(Oxford Handbooks) Serena Olsaretti - The Oxford Handbook of Distributive Justice-Oxford University Press (2018)

The Oxford Handbook of Distributive Justice
Oxford Handbooks Online

Edited by Serena Olsaretti
Print Publication Date: May 2018 Subject: Philosophy Online Publication Date: Jun 2018
(p. iv)
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 authors have 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
Page 1 of 2
PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). © Oxford University Press, 2018. All Rights
Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in
Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).
Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

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.
Page 2 of 2

Contributors

Contributors
(p. ix) 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.
Page 1 of 9

Contributors
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 (p. x) 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).
Page 2 of 9

Contributors
Paula Casal is ICREA Professor at Pompeu Fabra University, Barcelona, associate

editor 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).
Page 3 of 9

Contributors
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.
(p. 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
Page 4 of 9

Contributors
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
Philosophy. He is the author of sixteen books and the co-editor of four further
(p. xii)
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);
Page 5 of 9

Contributors
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
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.
Michael Otsuka is a Professor in the Department of Philosophy, Logic and

(p. xiii)
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.
Page 6 of 9

Contributors
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
Page 7 of 9

Contributors
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 (p. xiv) 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.
Page 8 of 9

Contributors
Page 9 of 9

Introduction: The Idea of Distributive Justice

Serena Olsaretti
Print Publication Date: May 2018 Subject: Philosophy, Social and Political Philosophy
Online Publication Date: Jun 2018 DOI: 10.1093/oxfordhb/9780199645121.013.38
Abstract and Keywords
This chapter introduces the idea of distributive justice. It identifies several different views
of what characterizes distributive justice, as opposed to other types of justice and to non-
justice-based moral demands. The preconditions of distributive justice, its primary subject
and its object, and its normative significance are discussed. The chapter then suggests
that bringing the diversity of usages of the concept of distributive justice into view helps
cast light on some of the many contemporary debates about distributive justice and its
limits. This chapter also introduces and outlines the different topics covered by the
different sections and chapters of the book.
Keywords: distributive justice, non-justice-based moral demands, preconditions of justice, subject of justice, object
of justice, normative significance of justice
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
Page 1 of 11

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
(p. 2) 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
Page 2 of 11

treating people equally necessary for settling fairly their competing claims? These
questions are the staple of many debates among theorists of justice.
Besides disagreeing on what justice demands, theorists of justice also disagree

(p. 3)
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 (p. 4) 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
Page 3 of 11

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 (p. 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.)
Page 4 of 11

(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 justice, 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 (p. 6) 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,
Page 5 of 11

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).
Theories of justice also take a stand on what the currency of distributive justice is:
(p. 7)
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
Page 6 of 11

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 (p. 8) 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
Page 7 of 11

justice (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 (p. 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.
Page 8 of 11

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.

Page 9 of 11

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. (p. 10)
Notes:
(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.
(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.
(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
Page 10 of 11

other types of justice (such as criminal justice), as well as to other non-justice-based

moral demands (such as the demands of beneficence).
(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.
Serena Olsaretti
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.
Page 11 of 11

Rawls on Distributive Justice and the Difference Principle

Rawls on Distributive Justice and the Difference
Principle
Samuel Freeman
This chapter analyzes Rawls’s complex account of distributive justice. Rawls’s difference
principle requires that economic systems be organized so that the least advantaged
members of society are better off than they would be in any alternative economic
arrangement. The following questions are addressed here: What constraints are imposed
by equal basic liberties and fair equality of opportunity on inequalities allowed by the
difference principle? What are the difference principle’s broad and narrow requirements?
Is maximizing the least advantaged position mandatory regardless of the inequalities
created, or is it optional so that a society can choose to limit inequalities permitted by the
difference principle? In what respect is the difference principle a reciprocity principle
and not prioritarian? What measures are required to realize the difference principle
under ideal conditions of a well-ordered society versus non-ideal conditions of an unjust
society? Why should property-owning democracy rather than welfare-state capitalism
satisfy Rawls’s principles of justice?
Keywords: John Rawls, distributive justice, economic justice, basic liberties, equality of opportunity, the difference
principle, property-owning democracy, welfare-state capitalism, prioritarianism, ideal and non-ideal theory
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.
Page 1 of 31

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 (p. 14) 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
Page 2 of 31

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 (p. 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.
Page 3 of 31

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,
(p. 16) (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).
Page 4 of 31

Regarding fair equality of opportunity: Rawls again says that economic inequalities
(p. 17)
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).
(p. 18)
Page 5 of 31

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
Finally, regarding inequality of opportunity for women due to childbearing and

(p. 19)
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).
Page 6 of 31

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).
Since the difference principle applies to institutions “at the front end” of this
(p. 20)
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
Page 7 of 31

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
(p. 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
Page 8 of 31

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:
Page 9 of 31

Table 1.1
Comma Market Propert Social Welfare “Libera Classic Liberta

nd socialis y- democr state l al rian
econo m owning atic capitali equalit liberal laissez-
my democr welfare sm y” laissez- faire
state acy state fair faire
socialis equal “syste
m opport m of
unity + natural
safety liberty”
net
Page 10 of 31

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 (p. 22) “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).
This addresses G. A. Cohen’s criticism that the difference principle is compatible

(p. 23)
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
Page 11 of 31

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.
It is not true, then, that local improvements to the least advantaged that allow
(p. 24)
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
Page 12 of 31

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 (p. 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
Page 13 of 31

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. (p. 26) 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 (p. 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
Page 14 of 31

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
Fig. 1.1 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
Page 15 of 31

feasible alternatives society must enact those measures that maximally improve the
position of the least advantaged.16
Rawls says various things that can seem to support both the Optional Inequality
(p. 28)
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.
The Optional Inequality interpretation of the difference principle must then be

(p. 29)
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
Page 16 of 31

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
(p. 30) (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
Page 17 of 31

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
(p. 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
Page 18 of 31

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 (p. 32) 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,
Page 19 of 31

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.
1.7 Ideal vs Non-Ideal Theory and the

(p. 33)
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
Page 20 of 31

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 justice
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
(p. 34) 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
Page 21 of 31

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 (p. 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
Page 22 of 31

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
(p. 36) 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.
Page 23 of 31

1.8 Property-Owning 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 (p. 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 welfare-state
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
Page 24 of 31

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?
One reason not to sacrifice the wealth and economic powers of the least
(p. 38)
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
Page 25 of 31

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 (p. 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.
Page 26 of 31

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.
Rawls, J. (2001). Justice as Fairness: A Restatement. Edited by Erin Kelly.

(p. 40)
Cambridge, MA: Harvard University Press.
Page 27 of 31

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.
Notes:
(1) “Justice as Fairness” (Rawls 1999b: 48). On the difference principle and the common
good, see also “Distributive Justice” (Rawls 1999b: 153).
(2) See Rawls (1999a: 6–7); see also, Rawls (1999b: 138) and Rawls (2001: 55–7).
(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).
(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.
(7) See Blinder (2013) on such practices, and measures needed to avoid them.
(8) For discussion, see Freeman (2014: 88–111).
Page 28 of 31

(9) 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).
(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.
(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).
(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).
(15) Rawls says that strict equality is irrational (2001: 151).
(16) 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
Page 29 of 31

one lower that provides them with less, since it is on or closer to a higher equal justice
line.
(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).
(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.
(20) See Parfit 1991 and 1997. “On the Priority View, we are concerned only with people’s
absolute levels” (Parfit 1997: 214).
(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.
(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.
Samuel Freeman
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).
Page 30 of 31

Page 31 of 31

Dworkin and Luck Egalitarianism: A Comparison

Richard J. Arneson
Ronald Dworkin is a founding father of what has come to be called “luck egalitarianism,”
a family of distributive justice doctrines that hold that the inequalities in people’s
condition that are brought about by sheer brute luck falling on them in ways that are
beyond their power to control should be reduced or eliminated, but that inequalities that
arise through people’s own fault or choice, such that they can reasonably be deemed
responsible for their condition, need not be reduced or eliminated. Dworkin himself has
come to embrace an alternative view, “justice as fair insurance.” This chapter
characterizes Dworkin’s view, compares it to luck egalitarianism, and criticizes both
doctrines.
Keywords: luck egalitarianism, Dworkin, the no-envy test, justice as fair insurance, responsibility, deservingness,
ex ante perspective, ex post perspective
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.
Page 1 of 26

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, (p. 42) 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
Page 2 of 26

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 (p. 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).
Page 3 of 26

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 (p. 44) 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.
Page 4 of 26

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 (p. 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.
Page 5 of 26

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, (p. 46)
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
Page 6 of 26

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 (p. 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
Page 7 of 26

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 (p. 48) 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.
Page 8 of 26

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 adds that even when an expensive taste is voluntarily cultivated, this
(p. 49)
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).
Page 9 of 26

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
is how badly off the person really is according to the correct objective account of
(p. 50)
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
Page 10 of 26

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.
The conviction that shapes Dworkin’s view is that provided the initial set-up is fair,
(p. 51)
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.
Page 11 of 26

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.
(p. 52) 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.
Page 12 of 26

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 (p. 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
Page 13 of 26

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 (p. 54) 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
Page 14 of 26

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 (p. 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
Page 15 of 26

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 (p. 56)
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
Page 16 of 26

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 (p. 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?
Page 17 of 26

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 (p. 58)
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 element
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
Page 18 of 26

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 (p. 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.
Page 19 of 26

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
(p. 60) 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
Page 20 of 26

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 (p. 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
Page 21 of 26

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 (p. 62) 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 justice 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
Page 22 of 26

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.
(p. 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.
Page 23 of 26

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
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.
Dworkin, R. (1981). “What Is Equality? Part 1: Equality of Welfare.” Philosophy and

(p. 64)
Public Affairs 10: 185–246; also “What Is Equality? Part 2: Equality of Resources.”
Philosophy and Public Affairs 10: 283–345.
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

Fleurbaey, M. (2008). Fairness, Responsibility and Welfare. Oxford: Oxford University

Press.
Hurley, S. (2003). Justice, Luck, and Knowledge. Oxford: Oxford University Press.
Page 24 of 26

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.
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.”

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.”

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.
Page 25 of 26

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.
Notes:
(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.
Richard J. Arneson
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.
Page 26 of 26

Equality Versus Priority

Michael Otsuka and Alex Voorhoeve
We discuss two leading theories of distributive justice: egalitarianism and prioritarianism.

The former holds that unchosen inequality is in itself bad because it is unfair; the latter
denies that inequality is in itself bad, but holds that a given increment in well-being has
greater moral value, the lower the level of well-being from which it takes place. We argue
that the most plausible versions of these views are “hybrids”: they are concerned with
both people’s expected well-being and their final well-being. We also argue that such
hybrid egalitarianism is superior to a hybrid prioritarianism because it more fully satisfies
a key requirement of distributive justice: respect for both the unity of the individual and
the separateness of persons.
Keywords: egalitarianism, prioritarianism, distributive justice, separateness of persons, risk
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.
Page 1 of 27

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 (p. 66)
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
Page 2 of 27

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 (p. 67)
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)
Table 3.1 Final well-being for the Two-Child Case with Certainty
Ann Ben
City 60 39
Suburb 70 30
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-
Page 3 of 27

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 (p. 68)
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.
Page 4 of 27

(1995: 23)
(p. 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
Parfit argues that many people who may take themselves to be egalitarians
(p. 70)
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
Page 5 of 27

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.
(p. 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 (p. 72) 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
Page 6 of 27

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.
Table 3.2 Final and expected well-being in the One-Child Case
Final well-being Expected well-being
p = 0.5 p = 0.5
Cathy able Cathy disabled
City 60 39 49.5
Suburb 70 30 50
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 (p. 73)
which inequality is not an issue, it is reasonable to accord equal moral weight to equally
Page 7 of 27

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
We emphasize that this second criticism of final-well-being prioritarianism is

(p. 74)
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
Page 8 of 27

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. (p. 75)
Page 9 of 27

Table 3.3 Final well-being for the Intra- versus Interpersonal Case
One benefited in suburb One burdened in suburb

(p = 0.5) (p = 0.5)
Denise Edmund Denise Edmund
City 60 60 60 60
Suburb in 70 + d 60 50 60
intrapersonal
scenario
Suburb in 70 + d 60 60 50
interpersonal
scenario
Page 10 of 27

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)
Unlike utilitarianism, final-well-being prioritarianism is sensitive to how well lives

(p. 76)
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
Page 11 of 27

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-Child 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:
(p. 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) t is insensitive to the availability of prudential justifications;

(ii) t is insensitive to the existence of competing claims.
Page 12 of 27

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 (p. 78)
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).
Page 13 of 27

Table 3.4 Final well-being for the Anticorrelated Case
Frank disabled Gwyneth disabled

(p = 0.5) (p = 0.5)
Frank Gwyneth Frank Gwyneth
City 39 60 60 39
Suburb 30 70 70 30
Page 14 of 27

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
(p. 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
Page 15 of 27

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 (p. 80) 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,
Page 16 of 27

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) failure to adequately respond to the presence of prudential justifications;

(ii) failure to adequately respond to the presence of competing claims;
(iii) failure to respond to unfair inequality.
We shall now examine to what extent egalitarian views can avoid these objections.
Page 17 of 27

Table 3.5 Final well-being for the Correlated Case
Helen and Isaac able Helen and Isaac disabled

(p = 0.5) (p = 0.5)
Helen Isaac Helen Isaac
City 60 60 39 39
Suburb 70 70 30 30
Page 18 of 27

(p. 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 (p. 82) 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.
Page 19 of 27

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.
(p. 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
Arneson, R. (1999). “Egalitarianism and Responsibility.” The Journal of Ethics 3: 225–47.
Page 20 of 27

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

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.
(p. 84) 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.
Page 21 of 27

