, while it is certainly true that people tend not to revise their linguistic attachments, it does not follow that they could not revise them if they tried. The fact is that most people identify with their attachments, even when those attachments are relatively unsuccessful. They would not revise them even if they could do so without cost. Being unwilling to do something is not the same, however, as being unable to do it. It is the latter condition that matters for responsibility, not the former. To be sure, people do not normally control their preferences and attitudes directly. But, over the course of their lives, they can influence their preferences by engaging with new options and forms of life. In a liberal society, where there are a plurality of valuable ways of life, someone who really chafes at having a relatively unsuccessful set of preferences has the freedom to explore alternatives. The second response is that luck egalitarianism is itself a contestable position. Suppose the previous observation is set aside and it is conceded that individuals are not generally responsible for their linguistic preferences. It would then be natural to suppose that individuals are not responsible for many of their preferences. For presumably, whatever

Language   615 reasons there are for doubting preference responsibility in the linguistic domain carry over into other domains as well. But then it seems that luck egalitarians will be on the hook for Harold and his expensive tastes (or at least that the case of Harold could be refined so that they are on the hook). This seems a significant embarrassment to the luck egalitarian view, and might lead one to question whether the choice/​circumstance distinction can bear all the weight that is being placed on it. As was noted earlier, the neutrality model does not depend on luck egalitarianism and would survive the rejection of that position.

28.5  Integrating the Two Frameworks Thinking about linguistic justice from within the instrumental framework generally, even if not invariably, counsels in favor of nation-​building. In nationally divided societies, there may be intense disagreement about which nation it is that should be privileged. Should Catalan be the common public language of Catalonia or Spanish? But the idea that the state, or a politically autonomous sub-​state unit, should seek to promote convergence by all its citizens on a common national language is shared in common by observers who adopt the instrumental approach. By contrast, the non-​instrumental framework, as theorized by the neutrality model, favors minority language rights. Such rights are an integral part of a framework that defines fair background conditions under which minority speakers with an attachment to their language can strive to satisfy their preferences. Depending on which approach to linguistic justice is adopted, then, rather different policy recommendations seem to follow. Given this apparent tension, it might be wondered whether and how the two frameworks can be integrated with each other. Can we say anything overall about what justice requires of language policy? A first point to note is that, in some cases, the tension between the two frameworks is more apparent than real. The state can recognize and protect minority language rights and expect that the values associated with nation-​building will be respected. In some of these cases, minority language communities are small and can offer only limited economic, social, and cultural opportunities to their members. In these cases, the state may be able to offer some set of minority language rights, confident that the minority will likely learn the dominant language anyways and use it in many contexts. In other cases, language minorities are rather large and contain a quite adequate set of economic, social, and cultural options. In these cases, public recognition of the minority language does not conflict with nation-​building values (e.g., economic opportunity, democratic self-​government, etc.) because the minority language community is capacious enough to allow for the realization of those values internally and there is no need for a common national language. The difficult cases are ones in which there is a genuine tension between the values promoted by nation-​building and the values realized by minority rights. By recognizing and protecting minority language rights, the state exacerbates efforts to realize the

616   Alan Patten nation-​building values. In our discussion of the instrumental framework, we have seen how this tension might arise. To grapple with this tension, we need to take a closer look at the value of neutrality, which I have suggested underpins the minority rights perspective. Neutrality is grounded in an idea of fairness: when the state extends fewer benefits to or imposes greater burdens on some object of preference than it does for objects of other preferences, then it denies bearers of that preference a fair opportunity to realize their conception of a good life.6 Once the value of neutrality is fleshed out in this way, however, it becomes apparent that neutrality is best viewed as a presumptive or pro tanto obligation of the state. There are weighty, fairness-​based reasons for the state to treat different preferences neutrally, but these reasons can, on occasion, be outweighed by other values. Neutrality has this presumptive character for at least three different reasons. First, violations of neutrality are not the only way of denying people a fair opportunity to realize their conception of the good. Another way of denying people this opportunity is by leaving them without access to a fair share of resources with which to pursue their preferences. These different elements of fair opportunity may compete with one another (we have seen how they might in the language case) and may need to be balanced appropriately. Second, fair opportunity to realize one’s conception of the good is surely not the only, or the supreme, value. At the very least, it is important that persons be put in a position where they can reflect critically about whether they have the right conception of the good. And, third, there may be other competing values as well, such as the perfectionist one of having a valuable conception of the good. Once the normative foundation for minority rights is seen to have a presumptive character, the path is open to balancing such rights against the values promoted by the instrumental framework. Because neutrality is based on a dimension of fairness, it is not a trivial concern that would be overridden by just any competing consideration. Minority rights should not be overridden because of moderate costs or because of added administrative burdens, nor are they defeated by citing highly speculative or improbably risks or vague national goals. The fact that identity is at stake for some individuals is also relevant to balancing the two frameworks. Just as the importance for believers of their religious convictions makes state neutrality with respect to religion a weighty concern, something like the same is true of language. Even if linguistic identity is not put on the same level as religious conviction, it is reasonable to suppose that the former shares at least some of the weightiness of the latter. On the other hand, if the arguments for the instrumental framework are sound, then instrumental considerations are sometimes very weighty too. It is plausible to think that they will be weighty enough in some instances to override neutrality. One implication of integrating the frameworks, for instance, is that the number of different languages in a single state that can be given extensive minority language rights is probably quite

6 

Again, see Patten 2012 and 2014: Ch. 4.

Language   617 limited. States cannot perform their core functions if they devote too high a proportion of their resources and energies to accommodating linguistic differences. A state that is committed to linguistic justice should, on balance, select a small number of languages and reserve its official support and protection for these. Another implication is that minority rights should be scrutinized very closely when there is a risk that they will exacerbate existing socioeconomic inequalities by segregating disadvantaged language minorities from the majority population. These points about balancing are meant to be illustrative rather than exhaustive.7 The more general claim is that someone concerned with linguistic justice does have some theoretical resources with which to integrate the instrumental and non-​instrumental frameworks we have been considering. These resources become apparent upon thinking about the value of neutrality that informs the non-​instrumental framework, and then asking how that value relates to other values and concerns that ought to be part of a conception of justice. The upshot is a coherent, if complex, theory of linguistic justice, one that supplies, in the appropriate contexts, an argument for minority language rights.

References Anderson, E. (2010). The Imperative of Integration. Princeton, NJ: Princeton University Press. Arneson, R. (2003). “Liberal Neutrality on the Good: An Autopsy,” in S. Wall and G. Klosko (eds) Perfectionism and Neutrality: Essays in Liberal Political Theory. Lanham, MD: Rowman & Littlefield, pp. 192–​218. Carens, J. (2000). Culture, Citizenship, and Community: A Contextual Exploration of Justice as Evenhandedness. Oxford: Oxford University Press. Cohen, G. A. (1989). “On the Currency of Egalitarian Justice.” Ethics 113(4): 745–​63. Cohen, G. A. (2004). “Expensive Taste Rides Again,” in J. Burley (ed.) Dworkin and his Critics. Oxford: Blackwell, pp. 1–​29. De Schutter, H. and Ypi, L. (2012). “Language and Luck.” Politics, Philosophy, and Economics 11(4): 357–​81. Dworkin, R. (2000). Sovereign Virtue. Cambridge, MA: Harvard University Press. Dworkin, R. (2004). “Ronald Dworkin Replies,” in J. Burley (ed.) Dworkin and his Critics. Oxford: Blackwell, pp. 337–​95. Green, L. (1987). “Are Language Rights Fundamental?” Osgoode Hall Law Journal 25(4): 639–​69. Kymlicka, W. (1989a). “Liberal Individualism and Liberal Neutrality.” Ethics 99(4): 883–​905. Kymlicka, W. (1989b). Liberalism, Community, and Culture. New York: Oxford University Press. Kymlicka, W. (1995). Multicultural Citizenship. Oxford: Oxford University Press. Kymlicka, W. (2001). Politics in the Vernacular. Oxford: Oxford University Press. Kymlicka, W. and Patten, A. (eds) (2003). Language Rights and Political Theory. Oxford: Oxford University Press.

7  One question that arises here is whether there is any general reason to prioritize the language claims of national groups over those of immigrants. For a cautious and qualified affirmative answer to this question, see Patten 2014: Ch. 8. See also Kymlicka 1995.

618   Alan Patten Laitin, D. (1993). “The Game Theory of Language Regimes.” International Political Science Review 14(3): 227–​39. May, S. (2001). Language and Minority Rights. Harlow/​London: Longman/​Pearson Education. Miller, D. (1995). On Nationality. Oxford: Oxford University Press. Patten, A. (2001). “Political Theory and Language Policy.” Political Theory 29(5): 683–​707. Patten, A. (2003). “Liberal Neutrality and Language Policy.” Philosophy and Public Affairs 31(4): 356–​86. Patten, A. (2009). “Survey Article: The Justification of Minority Language Rights.” Journal of Political Philosophy 17(1): 102–​28. Patten, A. (2012). “Liberal Neutrality:  A Reinterpretation and Defense.” Journal of Political Philosophy. 20(3) 249–​72. Patten, A. (2014). Equal Recognition: The Moral Foundations of Minority Rights. Princeton, NJ: Princeton University Press. Patten, A. (2017). “Protecting Vulnerable Languages:  The Public Good Argument.” Paper delivered to Workshop for Oxford Studies in Political Philosophy, Tucson, AZ. Rawls, J. (1999). A Theory of Justice, revised edn. Cambridge, MA: Harvard University Press. Rawls, J. (2005). Political Liberalism, expanded edn. New York: Columbia University Press. Réaume, D. (1991). “The Constitutional Protection of Language:  Survival or Security?,” in D. Schneiderman (ed.) Language and the State:  The Law and Politics of Identity. Cowansville: Éditions Yvon Blais, pp. 37–​57. Réaume, D. (1994). “The Group Right to Linguistic Security: Whose Right, What Duties?,” in J. Baker (ed.) Group Rights. Toronto: University of Toronto Press, pp. 118–​41. Réaume, D. (2000). “Official-​ Language Rights:  Intrinsic Value and the Protection of Difference,” in Will Kymlicka and Wayne Norman (eds) Citizenship in Diverse Societies. Oxford: Oxford University Press, pp. 245–​72. Taylor, C. (1992). “The Politics of Recognition,” in A. Gutmann (ed.) Multiculturalism and “The Politics of Recognition. Princeton, NJ: Princeton University Press, pp. 23–​73. Van Parijs, P. (2000). “Must Europe be Belgian? On Democratic Citizenship in Multilingual Polities,” in C. McKinnon and I. Hampsher-​Monk (eds) The Demands of Citizenship. London: Continuum, pp. 232–​56. Van Parijs, P. (2011). Linguistic Justice for Europe and the World. Oxford:  Oxford University Press.

Chapter 29

J ustice Acros s B orde rs Michael Blake

Questions of international wealth and poverty are now essential parts of the philosophical discussion of distributive justice. This is a comparatively recent development; when John Rawls re-​invigorated philosophical discussions of distributive justice, he was able to cordon off all questions of international justice to a few pages at the end of his book, and international distributive justice was neglected entirely (Rawls 1971). Since then, though, the field has grown rapidly, and now displays a striking degree of sophistication. As part of this development, those theorists who have discussed global distributive justice have had occasion to ask a variety of questions, and the questions they have asked have become increasingly complex. In particular, the earlier discussions of global distributive justice asked a comparatively simple question: do the norms of distributive justice, as developed within the context of the domestic state, apply outside that context? This question has, in recent years, developed into a more complicated, and multifaceted, question: what norms of distributive justice might we validly apply to the particular forms of institution and relationship found in the international realm? The shift to this latter question was occasioned by a growing willingness to consider the international realm as its own site of justice, without analogizing or reducing it to the institutions of the domestic state. The implications of this shift, however, have yet to be fully understood. In particular, I will argue that the specific questions we have begun to ask, to be answered well, require us to ask other questions, not all of which have yet received adequate philosophical attention. In particular, I believe that our considerations of global distributive justice must be answered with an eye towards questions of global political justice; towards, that is, the question of what would have to be the case for global political society to be legitimate in its implementation of global principles. This chapter is, therefore, both a look back and a tentative look towards the future. I will examine six questions that might be asked about distributive justice across borders. These questions are posed in what might be called the order of implication: answering any question well will require us to engage with the questions that follow it. I believe these questions are also arranged in roughly descending order of current philosophical attention; we have exceptionally well-​developed and sophisticated thinking about

620   Michael Blake the first few questions, that is, and still await more well-​developed theorizing about the last few questions. We can begin, then, with the question of global distributive justic­e, as it began to develop in the years immediately after Rawls’s A Theory of Justice was published: how, if at all, can we use these tools to analyze the justice of global wealth and poverty?

29.1  Can We Apply the Tools of Domestic Distributive Justice to the Realm of International Affairs? Questions of distributive justice have, historically, been asked only within the state; what discussions we have had of international morality have tended to reduce our international obligations to some comparatively weaker notions of charity or humanity (Barry 1982). In the 1970s, philosophers increasingly began to ask whether this limitation was ethically grounded, or whether we ought to make the world as a whole the site for our theorizing about distributive justice. Given that we already had a well-​developed set of ideas for thinking about distribution, the question naturally emerged: could we simply use these ideas internationally? For some philosophers, of course, the answer was yes, in a rather easy way; libertarian thinkers are as happy to deny distributive principles abroad as they are at home, and utilitarians like Peter Singer were able to extend the reach of their principles to the global community (Singer 1971). More interesting, though, was the question of whether the specific morality of inequality—​of, that is, not poverty taken as an absolute shortfall, but with reference to the gap between rich and poor—​could be analyzed internationally with the tools ready to hand. The first philosophers to address the morality of global inequality after Rawls did so in self-​conscious debt to the tools he had developed. There had been, of course, some classical and early modern discussions of justice and international affairs; the analysis of just war theory dates back to antiquity, and Kant’s notion of cosmopolitan right placed national self-​interest within the constraints of at least a limited notion of just­ ice (Kleingeld 2011). Rawls’s influence, however, was difficult to overstate in the years immediately after 1971, and many theorists began to argue that his ideas could not legitimately be limited to the context of an individual nation-​state, as Rawls himself assumed they could—​and, in so doing, they introduced a novel concern with international inequality to the field of political philosophy. The figures most important to the development of this methodology were Charles Beitz and Thomas Pogge, both of whom offered sophisticated accounts of how the Rawlsian account of distributive just­ ice could be applied to the realm of international production and distribution (Beitz 1979; Pogge 1989). Their arguments, while different at the level of detail, shared a similar

Justice Across Borders    621 architecture: both argued that Rawls’s principles ought to apply to any set of institutions that produced benefits and burdens, and that the global economic order was comprised of exactly that. If, in short, there were benefits and burdens to a global order, and if the benefits of that order were distributed to various players in the global game, why could we not use Rawls’s methods and principles to determine what fair principles of distribution would be for such a global order? These first thinkers argued that there was no principled reason to limit the application of Rawls’s principles of distribution to the local context, and generally took the label cosmopolitan as a description of their shared commitments. Their analysis was not universally accepted, and many philosophical critics responded that there was something different about the political and legal institutions of the domestic state—​something which rightly made Rawls’s principles of distributive justice applicable only within that realm, and not elsewhere. One common thread of criticism focused on what it was that the state actually did, for and to its own citizens: it coerced them, and sought to justify that coercion through some local principles of justice (Blake 2001; Nagel 2005; Risse 2006). Others argued that the state involved forms of reciprocity that could only be justified with reference to specific forms of distributive equality (Sangiovanni 2007). Still others asserted that the state was a site of self-​determination, and the primary consideration in international justice was not individual wealth but the collective freedom of the political community (Rawls 1999; Freeman 2006). Nationalists, finally, took the state to be a site of national self-​consciousness, which was the necessary site for any valid application of distributive principles (Miller 1995; Walzer 1986). All of these ideas were suggested or developed, in the years after the cosmopolitan movement took root, and—​ rather naturally—​acquired the label of noncosmopolitanism (or, sometimes, statism). The debate between these two camps continues, naturally, as people continue to argue about the relative merits of the two positions. My purpose at present is not to adjudicate between the camps, but to discuss something about the discourse as it has developed. The first thing to note is that it is a tremendously limiting way of framing the issue: it presupposes that there is a single set of principles called distributive justice, and that the only interesting question is whether that set shall be applied within the state or at the level of the globe. We can see, I think, why this discourse should have developed in this way; it should be equally clear that this is not the only way to think of the question of global distributive justice. As later sections will make clear, we are increasingly willing to accept that distinct distributive principles might apply in distinct contexts—​and we should not take this acceptance to entail that we are, as it were, denying the relevance of global distributive justice. The first noncosmopolitan writers were keen to refute cosmopolitan arguments; they did not, however, thereby refute the idea that some principles of global justice might apply to the international realm, and increasingly they are willing to accept as much (Risse 2012; Blake 2013). The idea that there is a single way of being in favor of global distributive justice—​that one must use the same tools at home and abroad, or accept that no tools can be used—​is an artifact of how the debate developed, and our way of framing this discourse has stunted that debate’s growth.

622   Michael Blake Evidence for this, I think, is seen in the shifting definitions of cosmopolitanism in recent years. Cosmopolitanism was first defined by its proponents as a commitment to the equal moral worth of persons (Pogge 1992; Sangiovanni 2007). The problem, of course, is that noncosmopolitans accept that commitment, and use it to ground their own reasoning about what is and is not true about global distributive justice (Miller 2003; Blake 2013). Cosmopolitanism has thus tended to mean a willingness to accept that there can be no differentiation between local and global duties of justice; in practice, though, this has tended to confuse as much as illuminate, since many prominent cosmopolitans accept the possible existence of distinctively local duties of justice, so long as these are not made stronger than transnational duties (Caney 2005). Thomas Pogge, recently, has argued that the core of cosmopolitanism is the existence of institutional agents who understand their mandates with reference to global, rather than local, constituencies (Pogge 2013). This is a promising development, but even here it seems that some people generally thought to be non-​cosmopolitans might in principle accept the necessity for global agents of that stripe (Risse 2012). In short, the term cosmopolitan—​and its contrary—​might be appropriate only within a discourse that is limited to whether or not we ought to extend distributive justice to the global level; it seems to presuppose a question to which yes or no is the appropriate answer. Increasingly, though, we are willing to ask a broader set of questions, ones not capable of being answered in this simple way. The question of global distributive justice, in other words, is increasingly a question of how, and where, and why distributive norms ought to apply—​not whether the norms we have close at hand ought to be made global. Once the question of global distributive justice has been rendered visible, as it were, we might well start to ask a different sort of question indeed. The debate, in short, is now characterized as pluralistic, rather than with reference to a simple dichotomy (DeBres 2012). This, however, changes the nature of the question we ought to ask about global distributive justice. We ought, then, to look directly at the sort of question we would ask, if we were inquiring about global distributive justice as its own distinct form of philosophical inquiry.

29.2  What Norms of Distributive Justice Apply at the Global Level The question considered here is what might be called the post-​cosmopolitan question. It is not post-​cosmopolitan because the arguments of the cosmopolitan have been refuted; these arguments are persistent, and powerful. (It is also not post-​cosmopolitan because cosmopolitanism has won at the level of political practice; the gap between political practice and even the most moderated form of cosmopolitanism is absurdly huge.) It is post-​cosmopolitan, instead, in that it does not think that the only relevant alternatives are either cosmopolitanism about distributive justice, or the denial of global principles

Justice Across Borders    623 of distributive justice; it is open, instead, to a wide variety of possible principles of distributive justice, in a number of distinct political and social contexts. As such, the post-​ cosmopolitan question can begin to ask a variety of smaller questions, the answers to which will produce distinct views about how to understand the nature of international distributive justice. I will focus here on five distinct questions, although no doubt many others could be added: (a) Are any relationships necessary for valid principles of distribution to be asserted? We can distinguish, here, between relational principles and non-​relational principles (Sangiovanni 2007). Relational principles demand that a particular relationship exists before some particular principle of distributive justice can be applied. Thinkers whose distributive principles rely only upon humanity, however, can insist that distributive principles must rest simply upon status as human, regardless of particular relationship (Caney 2005; Beitz 1983). (b) If there are some required relationships, what—​and where—​are they? There are two parts to any satisfactory answer to this question. The first argues for the particular relationship as a relevant site of distributive justice, following on the answer to our first question. The second seeks to establish where, in our present global circumstances, these relationships are instantiated. (c) Who are the agents whose rights are to be adjudicated by the principles we develop? Andrea Sangiovanni divides between globalists, who are interested in developing a set of principles on which individual persons can at the global level make claims upon one another; and internationalists, whose primary concern is on adjudicating the legitimate rights and duties of collective agents like states (Sangiovanni, 2007). The distinction highlights a simple fact about global distributive justice; there is a wide variety of potential agents whose claims might be considered in a discussion of global distributive justice, and we must ask not only what justice demands, but whose demands are to be pressed against whom, and there is a wide array of global agents—​from persons to states to transnational corporations—​to be considered in this task. (d) What is the proper currency for discussions of global distributive justice? The question here reflects the great difficulty in making comparisons, across cultures, of how different people’s lives are going. Distributive justice begins, not simply with the fact that some people’s lives are going poorly, but with a comparison between those people and others whose lives are going well. It thus requires some notion of what, precisely, we ought to examine to make these comparative judgments; and this has proven to be rather difficult. We might, for example, consider global equality with reference to Rawlsian primary goods, to capability sets, or to opportunities (Beitz 1979, Nussbaum 2006, Caney 2003, respectively.) None of these is obviously wrong, and none of them is without substantial problems—​ most of which emerge when we try to figure out how to compare the experiences

624   Michael Blake of persons in radically different social contexts, given the fact that most goods acquire meaning only within local cultural contexts (Miller 1995). (e) What distributive principles would be justified, in the international realm? This question is needed because most egalitarians accept a principle that isn’t quite reducible to equality. Rawls, after all, accepted deviations from equality when those could be shown to benefit the least advantaged member of society, domestically (Rawls 1971). Once we accept this, however, we can see that a wide variety of possible distributive principles—​from sufficiency, to Rawls’s own principles, to something more stringent—​might potentially be applied at the global level; global distributive justice does not necessarily demand the global equality of anything in particular. I cannot hope to discuss all the possible ways in which these questions might be answered; I  want to emphasize only that these specific questions must inform the broader question of global distributive justice, and that theorists of global distributive justice can be distinguished from one another based upon how they would answer them. Recent work in global distributive justice has begun to develop particular answers to these questions, and I want here to focus on four especially important ideas that have emerged. These ideas are notable in that they focus on what is particular to the global realm—​on, that is, the second question we have discussed: they begin by examining what institutions and agents exist within that realm, and then argue from those facts towards a particular view of global distributive justice. (1) Causation. We might think that distributive obligations emerge because the global order, understood as a particular set of institutional rules, has been imposed on the global poor, and is responsible for the perpetuation of their poverty; distributive obligations, on this account, might be made with reference to the specific facts of our shared and violent history, and to the ways in which that history has been enshrined in international law (Pogge 2002; Hassoun 2012). The distributive obligations produced here may or may not be cosmopolitan, in the sense discussed here. They will, however, undoubtedly regard the particular set of global institutions as a relevant relational set for the application of norms of distributive justice, and will likely condemn much of our current global distribution of resources. (2) Public reason. Even if the global context is not ruled by a government, it might nonetheless be true that it is a site within which reasons of a particular character are given in justification of particular forms of action. These facts might be sufficient to ground some particular forms of distributive principle (Cohen and Sabel 2006). What distributive principles are defended, again, may or may not be cosmopolitan; they will, however, begin with the fact that the world is now subject to the governance of (if not governed by) a set of global institutional relationships. (3) Cooperation. Whether or not the global community is a Rawlsian basic structure, it is undoubtedly a site within which goods are both created and distributed; we

Justice Across Borders    625 might develop principles that are appropriate for that site, and these might include constraints on relative wealth and poverty (Moellendorf 2009). The distributive norms developed by this methodology may be cosmopolitan or not. The basis of the application of those norms, though, is the fact of cooperative production of scarce and useful human goods; as such, there is a distinct relational basis for the application of these norms. (4) Self-​government. The state is morally significant because of how it relates to the wills of the individuals governed by that state; as such, some principles demanding at least the mitigation of resource shortfalls might be inferred when those shortfalls interfere with self-​government (Rawls 1999.) These views tend to produce anti-​cosmopolitan conclusions, given that the nature of the relationship specified as the basis for the application of distributive norms—​self-​government by a shared legal state—​is one that is generally thought inadequate as a description of the global institutional set. There is a very great deal to say about each of these ideas. There are, naturally, philosophical objections to each; each depends upon particular interpretations of the moral language of rights, justice, and equality. I want to note, however, that each also depends upon particular things being actually true about the relationships that do, in fact, hold within the global realm. For these theories to be materially applicable to global justice in the world as we know it, they must actually describe that world in an accurate way. This means, of course, that it is possible that some debates within political philosophy cannot in fact be solved by political philosophers: they require consultation with, and the willingness to be corrected by, empirical scholars (Blake 2012). The reason for this should be clear: what institutions we have matters, on most theories of global distributive justice, and we can only understand these institutions when we understand what they do and what they can do. To see this, note that there are two ways in which we might understand the World Trade Organization (WTO)—​a noble and a base interpretation. Noble: the WTO does, in fact, provide benefits to all member states, by adjudicating disputes between them, and setting rules for mutual benefit. Member states have some reason to regard the decisions and rules of the WTO as authoritative over their actions. Base: the WTO is, in fact, simply the name given to a set of trading policies and rules that benefit the Global North, and the United States in particular. Member states have no particular reason, other than self-​interest, to regard the decisions and rules of the WTO as having authority, any more than they would regard the trading proposals of (say) the United States as having legitimate authority. The point of this example can be brought out by two considerations. The first is that which interpretation is the better one matters, from the standpoint of justice. On the noble interpretation, it is possible—​if arguments that begin from public reason, for example, are defensible—​that the WTO is a site of justice, and that some specific distributive duties emerge only between states who are members of the WTO. (Note, here,

626   Michael Blake that the noble interpretation does not entail that the WTO is, currently, justified; it only entails that the WTO is the sort of institution to which questions of distributive just­ ice might form part of how a justificatory story would go.) If the WTO were to suddenly cease to exist, the duties of the states involved would change as well: they would lose, perhaps, one set of reasons to care about distributive justice. On the base interpretation, though, none of this seems right. The base WTO is incapable of creating any actual moral duties on the part of the states coerced by the Global North, and it would be hard to see how the WTO’s existence would significantly alter the rights and duties of the member states. The WTO is, on this interpretation, simply a gunman writ large, and the gunman’s threat is presumptively wrong, and it seems odd to think that the response to this wrong is to start looking at distributive justice. (It is not to say that the WTO is always wrong; sometimes gunmen are justified, although not as often as they claim.) The states in question, of course, might have rights of distributive justice grounded in some other way—​but they would have these whether or not the WTO itself collapsed, and the WTO itself does not add much to the story of distributive justice (Blake 2011). If we ground global distributive justice in global institutional relationships, then, it matters crucially what those global relationships are, and how they ought to be understood. The second thing to note, though, is that getting an adequate understanding of these relationships is exceptionally difficult to do. The question of whether the real WTO is more like the base or noble versions we have described is a hard question to answer; I suspect neither is a fully accurate portrait of the WTO we have built; I suspect that at different times it has been best interpreted in different ways, sometimes noble, sometimes emphatically base. Our judgment of the WTO should, though, track what the WTO is—​not what it might, in another world, be. If the WTO is generally simply the letterhead on which the United States makes its coercive trade proposals, then we ought to regard it as such, and our evaluation of the United States’s actions ought to simply abstract away from the WTO as an institution. If the WTO is frequently better than that, or if it begins to display some actual power to constrain even the superpowers that created it, then we might begin to think that there is some reason to regard it as an independent institutional form, relevant from the standpoint of distributive justice. The point is, though, that I simply don’t know, and that puts an understandable cramp in my ability to do philosophy. This means, then, that we have arrived at a third question, which necessarily involves the skill of people who are not philosophers: what sort of world are we actually living in?

29.3  What Sorts of Institutions do We Actually Share? I have argued elsewhere that there is an implicit empirical backdrop to our thinking about political justice; we know, or think we do, about what states (and people) are

Justice Across Borders    627 actually like, and what sorts of things states (and people) can be legitimately asked to do. I have also argued that these implicit empirical assumptions are rarely raised in the domestic context, because the very ubiquity of the state system hides from us the fact that our theorizing holds a host of assumptions about what we can expect from political society (Blake 2012). Domestically, that is, we all know (or think we do) what sorts of things a state can do, and what it can be held accountable for not doing; very few of us would blame the President of the United States for (say) being unable to effectively eliminate heartbreak within the American polity. Our arguments about political justice proceed against this backdrop of empirical agreement. Internationally, though, things are considerably more controversial, not just at the level of principle but also at the level of practice. There are at least two ways in which we might think that empirical facts are relevant to global justice: we have reason to ask what it is that international institutions actually have done, and what it is that they are capable of doing. Call these questions the question of culpability and capability, respectively. The issue of culpability can be dealt with by reference to the preceding discussion of the WTO. It matters, we think, whether or not the inequality we are trying to understand is rightly laid at the feet of the WTO itself, or at a powerful nation like the United States. Neither one, I should emphasize, tells us that the inequality is justified; they do, however, tell us markedly different things about how the inequality is to be evaluated, and rectified. If the WTO is itself an institution that can be regarded as having produced the inequality, by failing to be what it could and should have been—​then perhaps our best response is to say that the WTO’s members have specific duties of justice towards one another; perhaps, further, some idea of distributive justice could be developed from these materials. Thomas Pogge notes, for example, that the marginal members of the WTO are often not even wealthy enough to have regular representation at WTO meetings; it would not take a great deal of philosophy to argue that some constraint on that sort of inequality might be mandated simply in virtue of the WTO’s need to be justified (Pogge 2004). If the WTO is, instead, not really an institution at all, but a shell game—​if the will of the WTO is just the will of the United States, spoken through a proxy or puppet institution—​then our response to this will be correspondingly different. We might argue that there is something deeply unjust about an American government that went to the limits of its power in the attempt to extract concessions when bargaining with others (James 2012). We might condemn the American government for interfering with the self-​determination of foreign societies. We might think that the state system itself ought to be revised. We might think any number of things, in short, that will be subtly different in how the malign situation is described, whose behavior will have to change, and what sorts of distributive principle (if any) are invoked. I raise these issues, again, to point out that how we explain a given result will often implicitly involve an empirical statement about what is really going on in the world of global politics—​and that figuring out that question, if indeed it is capable of being figured out, is a question that cannot be asked by a philosopher using philosophical skills. We have reason to think that, at the very least, professional modesty should make us aware that we are liable to be corrected by those who understand global institutions

628   Michael Blake and economic relationships better than ourselves. The relevance to this for questions of global distributive justice should be obvious: the questions we must ask are now not simply more complicated than philosophers used to ask—​they are, instead, questions that philosophers themselves, in isolation from other academics and practitioners, cannot even in principle hope to answer well. This may mean that global justice is, in fact, not simply a set of philosophical questions: it contains a variety of questions, some philosophical, and some emphatically empirical. Political philosophers can, perhaps, provide clarity about moral arguments and commitments; they cannot hope to offer final or definitive policy prescriptions. I think something similar might be said about the issue of institutional capability. For us to assert that a given agent or institution is wrongful in failing to do a thing, we must be able to assert correctly that there is something that it both ought to have done and could have done. This is a fairly simple idea; the concept that ought implies can is hardly a new one. It is important, though, to recognize how important this idea will be for discussions of global justice. One very important idea here, associated both with the later Thomas Pogge and with some international organizations, is that the poverty of the Global South is not merely an accidental result of global capitalism; instead, the global institutions of law and politics have actively imposed a life of poverty on these people, such that the global system of politics represents the most large-​scale violation of basic human rights in human history (Pogge 2002; Watkins 2005.). Part of the way in which this system has ensured poverty, argues Pogge, is through rules that give any local despot the right to sell off minerals, regardless of whether or not that despot cares to respect the human rights of the local citizenry. On Pogge’s view, this is simply incentivizing despots, and encouraging ambitious warlords to do violence against those who might stand between them and political power. A rightful international system, which refused to give the right to trade in natural resources to such tyrants but only allowed such rights on condition of respect for human rights, would prevent such results; since we have not created that system, we all share in the blame for the avoidable deaths that our present system has produced (Pogge 2010; see also Buchanan 2004). I do not want to say that Pogge’s account here is wrong; I only want to say that it is contingent upon certain empirical facts being true—​and that I am not confident that they are. Look, for example, at the Dodd-​Frank Wall Street Reform Act of 2010, which prevented Americans from trading in Congolese minerals when such minerals were produced in such a manner that they fueled violence and conflict. This Act was, in a way, a move away from our current system, and towards a world more in line with Pogge’s principles of recognition. The Act, though, has not been effective at producing a more just distribution; instead, given that the Act made the market for natural resources less lucrative, mine owners moved their capital out of the Democratic Republic of the Congo, with the result that the miners themselves were made much worse off than they had been to begin with (Rubenstein 2014). I raise this to show a simple possibility: if ought implies can, we must have a can available to us—​a way in which the evil in which we are currently implicated could have been avoided. This empirical question, though,

Justice Across Borders    629 is rarely simple, and it is always possible that even well-​intentioned legislation (like the Dodd-​Frank Act) might in fact create a world worse than the status quo. This example is not effective if it is intended to disprove Pogge’s view; he has in mind, of course, a more global set of rules, on which global capital would find it considerably more difficult to simply pack up and leave. Pogge wants, in other words, to find an alternative global institutional system, within which actions such as those of the Congolese mine-​owners are effectively constrained. The main point, though, is not to say that Pogge and those who argue in this way are wrong, but to say that the truth of their argument depends upon whether or not they are able to describe a world in which the evil they have identified can be effectively avoided. I cannot hope, here, to describe fully what would count as an adequate description of such a world. I would expect, at the least, that it would include some description of how such a world would be stable—​how it would, that is, generate its own support amongst those ruled—​along with some notion of how we could effectively and justly move from where we are now to a more just global society. All this, though, depends crucially upon understanding what people are actually like, both as individuals and in the aggregate: a theory of justice for angels might be interesting as an exercise, but it cannot provide us much in the way of guidance. The truth of Pogge’s argument, in other words, depends upon empirical fact as much as philosophical ones, and empirical fact is a damned tricky thing to predict. The ideas I present here reduce to the following claim: when we argue about what relationships and institutions give rise to duties of distributive justice, we are always implicitly arguing against a backdrop of assumptions about the relationships and institutions we actually have. We are making assumptions about what results these institutions have actually produced; we are making assumptions about what sort of a world could be produced by changes in those institutions. This means, of course, that international justice is now a rather harder thing to argue about than it might at first appear. In the first place, it might be true that theorists of international distributive justice are not really disagreeing with each other, but arguing past each other; they are describing principles of justice for slightly different worlds, each with its own institutional capability and culpability. A further implication, of course, is that philosophy is now rather harder to do as a solo project; empirical social science may now be a necessary part of any valid answer to the question of what, in terms of global distributive justic­e, we are obligated to do (Cohen 2010). All of this, of course, means that we have occasion now to start asking questions that are more vexed than even the already-​quite-​vexed questions of global distributive justic­e. We cannot just ask what justice would demand; we have first to ask what we actually share, and listen to those people who have studied just that. We have to ask, further, what sorts of world we could share, given the need for clarity in understanding the space of possible alternatives. And, I will now suggest, we have to ask an even more difficult question, which is what we actually should share, at the global level. If there is more than one possible world that we could, in fact, aim at building, which is the one that ought to be our target and our talisman? This is, I believe, the next step for theorists of global distributive justice, and it is still relatively underexplored (but see Macdonald and Ronzoni

630   Michael Blake 2012; Caney 2005; Kuper 2004; Cabrera 2004; Held 1995). The reason why it is a necessary question is simple: if we have a particular principle of global distributive justice, it must be implemented by some set of institutions or other. We have reason to ask if the institutions we have now will do the job—​and, if they do not, what institutions could rightly exist that would.

