Political Legitimacy: NOMOS LXI 9781479869206

Citation preview

POLITICAL LEGITIMACY NOMOS

LXI

NOMOS Harvard University Press I Authority 1958, reissued in 1982 by Greenwood Press The Liberal Arts Press II Community 1959 III Responsibility 1960 Atherton Press IV Liberty 1962 V The Public Interest 1962 VI Justice 1963, reissued in 1974 VII Rational Decision 1964 VIII Revolution 1966 IX Equality 1967 X Representation 1968 XI Voluntary Associations 1969 XII Political and Legal Obligation 1970 XIII Privacy 1971 Aldine-Atherton Press XIV Coercion 1972 Lieber-Atherton Press XV The Limits of Law 1974 XVI Participation in Politics 1975 New York University Press XVII Human Nature in Politics 1977 XVIII Due Process 1977 XIX Anarchism 1978 XX Constitutionalism 1979 XXI Compromise in Ethics, Law, and Politics 1979 XXII Property 1980 XXIII Human Rights 1981 XXIV Ethics, Economics, and the Law 1982

XXV Liberal Democracy 1983 XXVI Marxism 1983 XXVII Criminal Justice 1985 XXVIII Justification 1985 XXIX Authority Revisited 1987 XXX Religion, Morality, and the Law 1988 XXXI Markets and Justice 1989 XXXII Majorities and Minorities 1990 XXXIII Compensatory Justice 1991 XXXIV Virtue 1992 XXXV Democratic Community 1993 XXXVI The Rule of Law 1994 XXXVII Theory and Practice 1995 XXXVIII Political Order 1996 XXXIX Ethnicity and Group Rights 1997 XL Integrity and Conscience 1998 XLI Global Justice 1999 XLII Designing Democratic Institutions 2000 XLIII Moral and Political Education 2001 XLIV Child, Family, and State 2002 XLV Secession and Self-Determination 2003 XLVI Political Exclusion and Domination 2004 XLVII Humanitarian Intervention 2005 XLVIII Toleration and Its Limits 2008 XLIX Moral Universalism and Pluralism 2008 L Getting to the Rule of Law 2011 LI Transitional Justice 2012 LII Evolution and Morality 2012 LIII Passions and Emotions 2012 LIV Loyalty 2013 LV Federalism and Subsidiarity 2014 LVI American Conservatism 2016 LVII Immigration, Emigration, and Migration 2017 LVIII Wealth 2017 LIX Compromise 2018 LX Privatization 2018 LXI Political Legitimacy 2019

NOMOS LXI Yearbook of the American Society for Political and Legal Philosophy

POLITICAL LEGITIMACY Edited by

Jack Knight and Melissa Schwartzberg

NEW YORK UNIVERSITY PRESS

• New York

NEW YORK UNIVERSITY PRESS New York www.nyupress.org © 2019 by New York University All rights reserved References to Internet websites (URLs) were accurate at the time of writing. Neither the author nor New York University Press is responsible for URLs that may have expired or changed since the manuscript was prepared. Library of Congress Cataloging-in-Publication Data Names: Knight, Jack, 1952– editor. | Schwartzberg, Melissa, 1975– editor. Title: Political legitimacy / edited by Jack Knight and Melissa Schwartzberg. Description: New York : New York University Press, 2019. | Series: Nomos ; LXI | Includes bibliographical references and index. Identifiers: LCCN 2018041770 | ISBN 9781479888696 (cl : alk. paper) Subjects: LCSH: Legitimacy of governments. | Sovereignty. Classification: LCC JC497 .P637 2019 | DDC 320.01/13—dc23 LC record available at https://lccn.loc.gov/2018041770 New York University Press books are printed on acid-free paper, and their binding materials are chosen for strength and durability. We strive to use environmentally responsible suppliers and materials to the greatest extent possible in publishing our books. Manufactured in the United States of America 10 9 8 7 6 5 4 3 2 1 Also available as an ebook

CONTENTS

Preface Jack Knight and Melissa Schwartzberg

ix

Contributors

xi

Introduction Melissa Schwartzberg

1

PART I. FOUNDATIONS OF LEGITIMACY: AUTONOMY, ASSENT, AND OBLIGATION 1. Legitimacy and Self-Determination Anna Stilz

9

2. In Defense of Functionalism Jonathan Quong

47

3. Is Political Legitimacy Worth Promoting? Amanda R. Greene

65

4. The Sovereign and the Republic: A Republican View of Political Obligation Ekow N. Yankah

102

PART II. THE JUSTIFICATION OF INSTITUTIONAL AUTHORITY 5. Political Legitimacy under Epistemic Constraints: Why Public Reasons Matter Fabienne Peter 6. Legitimacy as a Right to Err Daniel Viehoff

147 174

vii

viii

Contents

7. Official Intentions and Political Legitimacy: The Case of the Travel Ban Micah Schwartzman 8. The Political Legitimacy of International NGOs Jennifer C. Rubenstein

201 231

PART III. BELIEFS ABOUT LEGITIMACY AND COMPLIANCE 9. Evaluating Consensual Models of Governance: Legitimacy-Based Law Tom R. Tyler 10. On the Empirical Measurement of Legitimacy Jeffrey A. Lenowitz 11. The Empirical Study of Legitimate Authority: Normative Guidance for Positive Analysis Sanford C. Gordon and Gregory A. Huber 12. Trustworthy Government and Legitimating Beliefs Margaret Levi Index

257 293

328 362 385

PREFACE JACK KNIGHT AND MELISSA SCHWARTZBERG

This volume of NOMOS—the sixty-first in the series—emerged from papers and commentaries given at the annual meeting of the American Society for Political and Legal Philosophy, held in conjunction with the annual meeting of the American Philosophical Association, Central Division, in Kansas City, Missouri, on March 1–2, 2017. Our topic, “Political Legitimacy,” was selected by the Society’s membership. The ASPLP conference consisted of panels organized around three principal papers and commentaries: (1) “Legitimacy and the International States-System,” by Anna Stilz, with commentaries from Ekow Yankah and Jonathan Quong; (2) “The Public Reason Conception of Political Legitimacy: An Epistemological Defense,” by Fabienne Peter, with commentaries from Micah Schwartzman and Jennifer Rubenstein; and (3) “Evaluating Consensual Models of Governance,” by Tom Tyler, with commentaries from Daniel Viehoff and Jeffrey Lenowitz. The current volume features revised papers from all of the conference participants, as well as solicited papers from Sanford Gordon and Gregory Huber, Amanda Greene, and Margaret Levi. We are grateful to all of these authors for their excellent contributions. Thanks also to Arina Cocoru of New York University for her valuable assistance during the editorial and production phases of this volume. We wish to thank the editors and production team at New York University Press, particularly Ilene Kalish. On behalf of the Society, we express our gratitude to the Press for its ongoing support both for the series and for the tradition of interdisciplinary scholarship that it represents. We are also grateful to Brown University, Duke University, New York University, and Stanford University for subventions in support of this and future NOMOS volumes.

ix

x

Preface

Finally, we thank the members of the ASPLP council—President James Fleming; Vice Presidents Stephen Macedo and Michelle Moody Adams; former at-large members David Estlund and Deborah Hellman; former presidents Debra Satz and Nancy Rosenblum; and Andrew Valls (secretary-treasurer)—for their support and advice.

CONTRIBUTORS

Sanford C. Gordon Professor of Politics, New York University Amanda R. Greene Lecturer in Political Philosophy, University College London Gregory A. Huber Professor of Political Science, Yale University Jack Knight Frederic Cleaveland Professor of Law and Political Science, Duke University Jeffrey A. Lenowitz Meyer and W. Walter Jaffe Assistant Professor of Politics, Brandeis University Margaret Levi Sara Miller McCune Director of the Center for Advanced Study in the Behavioral Sciences (CASBS) and Professor of Political Science, Stanford, and Jere L. Bacharach Professor Emerita of International Studies, Department of Political Science, University of Washington Fabienne Peter Professor of Philosophy, University of Warwick Jonathan Quong Professor of Philosophy and Law, University of Southern California Jennifer C. Rubenstein Associate Professor of Politics, University of Virginia

xi

xii

Contributors Melissa Schwartzberg Silver Professor of Politics, New York University Micah Schwartzman Joseph W. Dorn Research Professor of Law, University of Virginia School of Law Anna Stilz Laurance S. Rockefeller Professor of Politics and the University Center for Human Values, Princeton University Tom R. Tyler Macklin Fleming Professor of Law and Professor of Psychology, Yale Law School Daniel Viehoff Assistant Professor of Philosophy, New York University Ekow N. Yankah Professor of Law, Benjamin N. Cardozo School of Law, Yeshiva University

INTRODUCTION MELISSA SCHWARTZBERG

A common theme, as this volume goes to press, is that the world’s leading democracies are facing a “crisis of legitimacy.” Although we can readily identify key challenges confronting democracy—in particular, threats to the integrity of elections, to the security of the rule of law, and to the freedom of the press—whether these problems undermine political legitimacy is much less clear. The murkiness derives in part from disagreement over the meaning of the term; legitimacy ranks among the most contested concepts in political thought. Yet it also remains among the most important. This volume reflects the cutting edge of responses to the perennial question of legitimacy, drawing, in the distinctive NOMOS fashion, from political science, philosophy, and law alike. The chapters cover a range of actors, institutions, and practices, and address legitimacy from both a normative and a sociological or descriptive standpoint. As a result, less typically for NOMOS volumes, Political Legitimacy features several empirical contributions as well, reflecting the importance of contemporary work studying citizens’ beliefs about compliance and justice, and the widespread view that legitimacy has both moral and descriptive dimensions. Indeed, several chapters herein explicitly take up the relation between normative and descriptive legitimacy; some authors, like Jeffrey Lenowitz, distinguish further between moral, sociological, and legal legitimacy, or highlight differences between the “firstperson” and “third-person” perspective, in Jennifer Rubenstein’s language. Most chapters in this volume approach the question from the normative perspective, examining the grounds on which a state or 1

2

Melissa Schwartzberg

institution’s claim to authority might be morally justified, which is not to imply that there is consensus either on definitions of legitimacy or authority, or on what justification would require. Several authors, including Anna Stilz, argue that legitimacy entails a state’s moral right to issue laws and policies, and to coercively enforce these laws; such legitimacy derives at least partially (on a Kantian account adopted by both Stilz and Jonathan Quong) from the state’s successful performance of key functions. Others hold that legitimacy denotes institutions that possess justified authority, authority supported by reasons, and, often, that such justified authority gives subjects moral, content-independent, and/or preemptive reasons for compliance. Institutions may possess legitimate authority, following Joseph Raz’s influential account, in his Morality of Freedom (1986), because in issuing directives, such institutions would enable subjects to better comply with reasons that apply to them anyway. Alternatively, they may possess legitimate authority because they act for the sake of those subject to their directives (and are, on Daniel Viehoff’s account, thereby insulated from bearing the costs of their errors). Or, on a public reason conception of legitimacy, they possess authority because, as in Fabienne Peter’s chapter, they yield decisions that derive—at least sometimes—from reasoned agreement, and because, as described in Micah Schwartzman’s contribution, they act for permissible intentions. Still others will insist that legitimacy rests on voluntary subjection to rule: For Amanda Greene, a political order is legitimate insofar as its subjects “willingly assent” to being ruled (assuming that such regime performs the minimal function of providing for all subjects’ basic security). In turn, some will specifically reject such an account; Ekow Yankah argues that, under a republican conception of legitimacy, such obligations are necessarily reciprocal, consisting in a shared aim of the common good rather than grounded on individual consent as such. From the standpoint of sociological legitimacy, which rests upon subjects’ actual beliefs about the justification of the regime, the circumstances under which legal authority elicits voluntary consent—and what signifies such consent—also remains disputed. (Indeed, following Margaret Levi, one might further distinguish between fully voluntary, “quasi-voluntary,” and contingent