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

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.
Page 22 of 27

Segall, S. (2015). “In Defense of Priority (and Equality).” Philosophy, Politics and
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.
Ubel, P. A., G. Loewenstein, D. Scanlon, and M. Kamlet (1996). “Individual Utilities

(p. 85)
are Inconsistent with Rationing Choices: A Partial Explanation of Why Oregon’s Cost-
Effectiveness List Failed.” Medical Decision-Making 16: 108–16.
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.
Notes:
(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.
(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
Page 23 of 27

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.
(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 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
Page 24 of 27

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.)
(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.).
(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.
Page 25 of 27

(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.
(11) These differences are also apparent in the Intra- versus Interpersonal Case.
(12) For further discussion, see Fleurbaey and Voorhoeve (2013) and Frick (2013).
(13) 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).
(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.
(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 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
Page 26 of 27

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.
Michael Otsuka
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.
Alex Voorhoeve
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.
Page 27 of 27

Sufficiency and Needs-Based Approaches

Gillian Brock
This chapter addresses several debates concerning needs and distributive justice, and
argues that the debate between egalitarians and sufficientarians has not been as
productive as it could have been. The issue is not whether we should prefer equality or
sufficiency, but rather what is the proper role for each in an account of distributive
justice. Several prominent debates concerning needs and distributive justice are also
addressed. Among needs-theorists there is convergence on what counts as a need and the
needs claims that give rise to appropriate claims 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.
Keywords: needs, human needs, sufficiency, distributive justice, equality, according to needs, justice, David Miller,
Harry Frankfurt, responsibilities
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
Page 1 of 25

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
(p. 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 justice 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.
Page 2 of 25

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
(p. 88) 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 justice. 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
Page 3 of 25

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?
The origins of the contemporary debates between sufficientarians and egalitarians

(p. 89)
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
Page 4 of 25

… 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 (p. 90)
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.
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
Page 5 of 25

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
Respect,” Frankfurt’s position is much less persuasive. Frankfurt makes several
(p. 91)
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.”
Page 6 of 25

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.
The proponent of sufficiency could respond to this attempted defense by pointing

(p. 92)
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
Page 7 of 25

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.
(p. 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
Page 8 of 25

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 (p. 94)
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
Page 9 of 25

with a capabilities-based approach. And the third question often invites analysis
(p. 95)
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 (p. 96) 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
Page 10 of 25

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.
A more complex version is Crisp’s (2003), which we could call “prioritarian

(p. 97)
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
Page 11 of 25

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
justice 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 (p. 98) 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
Page 12 of 25

Another way to embed sufficientarianism within a full account of distributive justice 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.
4.5 Which Needs Warrant Normative

(p. 99)
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
Page 13 of 25

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 (p. 100) 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
Page 14 of 25

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’ (p. 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
Page 15 of 25

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’ (p. 102) 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
Page 16 of 25

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
(p. 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.
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
Page 17 of 25

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 (p. 104) 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.
Page 18 of 25

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.
So, when goods are fully divisible and transferable, we might employ a strategy of
(p. 105)
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
Page 19 of 25

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-of- Justice 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,
(p. 106) 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 justice
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
Page 20 of 25

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.
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.
Page 21 of 25

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
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

Miller, D. (1999). Principles of Social Justice. Cambridge, MA: Harvard University Press.
Nussbaum, M. (1998). “Aristotelian Social Democracy,” in G. Brock (ed.) Necessary

135–56.
Page 22 of 25

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.
Reader, S. (ed.) (2005). The Philosophy of Need. Cambridge University Press,

(p. 108)
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.
Notes:
(1) For a good overview, see Patricia Springborg (1981).

Page 23 of 25

(2) 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).
(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.
(4) For excellent comprehensive treatment of the debate between sufficientarians,

egalitarians, and prioritarians, see Paula Casal (2007) and Liam Shields (2012).
(5) Some of this is argued for in Brock (2009), especially Ch. 12.
(6) 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).
Gillian Brock
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
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.
Page 24 of 25

Page 25 of 25

The Capability Approach

Ingrid Robeyns
This chapter analyzes the contribution of the capability approach to the literature on
distributive justice. The capability approach in itself does not provide a full theory of
distributive justice, but rather argues that the metric of distributive justice should be
functionings and/or capabilities. The chapter critically analyzes various issues that need
addressing when we adopt this metric, such as the questions of which capabilities should
be selected, and how they should be aggregated in order to make interpersonal
comparisons of advantage. Comparisons with other metrics of justice are also discussed,
such as Rawls’s social primary goods and welfarist metrics. The chapter concludes by
arguing that we should think of the capability approach to justice as a family of theories,
and describes which theoretical modules are needed for a full capabilitarian theory of
justice.
Keywords: functionings, capabilities, capability approach, metric of justice, Martha Nussbaum, John Rawls,
Amartya Sen
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
Page 1 of 21

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 (p. 110) 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 justice 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
Page 2 of 21

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.
5.2 Distributive Justice and the Metric of

(p. 111)
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
Page 3 of 21

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.
5.3 Functionings, Capabilities, and the

(p. 112)
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
Page 4 of 21

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 (p. 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 (p. 114) ends if
Page 5 of 21

all people have the same capacities or powers to convert those means into equal
capability sets. Capability scholars believe that these interpersonal 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 justice
(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 (p. 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.
Page 6 of 21

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: 215–16)
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 justice.
Sen believes that the more general problem with the use of primary goods is that it
(p. 116) cannot adequately deal with the pervasive differences among people. Primary
Page 7 of 21

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 (p. 117)
Page 8 of 21

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
justice 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.
Page 9 of 21

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
(p. 118) 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.
Sen and Nussbaum have said very little on the problem of expensive taste—
(p. 119)
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
Page 10 of 21

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 (p. 120) 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.
Page 11 of 21

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. (p. 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
Page 12 of 21

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
(p. 122) higher education accessible. Nelson (2008) concludes that there is no way in
which the capability approach can be neutral.
Page 13 of 21

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
(p. 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.
Page 14 of 21

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 nonideal 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 (p. 124)
capabilities on Nussbaum’s list. Which capabilities should justice-seeking organizations or
governments prioritize?
Page 15 of 21

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 (p. 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.
Page 16 of 21

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 (p. 126) 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
Page 17 of 21

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.
Page 18 of 21

(p. 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.
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

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
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.

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.

Nussbaum, M. (2011). Creating Capabilities: The Human Development Approach.

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.
Page 19 of 21

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).
(p. 128) 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: 359–71.
Williams, A. (2002). “Dworkin on Capability.” Ethics 113: 23–39.
Page 20 of 21

Wolff, J. and de-Shalit, A. (2007). Disadvantage. Oxford: Oxford University Press.
Notes:
(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).
(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.
(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.
Ingrid Robeyns
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
Page 21 of 21

Libertarianism, Left and Right

Joseph Mazor and Peter Vallentyne
This chapter explores the libertarian account of distributive justice. It explains the self-
ownership thesis and then considers criticisms relating to indeterminacy, small
incursions, and enforceable duties of assistance. It then turns to questions relating to
natural resources, including whether they are initially owned, the conditions under which
they may be appropriated, and the right vs left libertarian debate over whether or not
they should ultimately be owned in an egalitarian manner. The chapter then lays out the
libertarian account of just economic activity and considers criticisms based on the role of
luck, the moral limits of markets, discrimination, paternalism, unjust prices, alternative
notions of freedom, the plight of the desperately poor, and market failures. The chapter
then concludes by raising important topics for future research, including the moral status
of minors, future people, global justice, property rights in non-divisible natural resources,
and rectificatory justice.
Keywords: libertarianism, left libertarianism, right libertarianism, ownership, self-ownership, natural resources,
appropriation
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.
Page 1 of 26

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.
(p. 130) 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-Ownership

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.
Page 2 of 26

We will not take a position on which property rights are precisely necessary and
(p. 131)
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.
(p. 132) 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-
Page 3 of 26

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 (p. 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.
Page 4 of 26

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.
6.2.5 Full Self-Ownership and Small Incursions or Risks of

(p. 134)
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
Page 5 of 26

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 (p. 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 (p. 136)
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
Page 6 of 26

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.
Cohen (1995: 102–3) also considers a different type of initial ownership: Namely,
(p. 137)
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.
Page 7 of 26

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).
(p. 138) 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.
Page 8 of 26

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.
Nozick’s account of permissible natural resource appropriation is open to a wide

(p. 139)
variety of objections (Kymlicka 2001: 111–21), 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
Page 9 of 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.
(p. 140) 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
Page 10 of 26

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
(p. 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.
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
Page 11 of 26

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. (p. 142)
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.
(p. 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.
Page 12 of 26

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 (p. 144) 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
Page 13 of 26

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 (p. 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
Page 14 of 26

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).
It is difficult to overstate the economic problems that such market failures could
(p. 146)
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 justice 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
Page 15 of 26

these topics for future research, our focus will be on raising important questions and on
providing the reader with references to further reading.
(p. 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
(p. 148) 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
Page 16 of 26

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 (p. 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.
Page 17 of 26

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.
Kramer, M. H., Simmonds, N. E., and Steiner, H. (1998). A Debate over Rights:
(p. 150)
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.
Page 18 of 26

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.
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.
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.
Page 19 of 26

Quong, J. (2011). “Left-Libertarianism: Rawlsian not Luck Egalitarian.” Journal of Political

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.
Satz, D. (2010). Why Some Things Should not be for Sale: The Moral Limits of
(p. 151)
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
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.
Ethics 112(3): 529–57.
Vallentyne, P. (2003a). “Justice in General,” in P. Vallentyne (ed.) Equality and Justice.

New York: Routlege, pp. xi–xviii.
Page 20 of 26

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.
Notes:
(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.
(3) For a critique of the libertarian commitment to self-ownership, see Lippert-Rasmussen

2008.
(4) 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.
Page 21 of 26

(7) We set aside here the important question of what constitutes appropriate
compensation.
(8) 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.
(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 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.
(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.
(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.
Page 22 of 26

(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.
(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?
(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.
(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.
(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
Page 23 of 26

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.
(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.
(32) For an excellent discussion, see Wall 2009.
(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.
(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.
(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).
Page 24 of 26

(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.
(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.
(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.
Joseph Mazor
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.
Peter Vallentyne
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.
Page 25 of 26

Page 26 of 26

Desert-Based Justice

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, until the
1970s and 1980s, however, perhaps under the influence of Rawls’s desert-less theory,
desert was largely cast out of discussions of distributive justice. Now it is making a
comeback. This chapter considers recent research on the concept of desert, debate about
the conditions for desert, arguments for and against its requital, and connections
between desert and other distributive ideals. It suggests that desert-sensitive theories of
distributive justice, despite the challenges they face, have a promising future.
Keywords: desert, distributive justice, luck egalitarianism, neutrality, Rawls, responsibility, retributive justice
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.
Page 1 of 24

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
(p. 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
Page 2 of 24

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 (p. 154)
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.
Page 3 of 24

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.
(p. 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,
Page 4 of 24

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
(p. 156) 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
Page 5 of 24

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 (p. 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 (p. 158) encourages moral behavior and discourages immoral behavior, by
attaching rewards and penalties, respectively, to them.
Page 6 of 24

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
(p. 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:
Page 7 of 24

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,
(p. 160) 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.
Page 8 of 24

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 (p. 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
Page 9 of 24

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
retributive theory would be “intolerably revisionist” (2003: 69), but a desert-free
(p. 162)
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
Page 10 of 24

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).
It is plausible to regard the criminal law as codifying normative expectations

(p. 163)
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
Page 11 of 24

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.
It might be objected that, even if what I have said so far is right, it barely dents
(p. 164)
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
Page 12 of 24

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 (p. 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
Page 13 of 24

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 (p. 166)
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
Page 14 of 24

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 (p. 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
Page 15 of 24

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
(p. 168) 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.
Page 16 of 24

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.
She says that “it is not for society to decide which efforts are more, less, or
(p. 169)
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,
Page 17 of 24

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 (p. 170)
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
Page 18 of 24

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
(p. 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. (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.
Page 19 of 24

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
Chan, J. (2000). Legitimacy, Unanimity, and Perfectionism. Philosophy and Public Affairs
29(1): 5–42.
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
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.
Feldman, F. (1995). “Desert: Reconsideration of Some Received Wisdom.” Mind

(p. 172)
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.
Page 20 of 24

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
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.
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. (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.
Page 21 of 24

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.
(p. 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:
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.
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. (p. 174)
Notes:
(1) This section draws on Moriarty (2009).
Page 22 of 24

(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.
(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).
(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.
(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.
(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).
(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.
(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
Page 23 of 24

which is due to (dis)advantageous natural and social factors (1971: 311–12). For a critical
discussion of this argument, see Moriarty (2005).
(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).
Jeffrey Moriarty
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.
Page 24 of 24

Retributive Justice

Retributive Justice
Larry Alexander
The topic of this chapter is the relationship between retributive justice and distributive
justice. The author expounds his view that retributive justice should be noncomparative,
and that the currency of retributive desert should be suffering. Some theories of
distributive justice employ desert as a basis for distribution, whereas other theories of
distributive justice do not. The author explains his belief that retributive justice relies on
the notion of negative desert, but acknowledges that there are problems if not only
negative desert is to be punished but positive desert is to be rewarded. The challenge is
how to integrate retributive justice into the different theories of distributive justice.
Keywords: retributive justice, distributive justice, negative desert, positive desert, comparative justice,
noncomparative justice
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
Page 1 of 20

Retributive Justice
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
(p. 178) 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
Page 2 of 20

Retributive Justice
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].)
Although I shall not have the space to defend adequately the positions I take on
(p. 179)
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
(p. 180) 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.
Page 3 of 20

Retributive Justice
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 (p. 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
Page 4 of 20

Retributive Justice
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 (p. 182)
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
Page 5 of 20