29.4  What Sorts of Transnational Political Institutions are Required for the Rightful Administration of Distributive Justice? For relational views of distributive justice, the institutions we share are related to distributive justice in what might be termed the input end; we have to ask what we do, currently, share, and then see what duties we have. I believe we have reason to consider institutional form at the output end, as well; we need to ask what institutions would be needed to rightfully administer the principles of global justice we have identified. This, of course, makes things quite messy. Many of us, after all, think that what we owe to one another depends upon what sorts of institutions we now share; there may be some specific duties that emerge from shared citizenship, for example, and other duties that emerge from shared membership in particular transnational organizations. The duties we defend, though, might turn out to require new institutions, if these duties are to be fulfilled (Valentini 2011). This means, though, that we now have to work in both directions simultaneously; we must both respect and revise the institutions and relationships that ground our accounts of distributive justice. What we share affects what we owe to one another; what we owe to one another affects what we ought to share. I cannot wade into all the ways in which this process might prove to be difficult. I want, instead, to ask a fairly basic question about what sorts of institutions we might develop, if we defend norms of global distributive justice. Assume for the moment that we are defending globalist norms, on which all people everywhere are entitled to equality of something. They are able to make claims that are best understood as claims of justice, and they (rightly) want those claims to be defended institutionally, rather than simply relying upon the virtue of others. How could the global institutional system be organized, in order to respond to this global demand of justice? I think we might identify three representative positions, here; doubtless there are others, but these will do. (1) Minimal solution. The distributive principles we have in mind could best be administered by the states of the world as we know them. The fact that these states

Justice Across Borders    631 are limited in their territorial and personal jurisdictions does not preclude them from acting in the name of global justice (Beitz 1979). (2) Maximal solution. The distributive principles we have in mind could best be administered by some form of global political society, understood as having the right to directly and coercively act in favor of global equality (Cabrera 2004). (3) Moderate solution. The distributive principles we have in mind could best be administered by a global federalism, in which some sovereign authority exists both at the global and at the local level (Pogge 1992). All of these positions, to be sure, have their attractions; none of them is without difficulty. The problems emerge, I think, when we reflect that global distributive justice requires not just true principles of global distributive justice: it requires the institution that acts in defense of those principles to act rightfully in doing so. The fact that something might move us closer to a just distribution, after all, is not taken in domestic justice as a sufficient reason to regard an act or policy as rightful. We are morally precluded from publicly cutting off the heads of tax evaders, even if it could be shown that a vigorous program of public executions would produce a more just distribution of wealth; some things we cannot do. Some other things, only some legitimate agents can do; Warren Buffet doubtless has the money to hire mercenaries to impose the difference principle within Nigeria. He is, though, forbidden from doing so, both by Rawls and by common-​sense morality. Our question, then, is what institutions have the right to seek a just distribution, and what means they might employ to get us there. We therefore have reason to ask, not just about the distribution that is produced, but also about whether the institutions in question use coercive force rightly when they seek to bring that distribution about. On this score, I think the minimal solution fares poorly. On the one hand, it seems to make the successful pursuit of justice dependent upon the virtue of individual wealthy states; such states must hand over goods to the less well favored, and since there is no institution more powerful than those states, they must do so because of the moral rightness of the act itself. Virtue, though, is hardly the most effective long-​term basis for political equality; Kant’s political philosophy—​and the modern republican movement in political philosophy—​have convinced many of us that it is a bad thing, not just prudentially but morally as well, for our basic rights to await the voluntary choices of the powerful (Pettit 1999). More broadly, though, we can say that the effective administration of justice might require the effective ability to participate in electoral politics. Most of us think that political justice requires, at a minimum, that those who are coerced by the machinery of politics have their interests and rights represented in whatever legislative procedure gives rise to legal norms. To reduce states to mere administrative units, though, seems to fly in the face of this idea: states would be responsible for taxing only a limited subset of the world’s population, but would not have any special deliberative obligations to that subset. They would not be understood as communities of people ruling, and thereby imposing obligations upon, themselves; they would, instead, be simply sites within which a more abstract conversation about global right ought to be

632   Michael Blake pursued. It is not hard to see why this would pose some problems. It is one thing to tax Peter to pay Paul; those of us who are not libertarians accept, in principle, that this might be sometimes right to do. It is quite another to tax Peter and pay Paul, without giving both Peter and Paul the political rights required to negotiate the terms of that payment in the legislature. Democracy requires, at the very least, some idea that those whose lives are coercively organized by political institutions shall have some meaningful role in controlling how that institution is run. To reduce states to merely administrative agents, in contrast, seems to leave us without any meaningful role for democratic practice. States are, on this view, not required to offer special forms of justification to those over whom political power is exercised; they are, instead, obligated simply to do what is in the interests of global justice, broadly speaking. Democratic rights, though, seem essential, even in the justification of the most benevolent forms of distributive institutions. It is not surprising, in view of these worries, that many have started to be attracted to a more transnational form of democracy—​one in which both Peter and Paul might be given both substantive rights of distributive justice and procedural rights to speak within the same deliberative body. A perfect (and perfectly just) series of unelected despots might succeed in making the world’s economic distribution perfectly just; few of us would think they had the right to do so. The latter two solutions—​the moderate and the maximal—​are not without flaws, either, and these two solutions have many flaws in common. They begin, again, with the conditions of rightful political rule, and with whether or not such rule could be successfully organized at the global level. Take the maximal solution to begin with. Political philosophy has long been hostile to the idea of the world state; despite the obvious importance of global problems, such as climate change and control of warfare, few of us have been willing to accept the need for a single global political society (but see Einstein 1956). The reasons for our hostility begin with considerations of high theory, but also include elements of simple pragmatism as well. On the former, we can note that the successful administration of political justice includes not simply well-​designed rules, but also a population that sincerely accepts the authority of those rules, and is willing to voluntarily sacrifice so that the rule-​making authority is perpetuated. Rawls refers to this significant human achievement with the rather dull word stability; the idea, though, is a rather dramatic one—​namely, that societies can be destroyed not simply by bad rules, but by people who refuse to regard the rules as worthy of principled respect (Rawls 1993). If this is right, though, then we have to ask whether or not it is going to prove possible for people as we know them to have the personal attitudes towards a global polity that have ensured the survival thus far of individual local polities (Stilz 2011). I will discuss this question in more detail in a moment; right now, I want to notice only that these worries affect the moderate solution as much as the maximal one. There is no single moderate solution, of course; there are any number of distinct federal forms that might be developed to give specificity to this global ideal. I imagine, though, that they would all include some federal structure, on which some questions were to be dealt with by global political institutions, and others delegated to the subsidiary institutions of the national state. (We might, alternatively, imagine that there could exist a distinct

Justice Across Borders    633 global institutional set of diverse institutions such as the WTO and the United Nations; even this, though, would have to exist within some specified account of global federalism, so that these institutions do not simply become captured by the interests of superpowers like the United States.) The moderate solution thus limits the power of the global state with reference to some specified federal structure, on which some specified forms of power are reserved to the local polity. This is, I think, all to the good, but we still have to ask even about the rightfulness of this limited grant of transnational political authority. The problem is that we cannot quite rest assured that coercive political power, if granted to a global body, is capable of being exercised rightly, in accordance with the political and distributive rights of the world’s inhabitants; this problem persists even when the space over which that coercion is exercised is somewhat limited. I will say, once again, that I believe there are significant problems with the idea that political society could emerge at the global level that could be democratically legitimate; certainly, I do not believe that any institutions we have yet developed are both global in reach and democratic in their deliberative functions. Walter Sulzbach, writing at the end of the Second World War, argued that the success or failure of the United Nations depended upon whether individuals could be brought to regard themselves as citizens of the world, so much so that the assembly of the United Nations would be a site of multiparty democratic negotiations. Sulzbach’s idea, in brief, was that democratic legitimacy would require the existence of parties that conceived of themselves as representing the same set of people, but which differed at the level of practical and theoretical politics. Only this infrastructure could guarantee a discursive politics that actually produced democratic legitimation, and only a sense of national (or, in this case, transnational) identity could produce this sort of democratic politics (Sulzbach 1945; Blake 2015). Viewed through the lens of the past sixty years, the idea that the General Assembly of the United Nations could become a site for discursive politics of this sort seems almost absurdly ambitious; I believe, though, that Sulzbach was quite right about the need for democratic legitimation, and that this sort of legitimation will require the creation of a new form of consciousness. I should pause for a moment to emphasize why we are asking this sort of question in a handbook about distributive justice. The idea I am defending is that we cannot ask about questions of distributive justice except in conjunction with how these norms of distributive justice could be institutionalized and made real. This, though, means that we cannot ask about distributive justice and legitimate political authority in isolation from one another: the questions must be asked together. An institution that is tyrannical, in other words, cannot be defended as just, even if it produces distributive results we find pleasing; and we have, I think, an inadequate vision of how we might make the world more democratic and more equal in distribution at the same time. One response to this, of course, is to say that we ought to change both the institutions we have and the people we are at the same time. A global society with global legislative reach, for example, might well be democratically legitimate, if it were peopled by individuals whose primary allegiance was global, rather than local. Imagine, in other words, that people had the same loyalty towards humanity that they often exhibit towards their own local tribes and nations; imagine that people became, in Melville’s

634   Michael Blake lovely phrase, patriots only to heaven—​or, at least, to humanity (Melville 1851). Could we not build cosmopolitan citizens, as we build a cosmopolitan world? What is wrong with this, as an image of a future? The answer, of course, is nothing at all; it is a lovely image, and I think it is perhaps the most attractive vision we could have of a just world. My only worry is that this, as a vision, is so far removed from humans as we know them that we can gain no guidance from this form of theory. I think we might, at this juncture, identify a sort of negative nationalism that stands in the way of global justice. A positive nationalism identifies the national community as important, so much so that a specific concern for one’s fellow nationals is actually morally rightful (Walzer 1986; Miller 1995; Tamir 1993). A negative nationalism, in contrast, identifies nationalism as only an especially persistent human failing. On this account, we are simply bad—​predictably, persistently bad—​at being motivated by humanity in general. This does not make nationalism, or national preference, or the unwillingness to create and sustain transnational institutions permissible; it just makes these failings likely. We should bear this in mind in our theorizing. We can take facts about global institutional sets as subject, with some difficulty, to collective alteration; we have good reason to think that we should get our facts straight before remaking the world, but the world can (and should) be remade. Humanity, though, may be somewhat more recalcitrant. A political theory that ignores what human beings are like—​what they have been, and are likely to continue to be—​is unlikely to be of any use to us. I worry that this failing is so central a part of the human experience that we are likely to face almost insurmountable odds in developing a truly justified response to global inequality. A  truly just response—​just on both political justice, and on distributive justic­e—​might require truly transnational institutions, and those institutions might only be possible for people quite unlike ourselves. Liberal governance, again, requires the willingness to lose, to be motivated by the interests of those with whom one shares a political society, and the will to perpetuate that society when one might disrupt it through threats. It requires liberal citizens who care about that society, and are deeply attached to its institutions. The best-​case scenario for such a consciousness existing at the transnational level—​the European Union (EU)—​is facing a fundamental existential crisis; many individual European citizens are deeply ambivalent about the extent to which they are obligated to other Europeans (Kenny 2015). If the EU—​a group of developed societies, who share enormous cultural and political similarities, including a commitment to liberal democracy—​cannot create its own transnational consciousness, is there any hope for an institution including the whole world? There are times when philosophy should admit that it is working in the dark; this is one of those times. It is entirely possible that a political society could emerge at the global level; for it to work, it would require people to be not simply ruled by global norms, but to be motivated to perpetuate those norms. They would have to have a consciousness of themselves as motivated by humanity generally, rather than by any more local affiliation. They would have to have the opportunity to participate—​and participate on terms of global equality—​in the process of political self-​government. Whether this is possible

Justice Across Borders    635 or not is a question that cannot be answered by philosophy alone, and it is entirely possible that no answer can be provided by anyone. My own belief—​and it is only that—​is that we might, if we cannot rightly hope for a true transnational consciousness, face some rather unpleasant choices regarding global distributive justice. We can, perhaps, accept a somewhat illegitimate form of transnational coercion in the name of global distributive justice; proposals to have a transnational tax administered by a transnational body might fall under this heading (Pogge 2002). We could, instead, rely upon voluntary cooperation and individual agents being motivated by virtue, and hope that this tends to lead towards something like global distributive justice. Neither choice is especially pleasant. In both cases, we are far away from where we ought to be, and without any particularly good plan for how to get there. This leads, then, to the final question with which I think theorists of global distributive justice ought to be concerned. Whatever the ideal is to which we are willing to sign on—​whatever currency we accept, whatever relationship or institution we think gives rise to consideration of distributive justice—​we have to ask where we should go from here to get there. On one reading, this is a rather silly question: we move towards our ideal—​that is, after all, the point of having ideals. It may be, however, that the world is organized so that we cannot directly move towards where we would most want to be. I have already emphasized, I think, one way in which this is true: there may not be a global ideal that is compatible with people as we know them. I believe, though, that we may be even more limited than that in our attempt to make the world more just.

29.5  What Should We Do, Here and Now, to Bring about Distributive Justice? Imagine, once again, that we have a defensible ideal of global distributive justice. Imagine, further, that this ideal is one that does not involve the creation of political systems that will involve injustice; people have become such that they are able to be motivated by global concerns, and the global institutions we imagine can generate their own support. We have, in short, developed a plausible vision of what justice would look like at the global level. This is part, we might think, of ideal theory; we imagine that the institutions we propose are just, and that people can be counted on to be motivated by justice, and that they will voluntarily do their part to sustain these institutions (see Schmidtz, Chapter 15, this volume, for a more full account of ideal and non-​ideal theory). Ideal theory is not, and should not be, unaware of the imperfect and limited nature of human altruism, and human reason; Rawls’s notion of the burdens of judgment is not part of non-​ideal theory (Rawls 1993). Imagine, though, that we have identified a world that could be just, for creatures such as ourselves, and that we can be counted on

636   Michael Blake to sustain these institutions once they are in place. These institutions are, in other words, stable for the right reasons, and it only remains for us to figure out how to get from the unjust here to the fully just there. Is our problem now only one of political prudence? Do we need only discuss which of the many paths to justice is the most likely to be effective? I think we need to discuss a very great deal more than that. Sometimes, we may have to face particularly alarming choices about distributive justice that our focus on ideal theory has somewhat obscured. These choices emerge when there is no morally cost-​ free path from where we are now to where we ought to be. I believe there are at least two ways in which we might face such a problem. The first—​which I will call the problem of moral tragedy—​exists when there is no way towards justice that does not involve injustice. The second—​which I will call the problem of moral triage—​exists when there is a choice to be made now about whose claims in justice are most important. I will discuss both problems in turn. The concept of moral tragedy exists when we are in a morally sub-​par situation, and in which there are no morally permissible moves away from that situation. Imagine, for example, that we could only achieve equality of opportunity through the elimination of the family itself. We are not, therefore, commanded to eliminate the family, and move to a system of centralized child-​rearing; people have rights, and they include the right to raise children, even if the end result is that some children will necessarily end up unjustifiably disadvantaged as a result (Brock and Blake 2015; Rawls 1971; Fishkin 1984). This is tragic, I think, because the ideal of equality of opportunity cannot itself guide our actions; the only way we would have to achieve perfect equality would attack other deeply held moral norms, and we have to ask which norm is the one we are willing to abrogate. I think something similar might be true about global distributive justice. What if, for example, the only effective way to reduce African poverty—​to stop the violation of whatever norm of distributive justice we endorse—​is to violate certain important rights? Paul Collier, for instance, has argued for “over the horizon peacekeeping”—​on which wealthy Western societies reserve the right to simply kill African leaders who are insufficiently respectful of the rights of their citizens (Collier 2010). Even if this were effective—​and there are reasons to be hesitant—​the fact of its leading to good results is not enough for us to rest easy with the program. As William Easterly notes in reply, this is an explicit return to the colonial project, on which white Europeans propose to shoot Africans, and claim that the Africans themselves will ultimately benefit; it is hard not to notice the ethical problems with such a proposal (Easterly 2009). My broader worry, though, is that we might be in a situation in which any effective solution to global inequality might require treating some people unfairly. What do we do, when we can only make the world right by treating some portion of it wrongly? Moral tragedy, of course, is not an excuse for moral inaction; we might have an obligation to do what we can, to accept even the possibility of dirty hands in the course of working for equality (Walzer 1973). My only purpose here is to say that, at the very least, such questions will be a necessary part of any full response to the facts of global inequality.

Justice Across Borders    637 Moral triage, in contrast, involves two legitimate paths—​but a choice between the paths that requires us to decide whose claims are the most important. Sometimes we must ask whether we want to move closer to the long-​term goal of justice, for example, or whether we most want to alleviate those afflicted by injustice in the here and now. Take, for example, Mathias Risse’s discussion of fairness in trade (Risse 2007). We might, by making wages and living conditions better for coffee farmers in the developing world, make life better for those who are now most in need of global rectification. Risse criticizes the easy inference, though, that fair trade policies are therefore a good thing from the standpoint of global justice. It is at least possible, after all, that making things better for coffee farmers now also perpetuates the global economic relationships that ensure that the residents of those developing countries will continue to be coffee farmers in the years to come. We might, then, be making things better for the global poor in the short run, with the effect in the long run that global poverty itself is made more resilient. Neither Risse nor I want to make the claim that this is therefore a sufficient reason to condemn the Fair Trade movement generally, nor to condemn the practice of Fair Trade coffee. The question, instead, is more general. We have a genuine moral question about what to do when there are two pathways towards our distributive goals, and neither of them is without some truly awful implications. We have, I think, reason to worry that the road to global justice will involve a great many of these questions; we are, I believe, quite ill prepared to answer them. I cannot, at present, offer any general summing-​up of the literature on global distributive justice. I have argued, to begin with, that it has become remarkably sophisticated in the past few decades, as we have developed some exceptional and well-​developed analyses of how we might understand the distributive norms that rightly apply to the transnational realm. I have also argued that the questions raised by this literature must become broader still; theorists of global distributive justice must engage with empirical social science, and with questions of political justice. My final thoughts, as should be obvious from the preceding discussion, are not especially hopeful; I do not believe we have any easy pathway towards justice, and we are unlikely to work effectively for just­ice without having to make hard choices, and risk doing wrong even in the course of seeking what is right. I would, however, end by reiterating that nothing I have said here should be taken as a defense of the status quo. The world as we live in it is a site of tremendous injustice, on any plausible account of what global distributive justice would require. The fact that the way forward is full of moral danger does not give us reason to avoid the job.

References Barry, B. (1982). “Humanity and Justice in Global Perspective.” NOMOS XXIV:  Ethics, Economics and the Law. New York: Harvester Wheatsheaf. Beitz, C. (1979). Political Theory and International Relations. Princeton, NJ:  Princeton University Press.

638   Michael Blake Beitz, C. (1983). “Cosmopolitan Ideals and National Sentiment.” Journal of Philosophy 80(October): 591–​600. Blake, M. (2001). “Distributive Justice, State Coercion, and Autonomy.” Philosophy and Public Affairs 30(3): 405–​24. Blake, M. (2011). “Coercion and Egalitarian Justice.” The Monist 94(4): 555–​70. Blake, M. (2012). “Global Distributive Justice:  Why Political Philosophy Needs Political Science.” Annual Review of Political Science 15: 121–​36. Blake, M. (2013). “We are All Cosmopolitans Now,” in G. Brock (ed.) Cosmopolitanism Versus Non-​Cosmopolitanism. Oxford: Oxford University Press, 35–​54. Blake, M. (2015). “On Walter Sulzbach’s ‘Some Basic Problems of a League of Nations.’” Ethics 125(3): 833–​5. Brock, G. and Blake, M. (2015). Debating Brain Drain: May Governments Restrict Emigration? Oxford: Oxford University Press. Buchanan, A. (2004). Justice, Legitimacy, and Self-​Determination:  Moral Foundations for International Law. Oxford: Oxford University Press. Cabrera, L. (2004). Political Theory of Global Justice: A Cosmopolitan Case for the World State. London: Routledge. Caney, S. (2003). “Cosmopolitan Justice and Equalizing Opportunities.” Metaphilosophy 32: 113–​34. Caney, S. (2005). Justice beyond Borders. Oxford: Oxford University Press. Cohen, J. (2010). “Philosophy, Social Science, Global Justice,” in A. Jaggar (ed.) Thomas Pogge and his Critics. London: Polity Press, pp. 18–44. Cohen, J. and Sabel, C. (2006). “Extram Republicam Nulla Justitia?” Philosophy and Public Affairs 34(2): 147–​75. Collier, P. (2010). War, Guns, and Votes: Democracy in Dangerous Places. New York: Harper Perennial. DeBres, H. (2012). “The Many, not the Few:  Pluralism about Global Distributive Justice.” Journal of Political Philosophy 20(3): 314–​40. Easterly, W. (2009). “Forum: Development in Dangerous Places.” Boston Review, July 10. Einstein, A. (1956). “Towards a World Government,” in Out Of My Later Years: The Scientist, Philosopher, and Man Portrayed through His Own Words. New York: Wings Books. Fishkin, J. (1984). Justice, Equal Opportunity, and the Family. New Haven, CT: Yale University Press. Freeman, S. (2006). “Distributive Justice and the Law of Peoples,” in R. Martin and D. Reidy (eds) Rawls’s Law of Peoples: A Realistic Utopia? Oxford: Blackwell, pp. 243–​60. Hassoun, N. (2012). Globalization and Global Justice:  Shrinking Distance, Expanding Obligations. Cambridge: Cambridge University Press. Held, D. (1995). Democracy and Government:  From the Modern State to Cosmopolitan Governance. Stanford, CA: Stanford University Press. James, A. (2012). Fairness in Practice: A Social Contract for a Global Economy. Oxford: Oxford University Press. Kenny, M. (2015). “The Return of ‘Englishness’ in British Political Culture—​The End of the Unions?” Journal of Common Market Studies 53(1): 35–​51. Kleingeld, P. (2011). Kant and Cosmopolitanism: The Philosophical Ideal of World Citizenship. Cambridge: Cambridge University Press. Kuper, A. (2004). Democracy beyond Borders. Oxford: Oxford University Press.

Justice Across Borders    639 Macdonald, T. and Ronzoni, M. (2012). “Introduction: The Idea of Global Political Justice.” Critical Review of International Social and Political Philosophy 15(5): 521–​33. Melville, H. (1851). Moby-​Dick; Or, the Whale. New York: Harper and Brothers. Miller, D. (1995). On Nationality. Oxford: Oxford University Press. Miller, D. (2003). “Cosmopolitanism: A Critique.” Critical Review of International Social and Political Philosophy 5: 80–​5. Moellendorf, D. (2009). Global Inequality Matters. London: Palgrave MacMillan. Nagel, T. (2005). “The Problem of Global Justice.” Philosophy and Public Affairs 33(2): 113–​47. Nussbaum, M. (2006). Frontiers of Justice. Cambridge, MA: Belknap Press. Pettit, P. (1999). Republicanism:  A Theory of Freedom and Government. Oxford:  Oxford University Press. Pogge, T. (1989). Realizing Rawls. Ithaca, IL: Cornell University Press. Pogge, T. (1992). “Cosmopolitanism and Sovereignty.” Ethics 103(3): 85–​106. Pogge, T. (2002). World Poverty and Human Rights. Cambridge: Polity Press. Pogge, T. (2004). “‘Assisting’ the Global Poor,” in D. Chatterjee (ed.) The Ethics of Assitance: Morality and the Distant Needy. Cambridge: Cambridge University Press, pp. 260–​88. Pogge, T. (2010). “The Role of International Law in Reproducing Massive Poverty,” in S. Besson and J. Tasioulas (eds) The Philosophy of International Law. Oxford: Oxford University Press, pp. 417–​35. Pogge, T. (2013). “Concluding Reflections,” in G. Brock (ed.) Cosmopolitanism Versus Non-​ Cosmopolitanism. Oxford: Oxford University Press, pp. 294–320. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rawls, J. (1993). Political Liberalism. New York: Columbia University Press. Rawls, J. (1999). The Law of Peoples. Cambridge, MA: Harvard University Press. Risse, M. (2006). “What to Say about the State.” Social Theory and Practice 32(4): 671–​98. Risse, M. (2007). “Fairness in Trade I:  Obligations from Trading and the Pauper-​Labor Argument.” Politics, Philosophy, and Economics 3: 355–​77. Risse, M. (2012). On Global Justice. Princeton, NJ: Princeton University Press. Rubenstein, J. (2014). “Misuse of Power, Not Bad Representation: Why It is Beside the Point that No One Elected Oxfam.” Journal of Political Philosophy 22(2) 204–​230. Sangiovanni, A. (2007). “Global Justice, Reciprocity, and the State.” Philosophy and Public Affairs 35(1): 3–​39. Schmidtz, D. (2018). “Ideal Theory,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive Justice. Oxford: Oxford University Press, pp. 327–45. Singer, P. (1971). “Famine, Affluence, and Morality.” Philosophy and Public Affairs 1(3): 229–​43. Stilz, A. (2011). Liberal Loyalty. Princeton, NJ: Princeton University Press. Sulzbach, W. (1945). “Some Basic Problems of a League of Nations.” Ethics 55: 157–​66. Tamir, Y. (1993). Liberal Nationalism. Princeton, NJ: Princeton University Press. Valentini, L. (2011). Justice in a Globalized World. Oxford: Oxford University Press. Walzer, M. (1973). “Political Action: The Problem of Dirty Hands.” Philosophy and Public Affairs 2(2): 160–​80. Walzer, M. (1986). Spheres of Justice. New York: Basic Books. Watkins, K. (2005). Rigged Rules and Double Standards: Trade, Globalization, and the Fight against Poverty. Oxford: Oxfam International.

Chapter 30

Migrat i on Sarah Fine

30.1 Introduction How should we think about the relationship between the ethics of migration and the obligations of distributive justice? There certainly are many migration-​related issues that touch on questions of distributive justice, including for example the under-​theorized subject of the fair distribution of responsibilities to refugees (see Carens 2013: Ch. 10; Gibney 2015). Here I am going to focus on the fundamental issue of the relationship between the freedom to move across state borders and the demands of distributive justice. For some, the freedom to move across borders represents a key requirement of distributive justice, whereas others argue that the demands of distributive justice may justify more or less significant restrictions on international freedom of movement. In this chapter, after outlining the key terms, I will introduce the argument that the freedom to move across borders is a requirement of distributive justice. I will present different plausible versions of this argument, working through the main steps and scrutinizing its assumptions. I will also outline the key challenges it faces. Then I will address a set of arguments that point in the other direction, and which seek to illustrate that the obligations of distributive just­ ice may support limits on the freedom to move across borders. I will argue that those who look to distributive justice to provide us with definitive answers to questions about freedom of movement’s proper scope will be disappointed.

30.1.1 The Terms of the Debate: Migration and Distributive Justice Human migration is the movement of people, either across state borders, or within the borders of a state.1 Humans have always migrated, and will continue to do so. I will 1  Here I am using ‘state’ in the sense of ‘national’ states (e.g. France, Japan), rather than states within a federal system (e.g. Massachusetts, Bavaria).

Migration   641 be focusing primarily on people who move across state borders, with the intention to settle for a period of time in the new destination. That settlement may be shorter or longer term or even permanent, but it is important to note that, in practice, migration is frequently (and increasingly) ‘temporary, repeated, or circular’, meaning that migrants may return to their place of origin after a period of residence elsewhere, move on to another destination, or travel back and forth between two or more countries (Goldin et al. 2011: 2). Migration often is not a once-​and-​for-​all event. The total number of international migrants is currently only a relatively small proportion of the world’s population. In 2013, according to the United Nations’ migration statistics, an estimated 232 million people in the world were international migrants, and this equates to 3.2 per cent of the global population (United Nations, Department of Economic and Social Affairs, Population Division).2 However, the absolute number is growing, and is projected to continue growing in this century and beyond, for a range of factors, including the adverse effects of climate change, which is expected to make various territories inhospitable or uninhabitable over the coming years.3 Furthermore, as we know from the pervasive phenomenon of irregular migration (entering and/​or remaining within a state without the relevant authorization, including violating the terms of existing visas), immigration controls do not deter all those whom states wish to exclude.4 Migration is a subject of urgent relevance to the world in which we live. It poses numerous challenging, pressing questions for political philosophers, including questions about the freedom to move, settle, and acquire citizenship elsewhere, as well as the responsibilities of those who move, and the costs for sending and receiving countries (see the discussions in Fine and Ypi 2016). Distributive justice is a more contentious term. The ‘distributive’ part normally distinguishes the subject in hand from other justice-​related considerations. The focus is on the justice of certain kinds of distributions or methods of distribution or ways in which distributions have come about. Often in contemporary debates, distributive just­ ice is used relatively narrowly, to refer to the distribution of specifically socio-​economic benefits and burdens within a given context or contexts. Yet even from there, the variation in accounts of what distributive justice is and requires remains pronounced indeed. Are the relevant objects of distribution only material goods, such as income, or

2 

This figure is based on estimates of the number of people living outside the country of their birth, or, where that figure was not available, estimates of the number of people living outside the country of their citizenship. It is an estimate of international migrant numbers, and does not include internally displaced people. According to the Internal Displacement Monitoring Centre, since 2008, on average, 26.4 million people have been displaced per year by natural disasters alone. 3  We should be wary of alarmism about the figures. It is worth emphasizing that while the absolute number of international migrants is increasing, so too is the global population. In fact, the proportion of the world’s population who are international migrants has remained at around 3 per cent for over half a century (see Czaika and Haas 2014). 4  In 2012, there were an estimated 11.4 million irregular migrants just in the US (Zong and Batalova 2015). Irregular migration, also described inter alia as ‘illegal’, ‘unauthorized’, ‘undocumented’, and ‘clandestine’ migration, is a complex phenomenon. For further discussion, see Düvell 2008.

642   Sarah Fine also more abstract goods, such as opportunities or well-​being or capabilities (see Risse 2012:  5)? Which are the appropriate distributive principles (e.g. strict equality, need, desert)? How are those principles to be identified? Which kinds of contexts and/​or relationships are proper subjects of distributive justice? Are there different principles for and within different contexts? What exactly generates obligations of distributive justice? And so on. Much of the academic debate about distributive justice has focused on these sorts of questions. The way in which we answer these questions obviously will affect our understanding of the relationship between distributive justice and migration. That much seems clear when thinking about, for example, the precise content and scope of distributive just­ ice obligations. For example, if we think that we have stringent obligations of distributive justice to our fellow citizens but only a far more limited set of obligations to people beyond our borders, then we might think of our obligations of distributive just­ice to our fellow citizens as potential grounds for restrictions on immigration. This partly explains why many philosophers approach questions in the ethics of migration via prior discussions of the demands of justice. A common move is to take a stand on some of the central controversies in the debate about distributive justice (especially on content and scope), and then to think about the implications of these reflections on justice for questions in the ethics of migration (see e.g. Risse 2012: Ch. 8; Brock and Blake 2015). Rather than take that approach, I want to leave open controversial questions about the content and scope of distributive justice obligations, in order to retain as broad an audience as possible while we think through the structure of the relationship between distributive justice and the freedom to move across borders.

30.2  Freedom of Movement as a Requirement of Distributive Justice If we begin by surveying the world around us, in which opportunities are so unevenly distributed across different states, it is appealing to assume that there must be a close connection between the freedom to move across state borders and the pursuit of distributive justice. For a start, we cannot fail to notice that the countries in which we are born and in which we reside have a significant bearing on our prospects in life (see Sangiovanni 2011: 571; Mendoza 2015). A child born in Australia in 2012 had a life expectancy of eighty-two years, whereas a child born in the same year in Sierra Leone had a life expectancy of forty-five years (The World Bank a). That year in Australia, the infant mortality rate per 1,000 live births was four. In Sierra Leone it was 117 (The World Bank b). GDP per capita in Australia was 67,442 USD; in Sierra Leone it was 635 USD (The World Bank c). In 2011, 52.9 per cent of Sierra Leone’s population lived below the national poverty line (The World Bank d). These kinds of disparities in life prospects are replicated, to different

Migration   643 degrees, in comparisons between a variety of more and less prosperous countries, particularly between those of the Global North and the Global South. The causes of these extreme transnational—​and especially North–​South—​differences in prospects are many and multifaceted, and so addressing the causes with a view to improving the conditions of the world’s least advantaged people is no straightforward task. Nonetheless, it is safe to say that one obvious way for disadvantaged citizens of impoverished countries to improve their own and their family’s life prospects is via migration to more prosperous countries.5 In practice, substantial cross-​country differences in life prospects are themselves an important driver (among others) of international migration (Goldin et al. 2011: 220–​4). Moreover, remittances (money that migrants send back to people living in their former country) already dramatically outstrip the money that poorer countries receive in foreign aid and in foreign direct investment, indicating that migration is a significant contributor to development in poorer countries (see Oberman 2015; Provost 2013). At the same time, in our world, ever more desperate attempts to migrate seem to confront ever more determined attempts to control migration. To illustrate this, we need look no further than Australia—​a ‘traditional immigration country’—​which has become renowned for its strict, highly selective immigration and asylum policies, and in particular its dedication to deterring and preventing refugees travelling by boat from arriving on the mainland. In recent years its controversial practices have included turning back vessels containing asylum seekers (Pearlman 2014), holding boats indefinitely at sea (Whiteman 2014), and mandatory detention for ‘irregular maritime arrivals’, often for lengthy periods—​sometimes years. Australia makes use of offshore processing centres in Nauru and in Papua New Guinea, to reduce its own responsibilities towards the detainees, who then are not held under Australian law. Australia has formed agreements to resettle some refugees from its detention centres in other countries in the region, and has arrangements with transit countries designed to deter migrants leaving for Australia by boat in the first place (Fickling and Paton 2013; Crothers and Doherty 2014). In 2015, Australia’s then prime minister, Tony Abbott, urged the European Union to follow Australia’s lead in order to ‘stop the boats’ of refugees and other migrants travelling from Asia and Africa to Europe’s shores (AFP 2015). Additionally, to highlight the relationship in our world between the extent of one’s international freedom of movement and one’s country of citizenship, we know that opportunities to travel across state borders are unevenly distributed, and favour citizens of more prosperous and politically stable countries. Whereas holders of Finnish, German, Swedish, UK, and US passports are permitted to travel visa-​free to 174 countries, holders of Eritrean, Libyan, Sudanese, and Syrian passports can travel visa-​ free to just thirty-​eight countries. Australian passport holders can travel to 168 countries 5  See the discussion in Abizadeh (2016: 105–​6), where Abizadeh argues that ‘migrating to wealthier polities and partaking in their societies often represents one of the best opportunities for individuals from globally poor regions to alleviate their poverty and improve their relative material standing’.

644   Sarah Fine without having to acquire a visa, while a passport from Sierra Leone only grants visa-​ free travel to sixty-five countries (The Henley & Partners Visa Restrictions Index 2014). There is, writes one observer, a ‘strong correlation’ between a country’s ‘average wealth’ and ‘access to open frontiers’ for its passport holders (Calder 2013). Against that backdrop, then, we can easily make sense of Joseph Carens’s claim that wealthier states, whatever their other responsibilities and connections to the countries in question, are complicit in sustaining deep transnational disparities in life prospects in at least one highly visible way: via their immigration controls. These controls are usually designed to keep out those deemed undesirable in the prospective receiving states, including the global poor and unskilled, and to admit on favourable terms those deemed desirable, including the global wealthy and highly skilled (Carens 2013: 234; Shachar 2016).6 One of Carens’s most well-​known and oft-​quoted claims is that ‘citizenship in Western liberal democracies is the modern equivalent of feudal privilege—​an inherited status that greatly enhances one’s life chances’ (Carens 1987: 252). Immigration restrictions remain a means for protecting that inherited status and preventing others from enjoying the opportunities to improve their prospects. Ultimately, Carens argues that open borders ‘would be an important institutional feature of a just world’ and that ‘free movement . . . is part of what global justice requires’ (Carens 2013: 234). I am going to take that bold claim about the connection between free movement and justice as a starting point. I will focus on examining whether freedom of movement across state borders can be characterized as a requirement of distributive justice. Note that freedom of movement across borders could be a requirement of justice, without being a requirement of distributive justice in particular. Freedom of movement across state borders might be a requirement of distributive justice if it is (1) itself a requirement of distributive justice; or (2) necessary for the fulfilment of other requirements of distributive justice, but not itself part of what justice demands. In the course of this discussion we will think more about the nature and scope of the demand for ‘freedom of movement’, and how this connects with the demand for ‘open borders’.

30.2.1 Freedom of Movement as Itself a Requirement of Distributive Justice One possibility is that freedom of movement across borders is itself a requirement of distributive justice, just as one might argue that ‘treating people as equals’ is a requirement of distributive justice. When we think carefully about the relevant requirements,

6  This view of immigration controls is not uncommon within migration research. As migration expert Stephen Castles puts it: ‘the perceived migration crisis is really a crisis in North-​South relations, caused by uneven development and gross inequality. Migration control is essentially about regulating North-​South relations’ (Castles 2004: 862).

Migration   645 perhaps we will see that the freedom to move across borders is just part of what distributive justice demands. And if the freedom to move across borders is just part of what distributive justice demands, then we will be well on our way to defending the claim that open borders (or at least more open borders) are also a requirement of justice. The very mention of the freedom to move across state borders already presupposes the existence of states with borders. We might wish to start without that presupposition and imagine instead a world without borders (either with no states and no borders, or with states but without borders as we know them), and then simply ask how much freedom of movement is or may be a requirement of distributive justice. In a world without borders, would the freedom to move anywhere and everywhere (or at least, to anywhere that isn’t already private property, for example) be part of what distributive justice requires, or would something more limited suffice—​say, the maximum degree of freedom of movement across the globe compatible with everyone having that same amount? The virtue of starting without the presupposition of borders is that it forces us to address the question of why freedom of movement is valuable in the first place, and to think carefully about the significance of the borders in this context. For, as soon as we introduce the existence of states with borders into our account, there is a danger that the borders take on an unwarranted special significance in our minds with respect to questions of movement: should people be free to move within the borders but not necessarily beyond them? We are already familiar with a world in which we expect as a matter of right the freedom to move up to borders, but not necessarily the freedom to move beyond them. In principle, though, it is perfectly plausible that borders do not have any special significance in this regard, and that the freedom to move need not extend as far as the borders or may extend far beyond them. So let us consider why freedom of movement is valuable in the first place. There is clearly value in the very freedom to move around without external constraints (see Carens 1992: 26). But it seems as though freedom of movement is especially valuable for its role in enabling us to access other things that we value and/​or to which we are entitled. We need to be free to move to take advantage of the various opportunities available, to travel, experience new things, be with loved ones, associate with like-​minded people, go to college, practise our religion (Carens 1987: 258; Oberman 2016; Stilz 2016). And we need to be free to move away from unpleasant or potentially harmful situations, to extract ourselves from unhealthy relationships, and so forth. It seems reasonable to assume, given its crucial role in enabling access to other things that we value, that freedom of movement will figure somewhere, to some degree and in some sense, within most accounts of distributive justice. For example, in his theory of justice, Rawls includes freedom of movement as one of the primary goods, which are the ‘various social conditions and all-​purpose means that are generally necessary to enable citizens adequately to develop and fully exercise their two moral powers, and to pursue their determinate conceptions of the good’ (Rawls 2001: 57–​8). Let us assume, then, that it is not controversial to contend that ‘some’ freedom of movement will have an important place in a theory of distributive justice. I think we can even assume that some freedom of movement will be a ‘requirement’ of distributive justice.

646   Sarah Fine The next question is how much freedom of movement is properly characterized as a requirement of distributive justice. Is the freedom to move anywhere we want a requirement of distributive justice? Whatever our conception of justice, we should accept that some limits on the freedom to move wherever we want will be perfectly compatible with the demands of justice, such that the freedom to move anywhere does not itself appear to be a requirement of distributive justice. For example, it would not be difficult to justify some restrictions on access to dangerous places, such as active volcanoes, or to sites of outstanding natural biodiversity, or to world heritage sites which are threatened by the unregulated arrival of human visitors, such as the Galapagos archipelago or Machu Picchu. Regulations and restrictions on moving freely across the globe as we please may require justification, but those justifications may be forthcoming in a variety of cases.7 These justified regulations and restrictions will cover not just where we might travel but also where we might settle. For instance, even if you believe that people should be free to travel as extensively as possible, it would be difficult to deny that preventing people from settling in Machu Picchu is compatible with the demands of distributive justice (at least as long as there are plenty of available places in which people can settle that are not unique sites of world historic value). However, as soon as we introduce the notion that these exceptional and uncontroversial (and non-​ distributive-justice-related) restrictions may be compatible with the requirements of justice, we open up the possibility that various other kinds of restrictions might be perfectly permissible, too. So even if we start without the assumption of a world divided by borders, we quickly recognize that some restrictions on movement can be perfectly compatible with the demands of justice. We can go further than that, though. Let’s take another kind of freedom as an illustrative example. We might think of freedom of expression as a ‘requirement of distributive justice’ in some sense. Yet some restrictions on freedom of expression may be justified with reference to other internal demands of distributive justice itself. For instance, we might argue that it is legitimate to limit the freedom to engage in hate speech about members of religious minorities, in order that the targets of such speech do not become the victims of various kinds of related injustice. Similarly, it could be that some limits to freedom of movement may be justified in the name of fulfilling other demands of justice. In other words, even if we assume that some degree of freedom of movement is a ‘requirement of distributive justice’, a range of regulations on movement will pose no problems from the perspective of distributive justice, and others actually may be justified in the name of other demands of distributive justice. Therefore, at this stage, the answer to the question of how much freedom of movement is properly characterized as a requirement of distributive justice can be narrowed down to this: more than a negligible amount, but not as extensive as the freedom to move anywhere at all. Returning to a world of states with borders, while we cannot claim that the freedom to move anywhere we please must be a requirement of distributive justice, what about the

7 

For an illuminating discussion of legitimate restrictions on freedom of movement, see Stilz 2016.