Introduction

3

consent.) Tom Tyler argues that consent derives from subjects’ beliefs in the appropriateness of the authority, which in turn derives from their beliefs that the authority is disposed to act fairly. On these grounds, they perceive that they have an obligation to obey the law; they do so because of a sense of responsibility and the value of cooperation in managing social order. Other contributors, notably Jeffrey Lenowitz and Sanford Gordon and Gregory Huber, will suggest that demonstrating a causal relationship between subjects’ beliefs in legitimacy and their compliance with the law is in fact quite challenging, and instead seek alternative strategies for measuring legitimacy. Although these disagreements and others emerge in the chapters, the volume seeks to clarify the lines of argument across disciplines. As such, while the basic structure of the Political Legitimacy volume reflects the traditional division of the ASPLP and NOMOS into three core disciplines, the chapters are grouped thematically. The first section of the volume addresses basic questions of the legitimacy of the state and its regime, drawing on competing traditions in the history of political thought from Kant, Hobbes, and Aristotle. In “Legitimacy and Self-Determination,” Anna Stilz argues that Kant’s political theory foregrounds collective selfdetermination as well as justice in state legitimacy. Were justice the sole grounds on which a state possessed the right to govern its population and territory, as traditional liberal theories hold, colonialism could be justifiable. As such, the justice of a state’s institutions—i.e., the state’s “functionalist” capacity to perform morally mandated functions in a reasonably just fashion—is necessary but insufficient to realize that state’s legitimacy. Individuals also possess an interest in autonomy, in being the authors of their political institutions. As such, to be legitimate, the state must also enable individuals to act autonomously through collective participation in the choice of their political institutions. That is, it must ensure that institutions reflect a shared will formed under conditions of free deliberative reasoning. Jonathan Quong responds to Stilz in “In Defense of Functionalism,” holding that her account cannot explain the wrongness of colonialism by appeal to the way such acts prevent the colonized society from having institutions that correspond to their majority’s religion, culture, or conception of the good life—or to do so,

4

Melissa Schwartzberg

must depart from a Kantian commitment to state neutrality about the good life. Further, Quong argues, a functionalist approach to political legitimacy can indeed address the wrongness by appealing to the pro tanto wrongness of involuntarily changing people’s political status without their consent, waiver, or forfeiture. Quong defends this “individual status view” both as capable of meeting the demand to explain the wrongful nature of colonialism and as providing a more compelling explanation for the importance of democratic procedures—which Stilz argues are neither necessary nor sufficient to realize self-determination—as a means of allowing people to express their view of the establishment of political boundaries. In “Is Political Legitimacy Worth Promoting?,” Amanda Greene provides a new account of political legitimacy, drawing on Hobbes and Weber, to argue that a regime is legitimate insofar as it secures “quality assent” to its rule by its subjects. By assent, Greene holds that a subject must willingly accept her political subjection, grounded in the judgment that her subjection corresponds to some values that she can identify and affirm. But such acceptance depends upon a “quality” filter, which requires the provision of basic security to each member. Greene’s account is Hobbesian insofar as it demands peaceful order; legitimacy, on Greene’s account, is not possible in open conflict and threat of civil violence. Yet it is also Weberian in its emphasis on the acceptance of a system of command, sustained by a group’s belief in its validity. The achievement of political legitimacy, understood as quality assent, realizes three key political goods: the avoidance of political alienation, the establishment of durable security, and the alignment of values justifying the exercise of power. In contrast to the previous chapters, Ekow Yankah argues that attempts to ground legitimacy on Kantian or other liberal conceptions are necessarily fraught, and are incompatible with a natural and intuitive account of our lives in common. In “The Sovereign and the Republic: A Republican View of Political Obligation,” Yankah instead seeks to retrieve and defend a conception of Aristotelian republicanism, which grounds political obligation in our deep, inextricable moral and political connections; our well-being is bound up in the welfare of our community. As such, the justification of political and legal decisions at both the fundamental (the

Introduction

5

creation of a community) and prosaic (zoning norms) levels must rest on appeals to the common good, rather than to the aim of securing our individual rights or interests against each other or the state. The second section of the book takes up the reasons according to which institutional authority may be exercised. Fabienne Peter’s chapter, “Political Legitimacy under Epistemic Constraints: Why Public Reasons Matter,” defends the importance of a particular public reason conception of legitimacy. Via an examination of the epistemology of practical reasoning, Peter argues that sometimes permissible practical reasoning leads to disagreement about which outcome should be chosen, as in cases in which two people have justified beliefs about the correctness of opposing solutions and neither can provide compelling epistemic reasons for the other to change her belief. As such, in some epistemic contexts, political decisions must be justified to citizens in terms of public reasons, reasons that all could accept or that no one could reasonably reject. Peter holds, in fact, that this is the normal case of politics: We typically lack sufficiently robust knowledge of the correct political decision, and so justification in terms of objective reasons will be unavailable to us. Daniel Viehoff’s chapter, “Legitimacy as a Right to Err,” argues that legitimate institutions, which possess the authority to impose duties, are empowered both to make mistakes in the exercise of their power and to be insulated from bearing the costs of these mistakes in a way that other actors and institutions are not. That is, the legitimate exercise of normative power does not hinge upon these institutions reaching correct decisions; rather, it is conditional upon them acting for their subjects for their subjects’ sake. Insofar as this is the case, the subjects themselves, as a matter of fairness, should bear the costs of good-faith errors made by legitimate authority. But one should distinguish this form of legitimate authority from justified authority, in which a person may possess such normative power but must bear the costs of her own mistakes. When an institution no longer acts for the benefit of its subject, it may no longer be shielded from bearing the costs of such errors; indeed, subjects may rightly resist the imposition of such costs. In his chapter, “Official Intentions and Political Legitimacy: The Case of the Travel Ban,” Micah Schwartzman investigates the

6

Melissa Schwartzberg

moral conditions for the legitimate exercise of political power from a different vantage point. Drawing on the example of President Trump’s travel ban, in which courts held that the travel ban was unconstitutional because it was motivated by religious animus, Schwartzman asks whether we can give a moral justification for rejecting the legitimacy of an official action on the basis of the intention or motivation behind it. Whereas philosophers and legal scholars have argued that intentions are never directly relevant to the moral permissibility of actions, Schwartzman insists that this view cannot be sustained in reflective equilibrium through testing these arguments against cases of discretionary discrimination. He defends a doctrine of moral taint, which holds that past wrongful conduct by officials can yield appropriate skepticism about their later conduct; although this places a burden on these officials to demonstrate that their current actions are morally permissible, this burden can be satisfied through the acknowledgment of past wrongs and a credible demonstration of good faith. Shifting to a different institutional setting, in “The Political Legitimacy of International NGOs,” Jennifer Rubenstein addresses the question of how to conceptualize the legitimate exercise of power for large-scale, Western-based humanitarian and development INGOs such as Oxfam and Doctors Without Borders. The literature on INGOs tends to treat the question of legitimacy by asking whether such institutions exercise political power according to a set of criteria such as accountability, responsiveness, transparency, and efficiency. Rubenstein argues that this criteria-based approach neglects two crucial aspects of political legitimacy: that legitimacy entails (1) a “low bar” or minimum threshold concerning (2) the moral right to rule. According to Rubenstein, recognizing the political legitimacy of INGOs gives those subjected to their power moral reasons to cooperate with it, or not to interfere with others’ compliance with such rules, and helps to provide such subjects moral reasons to resist illegitimate INGOs. The final section of the volume turns to the empirical study of legitimacy and compliance, and on the relationship between what the authors describe as moral or normative accounts of legitimacy and sociological or descriptive legitimacy. Tom Tyler’s chapter, “Evaluating Consensual Models of Governance: Legitimacy-Based Law,” distinguishes two models of legal authority, one coercive and

Introduction

7

one consensual. Whereas the coercive model motivates compliance through law by the threat of sanctions, the consensual model focuses on personal values and the role such values play in eliciting consent to law; this latter model entails legal legitimacy. Tyler argues that legitimacy requires that people relate to authority in terms of their beliefs about their obligation and responsibility to defer to their decisions, and that they believe that such authorities consistently rely on fair procedures in rendering decisions. Comparing data from the European Social Survey and an American national survey, Tyler evaluates legitimacy in terms of the relative importance of a set of moral and social values and demonstrates that they shape compliance with law; further, Tyler holds that public perceptions of procedural justice constitute a key antecedent of legitimacy. In “On the Empirical Measurement of Legitimacy,” Jeffrey Lenowitz argues that Tyler’s findings distort the conception of legitimacy in potentially worrying ways. He examines Tyler’s claims—including that individuals are more likely to exhibit desirable legal behavior when they feel obliged to obey their authorities, trust these authorities, and believe them to share a common sense of morality—and demonstrates that they do not support Tyler’s causal claims as to the effect of such beliefs on sociological legitimacy, i.e., on individuals’ beliefs in the content-independent rightful authority of the law and its enforcers. Further, he shows that as Tyler measures only individuals’ perceptions of procedural justice, rather than the effects of institutions themselves, he gives authorities reason to focus on reforming such institutions so that they seem to be just, rather than become just. To avoid this outcome, Lenowitz asserts that moral argumentation remains indispensable for institutional change. In response to similar concerns about the inferences drawn from empirical research, Sanford Gordon and Gregory Huber analyze the difficulties associated with measuring beliefs in legitimacy and of determining the causal effects of such beliefs on compliance. In their contribution, “The Empirical Study of Legitimate Authority: Normative Guidance for Positive Analysis,” Gordon and Huber argue that isolating the effect of legitimacy as one among other motivations for an agent to comply with an authority may be challenging; moreover, both the decisions of other citizens and the

8

Melissa Schwartzberg

authority itself affect an agent’s intrinsic and extrinsic motivations to comply. They develop a simple, formal model seeking to identify the conditions in which changes in citizens’ perceptions of an authority’s legitimacy affect their willingness to comply, apart from any extrinsic motivations (including strategic considerations). In light of these challenges, they urge care in invoking the causal significance of legitimacy. Finally, in “Trustworthy Government and Legitimating Beliefs,” Margaret Levi examines the distinction between trustworthy and legitimate governments. Drawing on a wide range of empirical evidence, she argues that the two crucial aspects of trustworthy governments—their provision of promised goods and services and their procedural fairness in the determination and implementation of policy—do not fully explain the variation in beliefs about these governments’ legitimacy. Rather, legitimate governments also require widely accepted justifications for their members’ selection and their exercise of power, in particular. As in the previous chapters, Levi reflects on the challenges of distinguishing among the various conditions that may give rise to the behavioral response of compliance and highlights important new empirical results that shed light on these mechanisms. While Levi cautions that key economic transformations—including the decline in unionization—may threaten the shared justifications on which legitimacy rests, she identifies opportunities to reconstruct guiding principles in the process of establishing a new moral economy. Without reaching any consensus, then, the chapters in this volume highlight several avenues for future research. When political decisions fall short of procedural or substantive justice, or are supported by weak or unjust reasons, at what point may we identify deeper problems to political legitimacy? When do beliefs about unfair or disparate treatment of members implicate political legitimacy, and when might beliefs concerning the state’s treatment of noncitizens, within or beyond political borders, pose a challenge to legitimacy? Does resistance necessarily reflect a weakening of legitimacy, or might protest and contestation illuminate pathways for the reaffirmation of core values? These pressing questions remain on the agenda for scholars and citizens alike; the modest ambition of this volume is to provide support for a range of potential responses.