Retributive Justice
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 justice 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 justice, 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 (p. 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 justice)
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
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 justice 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
Page 6 of 20

Retributive Justice
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
8.4 Meshing Retributive and Distributive

(p. 184)
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.
Page 7 of 20

Retributive Justice
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?
Now one response to this apparent perversity is to attribute it to the use of jail
(p. 185)
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
Page 8 of 20

Retributive Justice
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 (p. 186) 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

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
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.
Page 9 of 20

Retributive Justice
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
justice 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, (p. 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
For our purposes here, however, these difficulties are somewhat beside the point.
(p. 188)
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.
Page 10 of 20

Retributive Justice
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.
(p. 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
8.4.3 Retributive Justice in Desert-Sensitive Patterned Theories of

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?
Page 11 of 20

Retributive 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 (p. 190) 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.)
Page 12 of 20

Retributive Justice
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?
Circumstantial luck should not affect negative desert. We may think Joe, had he
(p. 191)
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
Page 13 of 20

Retributive Justice
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 (p. 192)
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,
Page 14 of 20

Retributive Justice
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. (p. 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

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.
Page 15 of 20

Retributive Justice
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.
(p. 194) 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.
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:
Page 16 of 20

Retributive Justice
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
Notes:
(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.
(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.
(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 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
Page 17 of 20

Retributive Justice
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.”
(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.
(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.
(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).
(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.
(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
Page 18 of 20

Retributive Justice
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 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.
(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
Page 19 of 20

Retributive Justice
the desert base for rewards. I will nonetheless discuss the possibility that positive and
negative desert are asymmetrical.
(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.
(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.
Larry Alexander
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.
Page 20 of 20

The Good Society

The Good Society
Stephen Wall
This chapter discusses how considerations of the good, both the human good and the
impersonal good, can inform accounts of distributive justice for modern political
societies. Starting with the assumption that just political societies will aim to promote the
good of their members, it argues that considerations about good human lives are integral
to determining the content of justice. This chapter also argues that, while just political
societies must promote the good of their members, a good society realizes goods beyond
the good of justice. In particular, a good society supports impersonal goods and promotes
fitting attitudes toward them. The point of justice is the human good, but a just society is
perfected by its concern for the impersonal good.
Keywords: justice, pluralism, autonomy, human good, impersonal goods
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.
Page 1 of 20

The Good Society
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 (p. 196) 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.
Page 2 of 20

The 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?
Many philosophers answer this question by appealing to desires. The good for a
(p. 197)
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)
Page 3 of 20

The Good Society
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
(p. 198) 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
I have been discussing the dependence of the human good on human desires.
(p. 199)
Rejecting a strong desire-dependence condition, I have suggested that a weak desire-
Page 4 of 20

The Good Society
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.
This chapter assumes that autonomy is an aspect of the human good. Part of what
(p. 200)
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
Page 5 of 20

The Good Society
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
(p. 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
Page 6 of 20

The Good Society
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
(p. 202) 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.
Page 7 of 20

The Good Society
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,
Fig. 9.1 Autonomy and good lives
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
Page 8 of 20

The Good Society
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.
Our discussion of autonomy and its value point the way toward a response to the
(p. 203)
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
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?
Page 9 of 20

The Good Society
This is the currency question in the philosophical literature on distributive justice.

(p. 204)
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 (p. 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
Page 10 of 20

The Good Society
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
(p. 206) 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
Page 11 of 20

The Good Society
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 (p. 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
Page 12 of 20

The Good Society
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
(p. 208) 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
Page 13 of 20

The Good Society
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 (p. 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
Page 14 of 20

The Good Society
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 (p. 210)
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
Page 15 of 20

The Good Society
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 (p. 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.
Page 16 of 20

The Good Society
Arneson, R. (2006). “Desire Formation and the Human Good.” Royal Institute of
(p. 212)
Philosophy Supplement 59: 9–32.
Christman, J. and Anderson, J. (eds) (2005). Autonomy and the Challenge to Liberalism.
Dworkin, R. (2011). Justice for Hedgehogs. Oxford: Oxford University Press.
Nagel, T. (1979). “The Fragmentation of Value,” in T. Nagel, Mortal Questions.

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.
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): 359–71.
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.
Notes:
(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.
Page 17 of 20

The Good Society
(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.
(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.
(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.
(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.
(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.
(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.
(12) For a response to this objection see Wall 2013.
(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 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).
Page 18 of 20

The Good Society
(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.
(16) This is the issue that divides Dworkin from Cohen.
(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-autonomy-centered 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).
(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.
(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.)
Page 19 of 20

The Good Society
(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.
Stephen Wall
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.
Page 20 of 20

The Ethics of Care

The Ethics of Care
Virginia Hel
The focus of normative political theory in recent decades has been overwhelmingly on
distributive justice. Developed for institutions within national societies, questions of
justice and fairness have also dominated consideration of the global problems that
morality ought to address. For matters of war and peace, just war theory has been
central; for other issues, distributive justice. This “justice-dominated discourse,” greatly
influenced by the work of John Rawls, is now being challenged by the alternative outlook
of the ethics of care. Care ethics began to be developed in the last quarter of the
twentieth century by feminist moral and political theorists, and its development
continues. This chapter looks at this alternative view and some of its implications.
Keywords: ethics of care, global poverty, Kantian ethics, the liberal individual, John Rawls, rights, Sara Ruddick,
utilitarianism, violence
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
justice (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).
Page 1 of 24

The Ethics of Care
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
(p. 214) 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
Page 2 of 24

The Ethics of Care
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 justice 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
and rights. They more appropriately and fundamentally concern other aspects of
(p. 215)
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
Page 3 of 24

The Ethics of Care
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 (p. 216) 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
Page 4 of 24

The Ethics of Care
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 (p. 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
Page 5 of 24

The Ethics of Care
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.
(p. 218) 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
Page 6 of 24

The Ethics of Care
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
(p. 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
Page 7 of 24

The Ethics of Care
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 (p. 220) 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
Page 8 of 24

The Ethics of Care
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.
Care occurs between persons of very unequal power who did not choose each
(p. 221)
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.
Page 9 of 24

The Ethics of Care
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.
Care is based especially on actual experience. Every person has received care
(p. 222)
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
Page 10 of 24

The Ethics of Care
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).
As many of us defending the ethics of care have been saying for some time,
(p. 223)
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)
Page 11 of 24

The Ethics of Care
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 (p. 224)
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).
Page 12 of 24

The Ethics of Care
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 (p. 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
Page 13 of 24

The Ethics of Care
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 (p. 226) 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
Page 14 of 24

The Ethics of Care
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.
(p. 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.
Page 15 of 24

The Ethics of Care
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.
(p. 228) 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 justice 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
Page 16 of 24

The Ethics of Care
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).
(p. 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).
Page 17 of 24

The Ethics of Care
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 (p. 230)
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
Page 18 of 24

The Ethics of Care
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 (p. 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
Page 19 of 24

The Ethics of Care
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
justice of law and the use of force. There would be, in a caring world, far more
(p. 232)
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.

Press.
Page 20 of 24

The Ethics of Care
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.

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:
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.
Page 21 of 24

The Ethics of Care
Jaggar, A. M. (2009c). “Transnational Cycles of Gendered Vulnerability.”

(p. 233)
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.
Page 22 of 24

The Ethics of Care
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.
(p. 234) 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): 158–71.
Tronto, J. C. (2013). Caring Democracy: Markets, Equality, and Justice. New York: New
York University Press.
Notes:
(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
Page 23 of 24

The Ethics of Care
Marian Barnes. (2012). Care In Everyday Life: An Ethic of Care in Practice. (Bristol. UK:
The Policy Press).
(2) See section 10.7.
(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.
Virginia Hel
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.
Page 24 of 24

The Theory and Politics of Recognition

Colin Bird
This chapter investigates the relationship between the so-called ‘politics of recognition’
and the philosophical discussion of principles of distributive justice. It argues that the
literature has failed to distinguish clearly between three forms of recognition potentially
relevant to distributive justice: status-recognition, authenticity-recognition and worth-
recognition. Each of these forms of recognition is explored, and their various possible
links to arguments about the requirements of justice are distinguished and critically
discussed. Against much conventional wisdom, the chapter suggests that models of
recognition built around the recognition of ‘equal status’ need not be problematically
‘difference blind’; that claims about authenticity-recognition have a more tenuous relation
to discussion of (distributive) justice than many suppose; and that disadvantaged
individuals’ need for respectful recognition is not reducible either to claims about their
moral status or to demands that identity be authentically expressed in social discourse.
Keywords: recognition, respect, identity, equality, Axel Honneth, Charles Taylor, Will Kymlicka
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
Page 1 of 23

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
(p. 236) 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
Page 2 of 23

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 (p. 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
Page 3 of 23

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
(p. 238) (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
Page 4 of 23

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.
In what follows, I will consider these various possibilities in the context of status-
(p. 239)
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
Page 5 of 23

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
(p. 240) 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-Recognition
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,
Page 6 of 23

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 (p. 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.
Page 7 of 23

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
truism that individuals’ standing as moral equals must be somehow connected
(p. 242)
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
Page 8 of 23

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 (p. 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
Page 9 of 23

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 (p. 244) 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
Page 10 of 23

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,
one’s identity is not a simple datum, frozen in time and space, waiting to be
(p. 245)
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.
Page 11 of 23

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.
In arguments along these lines, and as our earlier discussion of stereotypes

(p. 246)
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
Page 12 of 23

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 (p. 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
Page 13 of 23

(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 (p. 248) 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
Page 14 of 23

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.
(p. 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
Page 15 of 23

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 (p. 250) 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-Recognition
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
Page 16 of 23

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
(p. 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?
Page 17 of 23

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 justice 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 (p. 252)
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
Page 18 of 23

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 (p. 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.
Page 19 of 23

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.

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.
Brown, W. (1995). States of Injury: Power and Freedom in Late Modernity.

(p. 254)
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.

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.
Page 20 of 23

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.
Page 21 of 23

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

Taylor, C. (1985b). “What’s Wrong with Negative Liberty”, Philosophical Papers vol. II.
Taylor, C. (1994). Multiculturalism: Examining the Politics of Recognition.

(p. 255)
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.
University Press. (p. 256)
Colin Bird
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.
Page 22 of 23

Page 23 of 23

Distributive Justice and Human Nature

Paula Casal
This chapter examines the relation between distributive doctrines and human nature. It
reviews various responses to the conservative view that many progressive social reforms
are doomed because of human nature. Some responses present a different view of human
nature while others stress that human nature can be modified because it is the product of
nurture or because it can be socially or biomedically altered. The chapter also offers an
evolutionary approach to values like ‘liberté, égalité, fraternité’ and discusses whether
human nature could have more than merely instrumental relevance to distributive justice.
For example, our endorsement, and interpretation, of moral demands regarding human
rights, human flourishing, and human capabilities, or even regarding the principles of
sufficiency, equality and priority, may depend on assumptions about human nature.
Keywords: human nature, human rights, evolution, liberty, fraternity, equality, priority, sufficiency
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?
Page 1 of 26

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 justice 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.
A different appeal to nature, made to deny that distributive justice favours

(p. 260)
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.
Page 2 of 26

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 (p. 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 Liberté
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.
(p. 262) 12.2.2 Fraternité
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
Page 3 of 26

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 (p. 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-
Page 4 of 26

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 pipefish 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 and Gintis 2011), (p. 264) 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
Page 5 of 26

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 (p. 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
Page 6 of 26

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 (p. 266)
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:
Page 7 of 26

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 (p. 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
Page 8 of 26

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’ (p. 268)
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
Page 9 of 26

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 (p. 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.
Page 10 of 26

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 (p. 270) 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
Page 11 of 26

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.
(p. 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
Page 12 of 26

(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 (p. 272)
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
Page 13 of 26

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 (p. 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
Page 14 of 26

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
(p. 274) 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,
Page 15 of 26

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
(p. 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
Page 16 of 26

others who have assets other than youth. So the relevant comparison is (p. 276) 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
Page 17 of 26

how to secure their protection (see Brighouse and Swift 2014). For example, they
(p. 277)
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
Page 18 of 26

the evidence of early women focusing on agriculture whilst men continued

(p. 278)
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 (p. 279)
Page 19 of 26

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.