Migration   647 freedom to move across state borders? At first it may not seem difficult to make the case for the freedom of movement across borders as a requirement of distributive justice. For one, I noted that we value freedom of movement in part because we need to be free to move to take advantage of the various opportunities available, be with loved ones, practise our religion, and so forth. Those opportunities, people, groups, and associations do not stop at the borders, and so why should our freedom to move be constrained by those borders (Carens 2013: 239; Oberman 2016)?8 There is no doubt that we as individuals have an interest in being free to pursue opportunities across borders that we value, and that interest may even ground various duties on the part of others. But should we think of the freedom to move across borders as a requirement of distributive justice? There may be various reasons to defend this freedom, and to want it to be as extensive as possible in as many cases as possible, without thinking of it itself as a demand of distributive justice. Moreover, since we have accepted that some restrictions on the freedom to move will be compatible with the demands of justice, I think we at least need to be open to the possibility that, in principle, there may be (justice- or non-​justice-​related) justifications for the regulation of movement across borders which are compatible with the requirements of distributive justice, such that freedom of movement across borders itself may not be a requirement of distributive justice. Again, this would not necessarily mean that regulation of movement across borders is desirable or even legitimate, all things considered, as there could be a host of reasons for objecting to it (perhaps including justice-​related reasons). But it is not the case that we can move easily from the uncontroversial assumption that some unspecified degree of freedom of movement is a requirement of distributive justice, to the claim that freedom of movement across state borders is a requirement of distributive justice.

30.2.2 Freedom of Movement and Equality of Opportunity Nonetheless, perhaps freedom of movement across state borders could be characterized more straightforwardly as a requirement of distributive justice indirectly, insofar as it is a necessary condition for the fulfilment of other requirements of distributive justic­e. One approach along these lines emphasizes the connection between the freedom to move across borders and a commitment to equality of opportunity, which many people believe to be a requirement of distributive justice. Carens, for example, argues that equality of opportunity is a key principle of distributive justice, and that freedom of movement across borders is ‘essential for equality of opportunity’, because ‘you have

8  In this context, both Carens and Oberman are discussing the question of whether that is a human right to immigrate, rather than whether the freedom to move across borders is a requirement of distributive justice.

648   Sarah Fine to be able to move to where the opportunities are in order to take advantage of them’ (Carens 2013: 227–​8). There are two common objections to this line of argument. Some opponents deny outright that equality of opportunity is a requirement of distributive justice. Obviously, if you do not accept that equality of opportunity is a requirement of distributive justice then you will not be moved by the rest of the argument. However, other opponents do accept that equality of opportunity is a requirement of distributive justice in some contexts and within certain relationships (say, within the state and/​or among compatriots), but argue that it is not a requirement of distributive just­ ice outside of those contexts, including beyond borders. According to Carens, the principle of equality of opportunity is ‘intimately linked to the view that all human beings are of equal moral worth’, and so we can move seamlessly between the demand for equality of opportunity within states, to the demand for equality of opportunity ‘for people across states’ (Carens 2013: 227–​8). In his view, the failure to defend the principle of equality of opportunity for all people anywhere looks like a failure to follow through on the basic commitment to moral equality. Yet, for many theorists of distributive justice, the argument for the principle of equality of opportunity as a demand of distributive justice within the context of the state is not premised directly and/​or entirely on a commitment to moral equality. They might be just as committed as Carens is to the idea that all persons are of equal moral worth, but they do not consider a global principle of equality of opportunity to be a necessary extension of that commitment. As Michael Blake puts it, ‘moral equality will demand different packages of rights and obligations in different institutional contexts’ (Blake 2008:  966). And once we separate the notion of equal moral worth from equality of opportunity, we can ask why we ought to conceive of all opportunities across the world as up for grabs universally, rather than, say, as the products of cooperation to which only the participations in the relevant cooperative scheme are entitled to lay claim (Pevnick 2011: 121–​2). But even if we assume with Carens that equality of opportunity at the global level is properly characterized as a requirement of distributive justice, can we use that to ground the argument that freedom of movement across state borders is a necessary condition for the fulfilment of the commitment to equality of opportunity? It seems that we can conceive of contexts in which the freedom to move across borders is not necessary in that way. Consider a hypothetical example in which all countries contained comparable and wide-​ranging opportunities. In such a case, people residing in different states could enjoy a global form of ‘equality of opportunity’ without the need for movement across state borders. The opportunities would not be absolutely identical, but we could picture a scenario in which the opportunities are ‘as good’ (for example, there are good schools and universities, jobs, hospitals, political options, and so on) so that the differences would not strike us as relevant to questions of distributive justice. In that kind of (profoundly idealized) context, unless we think that equality of opportunity should be understood as demanding that everyone has access to exactly the same set of opportunities, it does not appear as though freedom of movement

Migration   649 across borders must be a necessary condition for the fulfilment of the equalityof-opportunity requirement.9 In other words, we can imagine satisfying the demands of a global equality-of-opportunity principle without the need for global freedom of movement across borders.

30.2.3 Freedom of Movement and Just Entitlements Rather than focus specifically on equality of opportunity, we can conceive of a broader way in which freedom of movement across state borders might be a necessary condition for the fulfilment of other requirements of distributive justice. For one thing, it might be a necessary condition if it is the only way for people to access the goods (broadly conceived) to which they are entitled, according to the specifications of distributive justice. In the migration literature, this sort of argument is also often developed along egalitarian lines (e.g. Carens 2013: 227–​8), and then automatically encounters the same sorts of objections, that the same obligations of egalitarian justice do not obtain across borders (e.g. Wellman 2008: 15–​17; Wellman and Cole 2011: Ch. 2). But actually we need not take an especially demanding view of the content of distributive justice obligations across borders in order to get this argument off the ground. All we need to assume is that an account of distributive justice will include some principle about the kinds of goods to which people are entitled as a matter of justice (let’s call these their ‘just entitlements’). Then we can see that the freedom to move across state borders might be necessary for those who are unable to secure those kinds of goods for themselves and/​or their families in their current state of residence. In order to secure the kinds of goods to which they are entitled, people may need to move, at least temporarily, to another state. Now, a first way to try to head off this argument at the pass is to challenge the basic idea that there are any such things as obligations of distributive justice across state borders (e.g. Nagel 2005, especially pp. 128–​9). There may be humanitarian duties across borders, and there may even be other kinds of obligations of justice across borders, without obligations of distributive justice obtaining across borders. For example, we might think that there is something special about the institutional relationship shared by those who reside within state borders, and that this ‘something special’ is what generates obligations of distributive justice. According to this kind of view, then, obligations of distributive justice do not exist in the absence of that ‘something special’ which is only present within the confines of states.10 So Australia and/​or its residents may have humanitarian duties to assist people struggling to survive in Sierra Leone and elsewhere, but may not have obligations of distributive justice towards them. Then, even if residents of Sierra Leone are not able to access their just entitlements, it may not be the case 9 

In practice, of course, the world is not at all like that, as I will discuss in section 30.3.1. For Nagel, that ‘something special’ is the fact those within a state are ‘both putative joint authors of the coercively imposed system, and subject to its norms, i.e., expected to accept their authority even 10 

650   Sarah Fine that the freedom to move across borders must be granted as a requirement of distributive justic­e. If it is not a requirement of distributive justice that the citizens of Sierra Leone are granted the freedom to move across borders, then it seems Australia is under no obligation to open its borders to residents of Sierra Leone in the name of distributive justice.11 In response, we could try to defend the claim that freedom of movement across state borders is necessary for fulfilling the requirements of distributive justice, by seeking to defeat the argument that there are no obligations of distributive justice across borders. Indeed, it is going to be very difficult to specify the ‘something special’ about membership and/​or residence within a state in such a way that could be so distinctive and unique as to support drawing a sharp line: distributive justice obligations inside the state and no such obligations beyond borders (see Cohen and Sabel 2006). And even without any such thing as obligations of justice between non-​compatriots, it would be strange to maintain that obligations of distributive justice stop at the borders altogether, since we might think that at least some expatriates (for instance) retain some obligations of distributive justice to people in their former country of residence after moving beyond the state’s borders—​at least for a period of time (see Stilz 2016). However, the argument that freedom of movement across state borders may be a necessary condition for the fulfilment of other requirements of distributive justice is actually perfectly compatible with the claim that it is something about the shared institutional relationships between state residents which generates obligations of distributive justice. In other words, we do not need to defeat the argument of Nagel and his defenders first. Nagel himself offers us a way in here, with the notions of ‘secondary’ claims and ‘secondary’ responsibilities. As Nagel puts it, Everyone may have the right to live in a just society, but we do not have an obligation to live in a just society with everyone. The right to justice is the right that the society one lives in be justly governed. Any claims this creates against other societies and their members are distinctly secondary to those it creates against one’s fellow citizens. (Nagel 2005: 132)

Those secondary claims can get us quite far, though. It can be those secondary claims or responsibilities that trigger the demand for freedom of movement across borders in the name of distributive justice: There are good reasons, not deriving from global socioeconomic justice, to be concerned about the consequences of economic relations with states that are when the collective decision diverges from our personal preferences’. This is what triggers the specific demands of distributive justice, and it does not obtain outside states (Nagel 2005: 128–​19). But we could try to fill in the ‘something special’ in other ways. 11 

It might be under an obligation/​duty grounded in something other than distributive justice (e.g. humanitarian duties), though.

Migration   651 internally egregiously unjust. Even if internal justice is the primary responsibility of each state, the complicity of other states in the active support or perpetuation of an unjust regime is a secondary offense against justice. (Nagel 2005: 143)

Here Nagel is discussing ‘the consequences of economic relations’, but states could be complicit in the perpetuation of unjust regimes in a variety of other ways and thus guilty of a secondary offence against justice. In short, freedom of movement across borders may be a necessary condition for fulfilling other requirements of distributive justice, and may generate a prohibition against preventing movement across borders, where a failure to open one’s borders to people who are the victims of an unjust regime constitutes something like support for or perpetuation of that regime (e.g., because preventing movement across borders reduces exit opportunities, which in turn reduces efforts to undermine the durability of the regime). Thus, even if Australia does not have primary responsibility for the internal justice of Sierra Leone, preventing residents of Sierra Leone from crossing into Australian territory may constitute a ‘secondary offense against justice’; in which case, granting those residents the freedom to move across borders could be construed as necessary for fulfilling a requirement of justice. Indeed, even in cases that are not ‘egregiously unjust’, but, say, where states are failing to meet the demands of distributive justice for their residents through no obvious fault of their own, it may be the case that those residents need to exit in order to secure their just entitlements.12 In such situations, some freedom of movement across borders could be necessary for realizing the requirements of distributive justice. And, even if other states do not have primary responsibility for enabling those residents to secure their just entitlements, in order for exit to be a realistic option, those residents will require at least one other state (which is also able to provide access to just entitlements) to admit them, and other states on the route to allow them safe passage. Hence, even if we share Nagel’s view about the state-​centric nature of distributive justice, some freedom to move across borders still may be a necessary condition for satisfying the demands of distributive justice. In summary, as long as we accept the basic idea that an account of distributive just­ ice will include some principle about the kinds of goods to which people are entitled as a matter of justice, then the freedom to move across state borders might be necessary for those who are unable to secure their just entitlements in their current state of residence. Thus, we do not need to be egalitarians and/​or cosmopolitans in order to argue that freedom of movement across borders might be a requirement of distributive justice, under certain conditions.

12 

For a critical examination of the ‘liberal orthodoxy’ that there is a right to emigrate but no corresponding right to immigrate, see Cole 2000: Ch. 3.

652   Sarah Fine

30.3  How Much Freedom of Movement? So, let us assume that some freedom of movement across borders may be necessary in order for some people to secure their just entitlements, and thus that some freedom of movement across borders might be a requirement of distributive justice. Again we encounter the question of how much freedom of movement might be a requirement of distributive justice on these grounds. This way of characterizing the structure of the relationship between freedom of movement across borders and the demands of distributive justice only seems to require granting as much freedom to move as is necessary for people to secure their entitlements. That some freedom of movement may be necessary for people to secure their just entitlements does not seem to provide sufficient grounds for something as expansive as a universal moral right to freedom of movement across any and all state borders under any and all conditions (and here we can draw on a selection of arguments which David Miller mobilizes in a different, though related, context).13 It supports the freedom to move across borders where that is necessary for fulfilling the demands of distributive justice (and assuming that countervailing considerations do not push in the other direction—​a possibility we will explore in section 30.4). If the relevant distributive requirements could be met just as well or better without the need for border crossings, then freedom of movement across borders would not be a necessary condition for fulfilling the demands of distributive justice. Moreover, this does not establish a connection between freedom of movement across borders and distributive justice for those people who already have access to their just entitlements within their state of residence. And even where people are unable to secure the goods to which they are entitled as a matter of justice within their current states of residence, this is not sufficient to support a right to move anywhere and everywhere. It only supports the freedom to move in order to secure just entitlements.14 So, it could be compatible with the demands of distributive justice that a prospective migrant does not get to choose her own favoured country of destination, but is simply granted the opportunity to settle within at least one country that will enable her to secure her just entitlements. Furthermore, what is

13  Miller makes these sorts of arguments in the course of his critical discussion of instrumental strategies in defence of a human right to immigrate, i.e. attempts to defend a human right to immigrate ‘by showing that its recognition is instrumental to other human rights that have already found a place on the canonical list’ (Miller 2016: author’s emphasis). 14  On Miller’s view, with respect to the instrumental defence of a human right to immigrate, such an argument ‘holds only as long as we assume that the other human rights that provide the conditions for a minimally decent life cannot be secured without migration’. In addition, this defence ‘cannot be used to justify migration between societies all of whom already provide their members with an adequate range of opportunities, and in the case of those that don’t, it justifies only a more limited right to move to some society that does provide that range’ (Miller 2016: 22, author’s emphasis).

Migration   653 required to secure one’s just entitlements may not include the freedom to remain indefinitely within the receiving country. In line with this, perhaps disadvantaged residents of Sierra Leone are entitled to move (at least on a temporary basis) across state borders as a matter of distributive justice, but are not necessarily entitled to move and settle within whichever state they prefer. If Germany is willing to admit migrants from Sierra Leone and grant them access to their just entitlements, then this argument alone would not generate a right for them to move to Australia (or Canada or the UK) as a matter of distributive justice. And if a migrant can secure her just entitlements with a temporary visa rather than a right to long-​term settlement, then this argument does not generate a right to long-​ term settlement. Crucially, though, migrants do need to be able to reach the willing states. Unless they are able to travel directly from their current state of residence to the willing states without having to cross the borders of other states, then the migrants must also be granted the freedom to move through (even if not to settle or even linger in) the states on their route. This means that quite an extensive degree of freedom to move across state borders could be required as a matter of distributive justice, on this kind of argument, even if that is not accompanied by the freedom to settle in the state/​s of one’s choice.

30.3.1 Is it Necessary? The big question, then, is whether freedom of movement across borders is actually a necessary condition for fulfilling other requirements of distributive justice. One way of resisting the claim that freedom of movement across borders is necessary for people to access their just entitlements is to argue that, absent some powerful argument to the contrary, states and their residents are entitled to discharge their obligations in other ways, without granting the freedom to move across the borders to outsiders in general, or to those who are seeking to secure their just entitlements in particular (Miller 2005: 198; Walzer 1983: 48; Wellman 2008: 119–​120). In principle, states could indeed discharge their obligations in other ways. In some cases, people might not be able to access their just entitlements because their state of residence is relatively poorly resourced, or ill equipped to distribute the relevant goods. Other states could seek to meet their obligations via such means as contributions to foreign aid programmes; support for public and private investment in those countries; support for the establishment of stable, effective political institutions; leadership and other professional training programmes; and so forth. At the same time, or alternatively, states might choose to open their borders to some outsiders for a limited time period—​say, so that they may obtain access to education, or healthcare, or temporary/​seasonal work, but not offer the option of long-​term residence. Or instead, perhaps states might wish to discharge their obligations to outsiders via agreements with third countries who are more willing to admit migrants. Perhaps Australia could meet its obligations of distributive justice to

654   Sarah Fine outsiders by giving various forms of support to other countries willing to accept foreign migrants on Australia’s behalf. In the absence of a supplementary argument, it is difficult to see how freedom of movement across borders could be a necessary condition for fulfilling the requirements of distributive justice if the relevant requirements of distributive justice (however that is understood) actually could be realized without movement across borders. That is the case in principle. However, it is not the end of the matter, because we could try to mount the argument that, as a matter of fact, freedom of movement across borders is always likely to be necessary for fulfilling the demands of distributive justice—​other measures will not be enough on their own. That line of argument needs to respond effectively to at least two challenges: an empirical one and a normative one. The empirical challenge is to show that other measures will not be enough on their own, and so freedom of movement will remain a necessary condition for the pursuit of distributive justice. The normative challenge is to show that, as long as distributive injustice persists in the world, states or their residents will be obligated to continue taking action, rather than just to discharge a more limited set of obligations. Is it possible that there might be, in the words of David Miller, a ‘justice gap’ (Miller 2007:  276–​9)—​in this case, that some states could be doing as much as they are required to do or reasonably can be expected to do in the name of distributive justice, and yet that still some people’s claims of justice (to secure their just entitlements) may remain unmet? In response to the normative challenge first, it seems implausible that there is no ceiling on what can be expected of agents in the name of meeting the requirements of distributive justice, for otherwise we might spend our entire lives just trying to fulfil our obligations to others, and this would leave no room for anything else of value. It might even be plausible that all those states capable of responding to the demands of justice from outsiders could be doing as much as is required of them or can reasonably be expected of them without all legitimate claims of justice being met. Hence, in principle, there could be a justice gap, where some people’s legitimate justice claims remain unfulfilled. In that case we could be left in what seems to be a strange position. It could be correct that freedom of movement is necessary for the fulfilment of other requirements of distributive justice, but it also might be true that in some (rare?) circumstances there are no states that are actually required as a matter of distributive justice to allow movement across their borders (because they have already done as much as they are obligated to do). In other words, freedom of movement may be necessary for the realization of distributive justice, but this itself is not sufficient to show that all states, or even that any states, are required to open up their borders to immigrants in the name of distributive justice. Yet, given what we actually know about politics and the ways in which states currently operate, it is almost impossible to imagine a real world scenario in which we would be able to say that most (let alone all) of the appropriately resourced and positioned states have done as much as is required of them as a matter of distributive justice, such

Migration   655 that there is no requirement for most or any states to open up their borders to some immigrants in the name of distributive justice. Even if states are not required to respond to outsiders’ claims of justice without end, we seem to be on safe ground assuming that freedom of movement may be necessary for the realization of distributive justice, and that this is likely to require that at least some states open up their borders to immigrants in the name of distributive justice. On the empirical challenge, we will not be able to show conclusively that freedom of movement across borders is always going to be necessary for the realization of distributive justice, because we cannot make those kinds of conclusive statements about future unknowns. Nonetheless, once again it appears that current evidence is on the side of free movement as a necessary feature for the foreseeable future, as long as some people are unable to secure their just entitlements in their country of residence. This is because there are states which are unwilling or unable to provide access to the relevant entitlements for all people within their borders, and it would be an optimistic person who envisages a future in which that were no longer the case, however much foreign aid, investment, and intervention is involved. In short, we can confidently claim that some people will always have to move across borders to secure their just entitlements, and/​or to assist their friends and family at home (via remittances). So thus far, while it is plausible in principle that freedom of movement may not be necessary for the realization of distributive justice, in practice it is difficult to deny that some movement will be necessary and thus that some opening of borders might be a requirement of distributive justice—​at least in the world as we know it, which is the only one we have.

30.4  Countervailing Considerations The conversation about the relationship between freedom of movement across borders and distributive justice does not stop there, though, because so far I have focused on the idea that freedom of movement across borders might be a requirement of distributive justice on the basis of the justice claims of the prospective migrants. However, emigration itself may be problematic with respect to the demands of distributive justice, and might lead us to the conclusion that freedom of movement cannot be considered a requirement of distributive justice, on balance.

30.4.1 Emigration, ‘Brain Drain’, and Obligations of Distributive Justice Most obviously, in cases of countries where many people are living in poverty, it seems that certain levels and forms of emigration could make things worse for at least some of those who remain, even while potentially improving the prospects of those who

656   Sarah Fine leave. One serious issue is that of ‘brain drain’. Brain drain refers to the emigration of educated, professional, and skilled workers, and much attention is focused on the movement of the highly skilled group from countries in the Global South to wealthier countries in the Global North (see Collier 2013; Kapur and McHale 2006; and Ypi 2008). Political philosophers have long worried that, where large cross-​ country inequalities persist, opening up borders might not aid the cause of justice, and may even worsen the circumstances of the world’s worst-​off. They claim that it is usually not the least well-​off who are able to move to improve their prospects (e.g. Miller 2005: 198–​9). And when the more highly skilled, qualified members of poorer countries emigrate in large numbers, taking their prized skills and incomes elsewhere, this can leave the most disadvantaged people in those societies in an even worse position (Pogge 1997; Brock 2009; Pevnick 2011: 10). In these kinds of conditions, freedom of movement across borders may undermine the ability of developing states to deliver on their responsibilities to their own residents. This worry has led a number of political philosophers to argue that some restrictions on freedom of movement across borders might be justified in an attempt to prevent, or mitigate the effects of, brain drain. As Kieran Oberman has characterized that move:  ‘If anything can justify immigration restrictions, it seems brain drain can’ (Oberman 2013: 430). The adverse effects of brain drain might also ground a distributive justice-​based argument in defence of emigration restrictions. The most obvious question then is whether high-​skilled emigration actually has these unequivocally detrimental effects. There is some debate about the empirical claims involved in this conventional argument. For example, Oberman has pointed out that the effects of emigration from developing countries are much more complex and often more advantageous to those countries than the political philosophy orthodoxy tends to acknowledge or recognize, even in the case of skilled migration. As Oberman explains, surveying the relevant empirical literature: Skilled migration has a number of important benefits: it generates remittances . . ., establishes social networks  . . ., and diffuses knowledge  . . .  Perhaps the most important positive effect, however, is incentivising education. The possibility of earning large salaries abroad leads people to invest in education. If more people acquire skills in the hope of migrating than there are opportunities abroad, then a poor country will actually experience a brain gain, achieving higher numbers of skilled workers than if the borders had been closed. (Oberman 2015: 243)

As ever, in practice, the real-world picture is more complicated; while skilled emigration can have a variety of important benefits, some states experience fewer of these gains and may even experience detrimental effects. For example, Oberman emphasizes that ‘not all countries experience a brain gain’ and there is some evidence that ‘small states are much more likely to be losers than winners’. Therefore, ‘brain drain’ can be ‘a problem for some but not all countries’ (Oberman 2015: 243). So, while highly skilled

Migration   657 emigration actually can have various, diffuse benefits for sending countries, which may increase the likelihood of more people being able to secure their just entitlements, it also can have some detrimental effects on certain sectors of the population in certain sending countries. Those who believe that a fundamental requirement of distributive justice is to ensure that the least advantaged people (wherever they might be) are as well off as possible might think that the potential for emigration to make some of the world’s least advantaged people even worse off is a clear argument against characterizing the freedom to move across borders as a requirement of distributive justice (e.g. Pogge 1997). But we could be more nuanced, and stipulate that if and where emigration does have these detrimental effects on balance, freedom of movement across borders is not a necessary condition for the fulfilment of distributive justice demands, whereas some freedom of movement across borders will still be part of the story elsewhere. Thus, overall, we can easily imagine that the most appropriate response is to maintain that fulfilling the demands of distributive justice will continue to require the option of emigration from developing countries in many but not all cases, accompanied by more effective measures designed to mitigate possible adverse impacts of such emigration on sending countries, and indeed to reduce the need for emigration as a means for improving life prospects. Whatever happens, we need to take seriously the bare fact that where great disparities in life prospects persist across borders, people simply will continue to move (Goldin et al. 2011). In practice, therefore, attempting to tackle the problems that highly skilled emigration can cause for developing countries by trying to regulate movement is not going to be an effective response to the persistent distributive injustices on the ground. While, ideally, we might not want to think of emigration as a kind of technique for securing people’s access to their just entitlements (Oberman 2015), we cannot fail to notice that it takes on that sort of role in a variety of contexts now. Furthermore, as I have highlighted, we can be fairly confident that, however much foreign aid, investment, and intervention is directed at states with high levels of emigration, some of these states will repeatedly fail to secure the relevant entitlements for all the people within their borders. Hence, it seems reasonable to believe that some people will have to move across borders to secure their just entitlements. While on the subject of emigration, however, we also need to engage with the question of whether the freedom to emigrate allows people to evade their obligations of justice to their compatriots. On this note, Anna Stilz argues that since we have distributive justice obligations to fellow citizens, we cannot claim an unconditional right to freedom of exit which would allow us to renounce those obligations and that, under certain conditions, states may be entitled to impose regulations with regard to emigration, ‘designed to enforce distributive obligations to compatriots’. This might include paying an exit tax, for instance (Stilz 2016). Notice that this does not constitute an argument against considering freedom of movement across borders as an all-​things-​considered requirement of distributive justice, though. Instead, the idea of an exit tax offers a potential response to, and means of neutralizing, some of the problematic effects of emigration, as Stilz herself contends (Stilz 2016: 68). In the end, then, while in some contexts

658   Sarah Fine emigration may not be a net balance to sending countries and to their worse-​off members, we can continue to think of the freedom to move across borders as forming part of a package of necessary conditions for the pursuit of distributive justice.

30.4.2 Immigration and Distributive Justice There remains the question of whether the effects of immigration on receiving countries represent a challenge to the idea of freedom of movement across borders as a requirement of distributive justice. Consider two different but related, popular arguments along these lines. First, immigration might endanger the pursuit of distributive justice within receiving countries, where it adversely affects the conditions necessary for the maintenance of just institutions. For example, David Miller has long drawn attention to the issue of whether immigration, under certain conditions, might have a negative impact on the levels of social solidarity and trust between members within states—​ for instance, if increasing migration-​related cultural diversity within a state serves in various ways to erode the sense of common national identity which he considers a key condition for generating the requisite kind of solidarity under conditions of mass democracy (Miller 2003: 2008; Lenard 2012; Goodhart 2013). If the maintenance of social solidarity and trust are indeed key conditions for the successful pursuit of distributive justice (for instance, if they are necessary in order to encourage widespread support for social justice programmes, especially where those kinds of programmes require people to make compromises), and immigration at certain levels really does threaten to reduce social solidarity and trust to levels incompatible with support for and maintenance of just institutions, then we appear to have a distributive justice-based justification for immigration restrictions. Second, immigration might work against the domestic pursuit of distributive justice if it would have an unavoidably detrimental impact, all things considered, on the condition of the least well off members of that society—​for example, by increasing competition for employment and driving down wages (Macedo 2007; Goodhart 2013). Again, we would appear to have at least a prima facie distributive justice-based justification for restricting immigration. These arguments also face normative and empirical challenges. One normative challenge focuses on the question of whether states and their residents are entitled to prioritize the claims of their own residents over the claims of would-​be immigrants in this way, so that the distributive justice effects within the state take precedence and ground a justification for restricting freedom of movement across borders (Abizadeh 2016). A  second normative challenge takes the following form. Imagine that immigration can have certain detrimental effects on members of the receiving society and that these would be a concern with respect to the demands of distributive justice, but also that the effects could be neutralized or ameliorated by the enactment of domestic policies (say, related to housing or training) designed to respond appropriately to the changes. If the state did not enact the relevant policies and immigration did have detrimental effects,

Migration   659 do we have a distributive justice-based justification for immigration restrictions? Had the state instituted the relevant domestic policies, there would be no distributive justicebased concerns. The justice-based cause for complaint seems to be the state's failure to respond appropriately, rather than immigration itself (see Fine and Sangiovanni 2014: 206). A third normative question is whether the potential for migration to erode the current bases for social solidarity constitutes a legitimate justification for restricting migration. After all, it is not obvious that an unwillingness or lack of motivation to engage in solidaristic behaviour with other people is an acceptable justification for refusing to do so.15 The empirical challenge questions whether immigration does actually have the effects that Miller and others highlight and so actually lend support to these arguments in practice (Abizadeh et  al. 2015). On the fiscal side, for example, there seems to be little call for alarm about the effects of immigration on the public purse: according to the Organization for Economic Co-​operation and Development (OECD), over the past half-​century immigration seems to have had ‘a broadly neutral impact in OECD countries’, and ‘rarely exceeded plus or minus 0.5% of GDP’ (OECD 2014). There is still lively discussion, of course, about how the effects of immigration impact different sectors of society (Abizadeh et al. 2015). However, it does seem possible in principle (even if not at all probable in practice) that states could fulfil their distributive justice obligations to outsiders in such a way as to satisfy even the staunchest cosmopolitan egalitarian and yet still have a legitimate concern about the effects of immigration on their own worst-​off members, and thus that freedom of movement across borders in some cases could pose a problem from the perspective of the demands of distributive justice. Presumably, those cases would be rare indeed, but they might be possible. Once again, then, we arrive at the conclusion that while it looks as though some freedom of movement will be a requirement of distributive justice for the foreseeable future, we cannot claim that the demands of distributive justice support the freedom of everyone everywhere to move across any and all borders.

30.5 Conclusion At the end of this examination of the relationship between the demands of distributive justice and the freedom to move across borders, we reach a conclusion that will be unsatisfying to a range of different camps. I have argued that we cannot draw on the requirements of distributive justice to ground a right to move and settle anywhere that we please, because sometimes restrictions on movement will be perfectly compatible 15  See the discussion in Fine 2016 about the ways in which arguments regarding the supposed importance of racial homogeneity for the maintenance of functioning democracies have been used to defend racially discriminatory immigration restrictions.

660   Sarah Fine with the demands of distributive justice. While it is fair to predict that some freedom to move across borders will remain part of the package of necessary conditions for the fulfilment of distributive demands long into the future, nevertheless there are conditions under which the freedom to move across borders may need to be balanced against other considerations relevant to the realization of distributive justice. What is more, focusing exclusively on the demands of distributive justice will never provide us with definitive answers to the question of whether restrictions on the freedom to move across borders are justified, in general or in particular cases, all things considered. This is because such answers will need to take account of a host of other normative considerations which may be just as important or even more important than the question of what distributive justice demands. There are many more complex justice-​related factors to consider. Different states will face questions about issues of responsibilities to members of former colonies and to imperial subjects; about responsibilities to the victims of slavery and of conflicts in which those states have been involved, and to the descendants of people who have been driven from the state’s territory; about access to territory that has been procured by conquest and other illicit means; about responsibilities with respect to family reunification; and about which criteria for selecting between prospective entrants are legitimate, and so on. And even once we have accounted for those sorts of complex issues, we still need to consider questions about the extent to which states are entitled to prevent outsiders from entering their territories, and about which measures may or may not be permissible on that front. There are multiple non-​justice-related concerns to examine and evaluate, including questions about whether the freedom to move across borders should be considered a fundamental feature of a good life, and how we might balance that against other important interests. Questions about distributive justice will only get us so far; and, as it turns out, that is not very far at all.

Acknowledgements I would like to thank Serena Olsaretti for her immensely helpful comments on an earlier draft of this chapter, as well as for all her hard work as editor; Andrea Sangiovanni for very productive discussion of some of these themes; and Duncan Bell for his excellent comments on previous drafts. Versions of this paper were presented at the Bristol Political Philosophy Seminar Series, the Nuffield Political Theory Workshop, and the IMISCOE Conference, Measures of Control: Managing Migration in the 21st Century (with special thanks to commentators Nils Holtug and Godfried Engbersen). I am very grateful to the audiences on those occasions for the terrific questions and comments.

References Abizadeh, A. (2016). ‘The Special-​Obligations Challenge to More Open Borders’, in S. Fine and L. Ypi (eds) Migration in Political Theory:  The Ethics of Movement and Membership. Oxford: Oxford University Press, pp. 105–​24.

Migration   661 Abizadeh, A., Pandey, M., and Abizadeh, S. (2015). ‘Wage Competition and the Special-​ Obligations Challenge to More Open Borders’. Politics, Philosophy, and Economics 14(3): 255–​69. AFP (2015). ‘EU Should Follow Australia’s Example and Send Back Migrant Boats, says Tony Abbott’. The Telegraph, 21 April. Blake, M. (2008). ‘Immigration and Political Equality’. San Diego Law Review 45(4): 963–​80. Blake, M. (2015). ‘Migration’, in Jon Mandle and David A. Reidy (eds) The Cambridge Rawls Lexicon. Cambridge: Cambridge University Press, pp. 496–8. Brock, G. (2009). Global Justice: A Cosmopolitan Account. Oxford: Oxford University Press. Brock, G. and Blake, M. (2015). Debating Brain Drain: May Governments Restrict Emigration? New York: Oxford University Press. Calder, S. (2013). ‘Access (Almost) All Areas: The British Passport is No. 1 for Visa-​Free Travel’. The Independent, 2 October. Carens, J. H. (1987). ‘Aliens and Citizens: The Case for Open Borders’. The Review of Politics 49(2): 251–​73. Carens, J. H. (1992). ‘Migration and Morality: A Liberal Egalitarian Perspective’, in B. Barry and R. Goodin (eds) Free Movement: Ethical Issues in the Transnational Migration of People and Money. University Park, PA: Pennsylvania State University Press, pp. 25–​47. Carens, J. H. (2013). The Ethics of Immigration. New York: Oxford University Press. Castles, S. (2004). ‘The Factors that Make and Unmake Migration Policies’. International Migration Review 38(3): 852–​84. Cohen, J. and Sabel, C. (2006). ‘Extra Republicam Nulla Justitia?’ Philosophy & Public Affairs 34(2): 147–​75. Cole, P. (2000). Philosophies of Exclusion:  Liberal Political Theory and Immigration. Edinburgh: Edinburgh University Press. Collier, P. (2013). Exodus: How Migration is Changing our World. Oxford University Press. Crothers, L. and Doherty, B. (2014). ‘Australia Signs Controversial Refugee Transfer Deal with Cambodia’. The Guardian, 26 September. Czaika, M. C. and de Haas, H. (2014). ‘The Globalization of Migration: Has the World Become more Migratory?’ International Migration Review 48(2): 283–​323. Düvell, F. (2008). ‘Clandestine Migration in Europe’. Social Science Information 47(4): 479–​97. Fickling, D. and Paton, J. (2013). ‘Australia to Subject First Boat to New “Hardline” Refugee Policy’. Bloomberg, 20 July. Fine, S. (2016). ‘Immigration and Discrimination’, in S. Fine and L. Ypi (eds) Migration in Political Theory: The Ethics of Movement and Membership. Oxford: Oxford University Press, pp. 125–​50. Fine, S. and Sangiovanni, A. (2014). ‘Immigration’, in D. Moellendorf and H. Widdows (eds) The Routledge Handbook of Global Ethics. London: Routledge, pp. 193–​210. Fine, S. and Ypi, L. (eds) (2016). Migration in Political Theory: The Ethics of Movement and Membership. Oxford: Oxford University Press. Gibney, M. J. (2015). ‘Refugees and Justice between States’. European Journal of Political Theory. doi: 10.1177/​1474885115585325. Goldin, I., Cameron, G., and Balarajan, M. (2011). Exceptional People: How Migration Shaped our World and will Define our Future. Princeton, NJ: Princeton University Press. Goodhart, D. (2013). The British Dream:  Successes and Failures of Post-​War Immigration. London: Atlantic Books.

662   Sarah Fine Kapur, D. and McHale, J. (2006). ‘Should a Cosmopolitan Worry about the “Brain Drain”?’ Ethics and International Affairs 20(3): 305–​20. Lenard, P. T. (2012). Trust, Democracy, and Multicultural Challenges. University Park, PA: Pennsylvania State University Press. Macedo, S. (2007). ‘The Moral Dilemma of U.S. Immigration Policy: Open Borders Versus Social Justice’, in C. M. Swain (ed.) Debating Immigration. New York: Cambridge University Press, pp. 63–​81. Mendoza, J. J. (2015). ‘Does Cosmopolitan Justice ever Require Restrictions on Migration?’ Public Affairs Quarterly 29(2): 175–​86. Miller, D. (2003). ‘Social Justice in Multicultural Societies’, in P. Van Parijs (ed.) Cultural Diversity Versus Economic Solidarity: Is There a Tension? How Must it be Resolved? Brussels: De Boeck, pp. 13–​32. Miller, D. (2005). ‘Immigration: The Case for Limits’, in Andrew I. Cohen and Christopher Heath Wellman (eds) Contemporary Debates in Applied Ethics Oxford:  Basil  Blackwell, pp. 193–​206. Miller, D (2007). National Responsibility and Global Justice. Oxford: Oxford University Press. Miller, D. (2008). ‘Immigrants, Nations, and Citizenship’. Journal of Political Philosophy 16(4): 371–​90. Miller, D. (2016). ‘Is There a Human Right to Immigrate?’, in S. Fine and L. Ypi (eds) Migration in Political Theory:  The Ethics of Movement and Membership. Oxford:  Oxford University Press, pp. 11–​31. Nagel, T. (2005). ‘The Problem of Global Justice’. Philosophy & Public Affairs 33(2): 113–​47. Oberman, K. (2013). ‘Can Brain Drain Justify Immigration Restrictions?’ Ethics 123(3): 427–​55. Oberman, K. (2015). ‘Poverty and Immigration Policy’. American Political Science Review 109(2): 239–​51. Oberman, K. (2016). ‘Immigration as a Human Right’, in S. Fine and L. Ypi (eds) Migration in Political Theory: The Ethics of Movement and Membership. Oxford: Oxford University Press, pp. 32–​56. OECD (2014). ‘Policy Brief: The Fiscal and Economic Impact of Migration’, in OECD Policy Brief. https://​www.oecd.org/​policy-​briefs/​PB-​Fiscal-​Economic-​Impact-​Migration-​May-​ 2014.pdf/​(accessed 30 November 2017). Pearlman, J. (2014). ‘Australia Sends in its Navy to Push Asylum-​Seeker Boats Back to Indonesia’. The Telegraph, 7 January. Pevnick, R. (2011). Immigration and the Constraints of Justice:  Between Open Borders and Absolute Sovereignty. New York: Cambridge University Press. Pogge, T. (1997). ‘Migration and Poverty’, in Veit Bader (ed.) Citizenship and Exclusion. London: Macmillan, pp. 12–​27. Provost, C. (2013). ‘Migrants’ Billions Put Aid in the Shade’. The Guardian, 30 January. Rawls, J. (2001). Justice as Fairness:  A Restatement, ed. E. Kelly. Cambridge, MA:  Harvard University Press. Risse, M. (2012). On Global Justice. Princeton, NJ: Princeton University Press. Sangiovanni, A. (2011). ‘Global Justice and the Moral Arbitrariness of Birth’. The Monist 94(4): 571–​83. Shachar, A. (2016). ‘Selecting by Merit: The Brave New World of Stratified Mobility’, in S. Fine and L. Ypi (eds) Migration in Political Theory:  The Ethics of Movement and Membership. Oxford: Oxford University Press, pp. 175–​201.