PART I

FOUNDATIONS OF LEGITIMACY: AUTONOMY, ASSENT, AND OBLIGATION

1 LEGITIMACY AND SELF-DETERMINATION ANNA STILZ

How does a state acquire legitimate jurisdiction over a particular territory and its population? One story, to which I am sympathetic, holds that when people come into conflict about rights and resources, they are obliged to set up a state that can fairly resolve their disputes, since a state is necessary to the legitimate interpretation and enforcement of rights. Immanuel Kant, on whose writings I draw in this chapter, holds that we have an unconditional duty to enter the state, since a state is the condition of possibility of “conclusive” property rights (6:245).1 He also suggests that a person claiming property is permitted to coerce his neighbors to enter a civil condition along with him (6:255). This duty is binding on us independently of any special relationships we have, or any voluntary transactions we have engaged in. Yet this Kantian view is subject to an important objection, which holds that it is unable to provide any determinate account of the state’s proper boundaries. Critics argue that while the Kantian view may show that states are in general necessary to establish justice, the view cannot address which particular people or places should be included within any state’s jurisdiction, or indeed, why we should prefer a plurality of states to a single world state.2 Those who press this objection typically interpret Kant as offering a purely functionalist account of state legitimacy, which holds that a state has a right to rule a population and territory insofar as it performs morally mandated governance functions in a reasonably just manner.3 Functionalism comes in both maximizing and threshold variants. The maximizing variant holds that a particular state has a right to rule a territory if—compared to its rivals—it can do 9

10

Anna Stilz

best at delivering justice. The threshold variant holds that the state has a right to rule a territory if it achieves a decent level of success in that task.4 An important worry about functionalism, however, is that it may license colonialism or involuntary annexation.5 This chapter argues that the Kantian account of state legitimacy should not be assimilated to functionalism (on either variant), but should instead be understood as incorporating a (limited) claim to political self-determination. Functionalism and Colonialism The charge is not that functionalists will endorse colonial annexation in all cases. Colonialism and annexation frequently involve unjust violence and rights violations and may be ruled out on these grounds. Allowing just states to annex territory promotes war, which has predictably bad consequences for international peace and security. A functionalist may therefore endorse a general rule requiring respect for other states’ territorial integrity, except in cases of national self-defense or humanitarian intervention. Finally, functionalists would condemn most historical instances of colonialism. Colonial rulers dispossessed indigenous populations of their lands, they engaged in forced labor and economic exploitation, and they institutionalized systems of racial and cultural discrimination. Functionalist commitments to human rights and basic justice explain why these acts were wrong. Still, the functionalist does not rule out a benign colonial regime if it does a reasonable job of providing good governance. Indeed, functionalism might even be invoked to support “civilizing” colonialism: Particularly in the later colonial period, European colonizers often invoked arguments grounded in liberal principles to justify their practices.6 Colonial rule was defended on the basis that it would help abolish the slave trade in Africa, or that it would further the moral and material well-being of native populations, or advance commerce and development. The US occupation of the Philippines, for example, supposedly aimed to improve “the well-being, prosperity, and the happiness of the Philippine people” and to establish “an enlightened system of government.”7 Suppose a colonial regime were successful in living up to its “civilizing” ideology: It protected its subjects’ rights and delivered

Legitimacy and Self-Determination

11

enlightened governance to them. In that case, on either variant of functionalism, no claim to self-determination could be pressed against it. Ex hypothesi, the colonial regime meets the minimal justice threshold, and it may do better at securing justice than its rivals (also satisfying the maximizing criterion). Still, it seems that a population subjected to benign colonial administration would have a morally significant complaint: While not subject to grave injustice, they are denied self-rule. This violation of self-rule, I believe, is in itself a wrong, even when not accompanied by further violence or rights violations. The aim of this chapter is to understand the nature of the pro tanto wrong involved in the denial of political self-determination. Violations of self-determination can occur in other cases. Consider: Military Occupation. In 1945, the Allies occupied Germany through a just use of force. Suppose that instead of restoring the territory to the German people, the United States had annexed their zone of occupation, turning it into an additional state of the union. After annexation, the United States governed reasonably justly, protecting the Germans’ human rights and granting them rights of democratic participation in the now-unified polity. Would the Germans have had a claim to political independence? Humanitarian Intervention. Proponents of humanitarian intervention argue that it is permissible to intervene militarily in another state in cases of genocide, mass expulsions, or gross violations of basic human rights. They believe temporary foreign rule can be acceptable in the aftermath of a justified humanitarian intervention. Yet most people think occupiers are obliged to restore the country to independence once a decent domestic government can be established. Why do they have this responsibility? The best way to characterize our intuitions about these cases, I believe, is to hold that annexed, colonized, or occupied populations have a claim to govern themselves and their territory independently, and to order their political institutions as they choose. This claim to self-determination is defeasible and may sometimes be outweighed by competing concerns, as in a justified humanitarian intervention. But where weighty countervailing considerations

12

Anna Stilz

are not at stake, it ought to be respected. Functionalism, however, seems unable to account for self-determination. On its maximizing variant, no political group can ever claim a right—against a more just colonizing power—to govern itself independently. And while its threshold variant allows a political group that achieves decent rule to govern itself, its right to independence holds only pro tem and may be lost if in the future it becomes subject to reasonably just foreign rule. Two Dimensions of Legitimacy If we accept the claim to self-determination, we must view state legitimacy as having two distinct dimensions.8 Let me define “legitimacy” as the state’s possession of an exclusive moral right to make law and policy on behalf of a particular group and to use coercion or force to implement those laws and policies. While a theory of justice addresses the question: “What scheme of rules about rights ought a political community (ideally) to adopt and enforce?,” a theory of legitimacy addresses a different question, namely: “Who has the right to decide what scheme of rules a political community will adopt and enforce?” The concept of legitimacy is contested: Some see legitimacy as correlative to subjects’ obligations to obey the state’s directives, while others define legitimacy as a mere liberty to enforce, which imposes no obligations on subjects, not even the obligation to refrain from vigilante justice. The definition I adopt here is agnostic about whether a legitimate state imposes duties to obey the law, but it does hold that legitimacy correlates to some obligations on subjects, namely, the obligation not to interfere with or resist a legitimate state’s efforts to issue and enforce directives. While citizens of a legitimate state may not have a duty to obey an unjust law rather than engaging in civil disobedience or conscientious refusal, they do have an obligation not to compete with a legitimate state through revolutionary violence or vigilante enforcement, and not to obstruct it in carrying out its functions, including accepting punishment for disobedience. In theorizing state legitimacy, functionalism focuses on a (reasonably just) state’s role in providing justice-related benefits to its members.9 As institutional “takers,” individuals have interests in protection of their rights, a fair scheme of distributive justice, or

Legitimacy and Self-Determination

13

public goods that only a state can provide. This “taker” dimension of evaluation focuses on familiar aspects of the basic structure, e.g., “the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation.”10 It is concerned with the quality of state institutions, independent of anyone’s acceptance or endorsement of them. But on a self-determination theory, legitimacy also has a second, “maker” dimension: People have an interest in seeing themselves as the authors of their political institutions. It may be as important that their institutions reflect their priorities and values (in their role as “makers”) as that these institutions be good ones (from their perspective as “takers”). This second dimension of evaluation focuses not on the internal structure of the state itself (its qualities or characteristics), but rather on the relation between that state and those it rules. On one account of self-determination, the nationalist theory, cultural nations are bearers of self-determination rights. A nation is a group marked out by shared characteristics that expressly suit them for self-rule, particularly a common culture.11 On the nationalist view, a state has a right to rule a particular population and territory if (a) the majority of that territory’s population forms a nation, (b) the state is properly authorized by that nation, and (c) the state acts to reproduce the nation’s culture over time, granting it privileged status in state symbols, ceremonies, the public education system, and so on. A second account, the voluntarist theory, holds that collectives’ rights of self-determination are constructed on the basis of free consent of the governed. Voluntarists argue that subjecting people to political coercion without their consent objectionably infringes on their “natural freedom,” their ability to determine their lives by their own actions and decisions. A particular state has a right to rule a particular population only where individual subjects have actually historically consented to that state’s exercise of political power.12 Here I develop a third account of self-determination, which I call the political autonomy theory. This account draws in important ways on Kant’s political philosophy, especially his argument that the rules governing our collective political life ought not to be

14

Anna Stilz

enforced unilaterally. Kant instead holds that to have the right to make and enforce law and policy for a population, the state must represent its subjects’ omnilateral will: “Only the concurring and united will of all, insofar as each decides the same thing for all, and all for each, and so only the general united will of the people, can be legislative” (6:314). A key question is: What exactly makes for an omnilateral will? I suggest that an omnilateral will is an actual popular will. To be in a position to legitimately enforce law and policy, a state needs to reflect its people’s shared intention to cooperate together through particular institutions. In offering this reading, I hope to complicate our understanding of Kant’s political theory. Our most basic natural duties to others, for Kant, require us not only to set up a legitimate state, but also to impose that state’s political order in a manner that is sufficiently respectful of others’ capacity for rational judgment, including their capacity to make judgments about how they should be ruled. Once the conditions of an omnilateral will are unpacked, we will see that it is permissible to impose a political order on unwilling participants only in a narrow range of cases, and that the Kantian view does not sanction colonial annexation. Understanding the problem of unilateralism, in my view, also allows us to explain why a Kantian should prefer a system of states to a single world state. My argument proceeds in five steps:13 1. It is normally illegitimate to unilaterally use force or coercion to impose one’s judgments on other agents, including one’s judgments about justice, even if one’s views are correct. Instead, legitimate enforcement requires a public authority whose use of coercion represents an omnilateral, not a unilateral, will. While unilateral coercion is normally pro tanto wrong (with exceptions explained below), this wrong can sometimes be outweighed by other values. 2. Unilateral coercion is wrong because it threatens autonomy. Just as we have an important interest in personal autonomy, we have an important interest in political autonomy—that is, an interest in having our political institutions correspond to our judgments and priorities. Since omnilateral coercion is based on subjects’ shared intention to cooperate to estab-

Legitimacy and Self-Determination

15

lish justice, it is consistent with their political autonomy in a way unilateral coercion is not. 3. The state has reason to be responsive to the shared judgments only of a restricted domain of its subjects, whom I call cooperators. Cooperators are willing to recognize and respect others as possessing an equal right to autonomy and independence, and to acknowledge the political duties that flow from this recognition. No wrong is done when we deny political self-determination to those who reject the fundamental moral requirement from which its value is derived. 4. Shared political autonomy is possible when cooperators share a will. They share a will when there is an interlocking structure of joint intentions among them and common knowledge of this fact. The object of a people’s shared will need not be first-order policy decisions: It can be abstract ideals and political procedures, of the sort typically enshrined in a political constitution. 5. Colonialism and involuntary annexation are wrong because these acts conflict with cooperators’ political autonomy. This account is not reducible to either voluntarism or nationalism since cooperators’ shared will need not involve actual consent, nor need it involve a national identity. Justice in the State of Nature I begin by briefly recapitulating some core elements of Kant’s theory of legitimate state authority. As is well known, Kant holds that individuals are required to submit to a state: We are not permitted to remain in the “lawless freedom” of the state of nature. Three Kantian claims interest me here: First, the premise that we are subject to a fundamental natural duty of justice to respect others’ freedom-as-independence (Natural Duty); second, the claim that this natural duty grounds a derivative duty to enter a civil condition (State); and third, the claim that a legitimate state’s scheme of law ought to be imposed omnilaterally, not unilaterally (Omnilateralism). Kant begins from the idea that each individual has an innate right to freedom, which requires him to be independent from the will of

16

Anna Stilz

other persons. A necessary condition for autonomous, self-directed action is being free from the will of others, who might otherwise interfere with one’s capacity to set and pursue one’s own goals. To be independent, each agent must enjoy a secure sphere of personal self-determination within which others cannot interfere. The bounds of a person’s sphere are defined by her acquired rights, including the right to property, to enter into contracts, and to establish legally recognized personal relationships. Each individual has a fundamental, coercible natural duty of justice to respect others’ independence, by respecting their acquired rights (Natural Duty). This natural duty of justice is enforceable and binding on us without our consent. If you don’t perform your duty to respect my property, for example, I can make you do so, by demanding your compliance and, if necessary, by calling the police. Our natural duty of justice differs in this respect from purely voluntary obligations and duties of charity or beneficence. Kant further argues that we cannot fulfill the natural duty of justice without entering a civil condition (State). At least some of our general, coercible duties of justice—particularly the duty to respect other people’s property and contractual rights—are mediated by legitimate state institutions. Since our duties of justice are in this way mediated, we have a duty to comply with and support a legitimate state’s system of law, because this is necessary to treat other people justly. Why might duties of justice be institutionally mediated in the way Kant suggests? The main problem is that people disagree about which scheme of rules ought to guide their attempts to do justice to one another, and these disagreements require legitimate authority for their resolution: [B]efore a public lawful condition is established, individual human beings, peoples and states can never be secure against violence from one another, since each has its own right to do what seems right and good to it and not to be dependent upon another’s opinion about this. (6:312)

Due to disagreement, it would be difficult for even well-meaning individuals to carry out their duties of justice through private action in a state of nature.