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.
Page 20 of 26

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.
Casal, P. (2013). ‘Sexual Dimorphism and Moral Enhancement’. Journal of Medical

(p. 280)
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.
Page 21 of 26

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
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.
Milanovic, B. (2011). The Haves and the Have-Nots: A Brief and Idiosyncratic
(p. 281)
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.
Page 22 of 26

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.
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:

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.
Page 23 of 26

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.
Thornhill, R. and Palmer, C. T. (2000). A Natural History of Rape. Cambridge, MA

(p. 282)
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.
Page 24 of 26

Zerjal, T. et al. (2003). ‘The Genetic Legacy of the Mongols’. American Journal of Human
Genetics 72: 717–21.
Notes:
(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.
(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.
(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.
(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).
(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.
(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.
Paula Casal
Paula Casal is ICREA Professor at Pompeu Fabra University, Barcelona, associate

editor 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.
Page 25 of 26

Page 26 of 26

Political and Distributive Justice

Chad Van Shoelandt and Gerald Gaus
To most philosophers, unmet claims based on distributive justice imply a political

injustice—some have a complaint of justice against their political system. This article
explores a variety of views about how this connection may be grounded or qualified:
political institutions may be one tool among others to realize an independent good,
distributive principles might regulate the distributive activities of political institutions, or
distributive principles might apply in light of a special relation of a political institution
and its members. We also consider a view prevalent in the social contract tradition that,
in light of reasonable disagreement, one cannot demand that shared political institutions
conform to one’s own contentious distributive theory: members must seek terms with
which all can live, even though such terms may not be anyone’s most preferred
possibility.
Keywords: disagreement, distributive justice, institutions, political justice, social contract
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
Page 1 of 24

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
(p. 284) 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.
Page 2 of 24

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)
(p. 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
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 justice
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)
Page 3 of 24

(p. 286) 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 (p. 287) though they could is a defect of people, not of the theory.
Page 4 of 24

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:
(p. 288)
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
Page 5 of 24

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 justice. 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 (p. 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.
Page 6 of 24

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 (p. 290) 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
Page 7 of 24

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
(p. 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
Page 8 of 24

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)
This principle regulates decisions and actions, and is not in the first instance a
(p. 292)
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
Page 9 of 24

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.
13.1.3.3 The Existence of Political Institutions and the Value of

(p. 293)
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 (p. 294) for
Page 10 of 24

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 (p. 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
Page 11 of 24

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.
If Alf continues to hold a well-reasoned belief that DJ is violated, then (at least pro
(p. 296)
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
Page 12 of 24

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,” (p. 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,
Page 13 of 24

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]
(p. 298) 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,
Page 14 of 24

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
is what public morality requires. If the political process determines that the
(p. 299)
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 justice 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?
Page 15 of 24

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 (p. 300)
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-
Page 16 of 24

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 (p. 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
Page 17 of 24

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 (p. 302) 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
Page 18 of 24

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
(p. 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 justice, 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
Page 19 of 24

Christiano, T. (2008). The Constitution of Equality: Democratic Authority and its Limits.
Cohen, G. A. (2000). If You’re an Egalitarian, How Come You’re So Rich? Cambridge, MA:
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.
(p. 304) 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:
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.
Page 20 of 24

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.
Pettit, P. (2012). On the People’s Terms: A Republican Theory and Model of Democracy.
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.)

Press.
Rawls, J. (2001). The Law of Peoples. Cambridge, MA: Harvard University Press.
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.
Page 21 of 24

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.
Thrasher, J. and Vallier, K. (2013). “The Fragility of Consensus: Public Reason,

(p. 305)
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:
Notes:
(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.”
(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.
(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).
Page 22 of 24

(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).
(12) Though Cohen does not think this need mean a specifically political or institutional
project, given his support for an egalitarian ethos.
(13) See Hayek (1976: Ch. 10), particularly pp. 117 and 122. Cf. Lister (2013).
(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).
(15) Cf. Scheffler (2006: 124–5).
(16) 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.
(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).
(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).
Page 23 of 24

(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.
(23) Cf. Rawls (2005: 389).
Chad Van Shoelandt
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.
Gerald Gaus
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.
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).
Page 24 of 24

Consequentialism, Deontology, Contractualism, and Equality

Consequentialism, Deontology, Contractualism, and
Equality
Jonathan Quong
This chapter advances two main claims. First, that the distinction between
consequentialism and deontology, although widely adopted, is illusory and only serves to
obscure some of the genuine disputes underlying central debates in distributive justice.
Second, that 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—this may be a mistake, since this construal makes
each view less plausible than it otherwise might be. Instead, the chapter proposes 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 a contractualist account of what it is
for something to be just or unjust.
Keywords: agent-centered, agent-neutral, consequentialism, contractualism, democratic egalitarianism,

deontology, equality, luck egalitarianism, teleology
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.
Page 1 of 22

(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.
But many theories of distributive justice—for example, those that include

(p. 307)
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
Page 2 of 22

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.
The chapter is structured as follows. Section 14.1 explains why the distinction
(p. 308)
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
Page 3 of 22

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
(p. 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.
Page 4 of 22

(p. 310) 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 (p. 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
Page 5 of 22

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 (p. 312)
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
Page 6 of 22

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: justice
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 justice is achieved when
the equalisandum has been distributed in accordance with the (p. 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
Page 7 of 22

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).
In what follows, I offer an alternative map of the debate between democratic and
(p. 314)
luck egalitarians. The map is organized around three questions:

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.
Page 8 of 22

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 justice
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 (p. 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.
Page 9 of 22

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 (p. 316) 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
Page 10 of 22

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 (p. 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
justice. Proponents of democratic egalitarianism might argue that the current generation
Page 11 of 22

does, somehow, stand in a relationship with the future generation, and this relationship
(p. 318) 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:

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
(p. 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
Page 12 of 22

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 (p. 320) 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
Page 13 of 22

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 (p. 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.
Page 14 of 22

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
justice), 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
to the second question—what makes some things unjust—will end up with a view
(p. 322)
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 justice.
Page 15 of 22

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 justice.
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 (p. 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.
Page 16 of 22

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
what makes something unjust cannot be sustained, at least not when
(p. 324)
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
Page 17 of 22

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
There is, however, an important disanalogy between physical maps and

(p. 325)
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.
Page 18 of 22

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.

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. (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.
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.
Parfit, D. (2011). On What Matters. Vol. 1. Oxford: Oxford University Press.
Quong, J. (2011). “Left-Libertarianism: Rawlsian not Luck Egalitarian.” Journal of Political

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.
Page 19 of 22

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.
Tadros, V. (2011). The Ends of Harm: The Moral Foundations of Criminal Law. Oxford:
Tan, K. (2012). Justice, Institutions, and Luck: The Site, Ground, and Scope of Equality.
Notes:
(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.
(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).
(4) Sections 14.5 and 14.6 will offer some reasons to doubt this is, in fact, the best way to
understand contractualism.
(5) This point is made in Tadros (2011: 35).
(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
Page 20 of 22

converge. For an example of this latter response, see Stemplowska 2011. A notable
exception to this latter trend, however, is Tan (2012).
(7) Tan is an interesting exception to this general luck egalitarian view. See Tan (2012:
Chs 2–3).
(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).
(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).
(13) For discussion of these points, see Arneson (2011).
(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?
(16) Also see Anderson (1999: 322); or Tan (2012: 97).
Page 21 of 22

(17) 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.
(18) Anderson (1999); and Hinton (2001) both emphasize this idea.
(19) Consider, for example, how Rawls understands the principle of redress (see Rawls
1999: 86–7).
(20) For a different way of reconciling the two views, see Schemmel (2012).
(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.
(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.
(23) Though Derek Parfit rejects this view in Parfit (2011).
Jonathan Quong
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).
Page 22 of 22

Ideal Theory

Ideal Theory
David Schmidtz
There is a particular kind of ideal theory that does not work: namely, theorizing about
ideally just worlds. Such theorizing fails to bear on what we have reason to regard as an
ideal response here and now, and instead bears only on what would be ideal if we were to
depart far enough from the human condition as it really is. An alternative kind of ideal
theory, realistic idealism, is more useful: namely, theorizing about ideal responses to a
sober assessment of circumstances that demand a response here and now. Where there
are facts, where facts are 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.
Keywords: political idealism, political realism, ideal justice, Bernard Williams, political moralism, utopia,
compliance problem, justice, mutual advantage
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
Page 1 of 21

Ideal Theory
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)
As Rawls conceded, “Obviously the problems of partial compliance theory are the
(p. 328)
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
Page 2 of 21

Ideal Theory
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
Realistic idealism is theory about ideals, but does not start with ideals. Realistic
(p. 329)
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,
Page 3 of 21

Ideal Theory
there is no discernible defect in my plan to make lasagna for our dinner tonight. (p. 330) 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 (p. 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.
Page 4 of 21

Ideal Theory
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, (p. 332)
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
Page 5 of 21

Ideal Theory
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 (p. 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
Page 6 of 21

Ideal Theory
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
about what counts as an equal share, there is not always a settled truth about
(p. 334)
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
justice 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.
Page 7 of 21

Ideal Theory
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.
We can go badly astray if we strive for what Rawls called a “systematic grasp of
(p. 335)
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.
Page 8 of 21

Ideal Theory
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 (p. 336)
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.
Page 9 of 21

Ideal Theory
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.
One possible view here is that an ideal of unconditional giving “may play an
(p. 337)
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
Page 10 of 21

Ideal Theory
asking too much, but as openly inviting the free-riding that unconditional giving openly
(p. 338) 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.
We might say reciprocity is realistic while unconditional cooperation is utopian as

(p. 339)
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
Page 11 of 21

Ideal Theory
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
(p. 340) 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
Page 12 of 21

Ideal Theory
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.
If strict compliance is part of an ideally just society’s essence, that precludes

(p. 341)
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)—
Page 13 of 21

Ideal Theory
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 (p. 342)
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
Page 14 of 21

Ideal Theory
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 (p. 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
Page 15 of 21

Ideal Theory
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.
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.
Page 16 of 21

Ideal Theory
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.
Forcehimes, A. T. and Talisse, R. B. (2013). “Clarifying Cohen: A Response to Jubb

(p. 344)
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.

Page 17 of 21

Ideal Theory
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.
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.
Sleat, M. (2014). “Realism, Liberalism, and Non-Ideal Theory.” Political Studies.

(p. 345)
64: 27–41.
Spencer, H. (1871). Social Statics. New York: Appleton.

Page 18 of 21

Ideal Theory
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, &

Williams, B. (2005). In the Beginning was the Deed. Princeton, NJ: Princeton University
Press.
Notes:
(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.
(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 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
Page 19 of 21

Ideal Theory
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.
(4) See also Ronzoni (2012).
(5) See Larmore (2013) and Part 1 of Schmidtz (2006).
(6) 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.
(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).
(12) On the imperfect duty of reciprocity, see Part 3 of Schmidtz (2006).
(13) See Sleat (2014), and Jubb (2012).
(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).
(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).
Page 20 of 21

Ideal Theory
(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.
(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.
(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.
(22) I thank Robert van’t Hoff for the thought.
(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.
David Schmidtz
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.
Page 21 of 21

Constructivism, Intuitionism, and Ecumenism

Aaron James
Constructivism and intuitionism are often seen as opposed methods of justification in

political philosophy. An “ecumenical” view sees them as different but unopposed: each
style of reasoning can yield fundamental principles, for different questions of distributive
justice, and we can rightly take up different questions, with different, equally valid,
theoretical objectives, in hopes of cultivating a thousand blooming flowers. This chapter
develops this position with special interest in Rawls’s constructivism, his treatment of
reflective equilibrium, self-evidence, and “moral geometry,” and his evolving dialogue
with the intuitionist Henry Sidgwick. Rawls’s main difference from Sidgwick lies in the
way he frames the question of right or justice in the first instance. This brings out both
the possibility and the attractions of the ecumenist conception in political philosophy.
Keywords: constructivism, intuitionism, distributive justice, justification, self-evidence, moral geometry, reflective
equilibrium, John Rawls, G. A. Cohen, Henry Sidgwick
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
Page 1 of 22

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.
In this discussion, I elaborate this neglected position. I pay special interest to

(p. 347)
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
justice 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 (p. 348) 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.
Page 2 of 22

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.
(p. 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.
Page 3 of 22

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-Lake 2002: 18–23). The fairness (p. 350) 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)
Page 4 of 22

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.”
(p. 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
Page 5 of 22

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 (p. 352) 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.
Page 6 of 22

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 (p. 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
Page 7 of 22

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
(p. 354) (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 justice
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.
Page 8 of 22

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 (p. 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
Page 9 of 22

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 (p. 356) 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.
Page 10 of 22

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 (p. 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
Page 11 of 22

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 (p. 358) 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”:
Page 12 of 22

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
(p. 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).
Page 13 of 22

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
(p. 360) 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
Page 14 of 22

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 (p. 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).
Page 15 of 22

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 (p. 362)
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
Page 16 of 22

1999b: 306). Yet this remark is easily regarded as passing rhetoric. Certainly little
(p. 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
(p. 364) of distributive justice that simply aren’t about the possibility of autonomous
Page 17 of 22

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.
Page 18 of 22

The argument from autonomy thus turns out to be necessary, but only
(p. 365)
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.
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

Raphael, D. D. (1991). British Moralists, 1650-1800. Volume 1: Hobbes–Gay Indianapolis,

IN/Cambridge, MA: Hackett Publishing Co.
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.)