Migration   663 Stilz, A. (2016). ‘Is There an Unqualified Right to Leave?’, in S. Fine and L. Ypi (eds) Migration in Political Theory: The Ethics of Movement and Membership. Oxford: Oxford University Press, pp. 57–​79. The Henley & Partners Visa Restrictions Index 2014  (2014). Available at https://www. henleyglobal.com/files/download/hvri/HP%20Visa%20Restrictions%20Index%20141101. pdf​(accessed 28 February 2018). The Internal Displacement Monitoring Centre  (2015). ‘Global Estimates 2015:  People Displaced By Disasters’. Available at http://​www.internal-​displacement.org/​assets/​library/​ Media/​201507-​globalEstimates-​2015/​20150713-​global-​estimates-​2015-​en-​v1.pdf/​ (accessed 30 November 2017). The World Bank (a), GDP per capita. Available at http://​data.worldbank.org/​indicator/​ NY.GDP.PCAP.CD/​(accessed 30 November 2017). The World Bank (b), Life expectancy at birth. Available at http://​data.worldbank.org/​indicator/​SP.DYN.LE00.IN/​ (accessed 30 November 2017). The World Bank (c), Mortality rate, infant. Available at http://​data.worldbank.org/​indicator/​ SP.DYN.IMRT.IN/​(accessed 30 November 2017). The World Bank (d), Sierra Leone. Available at http://​data.worldbank.org/​country/​sierra-​ leone/​(accessed 30 November 2017). United Nations, Department of Economic and Social Affairs, Population Division (2013). ‘International Migration 2013 Wallchart’. Available at http://​www.un.org/​en/​development/​ desa/​population/​migration/​publications/​wallchart/​docs/​wallchart2013.pdf/​ (accessed 30 November 2017). Walzer, M. (1983). Spheres of Justice:  A Defense of Pluralism and Equality. New  York: Basic Books. Wellman, C. H. (2008). ‘Immigration and Freedom of Association’. Ethics 119(1): 109–​41. Wellman, C. H. and Cole, P. (2011). Debating the Ethics of Immigration:  Is There a Right to Exclude? New York: Oxford University Press. Whiteman, H. (2014). ‘Australia Admits Holding 153 Sri Lankan Asylum Seekers at Sea’. CNN, 9 July. Ypi, L. (2008). ‘Justice in Migration:  A Closed Borders Utopia?’  The Journal of Political Philosophy 16(4): 391–​418. Zong, J. and Batalova, J. (2015). ‘Frequently Requested Statistics on Immigrants and Immigration in the United States’. Migration Policy Institute, 26 February. Available at https://www.migrationpolicy.org/article/frequently-requested-statistics-immigratnts-andimmigration-united-states-0 (accessed 28 February 2018).

Chapter 31

Clim ate C ha ng e Simon Caney

Climate change raises many questions of distributive justice. The increase in temperatures, the rising sea levels, and the severe weather events associated with climatic changes will all have dire consequences for human and nonhuman life. For example, climate change jeopardizes many people’s access to food. The increase in temperatures will lead to desertification and crop failure, and rising sea levels will also disrupt agriculture. Climate change also has disastrous effects on people’s health: it will see an increase in the spread of food-​borne, water-​borne, and vector-​borne diseases, as well as other infectious diseases. In some cases it will result in loss of life (e.g., through heat stress and disease). Severe weather events will wreak havoc with people’s homes, livelihoods, and the infrastructure.1 In the face of these challenges we are confronted by many questions of distributive justice. Who is entitled to what level of protection? Who is responsible for bearing the burden of addressing climate change? How should rights to emit greenhouse gases be distributed? This chapter examines some of these issues. The chapter is structured as follows. I begin, in section 31.1, by delineating two distinct questions of distributive justice raised by climate change. Stated very roughly, one question concerns how much protection is owed to the potential victims of climate change (the Just Target Question), and the second concerns how the burdens (and benefits) involved in preventing dangerous climate change are distributed (the Just Burden Question). In section 31.2, I focus on the first of these questions, the Just Target Question. The rest of the chapter examines the second question, the Just Burden Question. To answer this question, I argue, it is necessary to address two important methodological questions. sections 31.3–​31.5, thus, set out and explore these two methodological issues. Having done so, the chapter then turns from methodological issues to substantive analysis, and in section 31.6 it examines three principles of distributive justice that, it has been suggested, should determine how the burden of addressing dangerous climatic changes should be distributed. 1 

These impacts are comprehensively documented by the Intergovernmental Panel on Climate Change. Its reports—​including the most recent one (the Fifth Assessment Report)—​can be found here: http://www.ipcc.ch.

Climate Change   665

31.1  Two Questions Let us begin then by identifying the questions of distributive justice that arise. As we have noted, I think we can identify at least two important questions of distributive justice.

31.1.1  Identifying the Just Target One important question concerns what balance to strike between, on the one hand, ensuring that moral subjects are not harmed by climatic changes,2 and, on the other hand, ensuring that the policies that would be required to prevent harmful climatic changes do not impose excessive burdens on people. Climate change is not an all-​or-​ nothing affair. There are different degrees of change. According to one widely held view, climate policy should aim for a world in which the global mean temperature is no more than 2°C higher than pre-​industrial times. Some, though, think that this is too lax and call for more stringent targets—​such as one in which global mean temperatures are no more than 1.5°C higher (a goal affirmed in Article 2 of the Paris Agreement). Alternatively, we might aim for a less demanding target, calling, for example, for temperature increases of no more than 3°C.3 These different targets will result in different impacts: the higher the permissible increase, the more dire the outcomes. At the same time, the more we aim for a smaller increase in global mean temperatures, the more onerous the burden that it will place on the duty-​bearers. The burdens involved in preventing harmful climate change can be quite demanding. I  will outline the kinds of burdens more systematically when discussing the second question. Here, some examples will suffice to give some sense of what is at stake. To take one major burden: climate policies normally impose burdens on those allocated responsibilities to prevent harmful climatic changes, such as requiring them to forego certain activities so that they keep their emissions of greenhouse gas within an acceptable limit.4 In addition to this, climate policies may also have an opportunity cost, and some would benefit if the resources devoted to combating climate change were spent elsewhere. For example, some might be worse off as a result of climate policies because the funds being used to enable

2 

Note that I use the phrase ‘not harmed by climatic changes’ as a shorthand to refer to actions which either prevent the climatic changes from occurring (often termed ‘Mitigation’) or which prevent the changes that do occur from being harmful (often termed ‘adaptation’) or by some combination of the two. For further discussion of Mitigation and Adaptation see section 31.1.2. 3  Note, one might also query the assumption that the target should be defined in terms of temperature increases at all, and, if it should, whether it should be defined in terms of global mean temperature increases. 4  I say ‘normally’ because in some cases the discharging of a responsibility can result in co-​benefits (including for those discharging the duty). A co-​benefit is a beneficial side-​effect. For example, levying carbon taxes may discourage people from driving, and as well as mitigating climate change, might also have beneficial effects on physical fitness and air quality.

666   Simon Caney people to live with climatic changes, or the funds needed to support clean technology, could have been spent on other things, such as research into infectious diseases. We thus face a situation in which permitting climatic changes imposes harms on people (the magnitude of which will vary depending on just how severe the changes are and how much is spent on adaptation); and yet preventing harmful climatic changes also imposes burdens on people (the magnitude of which will vary depending on how ambitious are the climate policies). With this in mind, we can now see that one question of distributive justice is: Q1: The Just Target Question: what is the fair distribution of burdens and benefits between on the one hand, those who will bear any impacts of climatic changes, and, on the other hand, those who will be adversely affected by the implementation of a policy that prevents or minimizes harmful climatic changes?

31.1.2 Sharing the Burden Suppose that we choose a certain target for climate policy. Suppose, for example, that we accept the conventional 2°C target. Achieving this goal will incur considerable costs. Given this, we now face a second question of distributive justice, namely Q2:  The Just Burden Question:  what is the fair distribution of the burdens (and benefits) of adopting policies that address climate change? As I have already noted, there is a variety of different burdens. We can identify three distinct kinds of burden that are relevant to Q2. The first two correspond to two kinds of responsibility. Thus, one kind of burden is the cost of what climate scientists refer to as mitigation, where this is understood as reducing the extent to which humans affect the climate system by affecting the volume of greenhouse gases in the atmosphere. Mitigation can take the form of lowering the flow of greenhouse gas emissions or increasing the extent to which greenhouse gas sinks absorb greenhouse gases. Mitigation is often, although not always, burdensome to those who engage in it. A second burden is that involved in enabling adaptation. I define adaptation as changes in the social, economic, and political systems that reduce the extent to which climatic changes undermine people’s entitlements. Adaptation can take many forms. For example, to prevent increased heat from killing people from heat stress, one might design buildings with better access to cool air; to prevent the dire effects of storm surges one might install better weather warning systems and more robust sea walls; and so on. The key point is that mitigation addresses the prospect of dangerous climate change by minimizing the climatic changes, whereas adaptation addresses it by reducing the harmful impacts of those climatic changes on people’s entitlements. Again, those engaging in adaptation policies normally incur costs.

Climate Change   667 We should record a third kind of burden. The first two concern burdens that are borne by agents when they discharge some kind of climatic responsibility. Sometimes, however, the policies enacted to combat climate change may impose costs on persons other than those implementing the policy: they sometimes impose harms on third parties. For example, using nuclear energy instead of fossil fuels may put the health of some (including those not duty-​bound to reduce emissions) at risk. These burdens too need to be borne in mind.5 With this taxonomy of burdens in mind we can now return to the initial formulation of the Just Burden Question. Its reference to the ‘the burdens (and benefits) of adopting policies that address climate change’ should be understood to refer to these three kinds of burden. Thus, the question posed by the Just Burden Question is: What is the fair distribution of these three kinds of burdens (and any associated benefits) that result from mitigating and adapting to climate change? A number of different principles of justice have been proposed. Three in particular are worth noting at the outset. First, some argue that the relevant principle for distributing the burdens involved in preventing dangerous climatic changes focuses on who has brought this problem about. They hold that those who have caused the problem should pay. This is often referred to as the Polluter Pays Principle.6 Others have defended a second principle, the Ability to Pay Principle. This holds that agents should bear the burden in proportion to their ability to pay. A third principle that has been invoked maintains that the burden should be borne by agents in accordance with the extent to which they have benefited from greenhouse gas emissions—​the Beneficiary Pays Principle (Caney 2005: 756). These principles will be discussed later on, but it is useful to introduce them here so that when we consider the methodological issues we are aware of what kind of principles are most commonly invoked. Doing so will also help illustrate the methodological issues that arise. Before moving on, one particular question concerning the Just Burden Question should be mentioned. The question I have in mind is one that arises only in the context of mitigating climate change. Since mitigation involves limiting the emission of greenhouse gases it raises a particular question that does not arise in the case of adaptation:  what (if anything) would be a fair distribution of greenhouse gases?7 Some reply that a fair distribution is one in which rights to emit greenhouse gases are shared 5 

For completeness’ sake we should also note a fourth kind of burden, namely Compensation. Suppose that people do not engage in sufficient mitigation (so create climatic changes) and do not provide adequate adaptation (so the climatic changes leave people unable to enjoy their entitlements). Then those who lack their entitlements have been harmed and are entitled to compensation. In the international negotiations this is now discussed under the heading of ‘loss and damage’. Since this Handbook is a volume on distributive justice, rather than corrective justice, I shall not, however, discuss compensation burdens further. 6  Some think that this term should not be employed to refer to the view that those who have polluted in the past should pay, arguing that the ‘Polluter Pays Principle’ is an entirely forward-​looking principle (Ally and Beckerman 2014: 90–​1; and Shue 2014: 184). 7  Space precludes examining this issue here. For discussion, see Caney 2012a.

668   Simon Caney equally (Meyer 2000). Others argue that a fair distribution is one that best enables the least advantaged to develop. They defend what has been termed the ‘Greenhouse Development Rights’ approach (Baer et al. 2008). Others hold that a fair distribution of emission rights would mirror the current, or past, distribution of emissions. This position—​known as ‘grandfathering’—​allocates emissions to reflect the level of emissions that agents had at some point in the past (Bovens 2011). This brief list is not exhaustive, but it gives some sense of some of the positions espoused. One final clarificatory point is in order. In distinguishing between the two questions I am not claiming that they should be treated wholly separately. One can quite coherently think that there are two distinct questions but also that the right way to answer them is to draw on a common normative theory. Indeed, as will emerge, this is precisely what I will argue. Consider an analogy from the ethics of war. It is common to distinguish between (at least) two questions about the ethics of waging war—​when to wage war (jus ad bellum) and how to wage war (jus in bello). While some—​such as Michael Walzer—​think that the answers to the two questions should be derived independently from each other, others such as Henry Shue, argue for grounding both in a common normative theory.8

31.2  Question 1—​What Level of Protection? Having identified two key questions, I will begin with the Just Target Question. How should we go about answering this question? Of course, one familiar way of thinking about it is to employ cost-benefit analysis. This would assess the harms (and benefits) caused by climate change and determine what costs to people’s well-​being or resources or whatever metric one employs would occur at what level of change. It would also calculate the costs and co-​benefits associated with mitigation and adaptation policies, again determining what the costs and benefits would be with more and less ambitious mitigation and adaptation policies. Having ascertained the costs and benefits of both climate change and also of climate change policies it would then, of course, recommend whichever policy results in the greatest net benefit. Some, such as Bjørn Lomborg (2001), have, in fact, reasoned in exactly this way. Lomborg’s analysis of the empirical data on climate change has been questioned by many (e.g. Cole 2003), but the aspect that is of particular relevance to our analysis here is his endorsement of a basically utilitarian approach to help think about the Just Target Question.

8 

See Walzer’s affirmation of ‘[t]‌he dualism of jus ad bellum and jus in bello’ (1977: 21: also pp. 123–​4 and 326) and Shue’s alternative (2005, especially pp. 737–​8).

Climate Change   669 Of course, one standard objection to utilitarianism is that it results in injustice, and thus many have developed accounts of distributive justice as an alternative to utilitarianism. These have, however, rarely (if ever) been applied to the Just Target Question. In fact, it is striking how little normative reflection there has been by philosophers on the Just Target Question.9 How then should we apply a theory of distributive justice to answer the Just Target Question? Any such account would have to have two features. First, since climate change is a global phenomenon (in the sense that the causes are spread throughout the world and the effects are too) applying principles of justice within societies would not be sufficient. To determine what target to aim for therefore requires a theory of global justice. Second, we need an account of justice that can cope with the intergenerational character of climate change. This holds for several reasons. In the first place, since climate change results at least in part from the emissions of previous generations, we need to know to what extent members of one generation can inherit obligations from the actions of previous generations. Furthermore, since many of the adverse effects of climate change fall on future generations, we need to consider what obligations one generation owes to succeeding generations. For example, may one generation discount the interests of future generations? In addition to this, when determining how to answer the Just Target Question, we also need to consider whether it is permissible to pass some of the burdens of combating climate change on to future generations (a position advanced by Broome 2012: Caney 2014). If, for example, it is not, then all the costs of mitigation fall on current generations. This, in turn, might entail that less mitigation is justified than would otherwise be the case. Answering the Just Target Question, thus, requires a theory of global and intergenerational justice. Deriving such a theory is, of course, highly challenging, and it is certainly beyond the ambitions of this chapter. So, rather than attempt to do that in the space available here, this chapter will merely illustrate the challenge by outlining two different accounts and exploring their implications. A sufficientarian approach:  consider, for example, a sufficientarian approach. Suppose, moreover, that it is claimed that this should apply to everyone in the world and that it should not be subject to any pure time discounting, and thus, that all persons in the world throughout time should be above this threshold. In answer to the Just Target Question: this would maintain, first, that the target of climate policy should ensure that all are above the specified threshold. (In parenthesis:  we might also note here that such an approach would also have implications for the Just Burden Question. In a world of reasonable abundance it entails that the ascription of responsibilities to combat climate change, and the selection of which climate policies to adopt, should leave everyone above the sufficiency threshold.) In short, so long as climate policies are implemented which ensure (a)  that there are not climatic changes which push people beneath the sufficiency level; and (b) that 9 

For one exception, see Heyward (forthcoming).

670   Simon Caney the costs of realizing (a) does not push people beneath the sufficiency level, then just­ ice would have been done. Depending on what the sufficiency level is—​and existing levels of abundance—​this might mean that a number of different climate targets satisfy the sufficientarian standard. At the same time, however, it might also be the case (especially with high levels of noncompliance and/​or with a high sufficiency level) that it is not possible to achieve both (i) the just target (so secure a safe climate system); and (ii) do so without adopting mitigation and adaptation policies that themselves push some people below the sufficiency threshold. In such a case, we need a further principle to help guide us. To get more of a sense of how to approach the Just Target Question, we might also consider other distributive principles. For example, some will object to a purely sufficientarian approach. Imagine, they might say, situations in which climate change greatly impacts on the standard of living of people yet, even with the onset of climate change, they are above the sufficiency threshold. Suppose, for example, that their property is harmed or that the company they have created is ruined by rising sea levels, but that they are above the sufficientarian baseline. Now, ex hypothesi, a purely sufficientarian approach to climate change policy is insensitive to the loss experienced by those in this situation. An egalitarian approach: with this in mind, some might propose an egalitarian principle. This would differ from a purely sufficientarian approach in several ways. First, when considering what the target should be, it would take into account, not simply the fact that climate change pushes people below some threshold, but also whether climatic impacts exacerbate existing inequalities or create new ones, including inequalities above any sufficiency threshold. Second, it would consider whether the burden associated with preventing climate change would worsen inequalities. Suppose, for example, that it is possible to prevent climatic changes that result in some being above the sufficientarian threshold but still having less than they would under an egalitarian world; and, suppose that it is possible to allocate the burden of preventing climate change in ways that would not undermine the cause of equality. Then egalitarians would do that. This might call for a more stringent target than a sufficientarian one (as well as a different regime of burden-​sharing). This discussion of two possible principles is necessarily brief. However, it brings to light several points that are worth emphasizing. The first is that answering the Just Target Question requires a comprehensive theory of global and intergenerational justice. Only with this can one determine what level of protection is required and what it is fair to ask others to do. A second lesson is that different principles of justice are very likely to propose different answers to the question of what constitutes a Just Target. Other things being equal, those who adopt a low sufficientarian threshold, for example, will impose less demanding targets than egalitarian ones that are concerned with inequalities above that threshold.

Climate Change   671 There is also a third lesson: namely that once we do draw on a particular principle of global and intergenerational justice to help specify the Just Target, this same principle also tends to have implications for the question of how burdens should be distributed. As we saw, if we think that everyone should be above a certain minimum standard of living then this implies, not just that the target should be set so that this is realized, but also that the distribution of burdens involved in combating climate change should also be determined in such a way that it does not push people beneath this minimum standard of living. It would be incoherent to employ sufficientarianism to specify a just target but to eschew it when allocating burdens. This then lends support to the view expressed at the end of section 31.1:  namely, the answers to the two questions that I identified in section 31.1. I might draw (at least in part) on a common normative framework.

31.3  Question 2—​Two Methodological Issues With this last point in mind, it is worth turning now from the Just Target Question to the Just Burden Question. How should the burdens involved in mitigating and adapting to climate change be distributed? What principles of justice should guide us here? A number of different principles have been proposed. However, before analysing these substantive criteria it is necessary to confront two important methodological issues concerning how to approach the burden-​sharing question.

31.3.1  D  istinguishing between Integrationism and Isolationism To introduce the first methodological issue we should recall that to address the first question, it is necessary to consider climate change in conjunction with other important issues (such as eradicating poverty or promoting development or meeting people’s health needs). We need to do so, in part, to know what level of protection people are entitled to and, in part, because combating climate change imposes opportunity costs and we need to judge whether those costs impose unfair burdens on others. Answering these questions requires us to draw on a more general theory of justice. This raises the following question:  when applying a distributive principle to the burdens involved in mitigation and adaptation, should we include all these other considerations here as well? That is, should we propose a principle in the light of other considerations (such as people’s entitlements to food or health or trade)? Or, should we seek to isolate the burdens involved with combating climate change and treat climatic

672   Simon Caney responsibilities (that is responsibilities to mitigate and adapt) in isolation? One can distinguish between three possibilities. Consider first: Isolationism: this holds that we should apply principles of justice to a good X in isolation of any other considerations. So, this holds that we should bracket out all other normative considerations (such as people’s entitlements to food or water or health or rights to develop), and should identify a principle that applies to this topic when considered in isolation from everything else. Isolationism takes a very radical ‘modular’ approach in the sense that it distinguishes between different domains (like health, or trade or climate) and applies a particular principle to each particular domain.10 As we shall see, a number of philosophers adopt an isolationist approach to the Just Burden Question. Isolationism can be contrasted with a second approach, which I term: Moderate Integrationism: this holds that we should apply principles of justice to a good X, but in doing so we should also take into account other considerations. By contrast with the first approach, this holds that we should factor in other considerations. However, it still holds that there is a specific principle that applies to the case of climate change.11 On this view it is normally fitting that burdens (and benefits) arising in this context are governed by a particular principle; but, it adds that this principle may be informed, or moderated, by other considerations. Consider an example that is far removed from climate change justice. One might think that the appropriate principle for a speeding offence is something like ‘Those who speed should be punished and the more that they exceed the limit the more severe the punishment.’ Now suppose that someone is speeding, we might think that this is the relevant principle; but we might also add that it can be qualified by taking other normative considerations into account. For example, we might take into account whether they had a pressing medical need or if they are being chased by someone seeking to hurt them. There is nonetheless a recognizable principle of justice that applies here and, so to speak, plays the dominant role. It is not that there is a general overarching theory and we just plug it into that. That leads on nicely to a third possible view, namely Strong Integrationism: this holds that we should treat X merely as one element in the total package of burdens and benefits and then this total package should be regulated by a general principle of justice (such as a global difference principle or a commitment to basic rights). 10  I borrow the language of ‘modularity’ from Jerry Fodor’s well-​known ‘modular’ account of the mind (Fodor 1983) (though, of course, the analogy is very rough and should not be interpreted literally). 11  For another example that seems to fit this mould, see David Miller’s treatment of fair trade (Miller 2010).

Climate Change   673 On this view, there is no specific principle that determines the distribution of X considered all on its own (isolationism); nor is it the case that there is a principle that generally governs the distribution of X but which also takes into account other considerations (moderate integrationism). There is just a more general principle that applies to a whole package of benefits and burdens and X is included within this package. So, when considering how the burdens associated with combating climate change should be distributed we need to consider whether we should adopt an isolationist, moderate integrationist, or strong integrationist approach. I have already argued that settling the Just Target Question requires us to draw on a more general theory of just­ ice that takes into account both the interests jeopardized by climate change and the interests that would be impacted if we adopt climate change policies, and which treats all of these interests together. This Strong Integrationism is necessary if we are to arrive at a just target—​one that gives a due concern to those affected by climate change but also one that does not make unjust demands on others who are burdened by climate change policies.12 And I further argued that the principles invoked to address the Just Target Question have implications for the Just Burden Question. This has the further implication that when considering the Just Burden Question we should draw on a more general theory of justice. In short, we should adopt a version of strong integrationism. This, however, goes against the grain of much contemporary theorizing about how to distribute responsibilities to prevent climatic changes that undermine people’s entitlements. So in section 31.5 we will consider why some defend an Isolationist approach.

31.3.2  Distinguishing between Holism and Atomism Before we do so we should note that we also face a second methodological concern. As we have noted, combating climate change involves two kinds of burden—​those associated with mitigation and those associated with adaptation. The question here then is this: Should we treat the burdens from mitigation and the burdens of adaptation together as a package and apply a distributive principle to that? Or, should we treat them separately and apply different principles to both? On one view, we should adopt the following: Holism: The distribution of mitigation responsibilities (and hence emission rights) and the distribution of adaptation responsibilities should be governed by a common principle of justice (or set of principles of justice).

12 

Hereafter my focus will be on the strong, rather than moderate, version of integrationism.

674   Simon Caney Some, however, deny this and affirm the following: Atomism: the distribution of mitigation responsibilities should be governed by one principle of justice (or set of principles of justice) and the distribution of adaptation responsibilities should be governed by another principle of justice (or set of principles of justice). In what follows I  shall consider the case for both, and outline why I  take a Holist approach to be more compelling.

31.4  Holism v. Atomism Let us start with the holism v. atomism issue. Although many adopt an atomist approach, few have provided a justification for this. One partial exception is Steve Vanderheiden. On his view, different principles of burden sharing apply to mitigation and to adaptation because the question of who should bear the burden of mitigation is a question of distributive justice, whereas the question of who should bear the burden of adaptation is a question of corrective justice (Vanderheiden 2008: 82, 222–​3 and 229–​30).13 Why accept this? Vanderheiden’s view rests on two further arguments, both of which seek to explain how mitigation and adaptation differ in morally relevant ways. Vanderheiden’s first argument runs as follows: Argument 1:  Adaptation Justice: past emissions cause harm and so adaptation is required to compensate for that harm. Because it is a matter of compensatory justic­e, the costs involved in providing adaptation should be governed by a backward-​ looking principle of compensatory or corrective justice to determine who should pay for this harm. (Vanderheiden 2008: 82, 222–​3 and 229–​30) If we turn now from adaptation to mitigation we encounter Vanderheiden’s second argument. This runs as follows: Argument 2: Mitigation Justice: The question of how to distribute the remaining budget of greenhouse gases among current and future emissions is not a question about compensating people for past acts that have caused climatic harms. As such, backward-​ looking considerations and principles of corrective justice are not appropriate here. Rather, what we face here is a question of how best to distribute a particular good, and hence it is governed solely by forward-​looking principles of distributive justice. (Vanderheiden 2008: 82, 222–​3 and 229–​30) 13 

Hyams makes similar assumptions (2009: 241). See also Risse (2012: 202–​6, especially p. 204).

Climate Change   675 Now on this basis, Vanderheiden then argues, first, that with respect to adaptation the appropriate principle of corrective justice requires that the costs of adaptation be allocated according to a variant of the Polluter Pays Principle. (I say a variant of the Polluter Pays Principle because he holds that it would be unfair to make people pay for pre-​1990 emissions on the grounds that they could be excused for not knowing the harmful effects pre-​1990, but that from 1990 onwards any ignorance is inexcusable Vanderheidehn (2008: 191).) And he argues, second, that with respect to mitigation, the appropriate principle of distributive justice requires that the right to emit greenhouse gases should be distributed according to broadly egalitarian principles of justice (Vanderheiden 2008: 226–​7). Now, if these two arguments were correct, then they would establish a morally relevant difference between adaptation responsibilities and mitigation responsibilities, and different principles of justice would apply to each. Unfortunately, however, both arguments are incorrect. Consider Argument 1 first. First, this argument falsely presupposes that adaptation is a form of compensation (and hence should be governed by principles of corrective justice). But this elides the distinction introduced in section 31.1 between adaptation and compensation. For adaptation, as I defined that term, refers to policies that prevent changes in the earth’s climate from inflicting harm on others. To give some familiar examples: adaptation involves building seawalls to prevent rising sea levels from having harmful effects; inoculating people against infectious diseases that they might otherwise acquire were it not for climatic changes. Thus understood, adaptation is not a matter of compensating people for harm inflicted on them. Rather it—​just like mitigation—​is a way of ensuring that one group of people (those currently emitting) does not harm others. There are, then, two ways that people can discharge a duty not to harm others—​(i) reducing emissions and/​ or enabling others to reduce emissions (mitigation); and (ii) designing social and economic and physical infrastructures such that any climatic changes that occur do not cause harm (adaptation). The claim that adaptation is a matter of corrective justice—​ and hence that it should be governed by a different principle of justice to mitigation—​is, therefore, false. Consider now the second problem. Argument 1 presupposes that where some engage in a harmful activity, then justice requires that the people who engage in that activity (and only they) should perform duties of corrective justice. It is on this basis that Vanderheiden holds that adaptation funding should be governed by a variant of the Polluter Pays Principle. But why assume that only backward-​looking principles of just­ ice are relevant for adaptation funding? To see why one might legitimately ascribe responsibilities to enable adaptation to people other than those who have emitted high levels of greenhouse gases, consider a familiar liberal egalitarian view like Ronald Dworkin’s (Dworkin 2000). This affirms two principles. The first is that persons are entitled to a fair set of opportunities. In Dworkin’s view, this mandates equality of resources. The second is that persons can be held accountable for their choices. Thus, if people who enjoy a fair set of opportunities bring misfortune upon themselves (or others) then, other things

676   Simon Caney being equal, they should bear the burden. Now suppose that someone, X, does engage in harmful behaviour. If we accept the liberal egalitarian view, it does not necessarily follow that X—​the polluter—​is the agent that should pay. For X may lack a fair set of opportunities, and others may be legitimately required to pay if they have more than their fair share of resources. For example, if a poor Indian who faces unjust circumstances emits high levels of greenhouse gases and thereby contributes to causing harm, then it is not necessarily true that the Indian emitter is the one that should pay for adaptation. Maybe those who have more than their fair share of opportunities ought to pay. If so, then a non-​backward-​looking principle is applying to adaptation—​namely the Ability to Pay Principle. My point, then, is that we lack any reason to think that adaptation funding should be governed solely by backward-​ looking considerations, and not at all by ones that attribute responsibility according to other criteria. Argument 1 is, thus, false. Consider now Argument 2. Recall that the claim here is that past emissions have no relevance for the distribution of remaining greenhouse gases (and that they matter solely insofar as they cause climate change). It is on the basis of this assumption that Vanderheiden says that the distribution of those greenhouse gases that may be permissibly emitted is a matter of distributive justice and not corrective justice. However, Argument 2 is mistaken. As Shue has pointed out in a different context, the emission of greenhouse gases in the past can have two distinct effects. One effect is that noted by Vanderheiden: high greenhouse gases can cause changes in the climate and thus call for adaptation (Shue 2014: 196). However, there is a second effect, namely that some past emissions use up a common resource and thereby leave less of this resource to other people (Shue 2014: 196–​200). The emission of greenhouse gases in the past, thus, has implications for the fair distribution of greenhouse gas emissions now. So backward-​looking considerations are relevant to the distribution of mitigation burdens. One response to this might be that it is unfair on those who emitted greenhouse gases in the past to make them pay twice—​once by debiting it from their quota of emissions and then a second time by charging them adaptation costs.14 One way to express the worry might be this: suppose that X emits a certain volume of greenhouse gas emissions, v, in the past but then X subsequently pays a sufficient amount of money into an adaptation fund such that any contribution that v makes to bringing about climate change is prevented from being harmful. X can then say: . . . the only reason we care about the volume of greenhouse gases that agents emit is to prevent harm but my past emissions v have not led to any harm. Therefore, v should not be debited from my current and future share of emissions. In terms of my contribution to causing dangerous climate change my behaviour (that is, emitting v but then

14 

I am grateful to Christian Baatz who in conversation, and in unpublished work with Konrad Ott, has raised this concern.

Climate Change   677 enabling and paying for adaptation to the extent that v results in no harm) is the same as if I have not emitted v at all. Given this, it is unfair to debit v from my share of emissions. This objection is, however, misplaced for it treats all past emissions in an undifferentiated way and ignores the possibility that some proportion of the past emissions might be relevant for mitigation and another proportion might be relevant for adaptation. To see this, suppose that one maintains that an agent X may permissibly emit up to 100 units in a ten-​year period, t1–​t10. Suppose, then, that in that period X emits 150 units. Suppose now that we are at t11 and are wondering how much X can emit in the future. Then it is reasonable to argue that X’s emission of 100 units should be taken into account when determining his future emissions. We can say that he used up his full quota and so has no spare left over to use up later on. This bears, then, on his Mitigation responsibilities, for if he had emitted less (say 80 units) then he would have some left (20 extra) to use, and thus would have had a less demanding responsibility to mitigate. Turn now to the 50 extra units above his fair share: these should be taken into account when ascribing adaptation responsibilities.15 In short, in cases like this, X’s past emissions can justify both some mitigation and also some adaptation. Vanderheiden’s strong claim that past emissions are not in any way relevant to mitigation responsibilities is, thus, incorrect, and the objection thus fails. In conclusion then:  we have seen that adaptation costs should not be governed uniquely by backward-​looking considerations like the Polluter Pays Principle, and may legitimately be governed by other considerations like people’s Ability to Pay. And, we have seen that mitigation costs can legitimately take into account backward-​looking considerations like people’s past pollution (à la Polluter Pays Principle), as well as other egalitarian principles of distributive justice. Finally, we have seen that a challenge to this second argument has proved unsuccessful. With this in mind, there is a case for combining the burdens involved in mitigation and those involved in adaptation and then allocating them together. In the absence of any reason to treat mitigation responsibilities and adaptation responsibilities differently, and given, moreover, the evidence that we have just seen of how backward-​looking and forward-​looking considerations apply to both mitigation and adaptation, I shall therefore assume that holism is correct.16 This has a major implication for climate justice for it means that we need just to construct a principle (or set of principles) to govern climatic responsibilities, and we do not need to construct two different kinds of account of distributive justice for the different climatic responsibilities.

15 

I set aside all complications here concerning levels of knowledge and whether X is an individual or country or another kind of agent (e.g. corporation). For brief discussion of some of the complications, see also section 31.6, and for fuller discussion, see Caney 2005, 2010. 16  For further discussion of these issues, and challenges to holism, see Jagers and Duus-​Otterström 2008 and Duus-​Otterström and Jagers 2012.

678   Simon Caney

31.5  Integrationism v Isolationism Having tackled one important methodological issue that we face when thinking about the fair way to apportion the burdens of climate change, we still face a second. This second one, recall, is whether we should treat climatic responsibilities in isolation from other normative issues (isolationism) or whether we should consider them in conjunction with other considerations (integrationism). When determining who should pay for climate change, should we factor out considerations like people’s other economic rights or entitlements? Or should we include them and adopt a kind of integrationism? Very many philosophers have adopted an isolationist approach. How might one ground this? We can distinguish here between principled and pragmatic arguments.

31.5.1 Principled Arguments for Isolationism Consider a principled argument first. One approach might be to draw on a pluralistic theory of justice of the kind defended by Michael Walzer in Spheres of Justice (1983). Walzer famously argues that different distributive principles apply to different goods. On his view, the appropriate principle for health is need; the appropriate principles for education are ability and willingness to work; the appropriate principle for money is the market; and so on. Walzer is, of course, not the only advocate of such a pluralistic theory, but his theory is the pre-​eminent one, and so I focus on his account to illustrate the point. Now this kind of view might seem hospitable to the isolationist position on climatic responsibilities. One might, that is, appeal to the idea that different ‘goods’ should each be distributed according to different distributive criteria to argue that the distribution of climatic burdens should be considered in isolation from other issues and according to its own particular distributive principle. For this strategy to work, however, one would have to establish two claims: first, that a pluralistic theory of the kind we have described is correct, and, second, that such a pluralistic theory entails that it is appropriate to treat climatic burdens in isolation. Let us suppose, for the sake of argument, that this kind of pluralistic approach is correct. Even if it were, it is unclear that it would support bracketing off the distribution of responsibilities to mitigate and adapt from all other questions of distributive justice. There are two central obstacles. First, for Walzer the distributive principle for a good—​ such as health—​depends on identifying the shared understanding of the nature of that good: the distribution is governed by what Walzer terms its ‘social meaning’ (1983: 19). Whether there are shared understandings for the goods that Walzer identifies—​like health, political power, membership and so on—​is hotly contested (Dworkin 1985), but even if there were for these goods, it is very hard to think of a global set of shared understandings about the nature of the climatic responsibilities. This is in part because the kinds of sacrifice required to prevent dangerous climate change are rather heterogeneous—​mitigation requires foregoing many different kinds of goods, as does

Climate Change   679 adaptation. In addition to this, unlike the goods that Walzer studies, there is no global consensus about or shared ‘social meaning’ concerning these goods from which one can read off who should mitigate or adapt and by how much. In addition to this, the Walzerian view depends on there being separate spheres. It can support isolationism only if the distribution of climatic responsibilities were correlated with one specific ‘sphere’ or ‘good’. If it was, then one could have a single principle that determined the fair distribution of mitigation and adaptation burdens on their own. In reality, however, the kinds of burdens involved in preventing dangerous climate change cross over into many of the different spheres of human life that Walzer discusses. To see this, consider one kind of sacrifice that would need to be made—​the reduction of greenhouse gas emissions. In practice, the distribution of greenhouse gases is determined (and in a Walzerian world would be determined) by the way in which many different spheres are organized. Consider, for instance, his examples of ‘medical care’ and ‘education’. Now Walzer holds that the distribution of medical care should be met according to a principle of need (1983: Ch. 3). However, to do so would have considerable implications for the distribution of greenhouse gas emissions. Energy is needed to construct hospitals, and to provide the heating and lighting and machinery to enable them to function. It is also needed to fuel ambulances, transport necessary medicines, and so on. So complying with Walzer’s account of just healthcare would have implications for the distribution of greenhouse gas emissions. Consider now a second different ‘sphere’—​education. To distribute education in the way that Walzer thinks appropriate (according to ability and interest) would also have considerable implications for the distribution of greenhouse gas emissions (1983: Ch. 8). Distributing education on Walzerian lines requires building schools and universities, heating and providing lighting for them, and providing power for computers and laboratories. It might also require paper, which requires a degree of deforestation, which, in turn, contributes to an increase in greenhouse gas emissions. One might continue this process further, noting how the distribution of other goods that Walzer analyses (such as ‘money’, ‘political power’, and ‘leisure’) according to their own principles would also have implications for the distribution of greenhouse gas emissions, and indeed other kinds of goods that would be affected by climatic responsibilities. However, reflection on the two examples just given is sufficient to establish the key point: namely, the distribution of greenhouse gas emissions is not governed solely by any one sphere. Note that this problem is further compounded once we take into account other goods, the sacrifice of which is needed to prevent dangerous climate change. The upshot, then, is that one cannot appeal to a Walzerian pluralistic theory of justice to argue that climatic burdens should be distributed according to their own specific distributive principle.