Legitimacy and Self-Determination

17

In part, disagreement arises because of the many indeterminacies in our moral rights. Even if a person’s freedom requires that she be able to use goods essential to her self-preservation and life projects, this leaves open many other aspects of a scheme of property: whether private ownership of the means of production should be allowed; what definition of contractual transfer rights we ought to adopt; whether property and income should be subject to tax, and at what level; whether or not bequest should be allowed; which rights, if any, over intellectual property should be recognized; and so on. In addition, moral rights in the state of nature have many vague aspects. While it may be clear that driving with a blood alcohol level of 0.2 poses unacceptable risks to others, is it better to set the limit at 0.08 or 0.07 or 0? The answers to these questions may have an ineliminably decisionist element. (Compare: What precise formalities should be required to conclude a contract? How long a statute of limitations should we institute? And so on.) Finally, the protection that a body of rules provides for these claims is in part a matter of how the entire scheme fits together. Given that we have already adopted a rule in one area, the best way to secure some additional moral claim may be to adopt a rule that meshes in a certain way with our preexisting structures, even if we would not necessarily have chosen this rule ex ante.14 But private rights in the state of nature are problematic for reasons that go beyond their partial indeterminacy. Even where moral rights are determinate, individuals will disagree about them in particular cases. Consider an example: Almost everyone thinks, at a very abstract level, that individuals have a right to life. Many people believe that this can ground a claim to material resources essential for one’s survival. And many in turn believe this grounds a claim of justice to adequate health care. But not all support these conclusions. Christian Scientists think that poor health is actually a sign of erroneous spiritual beliefs that must be corrected through faith healing. And libertarians believe that while the right to life is important, it imposes only negative duties, not positive duties, on others. In a pre-institutional situation, Christian Scientists and libertarians would not recognize any duty to contribute to securing others’ interests in medical care. When people lack a common understanding of rights, then their attempts to respect the rights

18

Anna Stilz

that, by their lights, other individuals have will not strike these others as an effort to respect their rights. Kant argues that these problems of disagreement mean that acquired rights—including provisional property rights— are not legitimately enforceable in the state of nature. When two individuals—each acting to implement a set of rights that they believe, in good faith, protects the independence of all—disagree in their interpretations of justice, neither is required to submit to the other’s judgments. No one can claim an exclusive moral right to interpret and enforce justice for others, in a way that would put those others under an obligation not to interfere with, compete with, or resist his efforts.15 As Kant puts it, “rights are in dispute (ius controversum)” in the state of nature, since in cases of conflict, there is “no judge competent to render a verdict having rightful force” (6:312). Kant allows that where a civil condition is impossible, an individual would be authorized to use force in defense of her rights. A right-holder in a state of nature, say, would have permission to push an intruder off her land. But this permission extends only to force necessary to thwart violations of one’s own rights (it does not extend to the right to prevent violations of others’ rights, or to punish transgressors after the fact). Further, since each individual in the state of nature is an equally authoritative judge of justice, scenarios can arise in which two parties have a good-faith belief that the land belongs to them, and where neither is required to submit to the other’s judgment about this. As free and independent individuals, we are entitled to coerce in accordance with our own (good-faith) judgments, not someone else’s judgments, even when our judgments are mistaken. Kant’s reflections on private rights in the state of nature lead him to two conclusions. First, he argues that people living side-byside under these conditions have a derivative duty—grounded in their more fundamental natural duty of justice—to coordinate under public authority (State). If they are to discharge their duty to treat others justly, they need to settle on some unitary public scheme of rights, because without any shared rules, their self-determination will become precarious indeed. Imagine meeting someone for the first time where you have no settled expectation about what they take your property rights to be. Some people, under their

Legitimacy and Self-Determination

19

preferred scheme, might recognize you as having a right to what you take to be your house. But other people might not, and might indeed try to forcefully dispossess you of what, in their estimation, is not really your property. How could you securely exercise even a minimal degree of independent self-determination in this world? A world in which there is no commonly accepted scheme of rights—only a set of competing and conflicting interpretations—is a world in which no one’s external freedom is secure. The natural duty to treat others justly therefore grounds a derivative duty to coordinate, through the state, on a common scheme of rights that can be enforced in a unitary way amid disagreement.16 By monopolizing the enforcement of justice, the state overcomes the normative breakdown of the state of nature. It does so, however, by claiming powers to decide on others’ behalf that would not exist in that pre-institutional scenario. First, the state claims an exclusive right to enforce its interpretation of justice, prohibiting any competing efforts by private individuals. It thus asserts a superior standing to its subjects: It settles the matter for them, even where they disagree. In addition, the state legislates laws that go beyond natural duties—resolving indeterminacies in our rights and also creating new rights—and it criminalizes infractions of these laws, enforcing mala prohibita, not just mala in se. Finally, the state claims the right to punish offenders after violations have occurred, to deter others and provide them an incentive to conform to the law, ensuring security for all.17 Kant’s argument does not end at the duty to enter the civil condition, however. He also holds that we must coordinate on a common, public understanding of rights and duties in a particular fashion (Omnilateralism). The public scheme of rights ought not to be imposed unilaterally, since “a unilateral will cannot serve as a coercive law for everyone” (6:356). To demonstrate Kant’s position, imagine a private individual whose interpretation of justice is substantively correct—i.e., she holds justified, true beliefs about everyone’s moral rights—and who is endowed with sufficient power to successfully bring other people to coordinate around her view. Perhaps she has access to a new weapon no one else possesses. The details do not matter: She has some effective way of bringing others into line. Though she could successfully impose a perfectly just scheme of rights by

20

Anna Stilz

acting unilaterally, Kant claims that she has some pro tanto reason not to do that. The fact that her judgment is correct is not sufficient to give her the right to force others to comply with her decisions. To fully understand Kant’s account of legitimate authority, we need to understand the pro tanto wrong involved in unilateral enforcement. We also need to understand how this wrong might be overcome, that is, how an agent could be in a position to omnilaterally enforce a scheme of rules about rights. Here Kant argues: The legislative authority can belong only to the united will of the people. For since all right is to proceed from it, it cannot do anyone wrong by its law. Now when someone makes arrangements about another, it is always possible for him to do the other wrong; but he can never do wrong in what he decides upon with regard to himself (for volenti non fit inuria). Therefore only the concurring and united will of all, insofar as each decides the same thing for all and all for each, and so only the general united will of the people, can be legislative. (6:314)

Under what conditions could an agent who interprets and enforces justice be said to represent the “general united will of the people”? To qualify, I argue that such a governing agent must be authorized by cooperators who share a joint intention to act together to establish justice through particular institutions. Cooperators, recall, are those individuals who are willing to acknowledge the basic Kantian natural duty of justice, recognizing and respecting one another as bearers of a claim to freedom-as-independence (Natural Duty), and who adhere to the duties of basic justice and political coordination that flow from this commitment (State). Cooperators who share a will constitute a collectively self-determining people. Two Accounts of the Wrong of Unilateralism What exactly is wrong with unilateralism? I believe there are two problems, both usually present, but one more fundamental than the other. The first—in my view less fundamental—problem concerns inequality: Unilateral enforcement sets up a hierarchical relationship between the parties involved. Suppose your state-of-nature

Legitimacy and Self-Determination

21

neighbor goes ahead and uses her powerful weapon to enforce her objectively correct interpretation of private rights. Even though she implements a substantively just scheme, she wields unequal power over us. This might lead the rest of us to see and treat her as our superior. If so, then our social relations might fail to exhibit certain intrinsically valuable features—like parity of respect and status—that relations among equals ought to have. The basic thought is that unilateral enforcement sets up two classes of people: the powerful rulers, on the one hand, and the others, who must obey the decisions that the rulers make. If equal social relations are of intrinsic value, then your neighbor has pro tanto reason to refrain from acting unilaterally—i.e., to refrain from using her superior private power to implement the substantively correct decision. Recent arguments for the authority of democracy have appealed to this inequality worry about unilateralism, and to the role that democratic procedures play in overcoming it.18 On these accounts, since social equality is intrinsically important, we have pro tanto reason to make decisions democratically, and to treat those decisions as authoritative, even if, in doing so, we settle on a distribution of rights, opportunities, and resources that is less substantively just than we might have achieved through the imposition of one particular agent’s view. The pro tanto reason in favor of egalitarian authority might sometimes be outweighed by countervailing concerns (e.g., for the quality of political decisions), but it is an important reason nonetheless. Though social equality is important, I doubt it is our most fundamental worry about unilateralism. Note too that when our state-of-nature neighbor enforces her interpretation of justice, she does so in a way that ignores and supplants everyone else’s rational agency. An important part of treating others as independent persons involves respecting them as autonomous rational deliberators, who can reason for themselves how to act. Yet our imagined neighbor is insensitive to the need to engage her fellows in this way. She does not offer them any reasons that might lead them to share her point of view about what justice requires, nor does she inquire into, or respond to, their reasons for not sharing it. Instead, she imposes her judgments by force, leaving others powerless to do anything but go along. This, it seems to me, is a disrespectful method for organizing a society. It also fails to

22

Anna Stilz

acknowledge others’ claim to live in a social world that makes sense to them, that in some way reflects their own convictions about how society should be arranged. Of course, before coercing her fellows, the neighbor could offer them reasons, hoping to bring them to endorse her rule. Still, this may not suffice to treat others as autonomous co-deliberators. Consider three scenarios. In the first case, reflecting on the reasons she offers, her fellows come to see her rule as justified and form the intention to cooperate in carrying out her directives. Here, there is no disrespect to their rational autonomy. In the second case, while her fellows see the substance of their neighbor’s decisions as justified, they do not see it as equally justified that she be the one to make and impose these decisions. Why not instead act together to establish a just scheme of rights, rather than being subjected to their neighbor’s self-appointed rule? In the third scenario, her fellows do not come to endorse the substance of her decisions: Their moral judgments are simply too far removed from hers. Whatever justification there may be for the scheme she imposes, it is not one they can appreciate. Faced with either of these latter two scenarios, suppose that the neighbor goes ahead and coerces her fellows anyhow. If she does so, they will suffer three related harms. First, they are unable to act independently, on the basis of their own sense of justice, to establish and comply with a political order they can affirm. Instead, they see their activities coordinated by external threats they do not endorse (or—in the third scenario—even comprehend). Because of this, they are likely to find life under their powerful neighbor’s rule significantly alienating. Subject to the neighbor, I imagine I would feel as though a hostile force exercised near-complete control over my life. Finally, participants will not experience their political and social world as a cooperative enterprise, in which they have valuable relationships with other participants, or where they can feel at home. Because it involves these important harms, I believe unilateral coercion is normally pro tanto wrong. There is an important class of exceptions: The requirement to refrain from unilateral coercion is restricted to those cooperators who are willing to recognize and respect others as possessing an equal moral right to independence, and to cooperate in establishing a public authority to

Legitimacy and Self-Determination

23

specify and enforce this right. Some people—think of murderers or rapists—refuse to recognize others’ independence, even in a minimal way. Suppose I hold you back while you are trying to stab me—thus thwarting your unjust attempt to kill me. It is not reasonable for you to press as an objection to my coercion that you find it alienating. By attempting to murder me, you have denied me any agency independent from your will. When someone refuses to acknowledge very basic elements of personal autonomy, his denial of my moral status releases me, in turn, from any requirement to respect his independence, by refraining from unilaterally coercing him. Of course, more needs to be said about how to define the basic elements of independence that condition the requirement to refrain from unilateral coercion, and I return to this issue below. But in cases where people refuse to acknowledge others as bearers of a claim to independence, there is no moral loss in coercing them in ways that do not reflect their judgments. So long as people are willing to recognize others’ basic claim to personal independence, however, I believe it is pro tanto wrong to coerce them unilaterally. This is so even when their interpretation of the precise demands of mutual independence is mistaken. When other people attempt, in good faith, to respect the claims that, by their lights, I have—even when their interpretation of my claims is incorrect—I respond to them very differently than I do to people who fail to acknowledge that I have any claims at all. Though they are mistaken, they do not treat me with contempt. For that reason, their capacity for making their own political judgments places demands on me, and I owe them respect for that capacity. I have pro tanto reason to interact with them using persuasive means, rather than coercively. I should note that even in cases where unilateral coercion is wrong—in the case of cooperators who respect others’ independence and are willing to act together to establish a common public authority—sometimes this wrong can be morally outweighed by other important social values. Perhaps a functioning legal system cannot be achieved unless we subject some dissenters to unilateral coercion. If there is no other way to secure the essentials of justice, security, and public order, the wrong of unilateral coercion— though important—may be trumped by its strongly beneficial effects. As noted earlier, a justified humanitarian intervention or