Page 19 of 22

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.
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.
Stratton-Lake, P. (ed.) (2002). Ethical Intuitionism: Re-Evaluations. Oxford: Oxford

(p. 366)
University Press.
Weithman, P. (2011). Why Political Liberalism? On John Rawls’s Political Turn. New York:
Williams, A. (2008). “Justice, Incentives, Constructivism.” Ratio 21(4): 476–93.
Notes:
(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.
(2) Rawls (1971: 51 n26), says as much, with reference to Schneewind (1963).
(3) 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.
(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.
(6) As, for example, in James (2012) for the case of the global economy.
(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).
Page 20 of 22

(8) Cohen (2008: 3).
(9) Joshua Cohen (2001) may be read this way.
(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.
(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.
(14) Scanlon (1998) may be read this way.
Aaron James
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.
Page 21 of 22

Page 22 of 22

Conceptual Analysis and Distributive Justice

Matthew H. Kramer
This chapter ponders several understandings of conceptual analysis in the context of

debates over distributive justice. The chapter's first three main sections consider the
concept/conception distinction in its multilayered complexity in a couple of prominent
recent accounts of justice. The chapter explores how those theories of justice unfold over
several levels of increasing specificity. Thereafter, the chapter takes up the vexed
question whether expositions of the concept of justice can ever be austerely analytical or
formal rather than morally value-laden. After distinguishing between value-independence
and value-neutrality, the chapter argues that, although some possible accounts of justice
are at least partly value-neutral, no accounts of justice are ever value-independent.
Keywords: justice, distributive justice, John Rawls, Ronald Dworkin, Hillel Steiner, W. B. Gallie, conceptual
analysis, concepts, political philosophy
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.
Page 1 of 20

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 (p. 368)
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 justice
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
Page 2 of 20

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
(p. 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.
Page 3 of 20

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 (p. 370) 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 justice 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
Page 4 of 20

degree, they methodologically assimilate Rawls’s endeavors to those of other

philosophers who work in quite different areas of the discipline.3
(p. 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 (p. 372) 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
Page 5 of 20

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 justice 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.
17.3 Concepts Versus Conceptions:

(p. 373)
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
Page 6 of 20

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,
(p. 374) 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
Page 7 of 20

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 (p. 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
Page 8 of 20

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 (p. 376) 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
Page 9 of 20

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 (p. 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
Page 10 of 20

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
concrete matters in those reflections. Thus, for example, a philosopher who
(p. 378)
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 justice—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
Page 11 of 20

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.
The neutrality advocated by supporters of political liberalism is not value-neutral

(p. 379)
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
Page 12 of 20

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)
Steiner declares that his compossibility test “does exemplary service in filtering
(p. 380)
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
Page 13 of 20

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.
17.7 The Value-Dependence of Theories of

(p. 381)
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)
Page 14 of 20

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 (p. 382) 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:
Page 15 of 20

[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
(p. 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.
Page 16 of 20

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 (p. 384)
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 justice.
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 justice. 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.
Page 17 of 20

Carter, I. (1999). A Measure of Freedom. Oxford: Oxford University Press.
Dworkin, R. (1986). Law’s Empire. London: Fontana Press.
Dworkin, R. (1996). “Objectivity and Truth: You’d Better Believe It.” Philosophy
(p. 385)
and Public Affairs 25: 87–139.
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.
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. (p. 386)
Notes:
(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).
(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 further sought to specify conditions necessary and
Page 18 of 20

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.
(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).
(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.
(6) My thinking about this distinction has been sharpened by some exchanges with Ian
Carter, to whom I am grateful.
(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.
(8) See, for example, Dworkin (1996: 129–39). See also Kramer (2009a: 56, 88–99, 179–
81).
Matthew H. Kramer
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
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).
Page 19 of 20

Page 20 of 20

Gender

Gender
Anca Gheaus
This chapter discusses gender in relation to the most influential current accounts of
distributive justice. There are various disparities in the benefits and burdens of social
cooperation between women and men, in the family, the workplace and more widely.
Which of these disparities, if any, one identifies as indicative of gender injustice will
depend on the theory of distributive justice that one endorses. Theoretical decisions
concerning the role of personal responsibility, the goods whose distribution is relevant for
justice, and the site of justice—institutions-only or individual behaviour, too—all influence
how one thinks about gender justice. New directions for research on this topic are
suggested.
Keywords: gender, women, distributive justice, gender justice, gender inequality, family, discrimination, individual
responsibility, choice, implicit bias
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
Page 1 of 28

Gender
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 (p. 390)
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 justice 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
Page 2 of 28

Gender
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 (p. 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
Page 3 of 28

Gender
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)
This definition has the advantage of capturing what is wrong with many gender
(p. 392)
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,
Page 4 of 28

Gender
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 (p. 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
Page 5 of 28

Gender
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.
Some believe that behaviour caused by gender norms that significantly harms
(p. 394)
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
Page 6 of 28

Gender
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
(p. 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)
Page 7 of 28

Gender
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
(p. 396) 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.
Page 8 of 28

Gender
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 (p. 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 justice—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
Page 9 of 28

Gender
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.
Not only do women have a lesser share of economic and status-related benefits
(p. 398)
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).
Page 10 of 28

Gender
Here is a way to challenge this argument. Many, although not all disadvantages
(p. 399)
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.
Page 11 of 28

Gender
18.2.4 Are all Gendered Behaviours, in Themselves, Equally

(p. 400)
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 (p. 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
Page 12 of 28

Gender
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.
For the time being it is likely that there exist both biological and social facts
(p. 402)
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
Page 13 of 28

Gender
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 justice 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 (p. 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.
Page 14 of 28

Gender
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 (p. 404)
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.
Page 15 of 28

Gender
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 (p. 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
Page 16 of 28

Gender
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 justice, 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 justice.
Recognizing the family as part of the basic structure does not entail that it has to be
(p. 406) 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 justice 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).
Page 17 of 28

Gender
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 (p. 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-
Page 18 of 28

Gender
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 (p. 408)
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
Page 19 of 28

Gender
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.
18.3.3 Is There a Duty to Resist Gender

(p. 409)
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
Page 20 of 28

Gender
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.
18.4 The New ‘The Personal Is Political’:

(p. 410)
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
Page 21 of 28

Gender
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).
If empirical research on implicit bias, stereotype threat, and gender is correct, it

(p. 411)
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 gender-
egalitarian 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.
Page 22 of 28

Gender
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.
(p. 412) 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
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
Page 23 of 28

Gender
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

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.
Gheaus, A. (2009). ‘How Much of What Matters can We Redistribute? Love,

(p. 413)
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.
Page 24 of 28

Gender
Gheaus, A. and Robeyns, I. (2011). ‘Equality-Promoting Parental Leave’. Journal of Social

Gilligan, C. (1982). In a Different Voice: Psychological Theory and Women’s Development.

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:
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
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.
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
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.
Page 25 of 28

Gender
Neufeld, B. (2009). ‘Coercion, the Basic Structure, and the Family’. Journal of Social
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.
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.
Pinker, S. (2008). The Sexual Paradox: Extreme Men, Gifted Women and the Real
(p. 414)
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.
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:
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.
Page 26 of 28

Gender
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.
University Press.
Notes:
(1) This means ‘by definition’.
(2) On the normative importance of ‘the natural’, see also Casal 2018 (Chapter 12 in this
volume).
(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.
(4) For a similar line of reasoning, see Olsaretti (2009).
(5) I dispute that only facts that we can change can constitute injustices (Gheaus 2013).
(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).
(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)
(9) The payment of a universal basic income to all citizens is another policy suspected to
have a similar effect (Robeyns 2001).
(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.
Page 27 of 28

Gender
(11) There is some evidence that this is the case in the United States (Gerson 2010).
Anca Gheaus
Anca Gheaus is Ramon y Cajal researcher at the 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.
Page 28 of 28

The Family

The Family
Colin MacLeod
This chapter examines various recent challenges that the family poses for the articulation
of a comprehensive account of distribution that is broadly egalitarian in character, a topic
largely ignored by many prominent theorists. It considers how opportunities to form
families should be distributed, who should bear the costs of rearing children, and how the
justice-based entitlements of children should be understood in relation to their parents
and other adults. Attention is given to the tension between equality and parental
partiality and to the way in which justice should be concerned with the provision and
distribution of intrinsic goods of childhood.
Keywords: distributive justice, parental partiality, goods of childhood, right to parent, dual parental rights,
egalitarianism, relational equality
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
Page 1 of 25

The Family
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
In what follows, I will consider ways in which contemporary theorists have begun
(p. 416)
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’
Page 2 of 25

The Family
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).
(p. 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 (p. 418) 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
Page 3 of 25

The Family
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 justice 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.
(p. 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
Page 4 of 25

The Family
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 same-
sex couples to parent on the grounds that children have a right to be raised by both a
mother and a father (Sommerville 2007). On this view, same-sex couples may be
(p. 420)
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
Page 5 of 25

The Family
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
(p. 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
Page 6 of 25

The Family
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 (p. 422) 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
Page 7 of 25

The Family
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
being able to nurture children adequately. The standard of parental adequacy can
(p. 423)
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
Page 8 of 25

The Family
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
(p. 424) 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
Page 9 of 25

The Family
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
(p. 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
Page 10 of 25

The Family
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 (p. 426) 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
Page 11 of 25

The Family
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.
19.5 Justice-Based Entitlements of

(p. 427)
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
Page 12 of 25

The Family
that children have to valuable resources and opportunities. This is a matter of identifying
(p. 428) 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 (p. 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
Page 13 of 25

The Family
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 adult-oriented
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.
(p. 430) 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
Page 14 of 25

The Family
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
(p. 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
Page 15 of 25

The Family
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 (p. 432) 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
Page 16 of 25

The Family
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 (p. 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 justice 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
Page 17 of 25

The Family
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 (p. 434) 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
justice 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.
Page 18 of 25

The Family
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.
Arneson, R. (2018). “Dworkin and Luck Egalitarianism: A Comparison,” in S.

(p. 435)
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
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.

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.
Page 19 of 25

The Family
Clayton, M. (2018). “Education,” in S. Olsaretti (ed.) The Oxford Handbook of Distributive

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
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.
Gheus, A. (2018). “Gender,” in Serena Olsaretti (ed.) The Oxford Handbook of

(p. 436)
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.
Routledge.
Kymlicka, W. (2002). Contemporary Political Philosophy: An Introduction. Oxford: Oxford

University Press.
Page 20 of 25

The Family
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
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.
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.
Page 21 of 25

The Family
Nussbaum, M. (2006). Women and Human Development: The Capabilities Approach.

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.
(p. 437) Rawls, J. (1993). Political Liberalism. New York: Columbia 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.
Page 22 of 25

The Family
Woodcock, S. (2009). “Five Reasons why Margaret Somerville is Wrong about Same-Sex
Marriage and the Rights of Children.” Dialogue 48: 1–21.
Notes:
(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).
(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).
(3) See Gheaus 2018 (Chapter 18 in this volume).
(4) Shiffrin’s position is associated with strong anti-natalism but she does not argue that
procreation is always wrong (1999: 139).
(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).
(6) See Clayton 2018 (Chapter 20 in this volume).
(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).
(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.
(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.
(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
Page 23 of 25

The Family
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.
(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.
(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.
(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.
(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.
(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).
Page 24 of 25

The Family
(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).
Colin MacLeod
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).
Page 25 of 25

Education

Education
Matthew Clayton
This chapter discusses the central questions about the content and distribution of
education debated by philosophers in recent years. How should educational opportunity
be distributed between individuals? Should society aim to achieve equal opportunity, or
should it allow departures from equality provided the least advantaged are thereby
helped or everyone enjoys an adequate education? Should society seek to eliminate or
temper only inequalities that are caused by class differences, or also those caused by
individuals’ genetic endowments? Education is not merely a good to be distributed; it is
also a vehicle for shaping individuals’ beliefs and desires. Is it permissible for the political
community to raise children to have a sense of justice, and for the community or parents
to get children to adopt a particular conception of the good life, such as particular
religious convictions? Finally, may parents determine the kind of education that their
child receives?
Keywords: education, educational justice, equality of opportunity, meritocracy, parental rights, moral education,
partiality, family
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
Page 1 of 24

Education
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 (p. 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
Page 2 of 24

Education
second is a conception that permits educational inequalities that are beneficial to less
advantaged citizens.4
(p. 440) 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
Acknowledging the force of this problem, some have entertained the idea of
(p. 441)
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
Page 3 of 24

Education
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 (p. 442) 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
Page 4 of 24

Education
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 (p. 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.
Page 5 of 24

Education
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 (p. 444)
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 justice—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
Page 6 of 24

Education
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, (p. 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
Page 7 of 24

Education
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
(p. 446) 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
Page 8 of 24

Education
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 (p. 447) for
educational policy might vary according to the larger conception of distributive justice 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
Page 9 of 24

Education
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
(p. 448) 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
Page 10 of 24

Education
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
(p. 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: 66–71). 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.
Page 11 of 24

Education
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 (p. 450) 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 curriculum
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
Page 12 of 24

Education
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
(p. 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 (p. 452)
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
Page 13 of 24

Education
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; (p. 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,
Page 14 of 24

Education
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 (p. 454) 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.
Page 15 of 24

Education
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 (p. 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
Page 16 of 24

Education
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 (p. 456) 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
Page 17 of 24

Education
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.
(p. 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
Page 18 of 24

Education
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.
Brighouse, H. and Swift, A. (2006b). ‘Parents’ Rights and the Value of the Family’.
(p. 458)
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.
Page 19 of 24

Education
Casal, P. (2018). ‘Distributive Justice and Human Nature’, in S. Olsaretti (ed.) The Oxford
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.
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.
Page 20 of 24

Education
Macleod, C. (2002). ‘Liberal Equality and the Affective Family’, in D. Archard and C.
Press, pp. 212–30.
Macleod, C. (2010). ‘Justice, Educational Equality, and Sufficiency’. Canadian Journal of

Philosophy 40(sup1): 151–75.
Macleod, C. (2018). ‘The Family’, in S. Olsaretti (ed.) The Oxford Handbook of

(p. 459)
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
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.
Page 21 of 24

Education
Notes:
(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).
(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).
(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).
(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).
(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
Page 22 of 24

Education
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.
(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.
(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).
(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.
(16) For further discussion of different conceptions of parental authority, including his
own ‘creative self-extension’ account, see Macleod (2018).
(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).
(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.
(19) For discussion of the differences between consequentialist and non-consequentialist

conceptions, see Kagan 1989.
Page 23 of 24

Education
Matthew Clayton
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).
Page 24 of 24

Health

Health
Shlomi Segall
The chapter discusses the evolution of theories of justice in health and healthcare. It
traces Norman Daniels’s Rawlsian account, as well as the criticism it received. It then
goes on to discuss two rival theories that sprang in opposition to Daniels’s, namely a
sufficientarian family of theories and luck egalitarian justice in health. Special attention is
devoted to three focal questions: the pattern of justice in health, its currency, and its
scope, that is, the what, how, and who. Under the latter, the chapter discusses the
requirements of global justice in health, and investigates what temporal unit is
appropriate in thinking of just healthcare.
Keywords: health, John Rawls, justice, luck egalitarianism, Norman Daniels
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
Page 1 of 20

Health
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?
In this chapter I will attempt to examine justice in health by weaving both

(p. 461)
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
Page 2 of 20

Health
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 (p. 462) 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 justice
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).
Page 3 of 20

Health
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 (p. 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
Page 4 of 20

Health
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 (p. 464) 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
Page 5 of 20

Health
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 (p. 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.
Page 6 of 20

Health
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
(p. 466) 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
Page 7 of 20

Health
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.
21.4 Alternatives to Rawlsian Justice in

(p. 467)
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
Page 8 of 20

Health
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 (p. 468) 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
Page 9 of 20

Health
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 (p. 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 (p. 470) 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
Page 10 of 20

Health
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.
(p. 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
Page 11 of 20

Health
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 (p. 472) 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.
Page 12 of 20

Health
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 (p. 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
Page 13 of 20

Health
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.
(p. 474) 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
Page 14 of 20

Health
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 (p. 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
Page 15 of 20

Health
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 (p. 476) 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,
Page 16 of 20

Health
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.