31.5.2  Pragmatic Arguments for Isolationism At this point it is worth noting that some adduce more pragmatic considerations in defence of isolationism. Some grant that, in principle, climatic responsibilities should

680   Simon Caney be treated as part of a more general theory of distributive justice. However, they argue that there are pragmatic reasons for treating it in isolation from them. Many argue, for example, that treating climate change in light of a general theory of distributive just­ ice that includes all normative issues will impede progress on reaching any agreement on climate change. Since there is deep and intractable disagreement about the nature of global and intergenerational justice there is a case, this pragmatic argument runs, for treating climate change in isolation from more general issues (Meyer and Roser 2006: 239).17 Three points bear noting here. The first is that the integrationist might happily respond that their claim is about the nature of justice and that if, in practice, we have to bracket off other issues in order to reach agreement that does not undermine their core claim. It does not show that climate justice is isolationist in character: it just shows that we should act if it were in order to get the best results. Second, integrationists might further add that in order to decide to bracket off other issues (like development or rights to procreate) on the grounds that that enables an effective climate agreement that would otherwise not be possible, we need to know whether the contribution to realizing justice that this realizes is worth it. Furthermore, to do this requires taking into account any loss to the realization of justice that pursuing an isolationist position incurs. For example, if it means that we have to frame a climate treaty in ways that ignore many people’s interests in development (and so if more people are prevented from escaping poverty than might otherwise have been the case), then we need to know whether the gain to justice by preventing climate change outweighs the loss to justice from the mitigation policies that thwart development. And this requires a general theory of justice. Put otherwise: the decision to promote preventing climate change over everything else requires an all-​things-​considered view, and that requires a general theory of justice that integrates climate change, development, and all the other phenomena which impact on core interests. We are led back to strong integrationism. My third point returns us to the pragmatic level. Although the argument given above assumes that reaching an agreement on climate change requires—​in practice—​ bracketing out other issues, that assumption needs argument and evidence. It is not self-​ evidently true and stands in need of empirical support. Indeed there is a widely shared view among negotiation theorists—​such as Scott Barrett, Howard Raiffa, and James Sebenius—​that isolating an issue can, in some circumstances, lead to deadlock, and that making a link to other issues can, in some circumstances, facilitate an agreement that would otherwise be unattainable (Barrett 2003:  Chs  12 and 13; Raiffa 1982:  13, 183 and 285–​7; Sebenius 1984:  192–​200). This is also borne out by the negotiations surrounding other environmental problems, such as the treaty on ozone depleting substances, in which agreement was reached when the negotiations included not just CFCs but also finance and technological substitutes (Barrett 2003; Benedick 1998). The 17 

For a distinct pragmatic argument, see Meyer and Roser (2010: 233). Gosseries adopts an isolationist position for both ‘methodological’ and ‘political’ reasons (2005: 283).

Climate Change   681 point, then, is that one cannot make an a priori assumption that isolationism facilitates reaching an agreement (Caney 2012a: 278–​80).

31.6  Question 2—​The Just Distribution of Burdens So far we have considered whether principles of distributive justice should be applied to climatic responsibilities in an atomistic or in holistic way, and we have also considered whether these principles should be treated in isolation from other normative concerns or in an integrated way. We have done so because it is not possible to engage in any substantive analysis about how principles of distributive justice apply to climate change without resolving these issues. With these methodological enquiries now complete we can now finally consider the Just Burden Question. As I noted at the start of this chapter, several principles are commonly defended. One that is commonly adduced is the Polluter Pays Principle (PPP). It is widely recognized that this should play an important role in determining responsibilities. However, a number of problems arise, both concerning the application of this principle to the case of climate change and also concerning how much weight to put on this principle in general. Consider the first kind of problem first. One objection to applying the PPP to climate change is that some of the polluters are no longer alive and so any burden resulting from their emissions needs to be attributed to others. Since people have been emitting greenhouse gases in large (and increasing) amounts since the Industrial Revolution, the PPP seems incapable of addressing many of those emissions (the problem of past emitters).18 One response to this objection is to note that the problem is less acute if we ascribe agency and hence responsibility to groups which persist through time (such as nations or states or firms). Whilst individuals alive now were not, of course, alive during the Industrial Revolution, many nations and states were, and have persisted throughout time. So, one question here concerns the nature of the agent that is said to be subject to principles like the PPP. If we insist on individuals, then the objection retains its force. However, if we think that groups can be, then we can say that the UK, for example, emitted then and so the UK can pay now. Consider now a second objection that is often levelled against applying the PPP to climate change. Some will argue that the PPP can apply only to actions that agents perform in situations in which one could reasonably expect those agents to be aware of the harmful nature of their actions. They then argue that some emitters were excusably

18 

For discussion, see Caney (2005: 756–​60); Gosseries (2004: 41–​2).

682   Simon Caney ignorant of the harmful effects of greenhouse gas emissions, and thus they conclude that the PPP cannot cover all emissions. Some who have emitted greenhouse gases (including some currently alive) should not be held liable for those emissions (the problem of excusable ignorance). This objection has elicited a number of different kinds of response. One reply is provided by Henry Shue, who argues that, whilst it might be unfair to punish the agent for actions which have harmful effects that one could not expect them to have foreseen, it does not follow, and it is not true, that they should not bear the burden of those effects (Shue 2014: 185). A second response focuses on the concern that imposing a burden on those who are excusably ignorant of causing harm pay is being unfair to them. This does sound unfair to the polluter. However, it does so because it focuses exclusively on the burden being imposed and it ignores other considerations (most notably whether they benefit from this harmful activity as well), and it seems questionable to focus solely on the potential burden to the agent that stems from their engaging in those actions and to ignore entirely the benefit that they also derived from those very same actions. Suppose that they do indeed benefit from emitting greenhouse gases: then this changes the situation and it might not be unfair to impose at least some costs on them—​depending on how well off they are in comparison to how well off they are entitled to be (Caney 2010: 208–​10; Gosseries 2004: 40–​1; Shue 2014: 186). The objections from past emissions and from excusable ignorance have raised questions about the application of the PPP to the case of climate change. We might, however, also wonder just how much weight we wish to attribute to the PPP itself. If we consider why agents should pay, the underlying idea seems to be that agents should be held responsible for their choices. As such, the PPP is drawing on the principle of individual responsibility that liberal egalitarians like Dworkin and Rawls affirm in their theories of distributive justice. This, however, draws attention to an important point about the role of the PPP, for Dworkin and Rawls insist that individuals should be held accountable for their choices within a background scheme of justice (Dworkin 2000: 5–​7 and Ch. 2; Rawls 1999: 241–​2, 261–​2, 284, 369–​70). So whilst the idea that agents can be held accountable for their choices is an important one, it is also crucial that they have a fair share of resources and opportunities. This, of course, take us to integrationism (of a strong kind), because to specify the fair share of resources and opportunities that agents are entitled to requires us to draw on a general theory of justice which affirms their civil liberties, economic rights, and so forth. Suppose, for example, that someone affirms a sufficientarian theory of justice: then they will not hold that the global poor should be held accountable for the emissions that they need to cover their basic needs. They would affirm a principle that applies only to the emission of greenhouse gases that are in excess of the amount required to meet core needs (Caney 2010: 218). It would seem very implausible to apply a PPP in an isolationist way, and thereby to ignore the claims of the least advantaged in the world to use electricity, say, to meet their own basic needs. Reflection on the PPP thus brings out the appeal of an integrationist approach.

Climate Change   683 Consider now a second principle of justice often invoked in discussions of who should pay for the burdens of mitigation and adaptation. Many appeal to an Ability to Pay Principle (Shue 2014: 186–​9). To evaluate it we need to consider whether the claim would be that it applies to climatic responsibilities considered in isolation from everything else or whether it applies to other burdens too. Consider the isolationist variant first. As we saw earlier, we lack good principled or pragmatic considerations for adopting isolationism. Furthermore, even if we set that aside, it is not clear what reason we would have for thinking both that this is a sound principle and also that it would apply just to the burdens of addressing climate change. If there were a case for this principle then why would it be limited simply to the burdens associated with mitigation and adaptation, and not to the distribution of burdens and benefits more generally? This point is strengthened further when we actually examine the reasons that might be adduced for endorsing the Ability to Pay principle. Some, for example, might endorse it on utilitarian grounds, arguing that in virtue of diminishing marginal utility, maximizing utility requires imposing the greatest burdens on the most advantaged. Others might defend it on egalitarian grounds, arguing that it brings us closer to a more equal world. What is noticeable here though is that if one accepts either of these rationales then there is no reason to apply the Ability to Pay principle exclusively to climatic responsibilities. Rather, they entail applying the Ability to Pay Principle to other burdens too, and considering it in conjunction with other concerns, such as trade and development. Note, moreover, that if this is true and we adopt the Ability to Pay Principle understood in an integrationist way, then there is no uniquely correct way of distributing climatic responsibilities, because who does what on this issue depends on what else they are doing. If one wealthy agent is paying for a great deal of development then they might be required to pay less for climate change than another agent who is equally wealthy but is paying for less development. Thus far we have seen that in the case of both the PPP and the Ability to Pay Principle, analysing the justifications of, and limits to, these two principles brings out the case for an approach that puts climatic responsibilities into the context of a more general theory of justice. This integrationist conclusion receives further support if we consider a third principle of distributive justice that is sometimes mooted. As noted earlier, some argue that climatic responsibilities should be distributed, at least in part, according to whether one has benefited from the activities which cause dangerous climate change (the Beneficiary Pays Principle).19 But what significance should we place on the receipt of benefits from activities that produce greenhouse gas emissions? A number of questions arise. First, which greenhouse gas emissions should be governed by this principle? It is often invoked to deal with past emissions. So, should it just govern past emissions

19  For discussions, see Caney (2005: 756; 2010: 215ff), Gosseries 2004, Duus-​Otterström 2014, Page 2012.

684   Simon Caney (Duus-​Otterström 2014: 449)? Or, should it also govern the responsibilities concerning current emissions and future emissions? Second, and very closely related to this, how does the Beneficiary Pays Principle fit in with other principles? Is there a division of labour according to which the Beneficiary Pays Principle applies to some emissions, but that other principles (such as the PPP) apply to other emissions? If the answer to this is ‘yes’, then, what is the argument for the Beneficiary Pays Principle applying to some emissions and for it not applying to others? Or, does a plurality of principles apply to the same emissions? Both options face problems. The division of labour approach according to which the Beneficiary Pays Principle applies to some emissions but other principles (like the PPP) applies to other emissions runs into the following problem: if the fact that A benefits from B’s emissions generates responsibilities for A if B is no longer alive, then why would it not also generate responsibilities for A if B is still alive? If enjoyment of benefits from emissions is responsibility-​generating, then why would this apply to some emissions but not others? That seems puzzling. But then if we reject the division-of-labour approach, and we think that responsibility for the emission of some specific greenhouse gases is governed by a plurality of principles, including the Beneficiary Pays Principle among others, we then face the question of how much weight to accord to it and how much to other principles. We also face the methodological question of how we would answer this question: what would be the normative basis for determining how much importance to ascribe to the Beneficiary Pays Principle when compared to other principles? Suppose that we resolve these problems: we then face another issue. Perhaps the most serious problem with the Beneficiary Pays Principle is that the receipt of benefits is not sufficient to entail responsibilities to contribute to paying for the costs of mitigation and adaptation. To see one reason why this is the case, consider someone who benefits from the emission of greenhouse gases (and let us suppose that she does so voluntarily, even though this will often not be the case). However, suppose that this recipient is well below the standard of living to which she is entitled as a matter of justice (whether that is specified by a sufficientarian standard, a prioritarian one, an egalitarian one, or something else). Given that this agent still lacks what she is entitled to as a matter of justice, it would be quite unfair to burden her merely because she has benefited from greenhouse gas emissions and irrespective of her overall rights and entitlements. To illustrate: consider an Indian peasant who lacks access to electricity but who benefits from someone else using excessive electricity. It is true that she has benefited, but since she is beneath what she is entitled to on any plausible account of justice, it would be wrong to infer from her benefiting that she is obligated to bear a burden. This, of course, takes us back to integrationism: it shows that we need an account of individuals’ general rights and responsibilities and that someone having benefited is not sufficient to obligate her if that results in her having less than she is entitled to. Put otherwise: it shows that our focus should be not on what her standard of living is now and how much greater it is than it would be without benefiting. It should be on the standard of living to which she is entitled.

Climate Change   685

31.7  Concluding Remarks This brings us to the end of the analysis. The above has, hopefully, shed light on two questions that any account of distributive justice and climate change should address, exploring both the methodological issues raised by these questions and several key substantive proposals. As noted at the start, the fact that there are two questions does not entail that our answers to them should not be cut from the same cloth. And what the arguments in this chapter have suggested is that, both in defining the just target and also in identifying the just distribution of responsibilities, it is necessary to draw on more general theories of global and intergenerational justice. In the case of specifying the target it is hard to see how this could be done without taking into account the claims resulting from other issues and thus appealing to a general theory of justice. And in the case of the distribution of responsibilities we found that there were both no good principled arguments  or pragmatic arguments for developing a theory of climate responsibilities in isolation from other issues and considerations. In addition to this, reflection on the standard principles of climate responsibility further brought out the implausibility of any treatment not grounded in a more general theory of justice. To this we should add that any general theory of justice invoked to help answer the Just Target Question (e.g., one committed to a sufficiency threshold or to equality, say) will also have implications for the just distribution of responsibilities and so to the Just Burden Question. There is a further point to be made here, namely the fact that both the target and the ascription of responsibilities are grounded in a more general account of justice enables us to consider the question of how we should negotiate potential trade-​offs between them when there is incomplete compliance and it is not possible to realize both. In such circumstances we are faced with the choice of how much to stick to the target that we would select under full compliance (and get some to bear burdens in excess of their fair share) and how much we should stick to the original distribution of responsibilities and burdens (and thus have to adopt a less ambitious target). When faced with this dire choice, it is imperative to have a normative framework for adjudicating between the claims of would-​be victims and the claims of the putative duty-​bearers, and for striking a fair balance between them. And this is what a strong integrationist account can do, because it treats the just target and the just distribution of responsibilities in light of, and as parts of, a single unified theory of justice. The upshot of all this, then, is that an adequate treatment of distributive justice and climate change requires us to draw on a general theory of global and intergenerational justice, and we need to have a good understanding of that before we can provide the answers to the questions that this chapter has explored. One can, of course, appeal to mid-​level principles that would enjoy support from a variety of different theories of justice, and for practical purposes this is vitally important. In debates in the public forum and international negotiations such mid-​level principles can play a central

686   Simon Caney role.20 However, it remains the case that for a definitive answer to the questions examined here, we need to identify the best account of global and intergenerational just­ ice and start from there.

Acknowledgements I am grateful to Serena Olsaretti and Henry Shue for their illuminating comments. Research on this work was supported by the Oxford Martin School, and I  am grateful to it for its support.

References Ally, M. and Beckerman, W. (2014). ‘How to Reduce Carbon Emissions Equitably’. World Economics 15(1): 75–​103. Baer, P., Athanasiou, T., Kartha, S., and Kemp-​ Benedict, E. (2008). The Greenhouse Development Rights Framework: The Right to Development in a Climate Constrained World, revised 2nd edn. Berlin:  Heinrich Böll Foundation, Christian Aid, EcoEquity, and the Stockholm Environment Institute. Barrett, S. (2003). Environment and Statecraft: The Strategy of Environmental Treaty-​Making. New York: Oxford University Press. Benedick, R. E. (1998). Ozone Diplomacy: New Directions in Safeguarding the Planet, enlarged version. Cambridge, MA and London, UK: Harvard University Press. Bovens, L. (2011). ‘A Lockean Defense of Grandfathering Emission Rights’, in D. G. Arnold (ed.) The Ethics of Global Climate Change. Cambridge:  Cambridge University Press, pp. 124–​44. Broome, J. (2012). Climate Matters:  Ethics in a Warming World. New  York and London: W. W. Norton. Caney, S. (2005). ‘Cosmopolitan Justice, Responsibility, and Global Climate Change’. Leiden Journal of International Law 18(4): 747–​75. Caney, S. (2010). ‘Climate Change and the Duties of the Advantaged’. Critical Review of International Social and Political Philosophy 13(1): 203–​28. Caney, S. (2012a). ‘Just Emissions’. Philosophy & Public Affairs 40(4): 255–​300. Caney, S. (2012b). ‘Addressing Poverty and Climate Change:  The Varieties of Social Engagement’. Ethics & International Affairs 26(2): 191–​216. Caney, S. (2014).‘Climate Change, Intergenerational Equity, and the Social Discount Rate’. Politics, Philosophy & Economics 13(4): 320–​342. Cole, M. A. (2003). ‘Environmental Optimists, Environmental Pessimists and the Real State of the World—​An Article Examining The Skeptical Environmentalist: Measuring the Real State of the World by Bjorn Lomborg’. Economic Journal 113(448): 362–​80. 20 

This raises important questions I set aside here about how the norms that should govern reasoning in the public realm about justice and climate change might legitimately be different from the kind of norms that should govern philosophical analyses of the same subject matter. My focus here is on the second kind of enquiry. For discussion of the first see Caney 2012b.

Climate Change   687 Duus-​Otterström, G. (2014). ‘The Problem of Past Emissions and Intergenerational Debts’. Critical Review of International Social and Political Philosophy 17(4): 448–​69. Duus-​Otterström, G. and Jagers, S. C. (2012). ‘Identifying Burdens of Coping with Climate Change:  A Typology of the Duties of Climate Justice’. Global Environmental Change 22(3): 746–​53. Dworkin, R. (1985). ‘What Justice Isn’t’, in R. Dworkin A Matter of Principle. Oxford: Oxford University Press, pp. 214–​20. Dworkin, R. (2000). Sovereign Virtue: The Theory and Practice of Equality. Cambridge, MA: Harvard University Press. Fodor, J. A. (1983). The Modularity of Mind. Cambridge, MA: MIT Press. Gosseries, A. (2004). ‘Historical Emissions and Free-​ Riding’. Ethical Perspectives 11 (1): 36–​60. Gosseries, A. (2005). ‘Cosmopolitan Luck Egalitarianism and the Greenhouse Effect’. Canadian Journal of Philosophy Supplementary vol. 31 on ‘Global Justice, Global Institutions’, ed. D. Weinstock, pp. 279–​309. Heyward, C. (forthcoming). The Cultural Dimension of Climate Justice. Cheltenham: Edward Elgar. Hyams, K. (2009). ‘A Just Response to Climate Change: Personal Carbon Allowances and the Normal‐Functioning Approach’. Journal of Social Philosophy 40(2): 237–​56. Jagers, S. C. and Duus-​ Otterström, G. (2008). ‘Dual Climate Change Responsibility: On Moral  Divergences between Mitigation and Adaptation’. Environmental Politics 17(4): 576–​91. Lomborg, B. (2001). The Skeptical Environmentalist:  Measuring the Real State of the World. Cambridge: Cambridge University Press. Meyer, A. (2000). Contraction and Convergence:  The Global Solution to Climate Change. Foxhole, Devon: Green Books. Meyer, L. and Roser, D. (2006). ‘Distributive Justice and Climate Change: The Allocation of Emission Rights’. Analyse und Kritik 28(2): 223–​49. Meyer, L. H. and Roser, D. (2010). ‘Climate Justice and Historical Emissions’. Critical Review of International Social and Political Philosophy 13(1): 229–​53. Miller, D. (2010). ‘Fair Trade:  What does it Mean and Why does it Matter?’  CSSJ Working Papers Series, SJ013. Available at http://​www.politics.ox.ac.uk/​materials/​centres/​social​justice/​working-​papers/​SJ013_​Miller_​Fairtrade.pdf/​ (accessed 1 December 2017). Page, E. A. (2012). ‘Give It Up for Climate Change: A Defence of The Beneficiary Pays Principle’. International Theory 4(2): 300–​30. Raiffa, H. (1982). The Art and Science of Negotiation: How to Resolve Conflicts and Get the Best out of Bargaining. Cambridge, MA and London, UK: Harvard University Press. Rawls, J. (1999). A Theory of Justice, revised edn. Oxford: Oxford University Press. Risse, M. (2012). On Global Justice. Princeton, NJ and Oxford: Princeton University Press. Sebenius, J. K. (1984). Negotiating the Law of the Sea. Cambridge, MA and London, UK: Harvard University Press. Shue, H. (2005). ‘War’, in H. LaFollette (ed.) The Oxford Handbook of Practical Ethics. Oxford: Oxford University Press, pp. 734–​61. Shue, H. (2014). Climate Justice:  Vulnerability and Protection. Oxford:  Oxford University Press.

688   Simon Caney Vanderheiden, S. (2008). Atmospheric Justice:  A Political Theory of Climate Change. New York: Oxford University Press. Walzer, M. (1977). Just and Unjust Wars:  A Moral Argument with Historical Illustrations. Middlesex: Penguin. Walzer, M. (1983). Spheres of Justice:  A Defence of Pluralism and Equality. Oxford: Basil Blackwell.

Chapter 32

F u tu re Gene rat i ons Rahul Kumar

32.1 Introduction To illustrate the kind of context in which questions of intergenerational justice typically arise, consider the following example: Depletion: As the current generation, we must choose whether to deplete or conserve certain kinds of resources. If we choose to deplete resources, the quality of life for the next three generations would be slightly higher than it would have been if we had chosen to conserve. But it would later, for many generations, be much lower than it would have been if we had chosen a path of conservation. This would be because, at the start of this period, people would have to find alternatives for the resources that we had depleted.1 We choose to adopt a policy of depleting resources. Define a ‘generation’ as, roughly, those who exist at a given time and those who they will live to meet (Gardiner 2011). Future generations are those who will exist in the future but with whom direct contact is not possible because their lives will in no way overlap with that of anyone who now exists. We can, however, benefit or burden them through our current policy choices, such as the balance struck between present consumption and investment that will secure resources to be consumed by them in the further future. What this balance ought to be is a question of distributive justice between generations. Intuitively, choosing Depletion is unjust because it involves striking a balance between present consumption and investment that is unfair to future generations. This intuitive view gains support from two observations. First, the policy inflicts serious burdens on later generations, with no offsetting benefits or compensation, in order to secure small or moderate benefits for the current generation and their immediate descendants. As these benefits are not particularly important for enabling the current 1 

This example is adapted from Parfit 1984.

690   Rahul Kumar generation or their immediate descendants to live satisfactory lives, the importance of securing these benefits is not something that could, hypothetically, be offered to those who end up burdened in the further future as a justification for choosing Depletion. Second, to the extent that those who will exist in the further future need the earth’s resources to live satisfactory lives, they have as much of a right to their fair use as the current generation. If the current generation chooses Depletion, it will be taking unfair advantage of it just happening to have come into existence at a point in time before later generations, who will also need the available resources, have come into existence. The current generation can therefore choose Depletion without experiencing the consequences of doing so. They won’t exist in the resource-​depleted world, so they don’t have to face the victims of Depletion because those people have yet to come onto the scene. In this respect, adopting Depletion is at least psychologically easier than, for instance, affluent states choosing policies that inflict serious and uncompensated burdens on states whose populations live in a state of relative deprivation. Both policies are unjust because they are unfair, but in the latter case those who benefit from the policy choice also have to live with being confronted with the complaints and demands for compensation of its victims. Thus far, I’ve been drawing attention to aspects of intuitive thinking about intergenerational justice. It recognizes some asymmetries between what justice requires in relations between those who exist contemporaneously and between those who now exist and those who will exist in the further future. There are, for instance, fewer ways in which we can affect the interests of future generations than we can those of our contemporaries, so there are arguably fewer obligations of justice owed to future generations than by contemporaries to one another. But the asymmetries are not deep. In particular, intuitive thinking about intergenerational justice does not take the grounds of what justice requires with respect to future generations to be different in kind than what it requires with respect to the relations between contemporaries. Intuitive thinking about intergenerational justice, it turns out, is difficult to defend. First, if the basis of justice claims between contemporaries and between the current and future generations is one and the same, one might think that this could be argued for through some kind of social contract model of reasoning about justice. Such an approach is natural in this context as it aptly captures both the intuitively interpersonal character of justice claims between contemporaries, and the thought that certain choices are distributively unjust because they are unfair. But, in fact, there are real difficulties in extending social contract thinking about distributive justice from the case of relations between contemporaries and future generations with whom they will not overlap. This suggests that the basis of claims concerning what future generations are owed as a matter of justice cannot in fact be understood as continuous with justice claims between contemporaries. Second, intuitive thinking about intergenerational justice is vulnerable to what Derek Parfit calls the non-​identity problem (Parfit 1984). Consider the case of a fourteen-​year-​old girl who wants a baby. If she has one, she will be unable to give him a good start in life. If she

Future Generations   691 has her first baby at twenty-five instead, she will be able to give him a better start in life—​ but that would be a different baby. So whom is she harming by giving birth at fourteen? No one. Not the baby, as long as his life is worth living. Now, suppose we, who are currently alive, facing the choice between depleting or conserving resources, decide to go ahead and deplete non-​renewable resources, with the result that the lives of future people are much harder. It would seem that we have made things worse for those future people. But, in fact, as long as their lives are worth living, this is not the case—​because if we had acted differently, the world would have been different, and those particular people would never have existed (in the same way that if cars had not been invented most people alive today would never have been born). So, although we have made the world worse in the future, we have made life worse for no one. If we understand justice as having to do, in part, with stringent claims not to be harmed, or not to have one’s rights violated by the policies of non-​voluntary institutions that influence one’s life prospects, and whose violation may justify a legitimate demand for compensation, the non-​identity problem poses a fundamental challenge to the idea that there is a subject matter of intergenerational justice to be made sense of. As both who and how many in fact come into existence in the further future is not independent of policy choices made now, how can a choice like that of the choice between conserving and depleting resources be criticized on the grounds that it was unfair to, or harmed, or violated the rights of, those who will live in the further future? How can a person be thought of as a victim of injustice, or as having a legitimate complaint about, a past policy choice that may well entitle him to some kind of (possibly symbolic) compensation if it in no way adversely affected his interests because he in particular would not exist as a bearer of interests, had a different policy been adopted?2 The discussion of different ways of thinking about the grounds of obligations to future generations falls into four sections. Sections 32.1 and 32.2 examine the prospects of using the resources of a social contract approach to justice to understand the requirements of intergenerational justice. Section 32.1 examines how Rawls’s contractualist account of justice between contemporaries might be extended to the intergenerational case, while section 32.2 examines ways in which Gauthier’s contractarian account of justice could be deployed to make sense of what intergenerational justice requires. In both cases, I identify what I take to be important obstacles to successfully doing so. In section 32.3, I  examine the case for understanding the requirements of intergenerational justice as a matter of fulfilling the rights of those who will live in the further future. Though such an approach cannot readily do justice to the thought

2  Parfit’s own view is that the non-​identity problem does not alter the reasons we have to avoid making life worse for those who will live in the further future. In this discussion, I do not examine Parfit’s own, or any consequentialist, views concerning obligations to future generations. The consequentialist approach to thinking about such obligations is important, and arguably dominates the current literature. But consequentialists do not see themselves as advancing a distinctive account of what justice, as opposed to morality, requires with respect to future generations. My focus in this discussion is concerned only with the former.

692   Rahul Kumar that choosing Depletion is unfair, it does promise to be able to make sense of choosing Depletion being an injustice done to those who will live in the further future. I argue that the non-​identity problem poses a challenge to the rights-​based approach to intergenerational justice that it cannot readily overcome. Section 32.4 discusses the application of luck egalitarian thinking about distributive fairness between living individuals to the case of fairness across generations. I argue that though the non-​identity problem does not pose a challenge to the luck egalitarian approach, it cannot support intuitive convictions concerning obligations to save and invest now in order to secure a certain decent quality of life for those who will live in the further future. I conclude with a brief consideration of the question of the importance of whether obligations owed to future generations are obligations owed to them as a matter of just­ ice or a matter of interpersonal morality.

32.2  Rawlsian Contractualism and Intergenerational Justice Social cooperation, as Rawls sees it, is necessary for human beings to lead satisfactory lives in the circumstances of justice. The circumstances of justice include the rough similarity of human beings in their mental and physical abilities, moderate scarcity, and limited altruism. These conditions create the potential for a conflict of individual interests as well as mutually beneficial cooperation. The social role of principles of just­ ice are to set the terms of a fair system of cooperation over time, from one generation to the next (Rawls 1971; Rawls 1993; Rawls 2001). In particular, Rawls sees the primary role of principles of justice as regulating the basic structure of a stable, ongoing society. By the ‘basic structure’, Rawls has in mind those institutions that are necessary to social cooperation that extends over time, such as a framework specifying the procedures for legislating and enforcing laws and adjudicating legal disputes, rules governing economic exchange, such as property rights and contractual relations, and norms regulating permissible forms of family life (the latter being important to enable society to reproduce itself). Rawls’s well-​known argument is that the morally justified principles for this role are those that would be chosen by parties to a hypothetical agreement on principles, behind a veil of ignorance (the original position device). It is assumed that all parties to the agreement are free and equal citizens, each representing some segment of the general population. Further, each party is assumed to be equal in bargaining power and motivated only by a desire to do as well for themselves as they can; each is mutually disinterested, so psychological factors, such as envy, have no role to play in converging on particular principles. The veil of ignorance blinds the representatives to morally irrelevant information about themselves that is not relevant to fixing fair principles, such as

Future Generations   693 knowledge of their race, gender, age, class, life plans, religious views, etc., so the parties to the agreement are equal in bargaining power. Rawls stipulates that all the parties to the original position belong to one generation, but they do not know what generation they belong to. That information is morally irrelevant. As what is being selected are principles for the basic structure of an ongoing society that, by stipulation, goes on indefinitely through time, there are an indefinite number of generations stretching out into the future. The issue of intergenerational justice arises behind the veil in the form of the need for a principle of just savings that specifies what the balance should be between current consumption of resources and investment in capital that will generate resources for consumption by future generations. Specifically, the problem of what is owed to future generations arises in the form of the following problem: Rawls holds that there is a natural duty to set up and maintain just institutions. Just institutions are those that realize the principles of justice he argues would be chosen behind the veil of ignorance (the principle of equal liberty and the two-​ part principle of distributive justice that specifies, first, that the offices and positions associated with the basic structure be open to all under conditions of fair equality of opportunity, and second, that inequalities are only justified to the extent that permitting them is maximally advantageous to the worst-​off). But the extent to which a society is actually in a position to implement the two principles depends, in large measure, on its state of economic development. In the early stages of a society, the realization of just institutions may not be feasible. The best that can be done to comply with the duty to set up and maintain just institutions is to save part of the wealth created by economic cooperation so as to pass on a greater capital stock and more resources to the next generation. Over the course of several generations, sufficient resources and capital will have been accumulated for the realization and maintenance of a just basic structure (Rawls 2001). Once that stage is reached, the requirement that each generation save abates. The just (or fair) savings rate is itself something to be determined behind the veil of ignorance, behind which parties to the agreement do not know what generation each belongs to. In choosing a savings rate, therefore, they have to balance several considerations. Saving for future generations takes away from resources that would otherwise go to improving the prospects for the worst-​off in society at a given time. If the savings rate is low, just institutions may never be realized. But if it is very high, the requirement to save will place very serious demands on early generations, especially the worst-​off. Though Rawls offers no clear procedure for arriving at a definite savings rate, it is plausible that the considerations his argument identifies are all relevant. The problem with deploying the original position device in this way is that what justice requires is supposed to be an output of the kind of reasoning described by the original device, not an input to its construction. The parties to the original position cannot, therefore, deliberate about principles of justice with an eye to both doing as well for themselves as they can and making sure that they are able to comply with the natural duty of justice to set up and maintain just institutions. But if parties to the original position are simply concerned to do as well for themselves as possible, and they don’t know what generation they belong to, what motivation do they have to save? Assuming

694   Rahul Kumar generations that do not overlap, fair cooperation on terms of reciprocity appears not to be a possibility in the way that it is between members of a generation. One generation might save for the next generation, but the generation that saves cannot be benefitted by the next generation. So why would any given generation save at all? Whether the past generation saved is something it can do nothing about. It could save for the next generation, but it can clearly do better for itself by simply consuming all available resources and not saving anything at all. Rawls tries to circumvent this problem when he first discusses the issue of just savings in Section 44 of A Theory of Justice by introducing an arguably ad hoc motivational assumption: human beings have a natural motive to care for their children and their grandchildren. An intergenerational chain of concern within family units, then, motivates the requirement to settle on a just rate of savings for the future. The problem with this motivational assumption is not just that, in the context of the original position construction, it is ad hoc. It is also the wrong kind of reason to justify a requirement of social justice that each generation save resources and capital to pass on to future generations. All the argument shows is that the principles of justice converged upon behind the veil of ignorance will accommodate, to some extent, the natural concern individuals have for their immediate descendants (Barry 1977, 1989). The concern is accommodated insofar as parties to the original position allow resources that might have gone to raising the position of the worst-​off to be held back as savings for future generations. But the argument doesn’t show that each generation is in fact required as a matter of justice to save anything for future generations. If anything, it leaves how much is saved by each generation hostage to how much families happen to care about their immediate descendants. In Political Liberalism, Rawls briefly offers a different argument for the requirement that each generation save for future generations. He introduces a constraint on original position reasoning that requires the parties to ‘agree to principles subject to the constraint that they wish all preceding generations to have followed the very same principles’ (Rawls 1993). Parties to the original position, not knowing what generation they happen to be along the chain of generations, have a clear reason to agree to a principle requiring savings for future generations that is not excessively demanding. This way of reasoning is, however, very indeterminate. The parties might reason: ‘we might be the first generation, so we will have nothing to gain from the savings of the past generations and a great deal to benefit from consuming a great deal of resources and passing on nothing. But we might also be quite far down the line of generations, in which case enough will have been accumulated that very little needs to be saved for future generations. Or we could be a middle generation that is still quite reliant on the savings of past generations in order to maintain a just basic structure.’ The constraint Rawls introduces is appealingly plausible because it forces parties to the original position to take into account the interests of all generations. In particular, it focuses their attention on the question of what rate of savings would be most acceptable to the generation to whom it is least acceptable. But it is difficult to assess what kind of requirement of just savings it might yield. That makes it difficult to assess the overall plausibility of this strategy.

Future Generations   695 Finally, it is important that the way in which the question of intergenerational justice arises for Rawls is in the form of a requirement to save part of the fruits of cooperation in order to allow a just basic structure to be maintained over time. But the objection to not curtailing carbon emissions, thereby raising the likelihood of catastrophic climate change, is not one captured by the idea of a failure to save enough of the fruits of cooperation to maintain a stable and just basic structure.

32.2.1 Extending the Veil of Ignorance? The scope of Rawls’s treatment of intergenerational justice is limited. But that doesn’t discredit his approach to thinking about principles of intergenerational justice. Barry, for instance, suggests that we could formulate such principles by asking what principles would be chosen in an intergenerational original position. Taking up a suggestion that Rawls considers and rejects, he argues that we should conceive of those represented in the original position as representatives of different generations, each of whom is ignorant of what generation they represent. Thinking about intergenerational justice in this way looks to be a promising way of justifying the intuition that choosing Depletion is unfair to future generations. There are two ways this kind of argument might be run. The first appeals to a principle of fair play: if you benefit from a scheme of social cooperation that involves others accepting certain costs or burdens, it is unfair for you to accept the benefits of cooperation but not accept your fair share of those burdens or costs associated with it. In the context of intergenerational justice, the thought is that maintaining, for example, our lakes and fisheries is an intergenerational cooperative project. If, as is plausible, we benefit from the stewardship of previous generations that have preserved them in good order for us to use, it would be unfair for us to free-ride and not do our part to maintain them for use by future generations. Say we accept that it is appropriate to conceive of the maintenance of the ecosystem as a form of intergenerational cooperation for mutual benefit (it isn’t obvious that we should). The argument will still face two objections. First, in an intragenerational case of free-riding, one person reaps the benefits and allows the others to do the work. There is more work for the others to do because they are making up for the free-riders’ non-​ compliance with the demands of the cooperative scheme. Nothing like this is true in the intergenerational case. The costs borne by previous generations to maintain the ecosystem in good order are sunk costs. Whether or not our generation follows suit cannot change the burdens they bore. It is hard to see, therefore, how non-​compliance counts, in this case, as free-riding.3 3 

Here I am assuming that the cooperation (or form of reciprocity) in question is direct cooperation, in which individuals cooperate with one another in ways that are mutually beneficial. The line of reasoning in question may, however, fare better if indirect cooperation is assumed. Arguably, such an assumption is more appropriate in the intergenerational case than that of direct cooperation (Gosseries 2009). I will turn to this question in section 32.3.

696   Rahul Kumar Second, it is implausible that the injustice of choosing Depletion depends on whether or not the previous generations chose such policies. If we inherited water supplies and fisheries in very poor condition we should clean them up, both for our own sake and the sake of future generations. The reason to do so has nothing to do with doing our part in a fair scheme of cooperation. It is plausible that if the water supplies and fisheries were bequeathed to us in a healthy state, we have some obligation to maintain them because they were passed down to us in a good state. But to conclude that the objection to failing to do so has to do with unfairness is to assume what is contention. Principles of intergenerational justice, chosen behind the veil of ignorance in an intergenerational original position, need not be conceived of as principles governing intergenerational cooperation in order to make sense of the intuitive conviction that choosing Depletion is unfair to future generations. Rather, the unfairness can be characterized as the unfairness of a generation unfairly taking advantage of its knowledge of what generation it happens to be; were it to be deprived of this information behind a veil of ignorance, it would never agree to a principle of justice permitting it to choose Depletion. Imagine an intergenerational original position, where each of the parties to the agreement represents a particular generation. No one knows what generation they belong to or any particular facts about the identities of those who constitute their generation. All each of the representatives knows is that he or she represents a particular generation and is motivated to do their best for that generation. Subject to these constraints, the proposal goes, what would be agreed to is a principle requiring the current generation to do ‘what they reasonably can, subject to their needs, rights and other moral duties, to ensure that future people are capable of normal functioning and live lives with normal life expectancies’ (Reiman 2007). Because the current generation can, at little cost to itself, do a great deal to ensure that those who will live in the further future are capable of normal functioning by choosing to conserve rather than deplete, it would be unfair of them not do so, an injustice that wrongs (or violates the rights of) those who will live in the further future. This is a promising way of conceiving of the grounds of principles of intergenerational justice. But it is derailed by the non-​identity problem. The intergenerational original position assumes that all who do exist or will exist are represented as parties to the hypothetical agreement. Facts about the particular identities of those who constitute each generation are not admitted into the parties to the intergenerational original position, on the grounds that such facts are morally arbitrary from the point of view described by the intergenerational original position. That who the particular members are of each generation turn out to be will depend on the policy choices of earlier generations is therefore thought not to be a problem. But it is also true that how many generations there are is not independent of what policies are chosen; some generations that will exist if one policy is chosen will not exist if an alternative policy is chosen. It is here that the non-​identity problem becomes a serious worry.