24

Anna Stilz

military occupation may be a scenario of this kind, where there is a permission to unilaterally coerce despite the moral cost to selfdetermination. So the constraint on unilateral coercion can be lifted where its recognition would threaten decent governance or entail grave social harms. While self-determination is of very great weight, it is not the only value, and the requirement to respect it is not absolute. I return to this issue below. To sum up these reflections: On my view, there is a distinct autonomy worry about unilateral coercion. The principle of individual selfdetermination holds that, as rational agents, we ought to be able to direct our lives, to a significant degree, according to our own judgments. This idea grounds important basic liberties, like freedom of speech, conscience, association, privacy, and the freedom to form valuable intimate relationships. These private liberties enable us to carry out our own values in areas central to our personal identity and self-conception, by granting us a protected sphere of individual freedom. I believe the autonomy principle also has implications for the making and enforcement of political decisions. A scheme of rules about rights might be imposed in a manner that manifests proper respect for the autonomy of its subjects, or it might not. If a set of rules is forced on a population of cooperators over their explicit objections, and against their will, then its imposition fails to adequately respect their autonomy. This is so, I believe, even if that scheme of rules sets up an otherwise just distribution of rights, duties, and material advantages. Surely it would be better to decide and carry out the rules governing our common life through the deliberative agency of those subject to them, rather than via a process that circumvents subjects’ agential capacities. The idea that a legitimate procedure for imposing rules about rights must manifest appropriate respect for each person’s rational autonomy is still vague. There are many problems about how exactly to understand this requirement, especially when there is internal disagreement among members of a group about collective decisions, and I will return to these problems later. But for now, I note that if political autonomy is indeed an important value, then unilateral enforcement in the state of nature fails to respect it. I also think failure of respect for autonomy is at the core of our complaints about colonialism. Of course, colonialism was wrong for many reasons, including human rights abuses, economic

Legitimacy and Self-Determination

25

exploitation, and racism. But unilateral coercion was one of its key wrong-making features. A particularly destructive effect of “civilizing” colonialism, for example, was the forcible imposition, onto a subject population, of a social order that bore no relation to their own judgments about how they should be governed. Those who lived through this experience tell of a sense of powerlessness, and a loss of orientation and control.19 This produced lasting alienation among these peoples, a problem distinct from other abuses perpetrated by colonial institutions, and one that persists today in the form of great bitterness and resentment. Even in the best imaginable scenario, where colonial institutions are substantively just, still they deny the autonomy of colonized subjects, disregarding their claim to shape their common life on the basis of their own judgments. Colonialism and Democracy Of course, most historical cases of colonialism also raise concerns about social equality, since colonial rulers typically stood in relations of political superiority to those they ruled. Subject peoples were often denied input into political decisions, which were made by a separate class of rulers. One might therefore object that I have misidentified the wrong in colonialism. Perhaps what makes colonialism pro tanto wrong is not a violation of autonomy, but rather a violation of the subject people’s democratic rights to exercise a fair share of political power. Yet I doubt that political inequality is truly the fundamental concern here. One way to see this is to ask whether our objection to colonialism would be neutralized if the colonizers treated their subjects as political equals. Consider the following case: Democratic Incorporation.20 Suppose that instead of extending support to the 2011 Libyan revolutionary movement, France had overthrown Qaddafi’s regime, occupied the country, and annexed Libya’s territory, much as it annexed Algeria in 1830. Further suppose that after annexation, France governed Libya reasonably justly and extended its inhabitants democratic participation rights within a wider French republic. Imagine that there were no distinctions between French citizens and “the former Libyans” in terms of their democratic or other rights. Would the former Libyans have had a complaint?

26

Anna Stilz

If the imagined incorporation is objectionably unilateral—as I believe it is—then it is hard to see how the democratic equality argument explains this. Because Qaddafi’s regime threatened grave humanitarian abuses, there was arguably a right to intervene in Libya in 2011.21 Since Libya was a non-democracy prior to the intervention, on the democratic argument, each Libyan had an outstanding individual claim to be enfranchised in the political decisions governing him.22 But France responded to these claims by granting the Libyans democratic rights after the annexation. Indeed, there are real colonial cases that followed this model, cases in which colonizers democratically enfranchised their subjects. In the 1950s, for instance, France granted full citizenship— with suffrage rights—to all adult men and women in its former Algerian colony.23 A second case is the United Kingdom of Great Britain and Ireland: From the Act of Union in 1801 until its independence in 1922, Ireland formed an integral part of a wider Britain, electing its own MPs to the British House of Commons.24 One might object here that an imposed regime, like the one envisaged above, cannot be democratically legitimate. The thought is that the incorporation itself was not authorized through an egalitarian procedure, even if later decisions were. But suppose that just prior to incorporation, a referendum had been held in the combined Franco-Libyan territory, and that a majority (composed almost entirely of metropolitan French) had voted in favor. Would the annexation then be legitimate? If legitimacy simply requires that all involved individuals have equal influence over the decision, it seems it would: No individual has been disenfranchised here. Yet intuitively, the annexation still seems objectionable.25 It might be further objected that a combined referendum does not represent a fair distribution of political power. But what—in addition to equal influence over the decision—is required for a fair distribution of political power? Presumably the idea is that granting citizens from larger countries the power to vote on decisions about whether to incorporate smaller countries has the de facto effect of disenfranchising them, preventing them from exercising any effective say in this matter. One might therefore supplement the equal influence criterion with some “boundary-drawing” principle. One possible criterion is that those people who are primarily affected by a political decision

Legitimacy and Self-Determination

27

are entitled to exercise a greater share of political authority about that decision. Applied to cases of annexation, this might suggest that people who reside in a territory are presumptively entitled to exercise greater influence over decisions about that territory’s annexation. After all, these people are most strongly affected by proposals to sharply depart from existing political arrangements.26 I am skeptical that affectedness will justify our intuitions about democratic incorporation. Intuitively, a fair merger requires two separate referenda—one within French territory, one in Libya. But the affectedness criterion simply tells us that since both French and Libyan residents are affected, both should have a say. It cannot tell us why that process should take the form of two separate votes rather than a single one (at least not without importing prior assumptions, beyond affectedness, about normatively appropriate boundaries between groups). Further, affectedness produces counterintuitive conclusions when applied to cases of decolonization. While annexation is a sharp departure from the status quo, so too was decolonization. Should French citizens, for that reason, have been allowed to vote in a referendum on whether Algeria should remain within their Empire? While French imperialists were affected, I do not think they were entitled to a say. Finally, as already discussed in the Introduction, affectedness tells in favor of far-reaching revisions to existing boundaries. Advocates of global democracy, for example, commonly appeal to the fact that our decisions affect people beyond our borders. Given the economic and political power of the United States, many proponents of affectedness support extending everyone a vote in US elections. For this reason, affectedness seems unlikely to support claims to collective self-determination. The objector might finally reply that a fair distribution of political power requires some other boundary-drawing principle. Perhaps boundaries should be demarcated, so far as possible, to align with people’s shared preferences, in order not to create permanent minorities. I am sympathetic to this proposal. Relatedly, the existence of majority voting procedures strikes me as insufficient to realize self-determination in the absence of further conditions on the group constituted by those procedures, including whether it contains subordinated minorities. But this view seems to me indistinguishable from my own: To be legitimate, a state needs

28

Anna Stilz

to reflect its people’s shared will as to how they ought to be governed. If this view is correct, however, it has implications beyond political annexation and decolonization. Groups that persistently find themselves in the minority when it comes to fundamental issues about how to organize political life, and who seek a sphere in which to shape their institutions in accordance with their distinctive shared goals, will also have pro tanto claims to political autonomy. To sum up, then, I believe the autonomy worry about unilateralism—the worry that unilateral enforcement fails to adequately respect subjects’ claim to guide their lives by their own rational judgments—is best positioned to explain our concerns about colonialism. Individuals have an interest in political autonomy: in being subject to political institutions that in some way reflect their judgments and priorities. In the next section, I try to say more about what political autonomy is. While autonomy is familiar in the personal context, it is unclear how it might be extended to the context of collective decision-making. What Is Political Autonomy? To move forward, it may help to say more about the nature of autonomy in general. In the broadest sense, autonomy describes a person’s ability to freely conduct her life on the basis of her own judgments and values. But this capacity can be broken down into two distinct components. The first component is self-directed agency. We act in a selfdirected way when we act on motives that are in some appropriate sense “our own.” On prominent “hierarchical” accounts of autonomy, a desire counts as the agent’s “own” when that agent endorses its functioning from the perspective of a higher-order attitude that has the authority to “speak for” the agent. This higher-order attitude reflects her practical judgment, her capacity to weigh and engage with reasons. The self-directed person, in other words, has her own evaluative and moral judgments, and she identifies with some of her occurrent desires, and rejects others, on the basis of these values. We can define self-directed agency, then, as action that is reflectively endorsed by the agent on the basis of her judgments.

Legitimacy and Self-Determination

29

Self-directed action, on this view, is not the same thing as doing what the agent wants, since not everything we desire accords with our practical judgments. An “unwilling” addict who acts on his desire for a drug he prefers not to take is moved by a desire that does not reflect his values.27 In taking the drug, he is not selfdirected. It is only when I act on desires endorsed by my practical judgments that I act in a self-directed way. Yet self-directed agency is not sufficient for autonomy. Autonomy also has a second, thinking component. An autonomous agent holds her judgments in a way that is responsive to her own independent reasoning processes. Of course, many of our values are acquired as a by-product of socialization, not necessarily through conscious reflection. But normally, an autonomous agent is able to critically reflect on her values, and to freely endorse or discard them, if she chooses to do so. We hold our values because we see their objects as good, and we are able to raise the question of whether or not they truly are good, by deliberating about this. If, upon deliberation, we no longer see our evaluative and moral commitments as supported by reasons, then we can modify them. Therefore, an autonomous agent’s practical judgments are undergirded by, and responsive to, her own authentic reasoning. One way of interfering with an agent’s autonomy is to interfere with the authenticity of that agent’s reasoning processes, through methods like manipulation, deception, brainwashing, or mind control. The neuroscientist who implants new values into a subject, preventing her from calling those implanted values into question, makes her the instrument of his judgments, not her own.28 So autonomy has two components: (a) self-directed action that reflects the agent’s evaluative and moral judgments, where those judgments are (b) supported by her authentic reasoning. The importance of autonomy tells in favor of granting individuals a considerable degree of control over their personal lives, especially when it comes to core identity-related features of that life, such as one’s choice of religion, associates, sexual partners, occupation, whether or not to found a family, and so on. The value of autonomy is less commonly thought to bear on the choice or enforcement of other-regarding social rules. Yet there is a classic line of argument—dating to Rousseau—which holds that personal autonomy has an important political analogue. The basic idea is

30

Anna Stilz

that just as our central life commitments express our own judgments, so too should our political institutions. Rousseau held political autonomy to be a key element in the justification of democratic procedures, subjection to which, he argued, enabled the subject to obey laws she prescribed to herself. A familiar rejoinder to this Rousseauian argument points out that voting procedures are insufficient to render obedience to law compatible with individual autonomy. Citizens have only a tiny, negligible influence over political decisions, and this influence rarely if ever affords them the power to ensure that these decisions reflect their own judgments. Here, it seems, we are ruled by the majority, we do not rule ourselves. I agree with the rejoinder’s proponents that democratic voting procedures are neither necessary nor sufficient to realize self-determination. Still, I think Rousseau was right to hold that there is a political analogue to personal autonomy, though I propose to conceive it differently than he did. Why might autonomy be an important value in the political context? Consider that the characteristic means by which political rules are imposed and implemented is via coercion or force. Like others, I think the use of this type of means poses a special threat to autonomy.29 The problem is not simply that coercion limits my options or deprives me of freedom. Many permissible acts of others—and even some acts of nature—can limit my options and deprive me of freedom. Yet even when deprived of the specific options I might have preferred, I am usually still able to make self-directed choices, setting and pursuing my own ends among the options that remain. But most cases of coercion or force go beyond this, subjecting the coercee to the will of the coercer. A coercer intends for his target to perform a relatively specific course of action, and he deliberately changes the circumstances of her choice so that she has no reasonable alternative to doing this. He “makes her” do what he wishes. Normally, this deprives the coercee of self-directed agency, substituting the coercer’s judgments for her own. Thus, coercion presumptively threatens autonomy. Note, however, that the state is a comprehensively coercive institution. States impose a whole web of threats upon their subjects, which structure the very social world in which those subjects live and operate. Moreover, political coercion is inescapable: Individuals cannot generally leave their state or take refuge in other social