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.
Page 17 of 20

Health
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:
Buchanan, A. (1984). “The Right to a Decent Minimum of Health Care.” Philosophy and
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.
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
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.
Lenard, P. T. and Straehle, C. (2012). Health Inequalities and Global Justice.

(p. 478)
Edinburgh: Edinburgh University Press.

Page 18 of 20

Health
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.
Page 19 of 20

Health
Wolff, J. (2012). The Human Right to Health. New York: Norton.
Notes:
(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.
(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.
(3) http://www.infoplease.com/ipa/A0005148.html/ (accessed 20 November 2017).
(4) WHO Health Report 2011 < http://www.who.int/whosis/whostat/2011/en/index.html >.

* figures are for life expectancy at birth for 2009.
Shlomi Segall
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).
Page 20 of 20

Discrimination

Discrimination
Kasper Lippert-Rasmussen
This chapter first defines discrimination. It then argues that discrimination per se is not
unjust and, accordingly, that its relation to distributive injustice is considerably more
complicated than one might initially think. Specifically, the chapter critically assesses
views according to which discrimination is unjust per se, because it i) violates distributive
justice; ii) clashes with a norm of treating people with equal concern and respect; iii)
reduces the discriminatee’s deliberative freedom; iv) is incompatible with a social
relational egalitarian norm to the effect that citizens relate to one another as equals. All
of this is compatible with the claim that some instances of discrimination are seriously
unjust.
Keywords: Elizabeth Anderson, deliberative freedom, discrimination, distributive justice, equality, indirect
discrimination, site of justice, Sophia Moreau, social relations egalitarianism, respect
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
Page 1 of 21

Discrimination
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 (p. 480) 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
Page 2 of 21

Discrimination
customers tend to be sexist with the result that men outperform women in terms of
revenue generated (p. 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
Page 3 of 21

Discrimination
for focusing on this third kind of discrimination, which I shall label group (p. 482)
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’ (p. 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
Page 4 of 21

Discrimination
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.
Another important site of distributive justice is society, where this is understood

(p. 484)
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
Page 5 of 21

Discrimination
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 (p. 485) employees simply on account of gender acts in a way that clashes with
distributive justice, 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
Page 6 of 21

Discrimination
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
(p. 486) 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
Page 7 of 21

Discrimination
statistically correlated with other features, such as certain kinds of criminal behaviour.
While statistical discrimination often takes forms that render it unjust, it (p. 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 (p. 488) 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
Page 8 of 21

Discrimination
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
(p. 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
Page 9 of 21

Discrimination
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 (p. 490) 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?
Page 10 of 21

Discrimination
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
out that a test resulting in a higher percentage of women being admitted actually
(p. 491)
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
Page 11 of 21

Discrimination
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 justice

requires that no one is worse off than others (or significantly worse off than (p. 492)
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 justice 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
Page 12 of 21

Discrimination
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 (p. 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
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!
Page 13 of 21

Discrimination
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 (p. 494)
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
Page 14 of 21

Discrimination
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 (p. 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 justice
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
Page 15 of 21

Discrimination
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
(p. 496) 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.
Page 16 of 21

Discrimination
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.
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.
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.
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.
Page 17 of 21

Discrimination
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.
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:

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.
Notes:
(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.
(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.
Page 18 of 21

Discrimination
(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
(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.
(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).
(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.
(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.
Page 19 of 21

Discrimination
(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).
(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).
(16) Griggs v. Duke Power Co., 401 U.S. 424 (1971).
(17) Admittedly, interpretations of ‘disproportionate disadvantage’ other than the two

proposed here are available. For simplicity’s sake I ignore these.
(18) The point still goes through if we also assume that almost all women are better off
under the unequal distribution.
(19) For critical assessments of social relations egalitarianism, see Lippert-Rasmussen

(2015a–b; 2018).
(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.
Kasper Lippert-Rasmussen
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
Page 20 of 21

Discrimination
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.
Page 21 of 21

Race

Race
Bernard Boxill
This article examines the tendency of racial differences to support violations of

distributive justice. It is thus a discussion in non-ideal theory and avoids discussion of
ideal theory. It begins by showing that races are social constructions. Social constructions
are sets of mainly false beliefs, and racial differences support violations of distributive
justice because of the nature of these beliefs. The races as social constructions are
generated by ideologies that are developed for some political purpose such as a defense
of black slavery or the forced deportation of black people from America. Affirmative
action is defended on the basis of a backward looking appeal to compensatory justice.
The argument is that it can be designed so as to expunge the false beliefs that constitute
the social constructions of race, and educate the public about racism.
Keywords: justice, distributive justice, ideal theory, non-ideal theory, compensation, deportation, social
construction, ideology, race, slavery.
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,
Page 1 of 16

Race
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.
Let us begin with the concept of race, since the idea that racial differences tend
(p. 499)
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
Page 2 of 16

Race
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 (p. 500) 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
Page 3 of 16

Race
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 (p. 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
Page 4 of 16

Race
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 (p. 502) 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
Page 5 of 16

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 (p. 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
Page 6 of 16

Race
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. (p. 504)
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
Page 7 of 16

Race
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 (p. 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
Page 8 of 16

Race
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.
(p. 506) 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
Page 9 of 16

Race
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 (p. 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
Page 10 of 16

Race
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 (p. 508) 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
Page 11 of 16

Race
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 (p. 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
Page 12 of 16

Race
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 (p. 510)
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
Page 13 of 16

Race
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 (p. 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.
Page 14 of 16

Race
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.
Plato (1992). The Republic, trans. G. M. A. Grube, rev. C. D. C. Reeve. Indianapolis, IN:
Hackett.
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.
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: 1697–714.
Zack, N. (1993). Race and Mixed Race. Philadelphia, PA: Temple University Press.
Bernard Boxill
Page 15 of 16

Race
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.
Page 16 of 16

Work

Work
Paul Gomberg
Many workers are unjustly harmed by the work they do: their confidence in their abilities
and even their abilities may be undermined because their work is too simple and they
work under constant supervision and command. John Rawls’s political liberalism may
criticize such work arrangements as undermining the internal psychological resources
required for democratic citizenship, invoking values of freedom and equality implicit in
democratic culture. Criticism of stultifying work can also be grounded in a view of the
human good as entailing development of complex abilities, contribution of human abilities
to a larger community, and earning esteem for those contributions. The second, more
“perfectionist,” approach is superior to the Rawlsian approach: it provides a simpler
argument and avoids the impossibility of marrying a market economy to egalitarian
principles.
Keywords: Rawls, political liberalism, democratic citizenship, contributive justice, perfectionism, work, human
good.
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
Page 1 of 22

Work
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.
Work differs from other human activities in having a purpose beyond the activity
(p. 514)
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,
Page 2 of 22

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 (p. 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-
Page 3 of 22

Work
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,
(p. 516) 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
Page 4 of 22

Work
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.
In a related vein, Stanczyk argues that the fundamental interest freedom of

(p. 517)
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 (p. 518)
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
Page 5 of 22

Work
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 (p. 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
Page 6 of 22

Work
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 (p. 520) 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)

Page 7 of 22

Work
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
(p. 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
Page 8 of 22

Work
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 (p. 522) (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
Page 9 of 22

Work
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 (p. 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
Page 10 of 22

Work
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. (p. 524)
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.
Page 11 of 22

Work
24.3 Aristotelian/Marxist 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
(p. 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.
Page 12 of 22

Work
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 (p. 526) 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).
Page 13 of 22

Work
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
(p. 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;
Page 14 of 22

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 (p. 528)
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.
Page 15 of 22

Work
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 (p. 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,
Page 16 of 22

Work
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
(p. 530) 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.
Page 17 of 22

Work
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.
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.
Page 18 of 22

Work
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.
Kornhauser, A. (1965). Mental Health of the Industrial Worker: A Detroit Study.

(p. 532)
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:
Lee, R. (1979). The !Kung San: Men, Women, and Work in a Foraging Society. Cambridge:
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.
Page 19 of 22

Work
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

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.
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:
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.
Notes:
(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.
Page 20 of 22

Work
(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.
(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.
(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.
(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).
(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.
(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
Page 21 of 22

Work
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.
(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).
(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.
(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).
Paul Gomberg
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.
Page 22 of 22

Exploitation

Exploitation
Ben Ferguson and Hillel Steiner
Exploitation is commonly understood as taking unfair advantage. This article discusses

the various prominent accounts that have been offered of how an exchange, despite being
Pareto improving and consensual, can nevertheless count as unfair or unjust and, hence,
as presumptively impermissible. Does the wrongness of an exploitative transaction
consist in its compounding a prior distributive injustice, or in its deliberately profiting
from someone’s vulnerability, or in its commodification of that which should not be
commodified? How should responsibility for exploitation be assigned, and can this avoid
generating moral hazard? The accounts of exploitation analysed here are classified along
two dimensions—historical vs. ahistorical and intentional vs. non-intentional—in their
conceptions of unfairness, and the possibility of a hybrid account is explored.
Keywords: exploitation, taking advantage, responsibility, fairness, distributive justice, moral hazard, historical
injustice, vulnerability, intention.
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
Page 1 of 24

Exploitation
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.
After all, they reason, if everyone involved is better off transacting and everyone
(p. 534)
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.
Page 2 of 24

Exploitation
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, (p. 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.
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
Fig. 25.1 Forms of disadvantage 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
Page 3 of 24

Exploitation
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 (p. 536) 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’.
Page 4 of 24

Exploitation
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’ (p. 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
Page 5 of 24

Exploitation
transaction, and the claim in (2) that some transactions that are both Pareto improving
and consensual are, nevertheless, impermissible.
(p. 538) 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
(p. 539) misfortunes’ (Goodin 1987). All four situations are, he claims, ‘manifestations of
Page 6 of 24

Exploitation
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
(p. 540) 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
Page 7 of 24

Exploitation
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. (p. 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.
Page 8 of 24

Exploitation
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 (p. 542) 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
Page 9 of 24

Exploitation
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:
(p. 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.
Page 10 of 24

Exploitation
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 (p. 544)
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
Page 11 of 24

Exploitation
‘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 (p. 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.
Page 12 of 24

Exploitation
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) There 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 (p. 546)
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
Page 13 of 24

Exploitation
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
(p. 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
Page 14 of 24

Exploitation
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 (p. 548)
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
Page 15 of 24

Exploitation
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 (p. 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
Page 16 of 24

Exploitation
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.
So, for instance, an alternative reason why Carol did not outbid Alice might be
(p. 550)
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.
Page 17 of 24

Exploitation
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 (p. 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.
Page 18 of 24

Exploitation
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
(p. 552) 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 (p. 553)
belief—is ‘aware’—that she does so. When Alice sees that Bob is unjustly disadvantaged
Page 19 of 24

Exploitation
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
(p. 554) impermissibility that we call injustice; and (ii) whether we believe that injustices
are the only form of morally impermissible acts that warrant legal interference.
Page 20 of 24

Exploitation
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’.
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:
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.
Page 21 of 24

Exploitation
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.
(p. 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 &

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.
Notes:
(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
Page 22 of 24

Exploitation
hypothetical competitive markets. We discuss these and other accounts of fairness in the
sections that follow.
(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.
(4) Thanks to Mike Otsuka for drawing our attention to this point.
(5) What exactly showing proper respect for Bob requires, in terms of transaction prices,
is left unspecified in Sample’s account.
(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.
(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.
(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).
(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.
Page 23 of 24

Exploitation
(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.’
(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.
(17) Indeed, the awareness condition can be seen as an operationalization of Roemer’s

dominance condition.
Ben Ferguson
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.
Hillel Steiner
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.
Page 24 of 24

Public Goods

Public Goods
David Miller and Isaac Taylor
Supplying public goods generates costs and benefits that can be allocated among
individuals in different ways. Yet leading theories of justice within liberalism fail to offer
concrete guidance on how these allocations should be assessed. This chapter critically
examines a number of principles of distributive justice that might be used for this
purpose. It argues that the appropriate principle will depend on whether the public goods
in question are “essential public goods” (ones that are required by justice) or
“discretionary public goods” (ones that are not). While a number of principles have
previously been put forward for both categories, the chapter concludes that none of them
are free from counter-intuitive implications. Further work is needed in order to arrive at a
satisfactory way of assessing the distributive consequences of public good supply.
Keywords: public goods, distributive justice, essential public goods, discretionary public goods, liberalism
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
Page 1 of 22

Public Goods
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 (p. 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
Page 2 of 22

Public Goods
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 (p. 558)
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 (p. 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
Page 3 of 22

Public Goods
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 near-
unanimous 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
(p. 560) 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
Page 4 of 22

Public Goods
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.
A very different perspective is provided by libertarian theorists, who are sceptical

(p. 561)
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
Page 5 of 22

Public Goods
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).
There are a number of questions and problems for Schmidtz’s approach (Miller
(p. 562)
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
Page 6 of 22

Public Goods
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
(p. 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
Page 7 of 22

Public Goods
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 justice
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.
In response, note that at least half of the question of distributive justice remains
(p. 564)
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).
Page 8 of 22

Public Goods
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
publicly funded healthcare service. Alternatively, however, the same need could
(p. 565)
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
Page 9 of 22

Public Goods
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
(p. 566) 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
Page 10 of 22

Public Goods
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 (p. 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
Page 11 of 22

Public Goods
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 (p. 568)
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 (p. 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.
Table 26.1 Arts and television
Page 12 of 22

Public Goods
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
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).
Page 13 of 22

Public Goods
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
Net benefit 7.5 7.5
In this scenario, the cost of Betty’s taste for expensive art is partially offloaded
(p. 570)
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
Page 14 of 22

Public Goods
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 (p. 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?
Page 15 of 22

Public Goods
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 (p. 572)
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
Page 16 of 22

Public Goods
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 (p. 573)
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.
Table 26.4 Chris joins society
Adam Betty Chris
Benefit 40 20 0
Cost 10 10 10
Net benefit 30 10 –10
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 justice 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.
Page 17 of 22

Public Goods
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.
(p. 574) 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:
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.
De-Shalit, A. (1995). Why Posterity Matters. London: Routledge.
Dworkin, R. (1986). A Matter of Principle. Oxford: Clarendon 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.
Page 18 of 22

Public Goods
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
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:
Nozick, R. (1974). Anarchy, State, and Utopia. Oxford: Basil Blackwell.
Olsaretti, S. (2013). ‘Children as Public Goods?’ Philosophy & Public Affairs 41(3): 226–
58.
(p. 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.
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.
Page 19 of 22

Public Goods
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.