Future Generations   697 To make the problem concrete, assume that choosing to deplete will result in many more generations coming into existence than would come into existence if a path of conservation were chosen. Further, assume that those who will exist if resources are conserved will enjoy normal functioning and have normal life spans. If resources are depleted, it will still be the case that many of the generations who come into existence enjoy normal functioning and life spans. But many further generation will exist with shorter life spans and less-than-normal functioning (though their lives will be worth living). Some of the generations will exist regardless of whether resources are conserved or depleted. What policy is chosen will affect the identities of those who constitute them, but not the fact of them. But there is no clear rationale for excluding from behind the veil, and a good reason for including there, representatives of the generations who are only possible generations, or generations who will only exist if a particular policy is chosen. The positive reason for including representatives of the generations who are only possible can be illustrated as follows: consider two policies, A and B. A will result in seven generations coming into existence over the next several hundred years, after which the human race will go extinct. Policy B will also result in seven generations coming into existence. But rather than disappear, the human race will continue for several more generations in a state of unrelenting suffering and misery. Clearly, A is the right policy. The human race going out of existence is regrettable, but it is better than it carrying on in the form of lives lived that are not worth living. But if we just compare the implications of A and B for the generations who will exist regardless of which policy is chosen, there is no reason to prefer one to the other. The serious objection to policy B only comes into view when we include the representatives of those generations who will only exist if policy B is chosen.4 Return to the choice of whether to conserve or deplete resources. The generational representatives behind the veil don’t know what generation each represents. Nor does any one of them know whether or not they represent a generation that will definitely exist whichever policy is chosen, or a generation whose existence is contingent on that choice. If each is concerned to do the best possible for his or her generation, each will see that an outcome that is less objectionable from the point of view of the generation who has the strongest objection to it than any of the alternatives will obtain if a principle that requires resources be depleted is agreed upon. Though many generations will then exist that enjoy shortened lives and less-than-normal functioning, such an outcome is better for those who constitute those generations than the alternative of not existing at all. This

4 

It may sound odd to include representatives of generations who object to adopting the policy that will result in the generations they represent coming into existence. The representatives in question need not, however, be thought of as actually part of the generations on whose behalf they speak, so there is no reason to think of them as objecting to their own existence.

698   Rahul Kumar conclusion, however, contravenes the intuitive view that depleting resources is unfair to those who will exist in the further future.

32.3  Hobbesian Contractarianism and Intergenerational Justice According to the more libertarian strand of social contract thinking, ‘contractarianism’, norms or principles of justice are those that it is rational for self-​interested individuals to agree to be guided by in their conduct in order to enable mutually advantageous cooperative economic activity. As developed by Gauthier, the view asks one to imagine a state of nature along the lines described by Locke, in which all agents respect one another’s natural rights, and in particular, one another’s property rights (Gauthier 1986). There is no cooperative activity, but there is also no interference in one another’s lives and, in particular, there is no taking advantage of one another. ‘Not taking advantage of ’ is not equivalent to ‘doing nothing that harms’ another. I may open a business down the street from you, foreseeing that this will make your business less profitable. I’m only taking advantage of you if I open the business down the street from you in order to drive you out of business for my own gain (perhaps I want to open my own store in your location). Similarly, I don’t take advantage of you if your car breaks down along a deserted road and though I could help, I decide not to. I only take advantage of you in the state of nature if I have, for instance, arranged for your car to break down in a deserted area (perhaps I want to keep you away from a business meeting so I can do better for myself). The contractarian claim is that though exchange is possible in a state of nature in which we all respect one another’s natural rights, we can in fact all do better for ourselves by agreeing to constrain our rationally self-​interested behaviour by certain moral rules. Doing so enables forms of cooperative economic activity (involving, for instance, promises and contracts) that would not otherwise be possible. To build a barn together, for instance, requires that we both promise to show up every day to do our part in the construction of the barn, even when, on occasion, following through is other than what would be best from the point of view of individual self-​interest. To not show up would be unfair to the others, given our agreement. Principles of justice that we all agree to be bound by make this kind of cooperative activity possible. The obstacle to deploying this approach in the context of intergenerational justice is that the circumstances of justice, in which mutually beneficial cooperation is both possible and necessary, look not to hold in the context of relations between generations, as cooperative activity between non-​overlapping generations looks to be a non-​ starter. The activity of earlier generations can harm or benefit later generations, but later generations that do not overlap with earlier generations cannot harm or benefit earlier generations. As the contractarian approach holds that requirements of justice

Future Generations   699 have to do with the distribution of what it is we produce together (the cooperative surplus), there is a case for concluding that it will not be able to include future generations within the scope of the class of those who can be subject to unjust, or unfair, treatment by others. This conclusion can be resisted in a couple of different ways. First, it could be argued that some of our interests and personal projects are intergenerational projects, in the sense that the value of these projects for us now depends on their being carried on and developed by future generations (De-Shalit 1995;Thompson 2009; Scheffler 2013). Research, for instance, is often characterized as a matter of building on the insights of the past to advance human knowledge of the subject, with an expectation that generations after one’s death will take up the insights one has developed and build on them to deepen understanding of the subject. It isn’t, therefore, true that cooperation with future generations is not possible. But even if we accept the premise that many of our interests and personal projects are intergenerational in character, this isn’t enough to show that disjunctive generations can stand in cooperative relations with one another. First, all it shows is that our intergenerational projects (to the extent we have many at all) require the existence of (or at least, a belief that there will be) future generations to carry them forward. But it is reasonable to think that if they do so, they will do so because they find them interesting and worth carrying on. They may well do so because the kinds of projects that are passed down are the projects that are presented to them by the culture they inherit as projects worth pursuing. But there is no reason having to do with fair exchange, or cooperation, for finding them worthy of carrying on and developing. Taking up these projects and advancing them need not, for instance, contribute to the advancement of other projects or interests future generations have found to be worthy of pursuit. Some projects may, of course, contribute to the interests of future generations, and we may be justified in our confidence that future generations will take an interest in them because they have benefitted from their past pursuit (one might think of medical research in this way). But that isn’t enough to show that any of the reasons for pursuing them flows from an obligation to do one’s part in a cooperative enterprise. Second, when we cooperate, I do something for you in exchange for your doing something for me. But there is also usually some kind of mechanism to assure compliance with the agreement. If you don’t do as you say you may, for instance, be subject to social sanctions by others that you will find burdensome. There is no way of assuring compliance in the case of cooperating with future generations, as we are assuming that the earlier generations are gone by the time those who will live in the further future are on the scene. A more promising strategy for justifying the inclusion of future generations within the scope of justice appeals to the idea that the generations are linked by indirect cooperation. Distant generations may not interact with one another, but adjacent generations do; the old and the young, for instance, engage in mutually beneficial cooperation during the period in which their lives overlap. The elderly might, for instance, be dependent on the young for the maintenance of public services on which they depend, so engage in

700   Rahul Kumar cooperative activity by continuing to pay taxes for services they no longer need in order to secure the availability of those they do rely on. Further, it is also in the interest of the younger generation to cooperate with the older generation because the younger generation will have to cooperate with a younger generation at some point in its future. Their cooperating with the previous generation now sets a precedent, or expectation, for the preceding generation to cooperate with them when they are older. In this way, overlapping generations create a chain of interaction that links all generations together, including non-​adjacent generations (Gosseries 2009; Heath 2013). Let me mention just two objections to this strategy. First, though cooperation between members of the same generation can be mutually advantageous, the claim that adjacent generations must cooperate with one another requires defence. The older generation might require certain services in order to maintain themselves in old age. But if they accumulate enough, these services could be purchased on the open market. Or, foreseeing the need for future assistance when they are still young, they could invest resources into building automated systems and robotic caretakers who will assist them. There is also nothing stopping them from manipulating the younger generation from a young age such that they grow up believing that they have a moral duty to take care of the elderly, even at significant cost to themselves. There is no reason internal to the contractarian approach for taking doing so to be impermissible; shaping the younger generation is not, for instance, a matter of taking advantage of them. Second, the strategy appeals to the idea that each generation needs to cooperate with the generation that comes before it and the generation that comes after it, and that this creates a chain of cooperation between the generations that extends indefinitely over time. Heath illustrates the idea through the operation of an unfunded pension scheme (Heath 2013). Such a scheme transfers earnings from younger employees to older employees in return for a promise that even younger employees, many of whom have yet to join the workplace, will do the same. Present workers expect the workers who will join the system to contribute to the pension scheme, thereby funding the pensions of the present workers when they retire, because present workers expect the incoming workers to expect that they will one day have younger colleagues who will do the same for them, and there is no expectation that this scheme of indirect cooperation will at some point shut down. Can the unfairness of choosing Depletion be explained using a contractarian approach that appeals to indirect intergenerational cooperation? Consider the following example: generation A decides to bury nuclear waste in the ground, knowing that it will not have deleterious effects on the population for another 400 years. After 400 years, the containers holding the waste will start to erode, releasing radioactive toxins into the atmosphere that will seriously damage the health of anyone alive at that time under the age of forty (call them generation F). Why should generation A not pursue this course of action? The argument is that generation F, their health having been badly damaged, will not cooperate with generation E (they may simply be unable to). But in that case, it is not in generation E’s interest to cooperate with generation D. Their main reason for doing so is to maintain an expectation

Future Generations   701 of intergenerational cooperation so that generation F will, in time, cooperate with them. Continuing this line of reasoning shows that the chain of cooperation will simply unravel if generation A buries the waste, because any expectation of cooperation between generations collapse. The only way to stop it is for generation B to refuse to cooperate with generation A unless it agrees not to do what will eventually harm generation F. This line of reasoning gets the right conclusion: it is in the interest of all concerned to agree to a principle of justice that does not permit A to defer the burden of dealing with the waste to later generations. But, first, it is also not clear why the principle agreed to would not permit the burden of the waste to be deferred if it was done secretly, so as not to disrupt the expectation of ongoing intergenerational cooperation. Second, even if one judges the argument to be successful in reaching the right conclusion, it does not capture the intuition that the reason why choosing Depletion is unjust is that it is unfair to those who will be burdened by the leaking waste. Third, if who exists in each generation is not independent of whether or not the waste is buried, and those who live in generation B have lives well worth living, it is no longer clear that generation B has reason not to cooperate with generation A. Generation B might not be able to benefit from downstream intergenerational cooperation, but they might not exist at all if generation A chooses not to bury the waste. Finally, indirect cooperation can only explain A’s reason not to bury the waste if it has more to gain from cooperation with the subsequent generation than it does from burying the waste. But, intuitively, no such assumption is needed to justify A’s obligation not to bury the waste. A ought not to bury the waste even if it will be sacrificing its own interests by not doing so.

32.4  Rights and Future Generations The appeal of thinking about intergenerational justice in social contract terms is that it promises to allow us to make sense of a particular complaint that can be lodged by future generations against choosing Depletion, namely that doing so is unfair to them. What I’ve argued so far is that unlocking this potential by extending social contract reasoning, in either its Hobbesian or Kantian forms, from the intragenerational to the intergenerational case faces obstacles. They may well be surmounted in either version of the social contract account by theoretical innovations that remain true to its animating ideas. An alternative response to the problems to be faced in extending social contract reasoning to the intergenerational context is to focus on the idea that choosing Depletion does an injustice to those who will live in the further future. But the complaint of injustice is best understood not as a complaint that doing so is unfair to them. Rather, the injustice is one of violating the rights of those who will live in the further future. Woodward elegantly articulates the intuition to which the rights-​based approach appeals in his classic discussion of the non-​identity problem and the rights of future

702   Rahul Kumar people. Considering a case of the current generation burying toxic waste in the ground whose deleterious effects will not be felt for several hundred years, he writes: [It] seems plausible that the [people who are later killed by the leaking toxic waste], like other innocent people, have rights that others should not knowingly pursue policies that will kill, injure, or poison them or will create substantial risks of these results, at least when there is no weighty justification for such policies and alternative policies which involve no such risks are available. (Woodward 1986: 812)

The content of the right in question can be specified in a number of different ways. It could, for instance, be specified as a right to live a life above a certain threshold of well-​ being, a right to lead a life characterized by normal functioning, a right to the best start in life, or a right to the possibility of enjoying the same access to environmental and genetic resources needed for living a satisfactory life that have been enjoyed by previous generations (Woodward 1986; Meyer 2003; Velleman 2008). The general intuition to which such views appeal is that each person has a right not to exist in a state of being disadvantaged with respect to their prospects of living a satisfactory human life when ending up in that state was avoidable. In some cases this kind of right will underwrite claims, as a matter of justice, of already existing people to resources, freedom from interference, etc. In other cases of the kind Woodward has in mind, the right is a right of those who do not yet exist but will (or may) exist that choices not be made that will result in their existing in a disadvantaged state when that outcome is avoidable. The idea that individuals do have such a right is not ad hoc or under motivated. Much of the appeal of a rights-​based approach to intergenerational justice is that it appears to make sense of the thought that choosing certain policies, like Depletion or burying toxic waste in the ground knowing the consequences of doing so will likely be very bad for those who live in the further future is an injustice done to those people. As Mill says, ‘[j]‌ustice implies something which it is not only right to do, and wrong not to do, but which some individual person can claim from us as his moral right’ (Mill 1969 [1863]). It also looks not to be vulnerable to the challenge of the non-​identity problem, as the rights-​based approach claims that certain choices do an injustice to the right holders even if those who end up coming into existence are not worse off than they otherwise would have been had an alternative course been chosen. We can distinguish two versions of the rights-​based approach to intergenerational justice: the first holds that choosing Depletion violates the rights of those who will live in the further future, while the second identifies the objection to that choice in the sequence of events initiated by that choice that results in future people coming into existence in a state in which they both have a right not to exist, and that right not being fulfilled because they do exist in that state. The most serious problem faced by the first version of the rights-​based approach is that of specifying who the right holders are supposed to be. At the time of choice, there are no determinate right holders, because those who live in the further future have not

Future Generations   703 yet come into existence. Further, who will come into existence is not independent of the choice made. If there are any right holders at the time of choice, they are merely potential people. It may not make any sense to speak of the rights of merely potential people. But grant for the sake of argument that it does. There is a further problem. Because the choice in question is an identity-​fixing fact, the class of potential people includes those who will exist if resources are depleted, and those who will exist if resources are conserved. These are, at the time of choice, distinct sets of potential people. Now, stipulate that in choosing Depletion, one will bring into existence people whose lives are lived below a certain threshold of well-​being. They have a right not to be brought into existence when the lives they will live will be below the threshold. Is that a reason to not choose Depletion? Arguably, it isn’t, because it is very plausible that those who come into existence with lives available to them that are below the threshold will simply waive their rights. That is, it is plausible that were they, hypothetically, to be asked whether or not they were willing to waive their right not to be brought into existence in a certain state, they would simply say ‘yes’. Though their lives would be lived below the threshold, it is better for the person who lives to live that life than have never come into existence at all. This might sound like a bad argument because it sounds like those who live in the further future are being offered a coercive threat along the lines of ‘Waive your right or I won’t bring you into existence.’ It is reasonable to think that the waiving of a right in response to this kind of threat has no normative validity. But this misconstrues the argument. What is being said is not ‘Waive your rights or I will deprive you of existence.’ Rather, the point is that it is just a fact that those will exist if Depletion is chosen can only exist with their rights having been violated. That fact is independent of anyone’s will. Because it is just a matter of metaphysics that they cannot exist without their rights being violated, it is plausible they would, hypothetically, agree to waive their rights because a subpar life is better than no life at all (Parfit 1986). The second version of the rights-​based approach avoids the problem of speaking of the rights of merely potential people. It holds that choosing Depletion is not wrong because it violates a currently existing right. Rather, it locates the objection in bringing people into existence who will both have certain rights when they come into existence but whose rights will also be violated because they will exist in a state that is worse than that to which they have a right. This approach says that someone born into such a state can claim that the choice that resulted in their being brought into existence has been an injustice done to them, because the act that created them also created rights (that they bear) that will be violated. This is so even if they are not worse off than they otherwise would have been had that choice not been made and assuming there was another choice available, one that would have resulted in someone else coming into existence without violated rights. (There having been an alternative is important, lest the approach end up committing to the claim that it is always unjust to bring people into existence who only have lives available to them that are below a certain threshold of normal functioning, but still well worth living, even if they are the only people who could have been brought into existence.)

704   Rahul Kumar This approach is open to several objections. I’ll just mention two. First, it assumes that there is a very strong objection to creating a right that cannot be fulfilled, such that there is a decisive objection to bringing someone into existence who will then have a right that cannot be fulfilled. But, first, though it is presumptively wrong to create a right in a person that cannot be fulfilled, that presumption can be overcome if doing so is the only way to benefit the right holder. Say you need money to pay off a large medical bill whose arrival you are anticipating, so you decide to invest your money in a risky scheme that promises to triple your money in a very short period of time. Studying the details of the scheme, I conclude, correctly, that the offer is fraudulent. I beg you not to invest. The only way I can convince you is by promising to give you the needed money to pay the bill when it arrives. I don’t have the money. So you now have a right against me that can’t be fulfilled. I understand what I have done, but also, correctly as it turns out, foresee that once the panic passes you will see clearly that investing is a bad idea and will pursue some other means of paying off the bill. Though I have created a right in you that cannot be fulfilled, I am surely morally justified in doing so. I have done you no injustice. One can, then, accept that bringing a person into existence who has a life that is worth living but is below the level of, say, normal functioning, is to bring a person into existence whose rights cannot be fulfilled. But it doesn’t follow that that person can complain of the injustice done to them in bringing them into existence. For even an impaired life that is worth living is a significant benefit to the person whose life it is, and it isn’t possible for them to have been benefitted to a greater extent than they have been (because of the non-​identity problem). It can’t, therefore, simply be assumed that this benefit is not sufficient to justify the creating of a person who will bear rights that cannot be fulfilled. A further argument is needed to establish that doing so is to do an injustice to that person. Second, this version of the rights-​based approach appears to be just as vulnerable to the rights-​waiver argument as the first version. The reason not to choose Depletion is that doing so is unjust. It is unjust because it will result in individuals coming into existence in the further future whose rights will be violated. But the only way this can be avoided is by choosing what results in their not existing at all. It is plausible that those who will exist in the further future will prefer to exist as bearers of rights violated by the very circumstances of their existence rather than not exist at all. If they have no reasons that might be offered on their own behalf to insist on their rights rather than waive them, the only reasons not to choose Depletion will be impersonal ones, such as that the existence of people living in the world with violated rights makes the world impersonally worse. But this kind of impersonal reason is not the right kind of consideration to underwrite the charge that choosing Depletion is unjust. For that charge to stick it has to be the case that because depleting resources would be an injustice done to those who will come into existence if resources are depleted, the current generation is obligated not to deplete resources. But it is hard to see why those who would exist if resources are depleted would object to their Depletion. These objections do not establish that no version of the rights-​based approach will succeed in casting light on the basis of the intuitive judgment that an injustice is

Future Generations   705 done to those who will live in the further future when Depletion is chosen. But they do draw attention to certain obstacles that a successful rights-​based approach to intergenerational justice needs to overcome.

32.5  Luck Egalitarianism and Intergenerational Fairness The rights-​based approach aims to capture the idea that the duty not to choose Depletion is a directed duty, one owed to future generations. But it abandons the idea that choosing Depletion is unjust because it is unfair to future generations. Luck egalitarian thinking about distributive justice takes a different tack: it aims to preserve the idea that the objection to choosing Depletion has to do with fairness. It holds, however, that the unfairness in question should be understood in a telic rather than deontic sense. The objection to choosing Depletion, on this view, is not that in choosing Depletion the current generation violates a duty of fairness owed to those who will live in the further future. Rather, it is that Depletion will result in future generations facing worse life prospects than earlier generations through no fault or choice of their own. The luck egalitarian claim is that such a distribution is intrinsically unfair, regardless of how it came out. And because it is unfair, it is, ceteris paribus, unjust. Because luck egalitarianism locates the objection to choosing Depletion in a comparison between the life prospects of distinct generations, it is not vulnerable to the threat of the non-​identity problem. Say we grant that the choice of Depletion is an identity-​fixing fact. Further, say that those who will end up living in the further future will have a shorter life span than those who would have existed had resources not been depleted. Let’s also grant that existence is a benefit (a controversial assumption), such that we can say that, in one way, the choice of Depletion benefits those who live in the further future. Their expected life spans may be comparatively shorter than those of previous generations, but they are better off than if they had never existed. The luck egalitarian claim is that such intrapersonal considerations have no bearing on the interpersonal unfairness of those who happen to exist in different generations facing very different life prospects through no fault or choice of their own. By luck egalitarian lights, it is unjust to pursue Depletion. But there are obstacles to fully understanding the requirements of intergenerational distributive justice in luck egalitarian terms. Take a generation whose members enjoy a relatively short expected life span due to climate that is inhospitable to human life. Further, assume that expected life spans have been short for the past several generations. New technological breakthroughs now make it possible to change the climate, such that, within four generations, it will be significantly more hospitable to human life. This can be done at little cost to the existing generation.

706   Rahul Kumar Intuitively, taking available, and not particularly burdensome, measures that will help secure a decent quality of life for future generations is required as a matter of justice. But because luck egalitarians understand distributive justice to be concerned with how the members of different generations comparatively fare, there can, on its terms, be no obligation of distributive justice to do what will make the quality of life available to future generations better than that enjoyed by earlier generations (Meyer and Roser 2009). On this view, it is worse if a certain generation fares worse than another through no fault or choice of its own, even in a case like Depletion in which it is plausible (because of the non-​identity problem) that the inequality is worse for no one. Luck egalitarians are, for the most part, pluralists about justice. Distributive justice may support intergenerational levelling down because that is what fairness demands. But fairness is not the only value internal to justice, even if it exhausts distributive justic­e. Luck egalitarians are not, therefore, committed to the claim that justice requires that resources be destroyed rather than passed on to the next generation, if allowing them to be inherited by future generations will contribute to intergenerational inequality (Page 2006). Fairness may require the prevention of intergenerational inequality. But there could be other justice-​related reasons for permitting resources to be transferred to future generations that will leave them better off than their predecessors. What luck egalitarian thinking about justice appears not to be able to support is the intuitive conviction that justice requires that we balance current consumption with savings so as to secure a decent quality of life for those who will live in the further future, one that may in fact be better than we ourselves enjoy.

32.6  Conclusion: Is What We Owe Justice? Throughout this discussion I  have taken it for granted that obligations to future generations, of the sort relevant to the question of whether to conserve or deplete existing resources, are obligations of justice. In closing, I  want to briefly consider how secure this assumption is. There are at least three reasons for characterizing the obligations governing intergenerational relations as obligations of justice. First, such obligations are thought to underwrite particularly stringent claims of individuals against other individuals and institutional actors, such as (but not limited to) states. Second, intergenerational obligations are, intuitively, obligations owed to those who will live in the further future, as opposed to impersonal obligations to ensure that the future is one in which those who live have lives of a decent quality. Third, questions of intergenerational justice are sometimes conceived of as having to do with the fair allocation or distribution of a stock of resources, one that may or may not have been the product of cooperative activity, a share of which all of humanity (past and future) has a claim to.

Future Generations   707 The first and second reasons are not compelling reasons for understanding intergenerational obligations as obligations of justice. The idea that they are is in part a consequence of the philosophical literature contrasting duties of justice, which have to do with the stringent claims, or rights, of individuals, with humanitarian obligations characterized as imperfect duties that allow agents a degree of latitude to decide what they will do for others not associated with requirements of justice. But the contrast is unmotivated. There is no reason to treat interpersonal moral obligations that individuals owe to one another as in any way less stringent, or less specific in what they do and do not permit, than requirements of justice. The third reason better supports the idea that intergenerational obligations are obligations of justice, by characterizing the relationship that the current generation stands in to future (and past) generations as one that is appropriate to think of as regulated by requirements of distributive justice. Standing in a relationship to future generations of co-​participants in a beneficial cooperative system that relies on intergenerational indirect cooperation, for instance, or taking them to be co-​owners of the earth, suggest a relationship between the living and those who will live in the further future that requires the living to consider the importance of the fair claims of future generations when it comes to policy choices that may affect their interests (Risse 2012). I cannot here examine the grounds for different ways of characterizing the relationship in which the living stand to those who will live in the further future that cast that relationship as one appropriately governed by requirements of justice or, more specifically in some cases, distributive fairness. For present purposes, what is worth noting that is that the force of these proposals lies partly in characterizing the relationship between the living and those who will live as a richer, or more robust, relationship than just the bond of common humanity. This arguably provides a more powerful motivational basis for taking seriously the interests of future generations in current policy deliberations. I believe this way of arguing underestimates the motivational force of the relationship of common humanity. One of the important reasons for thinking that simply appealing to the humanity of those who will live in the further future is an important reason for giving their interests non-​trivial weight in deliberating about current policy decisions is that such an appeal is often characterized as simply an appeal to a locus of interests whose importance in our moral thinking has to compete with the demands of more concrete relationships that we stand in to our contemporaries. Common humanity, that is, is not usually characterized as itself the basis of an important human relationship. As I’ve argued in other work, this is a mistake. Drawing on Scanlon’s contractualist account of what it is for one person to wrong another, I’ve argued that morality requires individuals to regulate their conduct in light of principles that are justifiable to others for whom their conduct might have implications on grounds no one can reasonably reject (Kumar 2009). The contractualist claim, in particular, is that regulating one’s conduct by principles no one can reasonably reject is required to successfully navigate a kind of relationship one stands in to all other human beings—​including those who do not yet, but will, exist—​a relationship characterized by the value of mutual recognition (Scanlon

708   Rahul Kumar 1998, 2008). A valuable type of relationship is one in which certain distinctive values are uniquely realized in it for those involved in a relationship of that kind. To be successfully involved in such a relationship requires individuals to exhibit certain stable attitudes and intentions towards one another. Friendship, for instance, characteristically requires such things as attitudes of affection, concern, enjoyment of one another’s company, intentions to spend time together, and willingness and preparedness to adjust one’s own plans in order to support the other in various ways. These things are required in the sense that friends, insofar as they are friends, legitimately expect each other to have such attitudes and intentions. The moral relationship that binds all individuals to one another is one in which the value of mutual recognition is realized. It is not analogous to a relationship like friendship or that between lovers, insofar as the existence of the relationship does not depend on individuals having certain attitudes and intentions towards one another. Rather, the moral relationship holds between any two individuals in virtue of a vulnerability to one another’s attitudes, one shared by those who are both capable of, and care about, leading rationally self-​governed lives. The same capacity that allows an individual to reflect on the reasons there are for wanting their life to go in certain ways and actively govern it accordingly enables a consciousness of others’ attitudes towards them. What they are particularly vulnerable to is what Philippa Foot calls the possibility of ‘second order evil’: the consciousness, that can be experienced as oppressive, of the disregard or indifference of others to the importance to one of living a rationally self-​governed life (Foot 2002). The indifference or disregard in question is not a result of others doing what results in one’s plans being thwarted or one being harmed in some way. It arises, rather, from an awareness that the fact that your plans might be thwarted, or that the possibility of your being harmed as a result of another’s conduct either did not register in their deliberations as calling out for justification, or was noticed but judged not to matter. Whether my conduct is justifiable to you, given its implications for your life, not even occurring to me as something that needs to be addressed in my deliberations, represents a distinct kind of evil to which rationally self-​governing beings are vulnerable. This mutual vulnerability is the grounds for holding all individuals to stand in a certain kind of relationship to one another, the moral relationship. The idea that all individuals, in virtue of their shared humanity, stand in a valuable relationship to one another, or that it can be usefully deployed to help understand the relation existing individuals stand in to those who will live in the further future, is open to various forceful objections (Scheffler 2010; Gibb 2014). Mounting a proper defence of it and its value for thinking about intergenerational obligations goes beyond what I can pursue here. I’ve briefly touched on it in order to draw attention to a way of thinking about intergenerational obligations that does not characterize them as being obligations of justice, but promises to make sense of the idea that they are particularly stringent interpersonal obligations owed to those who will live in the further future (Kumar 2015). Summing up: I’ve here examined the prospects for extending ways of thinking about justice requirements between contemporaries to the context of relations between

Future Generations   709 generations. I have not examined the prospects for extending all promising accounts of justice between contemporaries. Rather, I have chosen to focus just on those that promise to do justice to the intuitive features of thinking about intergenerational just­ ice and to identify theoretical obstacles that need to be overcome in order to successfully deploy them as part of a better understanding of the grounds of obligations of intergenerational justice. Finally, I’ve briefly argued that the case for understanding intergenerational obligations as obligations of justice rather than interpersonal moral obligations is a matter that merits more discussion than it has received.

References Barry, B. (1977). ‘Justice between Generations’, in P. M. S. Hacker and J. Raz (eds) Law, Morality and Society. Essays in Honor of H.L.A. Hart. Oxford: Clarendon Press, pp. 268–​84. Barry, B. (1989). Theories of Justice. A  Treatise on Social Justice, Vol. I. London: Harvester-​Wheatsheaf. De-​Shalit, A. (1995). Why Posterity Matters: Environmental Policies and Future Generations. London and New York: Routledge. Foot, P. (2002). ‘Rationality and Virtue’, in P. Foot (ed.) Moral Dilemmas: and other Topics in Moral Philosophy. Oxford: Oxford University Press, pp. 159–​75. Gardiner, S. (2011). A Perfect Storm: The Ethical Tragedy of Climate Change. Oxford: Oxford University Press. Gauthier, D. (1986). Morals by Agreement. Oxford: Clarendon Press. Gibb, M. (2014). ‘Relational Contractualism and Future Persons’. Journal of Moral Philosophy 13(2): 135–​60 (online first). Gosseries, A. (2009). ‘Three Models of Intergenerational Reciprocity’, in Gosseries and Meyer (eds) Intergenerational Justice. Oxford: Oxford University Press, pp. 119–​46. Heath, J. (2013). ‘The Structure of Intergenerational Cooperation’. Philosophy and Public Affairs 41(1): 31–​66. Kumar, R. (2009). ‘Wronging Future People: A Contractualist Proposal’, in A. Gosseries and L. H. Meyer (eds) Intergenerational Justice. Oxford: Oxford University Press, pp. 251–​74. Kumar, R. (2015). Risking and Wronging Future Generations. (MS). Meyer, L. H. (2003). ‘Past and Future:  The Case for a Threshold Conception of Harm’, in L. H. Meyer, S. L. Paulson, and T. W. Pogge (eds) Rights, Culture, and the Law: Themes from the Legal and Political Philosophy of Joseph Raz. Oxford: Oxford University Press, pp. 143–​59. Meyer, L. and Roser, D. (2009). ‘Enough for the Future’, in A. Gosseries and L. H. Meyer (eds) Intergenerational Justice. Oxford: Oxford University Press, pp. 219–​47. Mill, J. S. (1969 [1863]). ‘Utilitarianism’, in J. M. Robson (ed.) Collected Works of John Stuart Mill, Vol. X: Essays on Ethics, Religion and Society. Toronto: University of Toronto Press. Page, E. (2006). Climate Change, Justice and Future Generations. Cheltenham: Edward Elgar. Parfit, D. (1984). Reasons and Persons. Oxford: Clarendon Press. Parfit, D. (1986). “Comments.” Ethics 96: 832–​72. Rawls, J. (1971). A Theory of Justice. Oxford:  Oxford University Press; revised 2nd edn, Cambridge, MA: Harvard University Press, 1999. Rawls, J. (1993). Political Liberalism. New York: Columbia University Press. Rawls, J. (2001). Justice as Fairness. Cambridge, MA: Harvard University Press.

710   Rahul Kumar Reiman, J. (2007). ‘Being Fair to Future People: The Non-​Identity Problem in the Original Position’. Philosophy and Public Affairs 35(1):69–​92. Risse, M. (2012). On Global Justice. Princeton, NJ: Princeton University Press. Scanlon, T. M. (1998). What We Owe To Each Other. Cambridge, MA: Harvard University Press. Scanlon, T. M. (2008). Moral Dimensions:  Permissibility, Meaning, and Blame. Cambridge, MA: Harvard University Press. Scheffler, S. (2010). ‘Morality and Reasonable Partiality’, in S. Scheffler (ed.) Equality and Tradition: Questions of Value in Moral and Political Theory. Oxford: Oxford University Press, pp. 41–​75. Scheffler, S. (2013). Death and the Afterlife. Oxford: Oxford University Press. Thompson, J. (2009). Intergenerational Justice: Rights and Responsibilities in an Intergenerational Polity. New York: Routledge. Velleman, D. (2008). ‘Persons in Prospect’. Philosophy and Public Affairs 36(3): 221–​88. Woodward, J. (1986). ‘The Non-​Identity Problem’. Ethics 96: 804–​31.

Index

Abizadeh, A.  287, 294n.17, 643n.5, 658–​9 abstraction  3n.6, 223, 229, 331, 377, 383, 608–​9 accommodation  204, 240, 243, 248–​9, 427, 576, 579–​80, 583, 586–​7, 589–​91 Adler, M.  66n.1, 69, 71, 250 affirmative action  240, 248, 505–​11 agency freedom  204–​5 agent centered theories  8, 310 agent neutral theories  8, 310 aggregation  19, 59, 110, 122–​4, 146, 269, 273, 357, 376, 389, 399, 405n.7, 439, 548, 629 Ahdar, R. 577 Alexander, L.  167, 178, 182–​3, 185, 328, 483 Allen, B.  393, 402 allocative principle of distributive justice  19–​20 Ally, M.  667n.6 Altman, A.  489–​90 altruism  51, 182n.10, 192, 218, 263, 265n.4, 267, 277, 635, 692 Anderson, E.  6, 43, 65, 96, 117, 119–​20, 125, 143, 167–​8, 200n.8, 287, 296, 307, 312–​14, 316nn.11–​12, 318–​20, 433–​4, 447–​8, 467, 471, 494–​5, 507, 509–​11, 600n.2 Anthony, L.  265, 279 Anticorrelated Case  78–​82, 83n.16 anti-​fetishist move  113 anti-​natalism  417–​18 anti-​paternalism  207n.19 anti-​perfectionism  450–​1, 454–​6 anti-​perfectionist  education 450 theory of justice  406 views 451 Appiah, A.  235, 585 Aristotelian  inclinations 268 perfectionism 524 principle  271–​2, 521

Aristotle  86n.2, 109, 152, 238, 241, 259, 346n.1, 350, 500–​2, 514n.1, 519n.6 Arneson, R.  48, 52, 67–​8, 82, 93, 111, 117, 120, 124, 135, 140, 142, 154, 166, 168–​70, 188n.14, 198n.6, 207n.19, 273, 314–​15, 316n.10, 317n.13, 389, 399, 428n.11, 440n.6, 469, 483, 493, 586, 607 Arnold, S.  518n.5, 522–​3 Arrow, K.  461–​2 assurance  289–​90, 556, 561 asymmetry 164, 177 of cultures  594 of desert  161–​2, 164, 170, 177, 186–​7, 189, 191–​3 of justice  690 of power  65 of treatment  422 Atkinson, A.  69n.4, 81n.14 atomism  673–​4, 681 auction  42, 50, 54, 56–​7, 137n.20, 548–​50, 551n.16, 559, 608–​9 authenticity-​recognition  237, 239, 243–​6, 248–​50, 253 autonomous  agents/​persons  147, 158–​9, 214, 220, 456n.18 behavior 394 choices 168n.8, 394 decision making  201 governance  363–​5 individuals 221, 230 self-​government  200 autonomy  100–​1, 124, 130, 159, 199–​203, 205–​7, 224, 242–​3, 297, 299, 301, 303, 356, 363–​5, 394, 421, 423–​4, 427, 430, 447–​51, 455–​7, 522 Avraham, R.  178n.2, 179n.5 Axelrod, R.  336nn.15–​16

712   Index Baatz, C.  676n.14 Baker, J.  101–​2, 400, 409 Bakunin, M.  289n.9 Banerjee, A. 475 Barclay, L. 121 Barnes, G. 144 Barnes, M.  220n.1, 222, 226 Baron-​Cohen, S.  395 Barrett, S.  571, 680 Barry, B.  243–​4, 581, 584, 620, 694–​5 basic liberties  13–​15, 18, 21, 28–​9, 32–​4, 38, 95, 122, 295, 429, 520, 522–​3, 604–​5, 608, see also liberties basic structure  4, 14–​15, 20–​1, 24–​5, 27, 33–​6, 86, 115–​16, 314, 327, 332, 341, 355, 402, 404–​5, 578, 624, 692–​5 Batalova, J.  641n.4 Beckerman, W.  667n.6 Bedau, H.  2, 4 behavioural  differences  266, 395–​7, 402 patterns 264 Beitz, C.  620, 623, 631 Beitzig, L. 263 Bell, A. G.  328 Benatar, D.  392–​3, 398, 418, 474, 475 Benedick, R. E.  680 Beneficiary Pays Principle  667, 683–​4 benefit principle  567–​7 1 Benjamin, J. 396 Benn, S. I.  291–​2, 294 Bergmann, B. 407 Berlin, I. 516 Bertram, C. 547 Bicchieri, C.  289n.11 biological  account of health  463, 470 capabilities 463 constraints 405 differences  395, 419–​20, 499–​500 diversity 563 endowment 440, 444 facts  392, 402–​3 families 419 make-​up  612 parents  418–​19 relation/​kin  416, 418

role in reproduction  391 theories 331 biology  265, 275–​7, 419–​20 Bird, C.  236, 249, 390, 410, 593 Blake, M.  288, 293–​4, 621–​2, 625–​7, 633, 636, 642, 648 Blinder, A.  19n.7, 36 Boehm, C. 264 Bognar, G.  73n.10 borders  8, 105, 145, 148, 293, 294n.17, 619, 640–​60 Botterell, A. 422 Bou-​Habib, P.  424, 456, 586 Bovens, L.  73n.7, 79n.13, 148n.43, 668 Bowden, P.  217–​19, 226 Boxill, B. R.  509 brain drain  655–​6 Brake, E.  400, 405–​6, 415n.1 Braybrooke, D.  99–​103 Brennan, G.  251–​2, 429 Brighouse, H.  116, 277, 333n.10, 398, 408, 416n.2, 419, 421–​3, 432, 441n.7, 444, 446, 448n.14, 449, 452–​6 Brock, G.  98n.6, 99–​102, 229, 474, 636, 642, 656 Bronfenbrenner, M.  544n.11 Broome, J.  69, 70n.5, 80, 82, 669 Brosnan, S. 264 Brown, C.  73n.8 Brown, G.  442n.9 Brown, W. 249 Bubeck, D. 226 Buchanan, A.  2, 467, 628 Buddhism 578 Buffet, W.  471, 631 Butler, J. 249 Cabrera, L.  630–​1 Campbell, T. D.  2, 3n.7, 381–​3 Cancian, F. 263 Caney, S.  622–​3, 630, 667, 669, 677n.15, 681–​2, 683n.19 capabilities  7, 87, 95–​6, 109–​10, 112–​14, 116–​26, 272–​4, 406, 428, 442, 463, 467–​8, 471, 642 capability/​capabilitarian  approach  7, 109–​22, 124–​6 as metric of justice  112, 119