Legitimacy and Self-Determination

31

associations to get away from it. So it is natural to think that subjection to such a comprehensively coercive institution raises a concern about its subjects’ ability to live their lives according to their own judgments. This is particularly true of alien political coercion, that is, political coercion that bears no relation to the priorities, values, and beliefs of those subjected to it. When one is pervasively subject to alien coercion, substantial aspects of one’s life can come to seem hostile, threatening, and completely beyond one’s grasp. In such a scenario, it can become difficult to maintain any sense of oneself as an agent who charts her life in accordance with her own purposes.30 The importance of the political analogue to personal autonomy, I believe, lies in mitigating this threat of alien state coercion. Where the state’s use of coercion reflects subjects’ own judgments as to how they should be governed, they are enabled to relate in a distinctive way to their state and to the constraints it imposes on them. Because the subject can see the point of these coercive demands—in terms of judgments she is herself committed to—she also sees reason to hold herself to such demands (though she may not always desire to do so). Here, the state is no longer an overwhelming, alien power, but rather a tool that allows her to more effectively carry out practical commitments that are her own. I suggest, then, that there is an important autonomy interest in enjoying what Niko Kolodny has termed correspondence. Someone enjoys correspondence when her political institutions match her judgments in some way.31 She lives under an institution that she endorses, accepts, or believes to be justified or appropriate. The importance of correspondence is not that congruence between subjects’ attitudes and their institutions is intrinsically valuable, but that where it exists, it allows for individuals to experience selfdirected agency, even while subject to political power. Of course, no individual’s personal priorities can be mirrored in each and every political decision. Yet diversity of political opinions does not make correspondence impossible. There is a secondorder sense in which an individual’s priorities are often reflected in group decisions: namely, when she shares a commitment to a valued cooperative enterprise, and to certain shared policies by which she believes that enterprise should be structured. A commitment to participating in collective political action, on this view, is important in enabling correspondence.32 Participants can freely accept their

32

Anna Stilz

group’s policies, and the outcomes that result from them, even when those outcomes diverge from their first-order judgments, so long as they share a commitment to the collective political venture and to the values and procedures that structure it. Consider a partnership undertaking a joint venture, say, running a coffee shop together. Several philosophers have offered theories of the agency involved in such cases, arguing that it is undergirded by a structure of joint intentions. As they cooperate over time, this group will likely develop some shared commitments about how their enterprise should run. This does not mean that the partners will converge in all their judgments. More probably, they will divide on some issues. Yet even when they do not converge, partners are often able to generate shared commitments—not reducible to their own judgments—about how their joint venture should go. Such commitments can unify a group even when they diverge in their views on specific issues. Once such shared commitments have developed, members will generally act on them, and feel entitled to expect the others to act on them. Because a group can arrive at shared commitments, they can develop a group standpoint. This standpoint consists of a set of core values and priorities about how to organize more specific plans of action. The group standpoint will not fully correspond to any particular person’s first-order views about what to value. Still, each member can accept the group’s standpoint and share in it, as long as he intends to participate in their joint venture. I believe there is a perfectly straightforward sense in which we can speak of this small-scale partnership as sharing a will. Their shared will is to cooperate together in running their coffee shop, and to license the values and priorities that make up their group standpoint. A shared will, then, is nothing more than an interlocking structure of cooperative intentions on the part of each participant, amid common knowledge on the part of all that those intentions obtain. Further, if they value their enterprise, members may feel slighted when their group’s standpoint is not respected by others. While I often disagree with my colleagues about whom to hire, I prefer that we make our own hiring decisions together—through our accepted consultation processes—even though that means recognizing some decisions with which I disagree. Indeed, I would

Legitimacy and Self-Determination

33

consider myself disrespected if the dean overruled our collective decision, even if the result was to impose my preferred candidate. Though our hiring decisions do not always correspond to my firstorder preferences, there is still an important, second-order sense in which my priorities are reflected in these decisions. I share a commitment to a valued cooperative enterprise together with these colleagues, and to certain shared policies by which I believe that enterprise should be governed. The object of a group’s shared commitments need not be first-order policy decisions: it can be fairly abstract ideals and political procedures. When individuals together share such collective commitments, I will say that they share a political will. Though there is more to be said about the notion of a shared will, I lack space to undertake that deeper investigation here. But for now, let me stress the important role that a group’s shared political will plays in ensuring correspondence for its members. When each member participates in a group’s shared will, and government imposes laws and policies that reflect that shared will, this use of political coercion will not be alien to the members. Though she does not personally endorse every political outcome, each member is governed by political institutions that she views as appropriate, since she endorses the higher-order values and procedures by which her joint political venture is conducted. Omnilateral coercion differs from alien coercion, then, because it is based on a joint intention—in which each cooperator shares—to act together to establish justice through particular institutions. Though correspondence is ultimately an interest of individuals, it can be furthered by an individual’s membership in a self-determining group, if she affirms her participation in the group and accepts the higher-order values and procedures that structure it. In this case, rule by the group’s institutions will be in the service of each individual’s interests in self-directed agency and in non-alienation. To be valuable, however, correspondence must meet four additional conditions. First, correspondence only serves autonomy under conditions where the subjects’ shared political will is freely formed in a manner that is accountable to their own deliberative processes. If rulers instead “implant” their subjects’ political judgments—through manipulation, deception, brainwashing, or

34

Anna Stilz

the like—then correspondence will not promote their autonomy. In this case, subjects will be instruments of their rulers’ judgments, not their own. Second, correspondence between a group’s shared will and their institutions must come about through some causal process. A group might enjoy correspondence merely due to the goodwill of a benevolent absolutist who gives the people what they think right. But to ensure political self-determination, a government must robustly reflect the citizens’ shared will. This requires some channel by which the people might revoke authorization of their government if it ceases to reflect their shared will. Third, correspondence is valuable only for a restricted domain of constituents and a restricted domain of judgments. The reason correspondence is valuable is that it promotes autonomy. But autonomy has a private as well as a public dimension, and political autonomy has moral weight only where it is consistent with basic elements of personal autonomy. The popular will that a state ought to represent is therefore a will that is shared among cooperators, those individuals who are willing to recognize and respect others as bearers of a claim to independence and to establish political institutions that will protect this claim. There is no requirement to refrain from unilaterally coercing non-cooperators. Finally, the state has reason to respond only to a restricted domain of the cooperators’ judgments, namely, their judgments concerning the appropriate way to establish justice, not their judgments concerning other matters, such as the good life. An omnilateral will is a will concerning how a group should coordinate to specify, interpret, and enforce the substantive rights defining the contours of their sphere of personal independence. It is not a will concerning religious, moral, or cultural values. In that sense, the state need not strive for congruence with all its subjects’ attitudes and judgments. Why care about correspondence, even for this restricted domain of agents, and for this restricted domain of judgments? Unless we suppose that the Kantian state reflects the shared will of its people, it is difficult to see how it could ever solve the problem of unilateral coercion characteristic of the state of nature. The problem of unilateral coercion is a problem of deciding for others what justice requires and compelling them to submit to one’s judgments.

Legitimacy and Self-Determination

35

This is supposed to be a problem even where the coercer’s view of justice is correct. But if unilateralism is a problem for private individuals in a state of nature, then it is also a problem for the state. What makes the state’s enforcement of a system of rights—even an objectively correct system of rights—any less unilateral than the powerful neighbor’s enforcement of justice in the state of nature? Only, I submit, the fact that the cooperators among its population affirm its procedures and institutions, judging this an appropriate way to enforce justice among themselves. In this way, they grant their state the standing to decide on their behalf. My political autonomy account overlaps in some ways with Michael Walzer’s ideas, so it may help to briefly discuss his view here. Walzer argues that “a state is legitimate or not depending upon the ‘fit’ of government and community, that is, the degree to which the government actually represents the political life of its people.”33 He holds that this explains why it is wrong to coercively interfere with another political community: Such interference forces its people to be governed by “someone else’s conceptions of political justice and political prudence.”34 Instead, he maintains that “people should be treated in accordance with their own ideas about how people should be treated.”35 I agree with Walzer that there is an important value for people in being governed in a way that reflects their own shared judgments, provided these judgments are consistent with basic claims to personal autonomy and independence. Like him, I have argued that state legitimacy requires a certain “fit” between government and its people’s shared political will. Unlike Walzer, however, I have argued that there is no reason to strive for “fit” with all possible political values. In particular, I do not believe Kantian political autonomy requires us to favor institutions that reflect a majority’s religious or cultural values. On my view, correspondence is valuable when it comes to judgments about how best to secure the conditions of mutual independence. Walzer, on the other hand, characterizes the interest in “fit” in a much broader way. He sees self-determination as the right of a community to be “governed in accordance with its own traditions” and links this to the existence of an “inherited culture.”36 To pull these threads together, then, let me define political autonomy as (a) a causal relation of correspondence (b) between

36

Anna Stilz

the shared political will of cooperators and their institutions, (c) when that relation is achieved under conditions that enable their free deliberative reasoning. Is Political Autonomy Possible? It is natural to object that, however desirable this ideal of political autonomy might be, it is unachievable. As Michael Blake notes, states “are composed of persons, who disagree and quarrel and find their own answers to a variety of disputed questions.”37 A good account of political autonomy must be compatible with these facts. Even if it is conceptually possible for a group to share a will, one might still object that it is extremely unlikely that an entire territorially defined population will achieve unanimity of this kind. There is a wide range of values and attitudes among us. Must a legitimate state be responsive to all these values and attitudes? No. As I have emphasized, there is reason to ensure “fit” between state institutions and the shared intentions only of a restricted domain of cooperators. Essential elements of cooperators’ commitments are: (1) a willingness to define a scheme of private rights that meets the conditions of what I shall call basic justice, and (2) an acknowledgment that this requires coordinating in a territorially defined state, which can enforce a unitary interpretation of the contours of those rights. Claims to political autonomy, then, are necessarily moralized claims. I also hold that even when unilateral coercion is wrong, because it disrespects a cooperator’s autonomy, sometimes this wrong can be outweighed. The requirement to respect self-determination is not absolute: It may be permissible to subject people to alien coercion if this is the only way to achieve social goals of truly overriding importance. I unpack these ideas by briefly outlining three paradigm cases of dissent, where there is some lack of “fit” between state institutions and the shared priorities of a subpopulation of the state’s constituents. In the first two cases, I believe it is permissible to subject dissenters to alien coercion, either because they fall outside the domain of cooperators or because their claims to autonomy, though important, are outweighed. In the third kind of case, however, I believe dissenters have a claim to an institutional means that would afford them greater political self-determination.