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.
Notes:
(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).
(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).
(3) Though not always. Conscription, for example, may be used to pick up some of the
costs of national defence.
(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.
(5) For a slightly different, tripartite, distinction, see Miller (1999a: 159–60).
Page 20 of 22

Public Goods
(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.
(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.
(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.
(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.
(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.
David Miller
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
Page 21 of 22

Public Goods
2013) and Strangers in our Midst: The Political Philosophy of Immigration (Harvard
Isaac Taylor
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
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.
Page 22 of 22

Cultural and Religious Minorities

Peter Jones
This chapter examines issues of distributive justice as they relate to religious minorities
and to cultural minorities belonging to polyethnic societies. In many instances,
distributive justice requires only that those minorities should enjoy the same freedoms
and opportunities as others. Sometimes, however, justice is thought to require different
rather than identical treatment, such as the exemption of minorities from laws or rules
with which others have to comply. The chapter examines the justice of exemptions and
other forms of accommodation that provide for the religious and cultural liberty of
minorities, and for their opportunities to access other sorts of good such as employment.
It also considers how distributive justice bears on the relative standing that a society’s
different religions and cultures should enjoy.
Keywords: religious minority, cultural minority, justice, human rights, freedom, accommodation, exemption,
discrimination, establishment, recognition
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
Page 1 of 24

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.
(p. 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
Page 2 of 24

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
(p. 578) 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
Page 3 of 24

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
(p. 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
Page 4 of 24

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.
On the other hand, how much and what sort of accommodation there should be is
(p. 580)
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 ‘more-than-
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
Page 5 of 24

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 (p. 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.
Page 6 of 24

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.
(p. 582) 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.
Page 7 of 24

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
(p. 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
Page 8 of 24

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.
One question that may affect our thinking on exemptions and equality is whether
(p. 584)
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
Page 9 of 24

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 (p. 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
Page 10 of 24

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 (p. 586) 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.
Page 11 of 24

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).
Even if there is reason to treat claims of religion or culture (or some of them)
(p. 587)
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?
Page 12 of 24

27.7 Accommodation and Non-Comparative

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 (p. 588) 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.
Page 13 of 24

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
(p. 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,
Page 14 of 24

particularly those of Quong and Parekh, could be mobilized in its defence. Here, (p. 590)
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.
Page 15 of 24

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, (p. 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
Page 16 of 24

organizations and individuals sometimes want to discriminate in relation to

characteristics protected by discrimination law. That is uncontroversial when a church
(p. 592) 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 (p. 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
Page 17 of 24

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
(p. 594) 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
Page 18 of 24

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:
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
Bou-Habib, P. (2006). ‘A Theory of Religious Accommodation’. Journal of Applied

Page 19 of 24

Cornelissen, G. (2012). ‘Belief-Based Exemptions: Are Religious Beliefs Special?’ Ratio

Juris 25(1): 85–109.
Eisgruber, C. L., and Sager, L. G. (2007). Religious Freedom and the Constitution.
(p. 595)
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.
Greenawalt, K. (2009). Religion and the Constitution, vol. II: Establishment and Fairness
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:
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.
Page 20 of 24

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.
(p. 596) 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.
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.
Page 21 of 24

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): 53–71.
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

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.
Notes:
(1) United States v. Seeger, 380 U.S. 163 (1965); Welsh v. United States, 398 U.S. 333
(1970).
Page 22 of 24

(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.
(3) Barry expressly denies that his position is based on the chance/choice distinction;
Barry 2002: 215–19, responding to Mendus 2002.
(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.
(5) Copsey v. WWB Devon Clays Ltd [2005] EWCA Civ 932.
(6) Begum v. Denbigh High School [2006] UKHL 15.
(7) Davender Kumar Ghai v. Newcastle City Council [2009] EWHC 978 (Admin); [2010]
EWCA Civ 59.
(8) This example is based on Panesar v. The Nestle Company Ltd [1980] IRLR 64 CA.
(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.
(10) For a sample of views on how that challenge should be met, see Miller 1995; Modood
2007; Parekh 2006.
Peter Jones
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.
Page 23 of 24

Page 24 of 24

Language

Language
Alan Patten
In the small but growing literature on linguistic justice, it is possible to identify two main
approaches. One framework 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. The present contribution explores both of these approaches
and argues that each makes a valid contribution to a theory of linguistic justice.
Keywords: language, linguistic justice, instrumental approach, non-instrumental approach, distributive justice
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,
Page 1 of 24

Language
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?
As soon as one formulates these questions about policy, however, it immediately

(p. 598)
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
Page 2 of 24

Language
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
(p. 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
Page 3 of 24

Language
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 (p. 600) 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.
Page 4 of 24

Language
(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 (p. 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
Page 5 of 24

Language
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
(p. 602) 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
Page 6 of 24

Language
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.
28.3 A Non-Instrumental Framework: The

(p. 603)
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
Page 7 of 24

Language
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 (p. 604) 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
Page 8 of 24

Language
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.”
Although each of these conditions needs further specification and refinement, a

(p. 605)
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
Page 9 of 24

Language
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
(p. 606) 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
Page 10 of 24

Language
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)
If linguistic neutrality is impossible or incoherent, then supplementing the

(p. 607)
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
Page 11 of 24

Language
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
(p. 608) 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 justice
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.
Page 12 of 24

Language
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 (p. 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
Page 13 of 24

Language
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
(p. 610) 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
Page 14 of 24

Language
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
(p. 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
Page 15 of 24

Language
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 (p. 612)
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
Page 16 of 24

Language
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 (p. 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
Page 17 of 24

Language
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 (p. 614)
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.
Page 18 of 24

Language
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
(p. 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
Page 19 of 24

Language
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
(p. 616) 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.
Page 20 of 24

Language
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
limited. States cannot perform their core functions if they devote too high a
(p. 617)
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.
Page 21 of 24

Language
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. (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:
(p. 618) 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.
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.

Page 22 of 24

Language
Patten, A. (2017). “Protecting Vulnerable Languages: The Public Good Argument.” Paper
delivered to Workshop for Oxford Studies in Political Philosophy, Tucson, AZ.
Press.
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.
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.
Notes:
(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.
(2) On “opportunity hoarding,” see Anderson 2010.
(3) I develop an account of neutrality in Patten 2012 and 2014: Ch. 4.
(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
Page 23 of 24

Language
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.
(5) For discussion of these special cases, see Patten 2014: Ch. 6.
(6) Again, see Patten 2012 and 2014: Ch. 4.
(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.
Alan Patten
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.
Page 24 of 24

Justice Across Borders

Michael Blake
This chapter examines how philosophical concepts of distributive justice ought to be

applied at the global level. There has been a great deal of philosophical interest in this
topic in recent years, and the field has quickly grown to include some sophisticated
analyses of how we might think about global distributive justice. This chapter examines
this field, and argues that it must become more sophisticated still in order to adequately
deal with the complexities of the global arena. In particular, the article argues that we
have reason to examine more precisely the nature of global institutions—what powers
they actually have, and what it is that they might plausibly hope to become—as a key
focus of our philosophical analysis. The relationship between political and distributive
justice, in particular, ought to be made a particular focus in our efforts to understand the
nature of global justice.
Keywords: global justice, distributive justice, poverty, equality, cosmopolitanism
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
Page 1 of 23

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 (p. 620)
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.
Page 2 of 23

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 justice
(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 justice 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 (p. 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
Page 3 of 23

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.
Evidence for this, I think, is seen in the shifting definitions of cosmopolitanism in

(p. 622)
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.
Page 4 of 23

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
(p. 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
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
Page 5 of 23

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 (p. 624) 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
Page 6 of 23

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
(p. 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
Page 7 of 23

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, (p. 626) 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 justice 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?
Page 8 of 23

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
(p. 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.
Page 9 of 23

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 (p. 628) 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
Page 10 of 23

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, (p. 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
Page 11 of 23

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 (p. 630) 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?
Page 12 of 23

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
(p. 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
Page 13 of 23

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 (p. 632)
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
Page 14 of 23

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 (p. 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.
Page 15 of 23

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 (p. 634)
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?
Page 16 of 23

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
(p. 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
Page 17 of 23

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 (p. 636) 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?
Page 18 of 23

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.
Moral triage, in contrast, involves two legitimate paths—but a choice between the
(p. 637)
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.
Page 19 of 23

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.
(p. 638) 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
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.
Page 20 of 23

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:
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.
Macdonald, T. and Ronzoni, M. (2012). “Introduction: The Idea of Global Political

(p. 639)
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. (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.
Page 21 of 23

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
Pogge, T. (2013). “Concluding Reflections,” in G. Brock (ed.) Cosmopolitanism Versus

Non-Cosmopolitanism. Oxford: Oxford University Press, pp. 294–320.
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.
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

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.
Page 22 of 23

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.
Watkins, K. (2005). Rigged Rules and Double Standards: Trade, Globalization, and the
Fight against Poverty. Oxford: Oxfam International.
Michael Blake
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.
Page 23 of 23

Migration

Migration
Sarah Fine
This chapter focuses on 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 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. After outlining the key terms, the chapter critically
examines the argument that the freedom to move across borders is a requirement of
distributive justice. It presents different plausible versions of this argument and then
addresses a set of arguments that point in the other direction, and which seek to
illustrate that the obligations of distributive justice may support limits on the freedom to
move across borders. Ultimately, the chapter argues 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.
Keywords: distributive justice, migration, immigration, freedom of movement, borders
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
Page 1 of 26

Migration
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 justice 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
(p. 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 justice 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 (p. 642) also more abstract
goods, such as opportunities or well-being or capabilities (see Risse 2012: 5)? Which are
Page 2 of 26

Migration
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 justice
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 (p. 643) degrees, in comparisons between a variety of more and
less prosperous countries, particularly between those of the Global North and the Global
South.
Page 3 of 26

Migration
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 (p. 644) 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).
Page 4 of 26

Migration
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, (p. 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
Page 5 of 26

Migration
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.
The next question is how much freedom of movement is properly characterized as

(p. 646)
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
Page 6 of 26

Migration
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
(p. 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,
Page 7 of 26

Migration
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 (p. 648) 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 justice 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
Page 8 of 26

Migration
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 (p. 649) across borders must be a necessary
condition for the fulfilment of the equality- of-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
Page 9 of 26

Migration
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
(p. 650) 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
Page 10 of 26

Migration
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
(p. 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.
Page 11 of 26

Migration
In summary, as long as we accept the basic idea 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, 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.
(p. 652) 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 (p. 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
Page 12 of 26

Migration
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 (p. 654) 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.
Page 13 of 26

Migration
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 (p. 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
Page 14 of 26

Migration
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 (p. 656)
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
Page 15 of 26

Migration
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 (p. 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
Page 16 of 26

Migration
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 (p. 658)
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
Page 17 of 26

Migration
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, (p. 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 justice-based
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
Page 18 of 26

Migration
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
with the demands of distributive justice. While it is fair to predict that some
(p. 660)
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
Page 19 of 26

Migration
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.
Abizadeh, A., Pandey, M., and Abizadeh, S. (2015). ‘Wage Competition and the
(p. 661)
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.