Index   713 theories  110–​12, 113n.2, 114, 117, 119–​21, 123–​6, 467 capitalism  14, 16–​17, 21, 23, 33, 36, 38, 501, 514, 530, 628 capitalist  exploitation  545–​7 free markets  63 societies 17, 530 system 23 welfare state  22–​3, 34, 37 capitalists  38, 144, 216, 544 Card, C.  222, 400 Carens, J.  529n.16, 598n.1, 606–​7, 640, 644–​5, 647–​9 Carnevale, A. P.  518 Carter, I.  113, 121, 376–​7 Casal, P.  91n.4, 93, 98n.6, 147n.40, 261, 263, 395n.2, 408, 425–​6, 443n.12, 448n.14, 564n.6 Castles, S.  644n.6 causation 182, 624 Cavalieri, P. 261 Chakraborty, A. 122 Chambers, C. 394 charity  2, 220, 264, 620 Chavez, A.  289n.11 child  allowances 25 bearing 100 custody 29 childcare  216, 218, 231, 263, 277, 394, 406–​8, 444, 514 childrearing burdens  19, 246 children/​infants  18, 66–​74, 76–​81, 83n.16, 112, 134, 181, 213, 218, 220, 224, 226, 249, 266, 268, 270, 276, 401, 405, 408, 417, 420, 422–​6, 429–​30, 432, 434, 442n.9, 444, 446, 451–​6, 475, 514, 527, 612, 636, 642 Christianity  237–​8, 247, 249, 578, 583, 588, 593 Christiano, T.  1n.2, 285n.3, 287n.5, 294, 298n.19 Christman, J.  200n.8 Chwe, M. S.-​Y.  250 citizenship  218, 236, 239, 246, 448, 450, 520, 524, 526n.13, 528–​30, 630, 641, 643–​4

civic  capabilities 471 entitlements 241 equality/​equals  253, 321 friendship 362 opportunities 583 recognition 252 rights 467 civil  liberties 242, 682 rights  241, 295, 482 society  214–​15, 227n.3, 246, 591, 604, 606 Claassen, R.  114, 121 Clarke, S.  346–​7, 349, 357 class  15–​18, 20, 22, 30–​1, 34, 36, 38, 54, 91–​2, 100, 118, 187n.14, 218, 235, 237, 263, 276, 291, 362, 383, 392–​3, 409, 416, 434, 439–​41, 445–​6, 510, 514–​15, 517–​18, 525, 548, 551, 600, 693, 699, 703 Clement, G.  224, 226 climate change  632, 641, 664–​73, 675–​6, 678–​83, 685, 695, 705 adaptation  274, 502, 665n.2, 666–​8, 670–​1, 673–​7, 679, 683–​4 climate/​climatic  justice 677, 680 policies  665, 669–​70, 673 responsibilities  667, 671–​2, 677–​81, 683, 685 targets 670 treaty 680 Coase, R. H.  335, 557–​8 Cochrane, A. 261 coercion  2, 4, 101–​2, 144, 200, 202, 259, 293, 381, 519, 586, 621, 633, 635 Cohen, G. A.  2n.3, 3n.5, 4–​6, 14, 16, 20, 23–​4, 41, 48, 67, 109, 119, 131, 136–​7, 139, 144, 165–​6, 204–​6, 211n.21, 265, 284–​7, 290, 295, 300, 307, 314–​15, 316n.10, 330n.3, 331, 338, 340, 346–​7, 355n.10, 355n.12, 356n.13, 391, 415, 446, 354n.8, 484, 486, 544–​5, 570, 609, 613, 624 Cohen, J.  354n.9, 629, 650 Coker, J.  70n.5 Cole, M. A.  668 Cole, P.  649, 651n.12 Collier, P.  636, 656 commensurability  103–​4, 192, 201n.10

714   Index communities  2, 43, 47–​8, 53–​4, 57, 60–​2, 100–​1, 155n.2, 163, 215–​16, 243, 248–​9, 251, 298, 300n.21, 332n.7, 333–​4, 341–​2, 369, 373, 378, 381, 416, 425–​6, 440, 468, 495, 513, 516, 518, 526–​7, 562, 568, 570–​1, 577, 586, 601–​3, 605, 607, 609–​11, 615, 620–​1, 624, 631, 634 compensation  19, 43–​4, 46, 48–​9, 52, 55, 57, 59, 61–​2, 75, 130–​2, 136–​7, 148, 154, 168, 185, 188, 192, 204–​6, 210–​11, 218, 240, 248, 260, 268, 398, 402, 420, 439, 441, 474, 507–​10, 513, 527, 567n.8, 612, 614, 667n.5, 674–​5, 689–​91 compensatory  affirmative action  507 justice  248, 507–​8, 674 measures 248 social insurance system  22 compliance  19, 33, 54, 129, 289–​91, 295, 327–​8, 335, 336n.15, 339–​42, 581, 685, 699 compossibility test  379–​80 concept/​conception distinction  367, 382 conceptions  3n.6, 15, 20, 24n.14, 28–​9, 83n.16, 245–​6, 249, 293, 295, 300, 352, 354, 359, 367–​8, 369, 373, 383, 452, 509, 541, 612 of advantage  205 of citizenship  236, 520, 528, 530 of coercion  381 of compensation  508 of contractualism  324 of culture  593 of desert  168, 171, 180 of equal opportunity  439–​42, 444, 454 of equality  13, 45, 47, 316n.10, 434, 484n.8 of exploitation  539, 544, 547, 549–​53 of fairness  47 of health  284 of human evolution  277 of human nature  259 of inequality  140 of justice  2, 7, 21, 31, 39, 45, 102, 117, 168, 219, 227, 238, 267, 287, 290, 300–​3, 307–​8, 312–​14, 316, 325, 327, 338, 341, 357–​8, 363, 368–​72, 378–​9, 382, 395–​7, 416, 431, 438, 445, 447, 519–​20, 524, 528, 530, 599, 604, 617, 646 of justification  318, 347 of law  219, 227

of liberty  130, 140n.27 of luck egalitarianism  315 of man  219 of morality  253, 457 of neutrality  607 of political morality  449, 457 of politics  457 of punishment  311 of religion  450, 456 of responsibility  160 of rights  147n.39, 219, 227 of salvation  247 of self-​ownership  133 of social reform  267 of society  360 of the good  20, 28, 47, 115, 168, 196, 199–​200, 202–​3, 206–​8, 378–​9, 405, 407, 411, 423–​4, 427–​8, 429n.13, 448, 495, 516, 520–​2, 529, 578, 583, 616, 645 of the good life  117, 195, 199, 405, 549, 606, 616 of the ideal  200 of the world  577 of value  548, 564 of welfare  45–​6, 204 or parental authority  453n.16 concepts  3, 89, 113, 120, 153, 156, 169, 355, 367–​8, 373, 382–​4, 439, 550n.15, 552, 584, 628 of choice  165 of compliance  339 of desert  103, 106, 152–​3, 160, 165 of dignity  250 of discrimation  480n.2, 496 of domination  547 of entitlement  153 of equality  103, 106 of esteem  251 of exploitation  543, 547 of freedom  381 of gender  392 of intention  547 of justice  2–​3, 5–​6, 8, 237–​8, 291, 367–​70, 372–​84 of law  371n.3, 381 of liberal individual  219 of life plans  476 of love  251

Index   715 of morals  169 of moral tragedy  636 of needs  88, 103, 106 of ownership  130 of personal identity  236 of political morality  381 of power  547 of race  499 of respect  251 of self-​ownership  132 of sexual harassment  397 of vulnerability  540 conceptual  analysis  8, 367, 369n.1, 370–​2, 381–​3 differences 117 explication 375, 377 explorations 381 independence 287 unity 384 Connolly, W. E.  249 consequentialism  8, 14, 20, 31, 59, 69n.4, 130, 158, 192, 221, 273, 306–​13, 325, 451–​3, 457 nonconsequentialism  308–​11 consequentialist  nonconsequentialists  8, 308, 457 punishment 192 views concerning obligations  691n.2 conservatism 259, 276 conservative  222, 264–​5 arguments 278 belief 266 conception of justice  267 doctrines 260 emphasis on human nature  268 implications 260 views  267, 276–​8 conservatives  23, 43, 259, 264–​7, 276–​9 constitutional  democracies 359 essentials 30, 355 rights  581–​2 constructivism  8, 346–​8, 351–​3, 369n.1, 502 constructivist  approaches 348, 370 forms of argument  346, 353 methodology 370

methods 346 reasoning 349, 354 contractarian  account of justice  691 approach 698, 700 contractarianism 434, 698 contractualism  306, 310n.4, 319–​25, 692 contractualist 33, 69n.4 account of justice  691 account of wrongness  707 analysis of wrongness  319 doctrine 273 ideal 323 justice 307 perspective 325 test  310, 319–​20, 321, 323–​4 cooperation  26, 101, 214, 218, 232, 238, 261, 263, 267, 274, 277, 287, 289, 294, 332, 333, 336, 338–​9, 360, 609, 624, 635, 648, 692, 694–​6, 698–​701, 707, see also economic cooperation, intergenerational cooperation, social cooperation Copp, D.  99–​101 cosmopolitan  care 229 principles 289 right and Kant  620 solidarity 266 cosmopolitanism  87, 213, 289, 485n.11, 621–​4 criminal  acts 192 behaviour 266, 486 justice 3n.8 law  42, 53, 56, 162–​3, 239 punishment 186 Crisp, R.  73n.7, 97, 272–​3, 468 Crothers, L. 643 Cudd, A. 409 Cudworth, R.  349n.5 Cullity, G.  556n.1 culpability  167, 180, 182, 190n.17, 627, 629 cultural  allegiances  576, 579, 590 animals 269 artefacts 557 assimilation 243

716   Index cultural (cont.) commitments  576, 583–​5, 589 communities 243 differences  243–​4, 576–​9 disadvantages 206 diversity  576–​7, 658 environment  196, 206–​8 explanation 274 freedom  578, 581–​2, 589 groups  236, 243, 577, 579, 583, 594 heritage and inheritances  240, 557 identities  243, 582, 585 integrity 594 liberty  578–​9 life 18n.6, 587 minorities 8, 576 norms 586 opportunities  13, 16–​18, 28–​9, 32, 34, 590, 600, 615 practices  423, 576, 578–​9, 581, 586 rights and claims  585, 594 similarities 634 societies 594 traditions 423 transmission 261 values 34 culture  16–​18, 179n.4, 210, 218, 222, 242, 248, 251, 264–​6, 270, 300, 358–​9, 361, 363, 499, 519, 521, 577–​82, 584, 586–​90, 592–​4, 602, 606, 623, 699 Czaika, M.  641n.3 Daniels, N.  260n.1, 462–​8, 470, 473, 476 Darity, W. 505 Darwall, S. 342 Davis, K. 265 Deci, E. L.  528 de Haas, H.  641n.3 de Jasay, A.  556, 561 deliberative  body 632 freedoms  479, 487–​9 functions 633 obligations 631 processes 120 democracy  13–​14, 17, 20–​4, 26–​7, 32–​8, 96, 113, 169–​70, 223, 226, 250, 298, 524, 530, 598, 632, 634, 658

democratic  citizens  27, 249, 434, 447, 520–​1, 524 citizenship  448, 520, 524, 526n.13, 528–​30 culture 519 decisions 27, 450 deliberation  120, 560, 599 egalitarianism  307–​8, 312–​23, 325, 467, 471 equality  120, 319, 321, 447, 467, 471, 495 governance 239 institutions 294, 633 interpretation  17–​18 legitimacy 633 majorities 573 procedures/​processes  298, 433, 450, 562 reciprocity  31, 33, 37–​8 virtues 449 welfare 27 welfare state  21–​2 deontic  distributive justice  479, 485, 487, 496 egalitarianism 485, 487 prioritarianism 273 unfairness 705 deontological  constraints 182n.10 grounds 221 perspective 325 theories  8, 307–​9 deontology  209, 306–​8, 310–​12, 325 Depletion  689–​90, 692, 695–​6, 700–​6 De Schutter, H.  598n.1, 611, 613 desert  2–​3, 7, 42, 44, 51–​2, 87, 103, 105–​6, 152–​71, 177–​82, 184–​93, 238, 269, 284n.1, 307, 383, 492, 642 and crime  91, 153–​4, 156–​7, 160, 164, 167, 170, 178, 179n.6, 180n.7, 182n.10, 185, 192, 239–​40, 245, 266, 275–​6, 375, 495, 566 desert-​based  approaches 87 distributive justice  186, 188, 193 entitlements  185–​6 goodness 157 justice 152 patterned theories  186, 189 punishment  185–​6 de-​Shalit, A.  114, 560, 699 determinism 49, 182n.10 De Waal, F.  261, 264, 275, 277

Index   717 De Wispelaere, J.  422 Diamond, J. 277 Diamond, M. 266 difference principle  13–​14, 16–​39, 115, 228–​9, 301, 355nn.11–​12, 378, 406, 461, 466, 481–​2, 498, 522n.7, 523–​4, 559, 564, 566, 578, 631, 672 disabilities  29, 38, 42, 60, 62, 66, 68, 71–​2, 73n.9, 78, 115, 117, 204, 409, 417, 421, 473 discretionary public goods  562, 566–​9, 571–​3 discrimination  9, 142–​3, 211n.21, 239–​40, 245, 260, 263, 266, 277, 291–​2, 320, 371, 391–​4, 402–​3, 407, 410–​11, 431, 440, 449, 462, 466, 470, 473, 479–​96, 507, 510, 513, 516, 576, 581–​2, 585, 589–​92, 601, 604, 659n.15 distributive  equality  45–​6, 350–​1, 428, 432n.16, 621 paradigm 5, 244 rules  111, 119, 122–​3, 125 Dodd–​Frank Wall Street Reform Act  628–​9 Doherty, B. 643 Doppelt, G.  515, 521 Doyal, L. 100 Doyle, O. 493 Dreier, J.  309n.3, 310 DuBois, W. E. B.  505, 510 Duff, A. 180 Duus-​Otterström, G.  677n.16, 683n.19, 684 Düvell, F.  641n.4 Dworkinian  approach 610 argument for linguistic evenhandedness  609 distributive justice  487n.14, 549, 609 envy test  123 equal status-​recognition  242 principles 57 resources 118n.3, 122 response 486 Dworkin, R.  2n.4, 3n.6, 7, 22, 35, 41–​63, 113, 118–​19, 122–​5, 137n.20, 167, 198, 206n.16, 242, 293–​4, 360, 373–​6, 379, 383–​4, 415, 439n.3, 466, 476, 484–​7, 559–​60, 562, 565n.7, 566, 573, 608–​10, 675, 678, 682 Easterly, W. 636 economic  advantages/​benefits  2, 129, 142, 145–​6, 149, 162–​4, 396, 641

burdens  129, 142, 149, 641 cooperation  29, 30, 659, 693 distribution  14, 16, 19, 22, 632 efficiency 34, 598 equality 89, 91 goods  4, 241, 449 inequalities  16–​17, 27–​8, 35, 95, 417, 431, 456, 559, 617 institutions  1, 19, 421, 448, 530 justice  15, 19–​21, 87, 650 opportunities  238, 242, 576, 612, 615 powers  16, 22–​3, 28–​30, 34, 36–​8 production  19, 21, 30, 447 prosperity 142, 145 reciprocity  30–​1, 37 relationships  530, 628, 637, 650–​1 systems  14–​15, 17, 19–​25, 27, 33–​6, 171, 320 economics  23, 69n.5, 81n.14, 110, 122, 125, 220, 548 economists  22, 144n.33, 145, 251, 505, 533, 544, 553, 556–​7, 559, 567n.8, 598, 609 economy  21–​2 , 35, 53–​4 , 63, 215, 251, 288, 335, 339, 351n.6, 426, 439, 443, 445–​6 , 524, 529–​3 0, 546, 551, 600, 604 ecumenism  346–​9, 353–​6, 361–​3 education  8, 17–​1 8, 22, 28, 66, 74, 91, 96, 100, 102, 114, 118, 121–​3 , 215, 218–​19, 231, 259, 266, 276, 278, 291, 316, 393, 421–​2 , 425, 427, 429–​32, 434, 438–​9 , 441–​5 7, 469, 494, 504, 507, 509–​1 0, 580, 582–​3 , 600–​1 , 607, 613, 653, 656, 678–​9 and sufficiency  433, 447, 469 educational  achievements  439–​41, 444 adequacy  447, 450, 452 advantages/​benefits/​goods  18, 440, 442, 444, 446–​7, 453–​4, 517 authority 451 background 517 choices 452 circumstances 423 consequentialism  451–​2 curricula 240, 579 development 17 disadvantages 441 fairness 444

718   Index educational (cont.) inequalities 439, 446 institutions  439, 447–​50 investments 262 justice  433, 438, 440, 444–​5, 447–​8, 452, 457 malleability 265 needs 421 opportunities  8, 13, 16–​17, 34, 321, 429, 438–​40, 444–​5, 454 outcome 452 performance 389 resources  429, 439–​0, 444 success  438–​9 egalitarian  accounts and approaches  44, 90, 322, 390, 430–​1, 447, 485, 535–​6, 541, 670, 692 claims  383, 542, 705 conceptions  24n.14, 236, 290, 308, 396, 416, 441, 454 distributions  28, 275, 285, 290, 313, 430 division of labor  403 ethos  23n.12, 285, 484 fairness 82 health 470 healthcare 462 justice  43, 87, 124, 307, 312–​14, 321–​2, 325, 428, 433, 443n.11, 469–​70, 472, 487, 494–​5, 530, 649 principles  22, 307, 313, 353, 358–​70, 675, 677 relationships  263, 317, 390 social perfectionism  519, 528 social relations  494–​5 society  13, 28, 290, 445, 560, 576 status-​recognition  241–​2 theories and views  7, 41, 51–​2, 59, 65, 68, 70n.5, 80–​1, 124, 186–​7, 301, 397, 407, 433, 445, 488, 492–​3, 536, 551, 573, 611, 615, 692, 675–​6, 705–​6 values 140, 330n.3 egalitarianism 65, 68, 81 democratic  307–​8, 312–​23, 325, 467, 471 deontic 485, 487 final-​well-​being  81–​2 gender  391, 407, 409, 411 in health  472 pluralist  67–​8, 70n.5, 81 political 65

Rawlsian 519 relational  96–​7, 307, 316n.12, 433, 491 social relations  494–​6 telic 485 time-​slice  475 see also luck egalitarianism egoism 218, 310 Einstein, A. 632 Eisgruber, C.  583, 585, 587n.4, 588 Elkin, S. L.  330 Elster, J.  536–​7, 545, 547, 548n.12, 553 emigration  651n.12, 655–​8, see also immigration, migration, movement across borders emotions  121, 215, 218, 221, 228, 261n.2, 398, 503 empathy  216–​17, 221, 224, 228, 259, 262–​3, 267, 277, 395 employment  8, 23, 52–​3, 91, 225, 276, 398–​9, 406, 440, 442, 445–​6, 454, 462, 513, 544, 581–​3, 590–​2, 658, see also unemployment Endogenous Instrumentalist View  287 Enlightenment 504 entitlements  31, 94, 106, 121, 126, 153, 184–​6, 192, 203, 239–​42, 244, 248, 251, 264, 349, 351, 406, 408, 421, 424–​5, 427, 430–​4, 549, 649, 651–​5, 657, 666–​7, 671–​3, 678, 684 environment  29, 32, 34, 100, 103, 121, 187, 196, 205–​8, 232, 250, 275, 298n.18, 337, 396, 440–​1, 444, 502, 509, 564, 603 environmental  adaptation 274 advantages 208 amenities 566 conservation 147 disadvantages  205–​8 goods  7, 196, 557, 560, 564n.6 philosophers 563 problems 526, 680 protection 147, 266 regulations 558 resources 702 triggers 279 environmentalists  560, 563–​4 envy  66, 272, 274, 692 envy test  44, 46, 123, 608 Epicurus 86n.2

Index   719 equal  basic liberties  13, 15, 18, 21, 28, 33, 38 benefit principle  569–​7 1 claims  138, 141, 203, 207 distribution  13, 24n.14, 41–​2, 89, 95, 104, 139n.25, 209, 274, 276, 351, 484, 486, 588, 608 liberties  14, 27, 122–​3, 521, 583, 693 opportunities,  see equality of opportunity, fair equality of opportunity, fair equality of opportunity principle, meritocratic equality of opportunity treatment  13, 580, 601 equality  3, 13–​14, 21, 26–​7, 38, 41–​2, 45–​6, 49, 56, 58–​9, 61, 65–​70, 81–​2, 88–​91, 93–​8, 103–​6, 109, 111, 115, 120, 123, 125, 165–​7, 186, 219, 239, 241–​2, 249, 253, 260–​2, 266, 268, 272, 274–​5, 288, 290, 293, 295, 298n.19, 306–​7, 312–​13, 315, 316n.10, 317, 320, 325, 333n.10, 340, 390, 397, 399, 402, 428, 431–​4, 439, 450, 455, 463, 476, 484n.8, 485n.10, 486n.12, 508, 518, 526, 528, 538, 569, 582, 584, 588, 592–​3, 621, 623–​5, 630–​1, 634, 636, 642, 670, 685 in health  472–​3, 475 of capabilities  118 of condition  60, 63 of economic powers  29 of income  29 of labour  544–​5 of liberties  38 of life prospects  91–​2 of moral standing  494 of opportunity (EOP)  28, 33, 47–​50, 140, 142n.31, 395, 399, 405, 419–​20, 439–​42, 443n.10, 444–​5, 447, 464–​8, 470, 482, 498, 515–​16, 583–​5, 590, 599, 636, 647–​9 of resources  22, 42–​3, 45, 48, 50, 53–​6, 92, 486, 559, 608, 675 of shares  55, 349, 351 of wealth  29 of welfare  28, 45, 47–​9, 118n.3, 211n.21, 473 see also democratic equality, distributive equality, economic equality Equality Act  592n.9 esteem  92, 181, 246, 251–​2, 513, 516, 521, 525–​7, 529 Estlund, D.  286, 287nn.5–​6, 329n.2, 337, 338n.20, 340

ethics  8, 110, 307–​8, 312–​14, 325, 359 of authenticity-​recognition  249 of care  8, 213–​17, 219–​31, 400 of justice  216–​17 of migration  640, 642 of war  668 ethnic  conflict 221 groups 237, 606 ethnicity  235, 403, 409, 411, 582n.2, 600 European Convention on Human Rights (ECHR)  577, 581, 588–​9 European  ancestry 472 countries 16, 582 jurisdictions  577, 581–​2 European Union  634, 643 evenhandedness  603, 607–​11 evolution  245, 260, 264, 277 evolutionary  data 260, 276 debunking explanations  260, 272 findings  265–​8, 276, 278–​9 origin 266 science  260, 266, 276, 278 scientists 259 ex ante vs. ex post distribution  45, 50–​3, 60, 63, 79, 291 exemptions  240, 580–​9, 591, 592n.9 Exogenous Instrumental View  283–​4, 286 expensive tastes  48–​9, 119, 424, 425n.9, 563, 569–​7 1, 614–​15 exploitation  8, 222, 277, 289, 390, 400, 475, 533–​53 externalities  22, 53, 56–​7, 558–​9, 564, 609–​10 Eyal, N.  51, 165–​6, 167, 474 Ezorsky, G.  185n.12, 515 Faden, R. 467 fair background  15, 312, 397, 584, 604, 606–​8, 611, 615 fair division  349, 353–​4, 356 of assets/​goods  29, 215 of labour  398 fair equality of opportunity (FEO)  13–​19, 22, 33–​4, 38, 301, 405–​6, 432n.17, 462, 464–​6, 468, 470, 498, 506, 523, 583, 585, 604, 693

720   Index fairness  13, 20, 25, 30–​1, 33, 47, 51, 53, 55, 62, 82, 86, 97, 110, 115, 117, 123, 140, 163, 167, 190, 208, 225, 251, 266, 274, 292–​3, 295, 333, 342, 349, 351, 353, 369, 391, 426, 434, 444, 466, 524–​5, 534–​8, 540, 543–​5, 551, 553, 561, 607, 609, 616, 637, 692, 705–​7, see also unfairness familial  goods 432 intimacy 432 relationship goods  452–​3 values 433 family  8, 14, 66, 213–​14, 216–​17, 220, 223–​4, 226–​7, 263, 265, 270, 278, 291, 300, 391, 393–​4, 402, 405–​7, 416, 419, 423–​5, 430, 440, 453, 467, 476, 481, 513, 517–​18, 528, 585, 591, 602, 612, 636, 649, 655 allowances  18–​19, 25 background  433–​4, 440 circumstances 18 identity 611 law  29–​30 life  246, 251, 403–​4, 408, 417–​18, 432, 643, 692 reunification 660 values 276 federal  forms 632 law 587 structure  632–​3 system 640n.1 federalism 631, 633 Feinberg, J.  152–​4, 161, 451, 588 Ferguson, B.  537, 542n.9, 552–​3 Ferzan, K. K.  167, 178, 182 feudal  exploitation  545–​6 privilege 644 society 475 Fickling, D. 643 Fine, C. 395 Fine, S.  641, 659 Finnis, J. 278 Firestone, S. 402 Fishkin, J.  441, 636 fittingness 153, 155 Flanders, C. 180 Fleischaker, S.  1n.1

Fleurbaey, F.  43, 45, 47, 51 Fleurbaey, M.  74, 79, 81–​2, 114, 168 fortune  41–​3, 53, 80, 184, 187, 190, see also luck, misfortune Frailberg, S. 419 Frankfurt, H.  87, 89–​93, 96, 99, 102–​3, 467 Fraser, N.  235, 241, 245, 390, 407 freedom  7, 20, 87, 94, 112, 116, 119, 121, 125, 131, 134, 142, 144, 184, 204–​5, 238, 243–​4, 265, 271, 306, 360, 376–​7, 381, 394, 405, 408, 450, 452, 479, 487–​9, 495, 528, 553, 578, 581, 587–​8, 592 from discrimination  604 from harassment  100 from interference  130, 702 from self-​monitoring  455 from subordination  523 of choice  181n.9, 391, 517 of conscience  13, 23n.12, 239, 449–​50 of contract  21 of exit  657 of expression  245, 646 of movement  520, 640–​69 of occupation  285, 393, 515–​17, 521 of political participation  95 of religion  576, 578, 582, 587–​90 of speech  605 of the political community  621 of worship  251 to achieve well-​being  122 to emigrate  657 to pursue rational aims  204n.13 see also unfreedoms Freeman, S.  20n.8, 23n.12, 37–​8, 143, 358, 523nn.9–​11, 524, 621 Frick, J.  79n.12 Fricker, M. 397 Fried, B.  132–​3, 145 Friedman, M.  22, 369 Fried, M. H.  514 future generations  8, 34, 38, 105, 317–​18, 322, 520, 669, 677, 689–​99, 701, 705–​7, see also intergenerational justice Galeotti, A. E.  585 Gallie, W. B.  383–​4 Galston, W. A.  330, 452

Index   721 Gardiner, S. 689 Gardner, J. 489 Gaus, G.  251, 294n.17, 298n.20, 301–​2, 557 Gauthier, D.  75, 415, 691, 698 gender  8, 235, 237, 389, 391, 396, 402–​3, 406, 411, 419, 484–​5, 487–​9, 516, 518, 581–​2, 592, 693 behaviour  395–​6, 400–​1 bias 217, 219 differences/​disparities  217, 266, 390, 396n.3, 397 discrimination  265–​6, 481 division of labour  399, 403–​9, 411, 419 egalitarianism  391, 407, 409, 411 equality 268, 295 fairness 391 hierarchy  57, 219, 223, 405 inequality 391, 419 injustice  389–​2, 401, 404, 406, 409 justice  219, 390, 394–​5, 399, 401, 403–​4, 406–​11 lifestyles  399, 405, 408–​9 norms  391–​403, 405–​6, 408–​10 roles 407, 411 schemas 396n.3, 398 segregation 393 socialization 394 stereotypes 409 unfairness 394 genetic  advantage and disadvantage  441–​2 basis 266, 274 differences 441 endowments/​resources  47, 440, 443, 702 factors 260 inequality 441 inheritances 440 lottery  142, 333n.8, 440–​3 luck 441 manipulations 267 parent 418 predispositions 53 tendencies 279 testing 263 Geras, N. 269 Gerson, K.  403, 411n.11 Gettler, L. T.  268 Gewirth, A.  100–​2

Gheaus, A.  394, 399–​400, 401n.5, 408, 419n.5, 429 Ghiselin, M. 277 Gibb, M. 708 Gibney, M. J.  640 Gilligan, C.  217, 223, 400 Gintis, H. 264 glad-​to-​have test  49 global  agents  622–​3 capital 629 capitalism 628 community 620, 624 difference principle  672 disparities 474 distribution of resources  624 distributive justice  105–​6, 619–​26, 628–​31, 635–​7 duties of justice  622 duty to care  229 economic activity  228 economic order  621 economic relationships  637 economy 351n.6 equality  623–​4, 631, 634, 648–​9 federalism 631, 633 freedom of movement  649 fund 475 health  474–​5 inequality  474, 620, 634, 636 institutional norms  634 institutional relationships  624, 626 institutional sets/​system  625, 629–​30, 633–​4 institutions  627–​8, 635 interactions  229–​30 justice  219, 228, 474, 621, 625, 627–​8, 630–​2, 634, 637, 644, 650, 669–​7 1, 680, 685–​6 order 621, 624 patent regime/​system  474–​5 political institutions  632 political justice  619 political society  619, 631–​2 poor/​poverty  229–​30, 620, 624, 637, 644, 682 population 641 principles 619

722   Index global (cont.) public goods  570–​1 rectification 637 redistribution 145 responsibility 213, 229 society  215, 223, 629, 633 wealth/​wealthy  620, 644 globalization 228 Godfrey-​Smith, P.  331 Goldin, I.  641, 643, 657 Gomberg, P.  513, 516, 518–​19, 524n.12, 525–​8, 529n.16, 530 good society  7, 195–​6, 199–​202, 205, 208–​9, 211, 520, see also just society Goodall, J.  261, 277 Goodin, R.  93, 101–​2, 536–​45, 552 goodness  69n.4, 156–​7, 170, 183, 195–​9, 201–​2, 209n.20, 250, 308–​9, 430 Gornick, J. C.  407 Gosseries, A.  680n.17, 681n.18, 682, 683n.19, 695n.3, 700 Gough, I. 100 Gould, C. C.  219 Gray, J. 259 Green, D.  95, 97, 598n.1 Greenawalt, K.  577, 593 Greenhouse Development Rights  668 Grisez, G. 278 Grunebaum, J.  130, 136 Gutmann, A. 235 Hakim, C.  389, 407 Hall, B.  419n.5 Hall, E. 330 Hamilton, M. A.  587n.4 Hannan, S. 422 happiness  44, 113, 118–​19, 122, 156–​7, 269–​70, 308–​9, 311, 518 harm  2, 41–​2, 50, 52, 56–​7, 99, 113, 148, 182, 188, 192, 206, 271, 275, 311, 313, 320, 394, 400, 406, 409, 418, 481, 483, 496, 507–​9, 513, 515–​21, 523, 525, 529, 533–​4, 543, 558, 576, 578–​9, 645, 665–​8, 670, 674–​7, 681–​2, 691, 698, 701, 708 Hart, H. L. A.  2nn.3–​4, 368, 370n.3 Haslanger, S.  391, 419n.5 Hauser, M.  264, 274 Hausman, D. M.  472

Hayek, F. A.  22, 289, 291–​2, 369 health  8, 28, 60, 66, 71, 91, 100, 115, 121, 123, 170, 215, 218, 266, 275–​6, 284–​5, 398, 460–​76, 513–​14, 564, 565n.7, 566, 588, 664, 667, 671–​2, 678, 700 healthcare  22–​3, 28, 38, 91, 100, 102, 215, 218, 231, 276, 398, 421, 425, 427, 429–​31, 439, 461–​70, 476, 513–​14, 516–​18, 565, 579, 590, 653, 679 Heath, J. 700 Hegel, G. W. F.  86n.2, 236, 246 Held, D. 630 Held, V.  217–​19, 223–​4, 226, 229, 231, 400 Henrich, J.  263–​4, 274 heterosexuals  278, 403–​4, 417, 419, see also homosexuals Heyward, C.  669n.9 Hilton, C. E.  278 Hinton, T.  316, 320n.18 Hirschman, A. 264 historical process view  15 Hobbesian  conceptions 219 contractarianism 434, 698 public reason  300–​1 self-​effacing authorization  299 social contract reasoning  701 society 179 Hobbes, T.  219, 227, 231, 259, 296–​300, 346n.1 Hochschild, A. R.  389 Hohfeld, W.  130n.5 Hohfeldian  claim-​rights  138n.24 liberties 130 rights 132 holism  673–​4, 677 holistic  approach 674, 681 distributive justice  161, 187 Hollis, A. 475 Holroyd, J.  394, 411 homosexuals  278–​9, see also heterosexuals Honneth, A.  235, 246–​7, 251–​2, 390 Hooker, B. 158 Hope, S. 332 Horta, O. 269 Hrdy, S. 263 Hsieh, N. 524

Index   723 Huemer, M.  294n.17 human  activity  134, 138, 216–​17, 514 agency  100–​1, 288 altruism 635 behaviour 266 capital  439, 444–​6 communities 516, 518 condition  329, 331–​6, 338–​9, 342–​3, 351, 359–​60, 365 dependency 526 desires  197–​9 dignity 124, 143 ethic 223 evolution 277 evolutionary science  266 expansion 277 experience  88, 94, 339, 634 flourishing  100, 423, 447, 454 functioning  98, 100–​1 good  47, 195–​204, 206, 209, 211, 272, 528, 625 history 263, 628 identity  241–​2 integrity 586 interdependence 229 knowledge 699 nature  8, 86, 259–​60, 265–​9, 272, 276–​9, 330, 443n.12, 524 needs  86, 88, 95–​6, 98–​101 procreation 418 psychology 340 race 697 reason  347, 353, 356, 359, 635 relationships  170, 217, 401, 707 rights  87, 95, 106, 125, 229, 272, 278, 581–​2, 588–​9, 628, 647n.8, 652nn.13–​14 sensibility 519 social cooperation  525 success 277 suffering 135 survival 395 sympathy 505 thought 519 vulnerability 124 welfare  47, 59, 132, 134, 292n.14 well-​being  91, 444 humanitarian  assistance 228

concerns 2 considerations 105 duties  649, 650n.11 grounds 330 obligations 707 principle 3 values 330n.3 humanitarianism 70n.5, 330 Hume, D.  3, 30, 221, 228, 332–​3, 346n.1, 483n.6, 503 Hurka, T.  156, 161, 177n.1 Hurley, S.  49, 168, 169n.9 Husak, D. N.  161, 178 Hyams, K.  82n.15, 674n.13 hypothetical insurance  42–​3, 53–​5, 60–​3, 566, 571–​3 ideal  circumstances  20, 27, 33, 351 conception 327 conditions  14, 27, 32, 35, 198, 370 justice 303 libertarianism 21 market 50, 63 prices 340 principles 221 process 19, 328n.1 responses  329–​31, 335, 338–​42 society  20, 32–​3, 285, 328n.1, 335, 339 theory  33, 55, 116, 142, 327–​8, 331, 337, 339–​41, 432n.16, 498, 635–​6 worker 407 world  98, 167, 328n.1, 329–​30, 360 identity  3, 5–​7, 131, 182n.10, 236–​7, 241–​5, 247–​51, 313, 484, 495, 570–​1, 582, 600, 602–​3, 610, 616, 633, 658, 696–​7, 703, 705 census-​identities  237, 244, 249 claims 242, 611 development 249 dimension of language  611 expression  248–​9 formation  245–​6, 249 groups  235, 240, 243–​4, 249 of children  423 of self-​definition  236 politics  236, 244, 246, 250, 601 recognition 585 ideology  499–​502, 504–​5, 508

724   Index immigration  294n.17, 577, 583, 602, 641–​4, 647n.8, 651n.12, 652nn.13–​14, 656, 658–​9, see also emigration, migration, movement across borders impersonal  goods  195–​6, 199, 205, 209–​11 obligations 706 organizations 591 value 7, 73 income  4, 14–​17, 19–​23, 25, 28–​31, 34–​8, 61, 69n.5, 89, 91, 93, 113, 122–​3, 162, 170, 177n.1, 244, 253, 285, 288–​9, 292, 306, 320, 329–​30, 333, 377, 389–​90, 406n.9, 424, 428–​9, 438, 439n.1, 445, 449, 461, 466, 468, 470–​1, 518, 520, 523, 547, 563, 572, 641, 656 infants,  see children/​infants inheritance  15, 17, 54–​6, 79, 240, 274, 322, 406, 433, 440, 499, 502, 504, 507, 644, 669, 696, 699, 706 insurance  51, 55–​6, 167, 215, 542, 560 and equal auction  50 approach/​model  54–​5, 60, 62–​3, 571 decisions 54, 62 fair  53–​6, 60–​3 markets  42–​3, 53–​4, 60–​3, 566, 571–​3 system 22 integrationism  671–​3, 678, 680, 682–​5 intellectual  capacities/​means  278, 450, 516 development  516–​17, 524 differences 500 inferiority 504 integrity 464 intelligence  261, 332, 499–​500, 518n.5, 520, 522 interfamilial  inequality 431 justice 430 intergenerational  chain of concern  694 character of climate change  669 cooperation  695–​6, 700–​1, 707 fairness 705 inequality 706 justice  136, 147, 669–​71, 680, 685–​6, 689–​92, 695–​6, 698, 701–​2, 705–​6, 709 obligations  706–​9

projects 699 relations 706 transfers 55 Intergovernmental Panel on Climate Change 664n.1 Internal Displacement Monitoring Centre 641n.2 international  freedom of movement  640, 643 inequality 620 institutions 627 justice  136, 147, 619, 621, 623, 629 law  231, 359, 624 migrants/​migration  641, 643 morality 620 negotiations 667n.5, 685 order 484n.7 organizations 628 poverty 619 production 620 realm  619, 621, 624 relations  213, 216, 226 trade 294 wealth 619 Intra-​versus Interpersonal Case  74–​5, 76n.11, 82 intuitionism  8, 346–​9, 353–​7, 363–​4, 347–​9, 354, 357, 361, 363 intuitive method  357 Islam  578–​9, 586 isolationism  671–​3, 678–​83 Jackson, B.  1n.1 Jagers, S. C.  677n.16 Jaggar, A. M.  219 James, A.  287, 290, 295–​6, 340, 347nn.3–​4, 351n.6, 353n.7, 359, 627 Jaworska, A. 271 Jefferson, T.  500–​5 Jensen, K. 275 Johansen, R.  73n.10 Jones, C.  101–​2 Jones, P.  240, 585–​6, 591, 593 Jouvenel, B. de  302 Joyce, R. 260 Jubb, R.  334n.13 Judaism 579