Legitimacy and Self-Determination

37

In my first paradigm case, dissenters lack the “fit” constitutive of political autonomy only because they refuse to acknowledge any duty to cooperate on minimally just terms. I believe there is some essential minimal content as to what could count as a reasonable interpretation of mutual independence. Where dissenters are unwilling to cooperate in a state that meets these minimal requirements, their claims to political autonomy have no weight because they are inconsistent with basic justice for others. Basic justice, as I define it, is a condition on political legitimacy: a condition for a group to have the right to make and enforce its own laws and policies. As I stressed earlier in this chapter, legitimacy and justice are distinct values. Sometimes it is impermissible to enforce even ideally just laws and policies, for example where their enforcement would involve colonial rule. Sometimes it is permissible for a state to enforce laws and policies that are not ideally just, so long as those laws are enacted in accordance with collectively authorized procedures. Still, the failure to acknowledge very basic elements of personal independence may undercut a group’s claim to decide collectively. A community respects basic justice, on my view, when it aims at securing mutual independence for its members, on a reasonable (though possibly mistaken) understanding of what that value means.38 Even where a community makes mistakes about justice, its laws and policies may be legitimate, as long as they reflect a good-faith attempt to define a scheme of rights that secures mutual independence for all. But some acts cannot be reconciled with even a minimal commitment to mutual independence. Basic justice sets that threshold. A minimally just state, on my view, must protect at least: 1. security rights, to freedom from torture, slavery, arbitrary imprisonment, and severe threats to personal integrity 2. subsistence rights, to an economic minimum capable of meeting basic needs 3. core personal autonomy rights, to freedom of conscience and thought, to personal property, and the freedom to form family relationships 4. the preconditions of collective self-determination, to free expression, free association, and public political dissent

38

Anna Stilz

Security and subsistence rights are prerequisites for autonomous agency of any kind. They define a minimum level of control necessary to lead a life of our own. These rights could be recognized—if imperfectly—even in a pre-institutional state of nature. Even in that scenario, others would have a duty to refrain from deliberately harming us in ways that infringe on bodily integrity or subsistence. Personal autonomy rights reflect the principle that each person should be free from the will of others when forming her own values and expressing these values in the conduct of her life. Personal autonomy is especially important when it comes to judgments about fundamental ethical obligations, in matters of religion and intimate relationships. Personal autonomy is a scalar value, and protections for it can be more or less robust. But no society that failed to protect religious freedom, freedom of thought, the right to family, and personal property, could be said to have even a minimal commitment to it. Finally, the preconditions of collective self-determination are meant to ensure that individuals have an opportunity to form their own autonomous political judgments. These rights ensure that citizens can develop their own political views and make informed assessments of their institutions. If members of a society are coerced or manipulated into their political values, outsiders have less reason to respect or defer to these commitments. Like personal autonomy, protections for collective self-determination can be more and less robust, but no society that controls its citizens’ access to information, prohibits free discussion and free association, or denies its citizens the right to publicly express political dissent safeguards these preconditions. On this interpretation, then, the Kantian view is committed to certain basic rights as threshold criteria for state legitimacy. Of course, most well-ordered states will implement substantive rights that go beyond these conditions, and outsiders and the international community may wish to promote and incentivize respect for a longer list of rights than this. But the idea is that a state must grant at least these minimal guarantees to have legitimate jurisdiction over its population and territory. It is crucial to keep this distinction between basic justice and other aspirations in view.39 For their claims to political autonomy to carry moral weight, then, claimants must be willing to cooperate together in a

Legitimacy and Self-Determination

39

legitimate state that can secure basic justice. We are not required, out of respect for their judgments, to allow dissenters to act in ways that clearly violate this moral requirement. To such dissenters— including fascists, racists, theocrats, and so on—we can only say: Your alienation must be discounted, because greater recognition for your values is not compatible with upholding basic justice for others. More difficult, perhaps, is the case of dissenters—such as anarchists—who may acknowledge duties of basic justice in principle but refuse to accept the derivative duty to support a state. If Kant’s argument for the injustice of anarchism is correct, however, we also have a reply to these dissenters: A territorial authority is necessary to specify, interpret, and enforce a common, public interpretation of rights. If a state is necessary to secure basic justice, then those who live alongside the anarchist have a claim against him to support and comply with that state.40 Finally, some dissenters are alienated because their values are incompatible with acknowledging others’ equal claims to political autonomy. At the moment of decolonization, for example, many British imperialists were disaffected at the thought that they could no longer continue living in the glorious British Empire. Should their disaffection count as a reason to coerce Indians to continue to be ruled by the Raj? No. While dissenters can permissibly be forced to cooperate against their will when their cooperation is essential to sustaining basic justice for others, that is the only reason they can permissibly be forced to cooperate. Since basic justice for the British could perfectly well be guaranteed within “little” Britain, imperialists had no right to coerce the alienated Indians to uphold their imperial identity. Using coercion to force a dissenter to uphold someone else’s identity, without any further justice-based rationale, is wrong. So if greater recognition for one person’s values would involve forcing the unwilling cooperation of another—under conditions where the latter’s cooperation is not morally mandatory—then the former’s dissent must be discounted. These three scenarios together make up our first paradigm: cases where dissenters in some way refuse to recognize the equal autonomy of others. Claims to self-determination must be understood as moralized claims: Not everyone has a claim to be coerced by

40

Anna Stilz

a state that reflects their judgments, because not everyone’s judgments are worthy of respect. Where dissenters refuse to acknowledge the requirements of basic justice, to recognize a duty to cooperate in a legitimate state, or to respect others’ equivalent claims to self-determination, then their values and priorities can be overridden without any moral loss. Let me comment briefly on a second paradigm where dissent should be discounted. These are cases where dissenters’ priorities are compatible with others’ rights in principle but not in practice, due to institutional infeasibility constraints. Here I appeal to the importantly territorial nature of our natural duties of justice: We cannot establish a unitary interpretation of property and contractual rights, enforce those rights, and punish violators, unless people who live in proximity and interact regularly are subject to the same institution. If each person signed up for the jurisdictional organization of his choice, then interactions between persons in the same contiguous space would continue to generate conflicts and disputes.41 If disaffected individuals refuse participation in any feasible institutional configuration, then their alienation should be overridden. To explain this, I stress once again that selfdetermination is not absolute: When respecting it would entail very grave social costs, it can be outweighed. In this case, there is indeed a pro tanto wrong in coercing dissenters who fail to participate in a group’s shared political will. But this wrong is outweighed by the fact that there may be no other feasible way to secure the essentials of justice, security, and public order. Where very weighty social goods cannot otherwise be achieved, it is permissible, all things considered, to subject some dissenters to alien coercion, despite the pro tanto wrong to them. Finally, there are some scenarios where I believe we have a duty to allow for the self-determination of those whose shared judgments do not “fit” with their institutions. This brings us to the third paradigm case, where alienated groups have a claim to an institutional response. Here I have in mind groups who have shared political commitments that are (a) consistent with the provision of basic justice for others, (b) who possess or can create a territorially organized structure of representation, and (c) whose dissent can be feasibly addressed, at reasonable cost, by granting them separate political institutions. In cases of this kind, I believe we compromise the

Legitimacy and Self-Determination

41

political autonomy of dissenters by coercing them within alien institutions that they reject. Because a variety of institutional configurations are consistent with our natural duty of justice, it is often not true that coercion by the current coercer is necessary for basic justice to be realized. The innate Kantian right to independence—the right to determine our lives in accordance with our own judgments— gives people standing to reject alien state coercion whenever that coercion is not necessary to securing basic justice for others. This means that people’s shared political will as to how they ought to be governed should be extended some respect, provided that is consistent with fulfilling basic duties of justice. Let me now sum up this account. To have an exclusive moral right to make and enforce law and policy within a particular territory, on my view, a state must show: 1. That it meets the requirements of basic justice, enacting a scheme of private rights that guarantees security, subsistence, a minimum of personal autonomy, and the preconditions for collective self-determination for all its subjects. 2. That its institutions reflect the common will of a significant majority of cooperators among these occupants—i.e., their actual shared commitment to a political endeavor, and to a set of procedures as to how that endeavor should be governed. 3. That those who dissent from this shared will lack any right to an institutional alternative, because (a) their claims are inconsistent with respecting others’ independence; or (b) their priorities cannot currently be accommodated by a feasible alternative institution, and so are outweighed. I believe this account provides a convincing, and broadly Kantian, response to the worries about colonialism and involuntary annexation with which we began. The essential thought is that our natural duty of justice has two dimensions: On the one hand, we are bound to respect others’ personal independence, which requires us to cooperate together in a legitimate state that can enforce private rights; but on the other, we are also bound to respect their political independence—the moralized claim, held by cooperators, to be ruled in a way that reflects their shared judgments.

42

Anna Stilz

This value of self-determination, in my view, provides us moral reason to favor a system of separate states to a single world state. While coercion by some minimally just state is necessary for securing private autonomy, by itself, this is insufficient to decide how many states we should have, or which particular people should be subject to which ones. Should we institute a unitary world state, or a plurality of states? A federal system? Internal autonomy for indigenous peoples or other minorities? The view I have developed suggests we should adopt the feasible institutional configuration that ensures the greatest “fit” between people’s convictions regarding how they should be governed, on the one hand, and the institutions that rule them, on the other. So long as they are willing to respect others as independent equals, people have a pro tanto claim to be ruled in a way they endorse. If basic justice can be provided within a plurality of distinct states, then those who prefer separate institutions will have a prima facie claim to establish them. Further, it is the violation of self-determination that explains what is wrong with colonialism and involuntary annexation. Securing basic justice for people in the metropole almost never requires coercively imposing a political order on unwilling subject peoples. Conclusion To conclude, I briefly characterize the relation between my political autonomy account of self-determination and more familiar voluntarist and nationalist views. Critics argue that insofar as Kantians try to provide principled reasons for favoring self-determination, they can do so only by importing elements of these other accounts. It is true that the political autonomy theory is sensitive to some concerns shared by voluntarism and nationalism. But the theoretical structure of the political autonomy account is significantly distinct from each of these rivals. The political autonomy account agrees with the voluntarist that people’s actual attitudes can matter for political legitimacy. But these attitudes have moral weight only where they are consistent with the prior Kantian duty to respect others’ equal right to independence by coordinating in a territorially defined state. Further, there need be no communicative act of consent on the part of individuals in order for a state’s rule over a particular population

Legitimacy and Self-Determination

43

to be legitimate. As long as the cooperators in the state’s population share a political will that supports the state’s institutions, and their shared will was formed under conditions of free deliberative reasoning, the state will be legitimate. With respect to nationalism, my account agrees that citizens’ identification with their political community—i.e., their shared commitment to political action—is important in enabling state legitimacy. But, unlike nationalists, I doubt that shared culture plays any necessary role in facilitating this identification. Culturally diverse groups can and often do identify as part of one joint venture. Further, I believe that nationalist models of political identity are deeply unfair in most contemporary states, which contain people from widely diverse political backgrounds. The more culturally “neutral” states become, I believe, the more likely they are to bring diverse citizenries to affirm their political cooperation together. Notes 1. References are to Kant’s “Doctrine of Right.” I cite Kant’s writings by the standard German edition of Kant’s works, Kant’s Gesammelte Schriften, edited by the Academy of Sciences (Berlin, Walter deGruyter & Co. 1900–). These numbers are widely noted in the margins of English translations. 2. See A. John Simmons, Boundaries of Authority (Oxford: Oxford University Press, 2016), ch. 3, and Margaret Moore, A Political Theory of Territory (Oxford: Oxford University Press, 2015), 97. 3. Simmons, Boundaries of Authority, 60. 4. See Allen Buchanan, Justice, Legitimacy, and Self-Determination (Oxford: Oxford University Press, 2004), 70. 5. For this charge, see David Miller, “Neo-Kantian Theories of SelfDetermination: A Critique,” Review of International Studies 42:5 (2016), 858–875. See also Simmons, Boundaries of Authority, p. 75. 6. For a similar critique, see Lea Ypi, “What’s Wrong with Colonialism,” Philosophy and Public Affairs 41:2 (2013), 168. 7. Neta Crawford, Argument and Change in World Politics: Ethics, Decolonization, and Humanitarian Intervention (Cambridge: Cambridge University Press, 1996), 236–239. 8. A. John Simmons, Justification and Legitimacy (Cambridge: Cambridge University Press, 2001), 122–157, also distinguishes a state’s legitimacy from what he calls its justification (roughly synonymous with what

44

Anna Stilz

I refer to as “justice”). For a similar argument, see Philip Pettit, On the People’s Terms (Cambridge: Cambridge University Press, 2012), ch. 3. 9. See Buchanan, Justice, Legitimacy, and Self-Determination, 247. 10. John Rawls, A Theory of Justice (Cambridge, MA: Belknap Press of Harvard University Press, 1999), 6. 11. David Miller, Global Justice and National Responsibility (Oxford: Oxford University Press, 2007), 124. See also Avishai Margalit and Joseph Raz, “National Self-Determination,” Journal of Philosophy 87:9 (1990), 439– 461. 12. Simmons, Justification and Legitimacy, 129. 13. Thanks to Jonathan Quong for helpful comments to which this summary is indebted. 14. Similar points are raised by Jeremy Waldron, “Special Ties and Natural Duties,” Philosophy and Public Affairs 22:1 (1993), 24, and Thomas Christiano, The Constitution of Equality (Oxford: Oxford University Press, 2008), 237. 15. See Louis-Phillippe Hodgson, “Kant on Property Rights and the State,” Kantian Review 15:1 (2010), 57–87. 16. For an account of legitimacy that emphasizes the need to coordinate amid disagreement, see Jeremy Waldron, Law and Disagreement (New York: Oxford University Press, 1999). 17. Kant claims that punishment is something only a superior can do to a subordinate (6:347). 18. Niko Kolodny, “Rule over None II: Social Equality and the Justification of Democracy,” Philosophy and Public Affairs 42:4 (2014), 287–336; Daniel Viehoff, “Democratic Equality and Political Authority,” Philosophy and Public Affairs 42:4 (2014), 337–375. 19. For powerful accounts, see Luther Standing Bear, My People, the Sioux (Lincoln, NE: Bison Books, 2013); Jonathan Lear, Radical Hope (Cambridge, MA: Harvard University Press, 2006). 20. I introduced this case in Anna Stilz, “Decolonization and SelfDetermination,” Social Philosophy and Policy 32:1 (2015): 10–11, and the next three paragraphs follow the discussion there. 21. I recognize that this controversial judgment could be challenged. But my argument does not hang on whether the particular intervention in Libya was justified. So long as the reader agrees that intervention in a non-democratic country can sometimes be justified, she should feel free to substitute a case she regards as meeting the appropriate criteria of justification. 22. Libya did have an organ of political representation under Qaddafi—the General People’s Congress. However, real power remained