Press.
Page 20 of 26

Migration
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.
Page 21 of 26

Migration
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.
Kapur, D. and McHale, J. (2006). ‘Should a Cosmopolitan Worry about the “Brain
(p. 662)
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
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
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.
Page 22 of 26

Migration
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
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.
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.
Stilz, A. (2016). ‘Is There an Unqualified Right to Leave?’, in S. Fine and L. Ypi
(p. 663)
(eds) Migration in Political Theory: The Ethics of Movement and Membership. Oxford:
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).
Page 23 of 26

Migration
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-and-immigration-united-states-0 (accessed 28 February 2018).
Notes:
(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).
(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).
Page 24 of 26

Migration
(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.
(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’.
(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).
(7) For an illuminating discussion of legitimate restrictions on freedom of movement, see

Stilz 2016.
(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.
(9) In practice, of course, the world is not at all like that, as I will discuss in section
30.3.1.
(10) 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 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.
(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.
(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
Page 25 of 26

Migration
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).
(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.
Sarah Fine
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.
Page 26 of 26

Climate Change

Climate Change
Simon Caney
This chapter considers two questions of distributive justice that arise when we face
dangerous climate change. The first (the Just Target Question) concerns what balance to
strike between ensuring that moral subjects are not harmed by climatic changes and
ensuring that the policies required to prevent harmful climatic changes are not unduly
onerous. The second (the Just Burden Question) concerns how the costs involved in
combating dangerous climate change should be distributed among duty-bearers. The
chapter identifies several methodological issues we need to confront to address these
questions. In addition to this, it outlines how one might answer the Just Target Question,
and evaluates several leading accounts of how to answer the Just Burden Question. One
central finding is that the issues of justice raised by climate change cannot be treated in
isolation but must be analysed as part of a more general global and intergenerational
account of justice.
Keywords: atomism, climate change, holism, integrationism, isolationism, polluter pays, ability to pay
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?
Page 1 of 26

Climate Change
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.
(p. 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
Page 2 of 26

Climate Change
elsewhere. For example, some might be worse off as a result of climate policies because
the funds being used to enable (p. 666) 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.
Page 3 of 26

Climate Change
We should record a third kind of burden. The first two concern burdens that are
(p. 667)
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 (p. 668) 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
Page 4 of 26

Climate Change
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.
Of course, one standard objection to utilitarianism is that it results in injustice,

(p. 669)
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
Page 5 of 26

Climate Change
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 (p. 670) the
costs of realizing (a) does not push people beneath the sufficiency level, then justice
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
Page 6 of 26

Climate Change
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.
(p. 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
Page 7 of 26

Climate Change
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 Distinguishing 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 (p. 672)
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
Page 8 of 26

Climate Change
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).
On this view, there is no specific principle that determines the distribution of X

(p. 673)
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 justice 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
Page 9 of 26

Climate Change
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).
(p. 674) 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
Page 10 of 26

Climate Change
a particular good, and hence it is governed solely by forward-looking principles of

(Vanderheiden 2008: 82, 222–3 and 229–30)
Now on this basis, Vanderheiden then argues, first, that with respect to
(p. 675)
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 justice are relevant
for adaptation funding?
Page 11 of 26

Climate Change
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 (p. 676) 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
Page 12 of 26

Climate Change
my contribution to causing dangerous climate change my behaviour (that is,

emitting v but then (p. 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.
(p. 678) 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
Page 13 of 26

Climate Change
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
(p. 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
Page 14 of 26

Climate Change
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
(p. 680) 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
justice that includes all normative issues will impede progress on reaching any agreement
Page 15 of 26

Climate Change
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 (p. 681) point,
then, is that one cannot make an a priori assumption that isolationism facilitates reaching
an agreement (Caney 2012a: 278–80).
Page 16 of 26

Climate Change
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 (p. 682)
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).
Page 17 of 26

Climate Change
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.
Consider now a second principle of justice often invoked in discussions of who

(p. 683)
should pay for the burdens of mitigation and adaptation. Many appeal to an Ability to Pay
Page 18 of 26

Climate Change
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 (p. 684)
(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)
Page 19 of 26

Climate Change
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.
Page 20 of 26

Climate Change
(p. 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 (p. 686) role.20
However, it remains the case that for a definitive answer to the questions examined here,
Page 21 of 26

Climate Change
we need to identify the best account of global and intergenerational justice 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.
Page 22 of 26

Climate Change
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.
Duus-Otterström, G. (2014). ‘The Problem of Past Emissions and Intergenerational

(p. 687)
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:

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.
Page 23 of 26

Climate Change
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:
Shue, H. (2005). ‘War’, in H. LaFollette (ed.) The Oxford Handbook of Practical Ethics.
Shue, H. (2014). Climate Justice: Vulnerability and Protection. Oxford: Oxford University
Press.
(p. 688) 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.
Notes:
(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.
(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.
Page 24 of 26

Climate Change
(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.
(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.
(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).
(9) For one exception, see Heyward (forthcoming).
(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).
(12) Hereafter my focus will be on the strong, rather than moderate, version of
integrationism.
(13) Hyams makes similar assumptions (2009: 241). See also Risse (2012: 202–6,
especially p. 204).
(14) I am grateful to Christian Baatz who in conversation, and in unpublished work with
Konrad Ott, has raised this concern.
(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.
Page 25 of 26

Climate Change
(16) For further discussion of these issues, and challenges to holism, see Jagers and Duus-
Otterström 2008 and Duus-Otterström and Jagers 2012.
(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).
(18) For discussion, see Caney (2005: 756–60); Gosseries (2004: 41–2).
(19) For discussions, see Caney (2005: 756; 2010: 215ff), Gosseries 2004, Duus-Otterström
2014, Page 2012.
(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.
Simon Caney
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).
Page 26 of 26

Future Generations

Future Generations
Rahul Kumar
The policies concerning, for instance, the mitigation of climate change that the current
generation chooses to adopt will have far-reaching implications for the lives of future
generations. What policies ought to be chosen depends, in part, on what justice requires
with respect to the protection of the interests of those who will live in the further future.
This chapter discusses the prospects for extending four prominent ways of thinking about
justice within a generation to the intergenerational context—Rawlsian contractualism,
Hobbesian contractarianism, the rights-based approach, and luck egalitarianism. It
argues that none of them offer a wholly satisfactory approach to intergenerational justice.
The final section of the chapter discusses whether obligations to protect the interests of
future generations are in fact best understood as obligations of justice.
Keywords: future generations, Rawls, Gauthier, contractualism, contractarianism, luck egalitarianism,

intergenerational cooperation, rights, climate change
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.
Page 1 of 24

Future Generations
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 (p. 690) 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.
Page 2 of 24

Future Generations
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 (p. 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
Page 3 of 24

Future Generations
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 (p. 692)
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 justice
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 justice 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
Page 4 of 24

Future Generations
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 (p. 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
Page 5 of 24

Future Generations
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
(p. 694) 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
Page 6 of 24

Future Generations
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.
Finally, it is important that the way in which the question of intergenerational

(p. 695)
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
Page 7 of 24

Future Generations
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
Second, it is implausible that the injustice of choosing Depletion depends on

(p. 696)
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.
Page 8 of 24

Future Generations
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.
To make the problem concrete, assume that choosing to deplete will result in
(p. 697)
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.
Page 9 of 24

Future Generations
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 (p. 698)
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.
Page 10 of 24

Future Generations
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 (p. 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
Page 11 of 24

Future Generations
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 (p. 700) 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.
Page 12 of 24

Future Generations
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
(p. 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.
Page 13 of 24

Future Generations
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 (p. 702)
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
Page 14 of 24

Future Generations
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 (p. 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
Page 15 of 24

Future Generations
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.)
(p. 704) 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
Page 16 of 24

Future Generations
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 (p. 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.
Page 17 of 24

Future Generations
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.
(p. 706) 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
Page 18 of 24

Future Generations
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.
(p. 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,
Page 19 of 24

Future Generations
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 1998,
2008). (p. 708) 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).
Page 20 of 24

Future Generations
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 (p. 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 justice 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.
Page 21 of 24

Future Generations
Gardiner, S. (2011). A Perfect Storm: The Ethical Tragedy of Climate Change. Oxford:
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. (2001). Justice as Fairness. Cambridge, MA: Harvard University Press.
Reiman, J. (2007). ‘Being Fair to Future People: The Non-Identity Problem in the
(p. 710)
Original Position’. Philosophy and Public Affairs 35(1):69–92.

Page 22 of 24

Future Generations
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,

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.
Notes:
(1) This example is adapted from Parfit 1984.
(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.
(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.
(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.
Rahul Kumar
Page 23 of 24

Future Generations
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.
Page 24 of 24

Index

Index
(p. 711) 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
Page 1 of 42

Index
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
(p. 712) 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
Page 2 of 42

Index
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 567–71
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
Page 3 of 42

Index
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 (p. 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
Page 4 of 42

Index
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
Page 5 of 42

Index
(p. 714)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
Page 6 of 42

Index
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 (p. 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
Page 7 of 42

Index
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
Page 8 of 42

Index
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 (p. 716)
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
Page 9 of 42

Index
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
Page 10 of 42

Index
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
(p. 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
Page 11 of 42

Index
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, 529–30,
546, 551, 600, 604
ecumenism 346–9, 353–6, 361–3
education 8, 17–18, 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, 441–57, 469, 494, 504, 507, 509–
10, 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 (p. 718)
inequalities 439, 446
institutions 439, 447–50
investments 262
justice 433, 438, 440, 444–5, 447–8, 452, 457
Page 12 of 42

Index
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
Page 13 of 42

Index
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
equal
(p. 719) basic liberties 13, 15, 18, 21, 28, 33, 38
benefit principle 569–71

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
Page 14 of 42

Index
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, 569–71, 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
Page 15 of 42

Index
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
(p. 720) 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
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
Page 16 of 42

Index
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
(p. 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
Page 17 of 42

Index
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
Page 18 of 42

Index
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, 669–71, 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 (p. 722)
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
Page 19 of 42

Index
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
Page 20 of 42

Index
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
(p. 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
Page 21 of 42

Index
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
Page 22 of 42

Index
of children 423
of self-definition 236
politics 236, 244, 246, 250, 601
recognition 585
ideology 499–502, 504–5, 508
(p. 724) 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
Page 23 of 42

Index
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
(p. 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
Page 24 of 42

Index
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
Page 25 of 42

Index
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 (p. 726)
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
Page 26 of 42

Index
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, 369–71, 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
Page 27 of 42

Index
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
(p. 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
Page 28 of 42

Index
MacPherson, C. B. 138
mammalian persons 261–2, 269–71, 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
(p. 728) 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
Page 29 of 42

Index
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, 666–71, 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
Page 30 of 42

Index
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
(p. 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
Page 31 of 42

Index
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
Page 32 of 42

Index
pluralism 168, 183, 196–8, 200, 201n.10, 203–4, 450

(p. 730) 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
Page 33 of 42

Index
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, 68–71, 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
(p. 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
Page 34 of 42

Index
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
Page 35 of 42

Index
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 (p. 732)
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
Page 36 of 42

Index
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
(p. 733) Sessions, R. 515
Sevenhuijsen, S. 218, 223, 226

sex 143, 236, 263, 278, 391–2, 394, 396, 402, 420
Page 37 of 42

Index
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
Page 38 of 42

Index
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
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
(p. 734) 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
Page 39 of 42

Index
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
Page 40 of 42

Index
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
(p. 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
Page 41 of 42

Index
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
Page 42 of 42

(Oxford Handbooks) Serena Olsaretti - The Oxford Handbook of Distributive Justice-Oxford University Press (2018)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Oxford Handbooks) Serena Olsaretti - The Oxford Handbook of Distributive Justice-Oxford University Press (2018)

Uploaded by

Copyright:

Available Formats

The Oxford Handbook of Distributive Justice

Oxford Handbooks Online

Great Clarendon Street, Oxford, ox2 6dp,

Oxford University Press is a department of the University of Oxford.

© in this volume the several contributors 2018

The moral rights of the authors have been asserted

First Edition published in 2018

All rights reserved. No part of this publication may be reproduced, stored in

You must not circulate this work in any other form

Published in the United States of America by Oxford University Press

British Library Cataloguing in Publication Data

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Library of Congress Control Number: 2017959074

Printed and bound by

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Oxford Handbooks Online

(p. ix) Contributors

Larry Alexander is the Warren Distinguished Professor at the University of San

Colin Bird is Associate Professor of Politics and Director, Program in Political

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Bernard R. Boxill is Professor of Philosophy Emeritus at the University of North

Gillian Brock is Professor of Philosophy at the University of Auckland in New

Simon Caney is Professor of Political Theory at the University of Warwick. He works

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Paula Casal is ICREA Professor at Pompeu Fabra University, Barcelona, associate

Matthew Clayton is Professor of Political Theory at the University of Warwick. He is

Benjamin Ferguson is Assistant Professor of Ethics at Vrije Universiteit,

Sarah Fine is Senior Lecturer in Philosophy at King’s College London. She

Samuel Freeman is Avalon Professor of the Humanities and Professor of Philosophy

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Gerald Gaus is the James E. Rogers Professor of Philosophy at the University of

Anca Gheaus is Ramon y Cajal researcer at teh Universitat Pompeu Fabra in

Paul Gomberg taught philosophy at the University of Missouri–St Louis 1971–1978

Aaron James is Professor of Philosophy at the University of California, Irvine. He is

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Peter Jones is Emeritus Professor of Political Philosophy at Newcastle University,

Matthew H. Kramer is Professor of Legal and Political Philosophy at the University

Rahul Kumar is Associate Professor of Philosophy at Queen’s University, Ontario. He

Kasper Lippert-Rasmussen is Professor of Political Theory at University of Aarhus,

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Jeffrey Moriarty is Associate Professor and Chair of Philosophy at Bentley

Serena Olsaretti is ICREA Research Professor at the Universitat Pompeu Fabra,

Michael Otsuka is a Professor in the Department of Philosophy, Logic and

Scientific Method at the London School of Economics. In addition to prioritarianism

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Alan Patten teaches political theory at Princeton University. He is the author of

Jonathan Quong is Associate Professor of Philosophy at the University of Southern

David Schmidtz is Kendrick Professor of Philosophy, Eller Chair of Service-Dominant

Chad van Schoelandt is an Assistant Professor at Tulane University. He works

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Hillel Steiner is Emeritus Professor of Political Philosophy at the University of

Isaac Taylor is a Scholar in Residence at the Center for Western Civilization,

Peter Vallentyne is Florence G. Kline Professor of Philosophy at the University of

Alex Voorhoeve is a Professor in the Department of Philosophy, Logic and Scientific

Steven Wall is Professor of Philosophy at the University of Arizona, where he is a

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Oxford Handbooks Online

Abstract and Keywords

As any student of contemporary political philosophy can attest, theorizing about

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

As a preliminary to that discussion, it is helpful to clarify how the variation I am drawing

Subscriber: Vrije Universiteit Amsterdam Library; date: 15 February 2019

Besides disagreeing on what justice demands, theorists of justice also disagree