Index   725 jurisdictions  43, 577, 581–​2, 631 Just Burden Question  664, 666–​7, 669, 671–​3, 681, 685 Just Savings Principle  38 Just Target Question  664, 666, 668–​71, 673, 685 Justification  238–​9, 241, 243–​5, 253, 349–​50, 361–​2 Kagan, S.  44, 51, 170, 190n.17 Kani, I.  356–​7, 363–​5, 501–​2, 504, 620, 631 Kant, I.  231, 251–​2, 288–​9, 297, 347, 501 Kantian  argument from autonomy  364 constructivism  346, 355–​6, 360, 362–​3 morality 214, 227 social contract reasoning  701 tradition 379 utilitarianism  214, 217, 227 Kantians 224, 253 Kelly, E.  116, 160 Kenny, M. 634 Kipnis, K. 266 Kittay, F.  217, 219, 226, 400, 415n.1, 525–​6 Kleingeld, P. 620 Klosko, G.  526–​7 Knight, C.  50, 188n.14, 189 Kohn, M. L.  515, 516n.2, 521–​2 Kolber, A.  185n.12 Kolm–​Atkinson social welfare function  69n.5 Kolm, S.-​C.  69n.4, 81n.14 Koppelman, A.  585–​6 Kornhauser, A.  521–​2 Kraut, R.  526n.14 Kukathas, C.  235, 243 Kumar, R.  105, 677, 707–​8 Kuper, A. 630 Kymlicka, W.  6, 139, 224–​5, 227n.3, 235, 242–​3, 416, 547, 577, 583, 594, 598n.1, 599, 601, 606–​7, 611, 613, 617n.7 labor  19, 22, 38, 138n.22, 140–​1, 145, 216–​20, 263, 265, 285, 396, 398–​9, 403–​5, 411, 419–​21, 427, 464–​5, 514–​15, 518–​21, 523, 525–​9, 534, 538–​9, 544, 684 market  275, 277, 389, 391, 393, 406–​9 relations 21 theory of value  544–​5, 548

laborers  36, 38, 534, 546 laboring classes  22 LaFollette, H. 421 Lane, R. E.  515, 525 language  8, 114, 240, 251, 384, 392, 401, 446n.13, 577–​8, 580, 597–​8, 600–​4, 606–​8, 611–​12, 614–​16 communities  601, 603, 605, 615 groups  600–​1 instruction 600, 602 minorities  600–​2, 615, 617 policies  598–​9, 601–​2, 609–​10, 613, 615 protection 613 rights  580, 598, 615, 617, 625 see also linguistic Larmore, C.  287, 295, 332n.5 law/​laws  5, 14–​15, 19–​21, 24n.14, 29–​31, 42, 53, 56, 162–​3, 214–​15, 219–​20, 223, 227, 229–​32, 239, 242, 245, 273, 276, 293, 298, 328, 339, 359, 370n.3, 381, 392–​3, 406, 411, 417, 450, 487, 516, 518, 558–​9, 562, 564, 566, 576–​7, 580–​5, 587–​92, 594, 624, 628, 643, 692, see also legal Law of Peoples  38 Leacock, E. 525 least advantaged group (LAG)  13, 16, 20–​38, 187, 228, 322, 441n.7, 444–​7, 624, 643, 657, 668, 682 Lecky, W. E. H.  505 Lee, R.  514, 525 legal  action 249 activity 215 barriers/​constraints  144n.33, 215 entitlement 406 exemptions  582, 589, 591 implications 214 institutions  4, 14, 215, 230, 369, 448–​9, 621 interference  553–​4 liberty  56–​7 norms, 586, 631 positivism and positivists  370n.3, 371n.3 reasoning 328 recognition 252 redress 316 remedies 230, 245 rights  131, 239–​40, 246, 278, 389, 451–​2

726   Index legal (cont.) status 236, 416 systems  214–​15, 294 theorists 167 see also law/​laws legislation  227, 295, 297, 392, 406, 580, 582, 629 legislative  body 21 changes 21 phase/​stage  30, 115 procedure 631 legitimacy  1–​2, 8, 144, 162, 243, 249, 288, 342, 360, 369, 391, 396–​7, 402, 407–​8, 422, 431, 449–​50, 557, 559, 573, 587, 604, 610, 619, 633, 646, 675, 677 legitimate  agents 631 aims 591 authority  2, 5, 283, 625, 633 claims of justice  654 coercion 259 complaints 608 forms of patental partiality  432–​3 government and function of government 102, 484n.8 partiality 486 political arrangements  449 political authority  633 political education  450 power  295–​6 public values  34 societies 369 state 101 state interventions  406 system of governance  378 Leibniz, G.  346n.1 Leigh, I. 577 Leiter, B.  586–​7 Lenard, P. T.  474, 658 leveling down  54, 61, 67–​8, 70n.5, 73n.8, 98n.6, 468, 492, 569, 706 Levy, N. 153 lexical  difference principle  24n.14 ordered principles  369 priority  273, 306, 462

Liao, M.  251, 419n.5, 421 liberal  accounts of justice  370, 375n.5, 390 citizens 301, 634 conceptions  227, 269, 300 democracies  116, 121, 169–​70, 250, 253, 398–​9, 634–​4 economic systems  15 equality/​egalitarian view  21–​2, 28, 611, 675–​6, 682 governance 378, 634 individual  219, 225, 227 neutrality 560 orthodoxy 651n.12 political justice  301 political morality  452 political philosophy  169, 214, 381 socialism  22–​4, 33 societies  30, 116, 403, 593, 614 theories  227, 369–​7 1, 549, 558–​60, 562, 573 values 106, 243 liberalism  20, 121, 222, 241, 300–​1, 370, 378–​9, 519, 528, 560, 694 libertarian  account of distributive justice  142–​4, 146, 148, 369 account of economic activity  149 account of property rights  141–​3 account of self-​ownership  141 beliefs/​claims  7, 102, 135, 130n.3, 143 duties 129n.2 insistence on voluntarism  561 laissez-​faire  21 notion of freedom  142 social philosophy  530 theories  129n.2, 140, 573 thinkers  449, 561, 620, 698 views  6, 129, 135n.16, 138, 141–​2, 144, 148 libertarianism  21–​2, 41, 49, 129–​32, 134, 142–​4, 146–​9, 369, 370, 434 libertarians  22, 43, 102, 132–​8, 139n.25, 141–​7, 561, 632 liberties  13–​15, 16n.3, 28, 30, 44, 102, 123, 130, 242, 306, 370, 428, 461, 520–​1, 550, 564, 606, 608, 682, see also basic liberties

Index   727 liberty  13, 21, 27, 43, 56–​7, 102, 130–​1, 136, 138, 140n.27, 180, 181n.9, 211n.21, 246, 261, 269, 292, 355n.11, 380, 452, 455, 516, 565, 578–​9, 581, 583, 589, 605, 693 Light, A. 557 linguistic  behavior  597–​8 competence 153 concerns  382–​3 considerations  381–​2 differences  600–​1, 617 diversity  597–​602 identity 616 justice  598–​9, 602, 609, 612, 614–​15, 617 neutrality 607 outcomes 604, 610 practices 597 preferences  604, 608–​9, 613–​14 usages 382 see also language Lippert-​Rasmussen, K.  58, 130n.3, 165, 441, 480n.2, 483, 485n.9, 493, 494n.19 Lloyd, S. A.  404–​5 Locke, J.  15, 138, 219, 227, 231, 298, 508, 698 Lockean  conceptions of man  219 libertarianism 41 libertarians 43 proviso  138–​9, 369 view on rights  148n.43 Lomborg, B. 668 Lucas, J. R.  159–​60 luck  brute luck  41, 47, 54, 57–​62, 125, 142, 187, 188n.14, 191n.18, 192, 316, 322–​3, 395, 440–​1, 469–​70, 482–​3 circumstantial luck  190–​3 option luck  41, 54, 57–​61, 125, 187, 188n.14, 192 luck egalitarianism  41, 44–​5, 51, 57–​61, 63, 125, 165–​8, 169n.9, 170, 187–​9, 307, 312–​18, 319n.17, 321–​3, 325, 390, 397, 434, 467, 469–​72, 494, 535, 611–​12, 614–​15, 705 luckism  43–​4, 471 Lynch, K. 400 Lyons, M. 400

Macdonald, T. 629 Macedo, S.  278, 416n.2 Mack, E.  130, 134, 146 Macleod, C.  403, 416, 421–​3, 428n.12, 429, 432–​4, 448n.14, 452, 453n.16, 454 MacPherson, C. B.  138 mammalian persons  261–​2, 269–​7 1, 274, 277 Markel, D. 180 Markell, P. 249 markets  22, 30, 34, 42–​3, 53–​5, 57, 60, 62–​3, 142–​3, 144n.33, 275, 294, 369, 415, 529n.16, 530, 534n.1, 548n.13, 566, 608–​9 Marmot, M. G.  466, 518 Marxist  account of exploitation  534n.1, 537, 543–​5 economists 544n.11 perfectionism 524 theory 548 Marxists  219, 263, 544, 547 Marx, K.  22, 86n.2, 109, 268–​9, 285, 537–​8, 543–​5, 548 Mason, A.  335n.14, 342, 394, 440 maximin  24–​5, 27, 33–​5, 186, 372, 461–​2 Maximization 69, 70n.5 May, S. 610 Maynard-​Smith, J.  264 Mazor, J.  138, 147n.41 McBride, C. 593 McFall, M. 421 McHale, J. 656 McKerlie, D.  67, 69, 70n.5, 72, 76, 333n.10, 475–​6 McLeod, O.  154, 163n.6, 190n.17, 422 McMahan, J. 270 Melville, H.  633–​4 mental  abilities/​faculties  18, 455, 518, 692 characteristics 502 competence 100 disabilities 115, 417 health 275 maturity 350 states  45, 113, 118–​19 meritocracy  17, 440n.5, 443 meritocratic equality of opportunity (MEO)  440–​5, 447, 450, 452, 454

728   Index meta-​ethics  111, 347, 353 Meyer, A. 668 Meyer, L. H.  680, 702, 706 Meyers, M. K.  407 migration  8, 261n.2, 474, 578, 594, 640–​3, 644n.6, 649, 652n.14, 656, 659, see also emigration, immigration, movement across borders Milanovic, B. 264 Miller, D.  43, 103, 140n.28, 153–​5, 157, 160, 169, 559–​60, 562–​5, 567–​70, 583–​4, 586, 594n.10, 601, 621–​2, 624, 634, 652–​4, 656, 658–​9, 672n.11 Miller, S. C.  213, 229 Mill, J. S.  32, 34, 152, 231, 246, 250, 265, 396, 452, 556, 702 minimalist  account of advantage  536–​7, 547, 553 approach to language  604 minorities  8, 37, 220, 238, 240, 396, 482, 495, 562, 567, 576–​7, 579–​81, 583, 590, 593–​4, 597, 599–​602, 607–​10, 613, 615–​17, 646 Mises, L. von  289 misfortune  15, 22, 42, 52, 55, 60–​1, 80, 184, 187, 225, 273, 539, 673, see also fortune, luck mitigation  625, 665n.2, 666–​7 1, 673–​7, 679–​80, 683–​4 Modood, T.  235, 593, 594n.10 modularity 672 Moore, M.  159, 170, 178 Moore, W. E.  265, 346 Moreau, S.  487–​9 Moriarty, J.  152n.1, 161, 168n.8, 187n.13 Morris, H. 160 Morse, S. J.  167 most advantaged group (MAG)  24–​7, 31, 36, 187, 683 movement across borders  640, 642, 644–​5, 647–​60, see also emigration, immigration, migration multicultural  character 236 societies  577, 579, 594 state 243 multiculturalism 244, 577 Munoz-​Darde, V.  432 Murphy, J. B.  523 Murphy, L. B.  20, 567–​9

Nagel, T.  4, 41, 66–​8, 70–​1, 76, 145, 209, 228, 260n.1, 270, 273, 290, 293–​4, 440n.6, 441n.8, 567–​9, 621, 649–​51 Narveson, J.  130, 135n.16, 136–​9, 141, 148n.45 nationalism  221, 235, 634 nationalists 621 nation-​states  136, 146–​7, 577, 620 natural  disadvantage  535–​6 duties  13, 20, 449, 693 endowments 15, 124 factors 395, 401 goods  210, 525, 527 inequalities 260n.1, 498 liberty  21–​2, 355n.11 order  87, 260, 265 origin  260, 265–​6 procreation 418 races  502–​3 resources  129, 135–​9, 141–​2, 145–​9, 628 rights 130, 698 talents 140, 142 needs-​based approaches  86 Neeson, J. M.  515 Nelson, E.  121–​2 Neufeld, B.  285n.4, 403, 406 neutralism  169–​70 neutralists  168, 169n.9, 170 neutrality  168, 170, 243, 347, 368, 376–​9, 382, 408, 411, 560, 592–​3, 616–​17 argument  168–​9 condition  606–​7 model  603, 608, 610–​13, 615 Noddings, N.  216–​17, 226 noncomparative  analyses 87 conception 180 decisions 589 desert  161n.5, 170, 177–​80, 189–​1 ideal 52 injustice 588, 591 justice  182, 587–​9 nature of prioritarianism  69 principles 272 nonideal  circumstances 20, 33 conditions  14, 22, 24, 27, 34, 401 theory  33, 54, 123, 327–​8, 341–​2, 635

Index   729 non-​identity problem  690–​2, 696, 701–​2, 704–​6 Nord, E.  73n.10 normal species function (NSF)  463–​5, 468–​9, see also species Norman, R. 65 normative  ethics  8, 307–​8, 312–​14, 325 force  90, 94, 99, 145 significance  3, 5, 102, 152, 157–​8, 265, 267 normativity  299, 301–​2, 330n.3 Nozick, R.  2, 15, 20n.9, 21, 41–​3, 55, 134–​5, 138–​9, 141, 144–​6, 148n.45, 184, 187, 203, 291–​2, 306–​7, 369, 415, 449, 482, 485, 507, 544–​5, 549, 561–​2 Nussbaum, M.  100, 102, 114, 117, 119–​21, 123–​6, 261, 268, 428n.12, 467, 583, 586, 593, 623 Oakeshott, M. 259 Oberman, K.  643, 645, 647, 656–​7 Ogilvie, M. 278 Okin, S. M.  404–​6, 415n.1 Olsaretti, S.  112, 144, 147n.40, 152–​4, 167–​8, 171, 177n.1, 187n.13, 188n.14, 191n.18, 398n.4, 404n.6, 408, 414, 426, 456, 557n.2, 588 O’Neill, O.  65, 73n.7, 100, 126, 222, 341, 523nn.10–​11, 523n.8, 524, 564 Organization for Economic Co-​operation and Development (OECD)  659 original position  33, 124, 219, 301, 328, 355n.11, 357, 360, 369n.2, 370–​2, 378, 461, 592–​3, 694–​6 Otsuka, M.  71–​2, 73nn.7–​8, 76, 78–​9, 130n.6, 132, 138n.23, 139–​41, 273, 539n.4 Ott, K.  676n.14 Overall, C. 418 ownership  130–​1, see also property rights, self-​ownership Page, E. A.  683n.19, 706 Parekh, B.  235, 249, 584–​5, 589, 594n.10 parental  adequacy 423 anti-​perfectionism  456 authority 453n.16 competency  420–​1 freedom 452

leave  393, 403, 407–​8, 421 licensing 422 love 429, 454 partiality  428, 431–​4, 443, 453–​5 rights  423, 451–​3 shaping of religious values  455 parents  18, 62, 218, 221, 263, 268, 270, 275–​6, 322, 393, 398, 403, 407–​8, 416–​27, 429–34, 438, 439n.1, 443–​8, 454–​7, 508–​9, 525, 611 Pareto  improving  24, 35, 533–​4, 537, 539, 542n.7, 548, 553 measures 23, 25, 35 Parfit, D.  31n.20, 67–​70, 73n.7, 79–​80, 120, 197–​9, 273, 325n.23, 485, 492, 569, 689n.1, 690, 691n.2 Parker, G. A.  264 Parr, T. 61 paternalism  7, 52, 113n.2, 121, 142–​3, 207n.19, 220, 225, 456 Paton, J. 643 Patten, A.  585, 598n.1, 606–​7, 610, 613n.5, 616n.6, 617n.7 Patton, C. 410 Pearlman, J. 643 Pennington, M. 342 perfectionism  121, 169, 451, 454–​6, 519, 524, 526, 528 perfectionist  account/​theory of justice  117, 406 approach 519 capability theory  120 education 450 goods  209–​10, 528 liberal 121 values 210 perfectionists  169, 378–​9, 450 Persson, I.  73n.7, 267, 274, 492 Pettit, P.  1n.2, 251–​2, 294, 298n.19, 631 Phillips, A. 114 Pickett, K.  275–​6 Pierik, R.  118n.3, 120, 123, 126 Piketty, T.  15, 264 Pinker, S.  259, 263, 395 Plato  86n.2, 346n.1, 347, 498 pluralism  168, 183, 196–​8, 200, 201n.10, 203–​4, 450

730   Index pluralistic  67, 81, 183, 622, 706 approach 678 distributive justice  58 egalitarianism  67–​8, 70n.5, 81 intrinsic goods  182 liberal democracy  169–​70 prioritarianism 83n.16 theory of justice  678–​9 Pogge, T.  116, 120, 228, 260n.1, 439n.4, 475, 620, 622, 624, 627–​9, 631, 635–​7 Pojman, L. P.  190n.17 Polanyi, K. 515 political  anti-​perfectionism  450, 456 authority  1, 295–​7, 299, 301, 303, 633 autonomy  449, 450, 455, 615 community  43, 47, 54, 60, 440, 468, 601, 621 conceptions  168, 300–​2, 358, 449–​50, 519 constitution  14–​15 culture  16, 358, 361, 363 education  449–​50 equality  16, 313, 317, 322, 631 freedom 376, 381 ideals  66, 313, 449, 484n.8 justice  34, 121, 283, 285–​8, 294–​5, 297–​9, 301–​3, 619, 626–​7, 631–​2, 634, 637 liberalism  121, 300, 378–​9, 519, 528, 559, 694 liberties  15, 16n.3, 38, 428, 520 life  8, 226, 230, 297–​8, 302–​3, 313 morality  301, 370, 380–​1, 449, 452, 457 office 114, 429 opportunities 392, 410 participation 16, 95 power  203, 238, 294–​5, 396, 452, 628, 632–​3, 678–​9 rights  38, 294, 632–​3 self-​government  634–​5 stability 598 status 87, 236 values  222, 284, 286, 324 politics  of identity  5, 244, 246, 250, 601 of recognition  235–​6, 244 Polluter Pays Principle (PPP)  667, 675, 677, 681–​4 pollution  131, 134, 146n.37, 275, 509, 558, 676–​7, 681

polyethnic societies  577–​8, 580, 586, 594 polygyny 263 Porter, T.  73n.7, 76 positivism 370n.3 Postema, G. 360 poverty  22, 29, 58, 93, 95, 110, 144–​6, 219, 222, 228–​30, 264, 275–​6, 290, 321–​2, 389, 540, 620, 625, 628, 636–​7, 642, 643n.5, 655, 671, 680, 624 Powers, M. 467 Prah Ruger, J.  467–​8 Principle of Abstraction  57, 608–​9 Principle of Correction  57, 610 Principle of Equal Consideration of Interests  291–​2 Principle of Independence,  57 Principle of Precedence  103 prioritarian  approach/​views  65, 71, 79–​80, 82, 186 considerations 14 interpretation 31 pattern of healthcare  462 principles  14, 273, 307 sufficientarianism 97 welfarist 62 prioritarianism  31, 59, 65, 68–​9, 70n.5, 82n.15, 87, 98n.6, 125, 192, 273, 370, 492 expected-​well-​being  77, 79–​80 final-​well-​being  71–​7, 79–​80 prioritarians  43, 59, 66, 68–​70, 73, 91n.4, 492 Priority View  31n.20, 68–​7 1, 73n.8, 79, 273 Prisoner’s Dilemma  336 privatization  557, 606–​7, 609 procreation  221, 262, 396, 398, 417–​18, 425–​6, 680 Proctor, D. 275 property rights  6, 21, 106, 129–​33, 135–​6, 140n.27, 141–​2, 144–​6, 148–​9, 293, 558, 692, 698 property-​owning democracy (POD)  14, 17, 20–​4, 26–​7, 33–​8, 113, 524, 530 Proust, M.  261n.2 Provost, C. 643 psychology  340, 405, 527, 690 Pummer, T.  182n.10 punishment  7, 130, 153–​4, 156–​7, 159–​64, 167, 170, 177–​82, 184–​7, 189, 191–​3, 218, 231, 250, 311, 420, 672, see also criminal justice, retributivism

Index   731 Quong, J.  139, 301–​2, 315n.9, 369n.1, 557, 583–​5, 589 Rabinowicz, W.  71–​2, 73n.7 race  179, 218, 235–​6, 320, 416, 481, 488, 498–​9, 501–​6, 508, 510–​11, 516, 518, 582, 591, 693, 697 Rachels, J.  153, 157, 159–​60, 164 racial  characteristics 591 construction  502–​3 differences  498–​9 discrimination  260, 371, 440, 482, 489, 494, 499, 510, 591 divisions 8 domination 316 homogeneity 659n.15 ideologies 499, 501 inequality 507, 511 injustice 509 integration 511 minorities  238, 240, 482, 590 segregation  506, 510–​11 racism  123, 263, 480, 501, 504–​5, 510–​11, 527, 601 Radcliffe-​Richards, J.  395 Rahul, K.  105, 677 Raiffa, H. 680 Rakowski, E.  147n.40, 408 Raphael, D. D.  1n.1, 349 Rawlsian  basic structure  624 constructivism 346 contractualism 692 distributive justice  549, 620 equal status-​recognition  242 equality of opportunity  467 justice  114, 117, 339, 397, 405–​6, 461–​2, 467, 469, 520, 608 justifications of freedom  521 liberals 301 objection to the capability approach  117 objection to utilitarianism  76 on work  519 overlapping consensus  121 political constructivism  300 political liberalism  519, 528 primary goods  400, 461–​3

property-​owning democracy  530 society 449, 583 views 5131,  525, 519, 529 Rawlsianism  528–​9 Rawlsians  114, 116–​17, 186, 301–​2, 394, 462, 516, 564 Rawls, J.  1–​2, 4–​7, 13–​39, 47, 62, 75–​6, 86–​7, 95–​7, 106, 110, 113–​17, 121–​4, 152, 161, 163, 168–​9, 186–​7, 199, 210, 213, 219, 227–​8, 235–​7, 242, 244, 260n.1, 268, 271–​2, 274, 287n.8, 288–​9, 295–​7, 300–​3, 306–​9, 314, 316n.12, 322n.19, 327–​8, 331–​3, 335–​6, 338n.19, 339–​42, 346–​7, 350, 352–​3, 355–​65, 368–​79, 389–​91, 402–​5, 415, 424, 426, 428–​9, 432n.17, 434, 439–​41, 443, 445, 448–​51, 455, 460–​2, 464–​6, 470, 498, 519–​24, 526, 528, 558–​60, 562, 564, 567, 573, 578, 585, 598, 604, 619–​21, 624–​5, 631–​2, 635–​6, 645, 682, 691–​5 Raz, J.  201n.10, 243, 381, 450, 456 Reader, S. 100 realism  287, 339, 347n.4, 353n.7, 355n.12 realistic idealism  328n.1, 329–​30, 337, 340 Réaume, D.  598n.1 Rechtstaat 253 reciprocity  3, 13–​14, 18, 20, 30–​1, 33–​4, 37–​8, 261, 274, 277, 333, 334n.12, 336, 339, 525, 621, 694–​5 recognition  4, 6–​7, 100, 155, 159, 164, 169, 224, 228–​9, 235–​53, 261, 353, 357–​9, 362–​3, 365, 373, 382, 390, 397, 410, 416, 427n.10, 430, 585, 592–​3, 608–​10, 615, 628, 652n.13, 707–​8 Reiman, J. 696 relational  egalitarianism  96–​7, 307, 316n.12, 433, 491 egalitarians  312–​13, 318, 434 equality  397, 433–​4 goods 397, 410 justice and injustice  287, 390, 410 persons  220–​1, 229 theories/​views  491–​4, 623, 630 religion  179, 222, 251, 278, 296, 298, 448, 450–​2, 456, 576–​94, 605, 616, 645, 647 religious  affiliations 235, 590 allegiances  576, 579, 590 beliefs  378n.7, 417, 450, 583–​4

732   Index religious (cont.) commitments  249, 301, 355, 450, 576, 583–​5, 589 communities 251, 611 discrimination  582, 590–​1 diversity 576, 579 equality 593 establishment 606 exemption  580–​1, 585, 592n.9 freedom  576, 578, 582, 587–​90 identities 582, 585 justice  604–​5 liberty  578–​9, 581, 583, 589 life 583, 587 minorities  8, 576, 590, 593, 646 practices  198, 423, 576, 578, 581, 586 school 453 Religious Freedom Restoration Act 1993 (RFRA) 587n.4  Rendall, M.  73n.7 responsibility  153–​4, 158, 162 responsibility-​based patterned theories  186–​7, 192–​3 retributive  desert  162–​5, 178–​82, 184, 186–​90 justice  161–​3, 170, 177–​8, 181–​5, 187–​9, 192–​3, 483 punishment  179–​81 theory 162 retributivism 181 retributivists  177–​80, 311 Richards, D. A. J.  223 Richards, N. 416 Richardson, H. S.  116–​17 Riddell, R. C.  95, 97 Ripstein, A. 47 Risse, M.  148n.42, 621–​2, 637, 642, 674n.13, 707 Roark, E. 137 Robeyns, I.  109n.1, 110, 116, 118n.3, 120, 123–​4, 333n.11, 394, 406n.9, 408, 415n.1, 428n.11 Robinson, F.  216, 223, 231 Robinson, T. 375 Roemer, J. E.  22n.11, 41, 62, 469, 537, 544n.11, 545–​8, 551–​2 Ronzoni, M.  331n.4, 629 Roser, D.  680, 706 Rose, S. J.  518

Ross, W. D.  152, 156–​7, 346, 349–​51, 357 Rothbard, M.  136–​8, 141, 139, 549 Rousseau, J.-​J.  86n.2, 339, 500 Rubenstein, J. 628 Ruddick, S.  216–​17, 226, 231, 400 Sabel, C.  624, 650 Sager, L. G.  583, 585, 587n.4, 588 Sample, R.  536–​8, 540–​5 Samuelson, P. A.  544n.11, 557 Sangiovanni, A.  621–​3, 642, 659 Sartre, J.-​P.  86n.2 Satz, D.  96, 143, 433–​4, 447–​8, 468–​9 Saul, J.  392, 410 Saunder-​Staudt, M.  221 Scanlon, T. M.  4, 18, 157, 199n.7, 205n.14, 353n.7, 273, 319–​20, 347n.4, 361–​2, 707 Scheffler, S.  43–​4, 49, 153, 160–​5, 186–​7, 190n.17, 288, 293n.15, 312, 314, 316nn.11–​12, 494, 699, 708 Schemmel, C.  322n.20 Schmidtz, D.  154–​5, 157, 159n.4, 332n.5, 333n.9, 333n.11, 334n.12, 561–​2, 635 Schmitt, C. 259 Schoeman, F. 423 Schokkaert, E. 122 Schooler, C.  515, 516n.2, 521–​2 Schouten, G. 403 Schwarzschild, M. 183 Sebenius, J. K.  680 Segall, S.  50, 73nn.7–​8, 465, 482n.3 Seglow, J. 585 self-​conception  15, 245–​6, 360, 363–​4 self-​esteem  92, 521 self-​evidence  347, 349–​50, 356–​9, 361, 364 self-​government  200, 240, 522, 594, 615, 625, 634 self-​ownership  129–​36, 140–​1, 147–​8 self-​relation  246–​7, 249 self-​respect  14–​15, 17, 26, 33–​8, 66, 117, 122, 235, 272, 390, 405, 520–​2, 523n.11, 524, 529, 586 Sen, A.  24n.14, 33n.22, 70n.5, 109–​10, 112–​17, 118n.3, 119–​20, 122, 124–​5, 204, 428n.12, 461–​2, 467, 473 Separability  69, 70n.5, 82n.16 separateness of persons  65, 73, 76–​8, 82

Index   733 Sessions, R. 515 Sevenhuijsen, S.  218, 223, 226 sex  143, 236, 263, 278, 391–​2, 394, 396, 402, 420 sexism  123, 237, 405, 480 sexist  conceptions 199 social norms  417 socialization 405 upbringing 391, 405 sexual  difference 392, 396n.3 discrimination  402, 440, 481–​2, 494 harassment 397 liberation 278 violence 398 Shachar, A. 644 Sher, G.  43, 51, 169, 188n.14, 223 Shields, L.  91n.4 Shiffrin, S. 418 Shorten, A.  581, 586 Shue, H.  100–​2, 667n.6, 668, 676, 682–​3 Sidgwick, H.  152, 346–​7, 350, 352–​3, 357–​9, 361, 364 Sieyes, I.  235–​6 Simmons, A. J.  557, 565 Singer, P.  261, 620 Skow, B. 170 slavery/​slaves  131, 143, 240, 485, 498, 500–​5, 508, 514, 660 Sleat, M.  334n.13, 217, 219, 221 Smart, J. J.  273 Smilansky, S. 153 Smith, A.  109, 515, 527, 556 social  circumstances  28, 220, 265, 356n.13, 399, 419, 423–​4, 429 class  15, 17–​18, 22, 38, 235 conditions  28, 245–​6, 356n.13, 401, 417, 430, 433, 495, 522, 525, 645 construction  266–​7, 499–​500, 502–​4, 511 constructivism 267, 502 contract  124, 130, 219, 297–​9, 302–​3, 339, 690–​1, 695, 701 cooperation  2–​4, 7, 14, 65–​6, 86, 195, 238, 269, 351, 368, 389, 404, 415, 428, 525, 527, 562, 564, 692, 695 equality  38, 65, 313, 322

goods  14, 28, 31, 36, 114, 122, 242, 320, 461–​2, 527 inequalities  16, 27, 65, 95, 260n.1, 430, 440–​1 labor  525–​6 life  8, 14, 18n.6, 30, 216, 265, 287, 295, 297, 302, 368, 374, 406 minimum  22, 25–​7, 32, 34, 38, 47, 95, 564, 604 mobility 445, 599 norms  4, 114, 123, 395, 400, 417, 425, 516 opportunities  392, 410, 576, 600, 612, 615 policies  29, 31, 92, 97, 407, 419, 422 practices  4, 116, 157, 237, 239, 248, 250, 265–​6, 351, 360, 419, 461, 517 relations  98, 120, 214, 306n.1, 320, 481, 494–​6 relations egalitarianism  494–​6 solidarity  601, 658–​9 status  13, 87, 236, 250, 393 utility 155n.2 values  2, 32, 69n.5 social process view  1 socialism  21, 22, 24, 33, 530 socialist  exploitation  545–​6 ideal 285 revolutions 530 society 285 system 23 socialization  47, 394, 403, 405–​7, 426, 516, 525, 529, 557n.2, 598 Socialization Principle  516, 525 sociolinguists 598 Sommerville, M. 420 Sparrow, R.  265n.4 species  121, 209, 262–​4, 269, 271–​5, 277, 463, 500, 560, 563, see also normal species functioning Spencer, H.  259, 265, 327 Springborg, P.  86nn.1–​2 Staats, C. 410 Stake, J. A.  264 Stanczyk, L.  516–​17 Statman, D.  178n.2, 179n.5 status-​recognition  236–​7, 240–​2, 248, 251, 253, 593 Steedman, I.  550n.15

734   Index Steiner, H.  130n.6, 132, 137n.19, 139–​41, 144n.34, 147n.39, 148n.42, 309n.3, 379–​81, 534n.1, 537, 545, 548, 550–​3 Stemplowska, Z.  166, 188n.14, 312n.6, 337n.17, 340, 542 Sterba, J. P.  102, 130n.4 Stiglitz, J. E.  144n.33, 145 Stilz, A.  632, 645, 646n.7, 650, 657 Straehle, C. 474 Stratton-​Lake, P.  349 Strawsonian conception of responsibility  160 Street, S. 260 Striffler, S. 515 sufficiency  86–​9, 91–​7, 98n.6, 106, 111, 123, 125, 261, 272–​3, 317, 467–​8, 471, 624, 669–​70, 682, 684–​5 sufficientarianism  44, 87–​9, 91, 96–​8, 106, 370, 448n.14, 467–​9, 492, 671 sufficientarians  43, 87–​9, 91n.4, 92–​3, 106, 468 Sulzbach, W. 633 Swift, A.  277, 333n.10, 337n.17, 398, 416n.2, 422–​3, 432, 441n.7, 444, 446, 448n.14, 452–​6 Sylvan, R. 563 Tadros, V.  156, 167, 311n.5 Taket, A. 474 Tan, K.-​C.  4, 166–​7, 312n.6, 313n.7, 315, 316n.12, 319n.16 Tawney, R. 66 taxation/​taxes  6, 15, 17, 21–​5, 27, 35, 43, 54, 56, 93–​4, 97, 121, 129, 154, 209, 249, 320, 406, 407n.10, 425–​6, 445–​6, 453, 495, 534, 547, 556–​9, 561, 563–​4, 567n.8, 568–​9, 571–​2, 581, 599, 605, 631–​2, 635, 657, 665, 667, 700 Taylor, C.  235–​6, 245, 249, 593, 598n.1 Taylor, I. 565 teleological  arguments 500 theories  8, 308, 409 teleology  308–​9 Temkin, L. S.  41, 67–​8, 70n.5, 82, 165–​9, 188n.14, 191n.18, 273, 475–​6, 492 Terkel, S. 515 Thompson, J. 699

Thompson, S. 593 Thomson, G.  99–​100 Thomson, J. J.  131, 271, 319 Thornhill, R. 263 Thrasher, J.  289n.10 Tomasi, J. 446 Treatment  238–​41, 244–​5, 252–​3, 464 Trigg, R.  586–​7 Trivers, R. L.  263 Tronto, J. C.  217–​19, 221, 223, 226, 231 Tully, J. 245 Ubel, P.  73n.10 unemployment  15, 36, 60, 515, 518, 560, see also employment unfairness  60, 67–​8, 82, 191, 210, 224, 241, 268, 394, 399, 401, 418, 535–​7, 539, 544, 552, 562, 605, 696, 700, 705, see also fairness Unger, P.  145n.36 United Nations (UN)  99, 389, 393, 633, 641 Universal Declaration of Human Rights  271 utilitarian  accounts of justice  75, 117, 186, 369 approach to political philosophy  235 approaches 217, 668 moral theory  227 utilitarianism  30, 66, 69n.5, 75–​7, 81, 146, 214, 227, 273, 308–​10, 319, 352, 357, 359, 369–​70, 375, 669 utilitarians  66, 71n.6, 186, 268, 620 utility  29–​30, 66, 70n.5, 82n.15, 155, 166, 186, 227, 268–​9, 289, 310, 361, 538, 546, 551, 567n.8, 571, 683 utopia 116, 335 utopian 286 ideal/​idealism  328n.1, 329–​31, 337 standards 337 Valdman, M.  200n.9, 201 Valentini, L.  337n.18, 630 Valian, V. 396 Vallentyne, P.  3n.7, 4n.9, 58, 82n.15, 120, 125, 129, 130n.4, 130n.6, 132, 138n.22, 139–​41, 143, 145n.35, 147, 148n.42, 168, 191n.18, 205n.13, 273–​4, 408, 425, 437, 442 Vallier, K.  289n.10, 302 value-​independence  368, 376–​7, 380, 382

Index   735 value-​neutrality  112–​13, 368, 376–​80, 382 Vanderheiden, S.  674–​7 van Parijs, P.  18n.6, 598n.1, 601, 610 Van Schoelandt, C.  285n.4 veil of ignorance  53–​4, 62, 370, 476, 692–​6 Velleman, D.  418, 702 Venkatapuram, S.  467–​8 Vernon, R. 422 Vilhauer, B. 154 violence  219, 221, 230–​1, 238, 259, 263, 265n.4, 276, 390, 398, 628 virtue theory  221 voluntarism 561 Von Neumann–​Morgenstern axioms  71n.6 Voorhoeve, A.  71–​2, 73n.7, 74, 76, 79, 82, 273 vulnerability  124, 216, 229, 246, 452, 536, 538–​40, 542–​3, 552, 708 Waldron, J.  243, 251, 586 Wallace, B.  267, 274 Wall, S.  132n.11, 143n.32, 200n.9, 203n.12, 526 Walt, S. 552 Walzer, M.  6, 236, 467–​8, 483n.5, 560, 562, 573, 621, 634, 636, 653, 668, 678–​9 Weinstock, D. 422 Weirich, P.  70n.5 Weithman, P.  289, 355 welfare  5, 7, 19, 28, 44–​50, 52, 59, 61, 68, 69n.5, 74, 81n.14, 113, 117–​18, 132, 134–​5, 140, 142n.31, 159, 163, 165–​6, 170–​1, 177, 182–​3, 195n.1, 199, 204–​5, 210, 211n.21, 226–​7, 253, 261, 267, 274, 278, 292n.14, 306, 312, 390, 406, 428, 442, 473, 482–​3, 487, 493–​4, 543, 547, 563, 569, 580 economics  69n.5, 81n.14, 110, 122, 125 state  21, 23–​4, 34, 36, 93 state capitalism (WSC)  14, 22, 36–​8 welfarism 63, 118 welfarist 62 doctrine 273

equal opportunity  47, 49 metrics  110, 118–​19, 428n.12 welfarists  48–9, 57, 62, 204, 273 well-​being  34, 52, 63, 66–​9, 70n.5, 71–​82, 83n.16, 87, 91–​2, 110, 112–​13, 116, 119–​20, 122, 199n.7, 204n.13, 218, 340, 390, 398, 405, 420, 429, 431–​2, 444, 456, 463, 540, 586, 642, 668, 703 Wellman, C. H.  649, 653 Wertheimer, A.  144, 534n.1 White, J. 447 Whiteman, H. 643 White, S.  163, 565–​6, 571–​2 White, T. 261 Wicksell, K. 559 Wiggins, D.  99–​103 Wilkinson, R.  275–​6, 466 Williams, A.  5, 49, 73n.7, 82n.16, 118n.3, 147, 273, 346, 408, 425–​6 Williams, B.  273, 332 Williams, E.  501, 505 Williams, J. 407 Williamson, T. 524 Wolff, J.  114, 346n.1, 467 Wolin, S. 235 Woodward, J.  701–​2 World Bank  642 World Trade Organization (WTO)  625–​7, 633 worth-​recognition  237, 239, 244, 246–​8, 250–​3 Wortman, J. 515 Wrangham, R. 277 Wright, E. O.  398, 408 Young, I. M.  5, 244, 390, 397, 416n.2 Young, M.  440n.5 Ypi, L.  598n.1, 611, 613, 641, 656 Zak, P. J.  267 Zerjal, T. 263 Zong, J.  641n.4