Legitimacy and Self-Determination

45

with Qaddafi himself: There was no right to form political parties or contest elections. 23. See Todd Shepherd, The Invention of Decolonization (Ithaca, NY: Cornell University Press, 2006), 19–54. 24. See John Ranlegh, A Short History of Ireland (Cambridge: Cambridge University Press, 1983). 25. For related discussion of the democratic “boundary problem,” see Arash Abizadeh, “On the Demos and Its Kin: Nationalism, Democracy, and the Boundary Problem,” American Political Science Review 106, no. 4 (2012), 867–882; Robert Goodin, “Enfranchising All Affected Interests, and Its Alternatives,” Philosophy and Public Affairs 35, no. 1 (2007), 40–68; and David Miller, “Democracy’s Domain,” Philosophy and Public Affairs 37, no. 3 (2009), 201–228. 26. I thank Jonathan Quong for suggesting this argument to me. 27. For this example, see Harry Frankfurt, “Freedom of the Will and the Concept of a Person,” in The Importance of What We Care About (Cambridge: Cambridge University Press, 1988), 17. 28. Cf. Michael Walzer’s worry about a Swedish government that introduces a wondrous chemical into Algeria’s water supply, turning the Algerians into Swedish-style social democrats. Surely the objection to this intervention is an objection to manipulation: It circumvents the Algerians’ ability to rationally form and implement their own political judgments. Michael Walzer, “The Moral Standing of States,” in Thinking Politically, ed. David Miller (New Haven, CT: Yale University Press, 2007), 231. 29. Many other thinkers have suggested this. See Frankfurt, The Importance of What We Care About, 42; Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986), 154; Philip Pettit, On the People’s Terms: A Republican Theory and Model of Democracy (Cambridge: Cambridge University Press, 2012), 147; Michael Blake, Justice and Foreign Policy (Oxford: Oxford University Press, 2013), 84; John Rawls, Political Liberalism (New York: Columbia University Press, 2005), 137. 30. Simmons, Justification and Legitimacy, 136. 31. Kolodny, “Rule Over None I: What Justifies Democracy?,” 199. 32. For a similar view, see Ronald Dworkin, Freedom’s Law (Cambridge, MA: Harvard University Press, 1996). 33. Walzer, “The Moral Standing of States,” in Thinking Politically, ed. David Miller (New Haven, CT: Yale University Press, 2007), 223. 34. Ibid., 224. 35. Walzer, “Nation and Universe,” in Thinking Politically, 198. 36. Walzer, “The Moral Standing of States,” 227, 224, 220. 37. Michael Blake, Justice and Foreign Policy, 32.

46

Anna Stilz

38. For a similar view, see Ronald Dworkin, Justice for Hedgehogs (Cambridge, MA: Belknap Press of Harvard University Press, 2011), 322. 39. For the idea that the function of human rights is to set bounds to states’ external legitimacy and can trigger duties of international assistance, interference, or intervention, see John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999), 79–80; Charles Beitz, The Idea of Human Rights (Oxford: Oxford University Press, 2009), 116; Joseph Raz, “Human Rights without Foundations,” in The Philosophy of International Law, ed. J. Tasioulas and S. Besson (Oxford: Oxford University Press, 2010). I particularly emphasize the role of basic justice as setting a threshold for coercive intervention. 40. For a similar line of thought, see Jonathan Quong, Liberalism without Perfection (Oxford: Oxford University Press, 2011), 128. 41. For similar points, see Christopher Wellman in Wellman and Simmons, Is There a Duty to Obey the Law (Cambridge: Cambridge University Press, 2005), 14–16, and Quong, Liberalism without Perfection, 129.

2 IN DEFENSE OF FUNCTIONALISM JONATHAN QUONG

Anna Stilz’s rich and excellent chapter begins with an apparent problem for the Kantian view of political legitimacy, at least when this view is understood as a functionalist view. To simplify somewhat, the functionalist view tells us that if a state performs various morally mandated functions (e.g., determining property rights, providing basic security) in a reasonably just manner, this is sufficient to establish that state’s legitimate authority with regard to some territory. A state is legitimate when it has “an exclusive moral right to make law and policy on behalf of a particular group and to use coercion or force to implement those laws and policies.” Although the functionalist view faces a number of important objections, the objection that is central to Stilz’s chapter is the following: The view cannot explain the importance of self-determination, in particular, it cannot explain the wrongness of colonialism or involuntary annexation in cases where this is achieved without violence or threats of violence, without violations of other basic civil or political liberties, and where the regime imposed performs the morally mandated functions in a reasonably just manner. I will call this the objection. Although Stilz is sympathetic to the functionalist picture of legitimacy, she agrees with the critics that the functionalist view, on its own, is vulnerable to the objection. She does not, however, seek to supplement the Kantian view with appeals to consent or nationality. Instead, she appeals to a Kantian conception of political autonomy in developing an account of self-determination. Stilz argues that just as we have an important interest in personal autonomy, we have an important interest in political autonomy—that 47

48

Jonathan Quong

is, an interest in having our political institutions correspond to our values and judgments. And shared political autonomy is possible when a group of people share a will. They share a will when there is an interlocking structure of joint intentions among participants, and common knowledge of this fact. The object of the shared will need not be first-order judgments. What is important is that there be correspondence between citizens’ values and judgments and the polity’s governing procedures and ideals. Colonialism and involuntary annexation can thus conflict with the political autonomy of the subjugated people by imposing on them ideals and procedures to which they don’t subscribe. This account of the wrongness of colonialism is not, moreover, reducible to either consent-based or nationalism-based explanations, since the shared will need not involve actual consent, nor need it involve a national identity. Stilz thus offers a distinctively Kantian account of legitimacy and self-determination that evades the objection to which the functionalist view is apparently vulnerable. I am not persuaded, however, that the objection to the functionalist account succeeds. There is a straightforward way for the functionalist to explain the wrongness of the cases of colonialism and annexation that Stilz has in mind, and thus there is no need to appeal to any notion of a shared will to handle these cases. I develop an objection to Stilz’s conception of political autonomy in the first section, before going on to explain how the functionalist view can effectively respond to the objection in the next section. I then consider some objections to my proposal, and the final section of the chapter offers some conclusions. An Objection It is very plausible to suppose that I have a morally legitimate interest in shaping my life: in self-directed action that is also selfreflective in the way described by Stilz. It’s less clear, however, that I have a morally legitimate interest that my political institutions correspond to my values or judgments. Whether I have such an interest depends on how exactly we formulate the notion of correspondence. In chapter 1, Stilz says, “Someone enjoys correspondence when her political institutions match her judgments in some way.” This occurs when “She lives under an institution that she endorses,

In Defense of Functionalism

49

accepts, or believes to be justified or appropriate.” If correspondence requires only that my political institutions can be justified to me in light of values or principles that I endorse—or that I would endorse at some level of idealization—then I agree that this is a morally legitimate interest. But when understood in this way, correspondence seems too weak for Stilz’s purposes. After all, a state that performs all the morally mandated functions—protects basic rights and freedoms, secures a fair distribution of resources, etc.—is justifiable to each citizen in light of values or principles that each endorses or would endorse at a suitable level of idealization. If Stilz’s account of legitimacy is distinct from the functionalist view, there must be more to correspondence than this. Suppose, then, that correspondence involves something thicker. Perhaps even when political institutions can be justified to everyone, there may be a lack of correspondence when those institutions have not been shaped by the shared will of the members of the polity. The members must correctly see the correspondence between the institutions and their judgments as having been brought about by, or caused by, their joint intentions to create or sustain a political community. For example, when Stilz fleshes out the notion of correspondence, she says, “correspondence between a group’s shared will and their institutions must come about through some causal process. A group might enjoy correspondence merely due to the goodwill of a benevolent absolutist who gives the people what they think right. But to ensure political self-determination, a government must robustly reflect the citizens’ shared will. This requires some channel by which the people might revoke authorization of their government if it ceases to reflect their shared will.” I’m not sure, however, that this causal process condition suffices to distinguish Stilz’s view from the functionalist view. After all, on any plausible functionalist view of political legitimacy, one of the state’s morally mandated functions will be the protection of individual rights, including the rights to vote and run for elected office in free and fair elections. The functionalist can thus also insist that there is a causal process that ensures political institutions will robustly reflect the will or preferences of citizens. To really distinguish political autonomy from the functionalist view, correspondence must go beyond the morally mandated functions of basic justice and good governance. For example, it

50

Jonathan Quong

could apply to citizens’ views about culture, morality, religion, or the good life. As a shorthand, let’s call all such matters non-justice issues. If political autonomy requires correspondence on non-justice issues, this would clearly distinguish Stilz’s view from the functionalist theory. But Stilz rejects this construal of correspondence and political autonomy. She says, “the state has reason to respond only to a restricted domain of the cooperators’ judgments, namely, their judgments concerning the appropriate way to establish justice, not their judgments concerning other matters, such as the good life. . . . In that sense, the state need not strive for congruence with all its subjects’ attitudes and judgments.” But if correspondence is limited to judgments concerning the appropriate way to establish justice, Stilz’s account is disabled from explaining the wrongness of colonialism or involuntary annexation by appeal to the way such acts prevent the colonized society from having political institutions that correspond to the local majority’s religion, culture, or conception of the good. At the very least, this conflicts with what many victims of colonialism or annexation believe is wrong about what is done to them. I suspect that in many cases, the victims of colonialism or involuntary annexation care about political self-determination precisely because they see self-determination as essential to ensuring that their political institutions reflect local views about culture, religion, or the good life. Stilz, on the other hand, must insist that the wrongness of colonialism or annexation in these cases is explained by a failure to respect the shared will of the local population with regard to matters of justice alone. Consider the following case: Just Annexation: Country A’s political institutions do very well in establishing justice for all its citizens. Country A involuntarily annexes country B without any bloodshed or violation of basic rights, and country A extends equal rights to the newly annexed members of country B. The members of country B do not object to the conception of justice (or the democratic procedures) embodied in country A’s institutions, indeed they endorse roughly the same conception and procedures. But the majority of the members of B are strongly opposed to the annexation. The basis of their opposition is grounded in the fact that they are a historically Catholic

In Defense of Functionalism

51

nation, and they wish their political institutions to reflect that fact. Insofar as there is a shared will or shared political autonomy among the members of B, it makes essential reference to the Catholic nature of their national identity.

Stilz’s account does not seem able to explain the wrongness of annexation in this case. I take this to be a serious problem for the account.1 Stilz could avoid this problem by modifying the conception of correspondence. Instead of saying that the state has reason to respond only to a restricted domain of judgments, namely, judgments concerning justice, Stilz could instead say that correspondence concerning matters of religion or the good life is also important so long as achieving such correspondence does not involve clear or serious injustice. Correspondence for religious extremists who want to prohibit the free exercise of other religions would require violating people’s basic rights, and thus this form of correspondence has no weight. But so-called modest forms of perfectionism might seem different. Suppose the majority of the citizens in country B wish to have Catholicism recognized as the official state religion, and they also wish to have Catholic churches and charities subsidized by the state, but they do not wish to restrict anyone’s freedom to practice their preferred religion. These policies might not be unjust, or at least not seriously unjust. Thus, on the modified conception of correspondence, country A acts wrongly by thwarting the political autonomy of the members of B with regard to religious policies that are not unjust, or at least not seriously so. The difficulty, however, is that the modified conception of correspondence sits very uncomfortably with a Kantian view about the role or function of the state. The state’s purpose, on the Kantian view, is to secure the conditions of mutual independence or equal freedom for all citizens. This isn’t possible in the state of nature for the familiar reasons Stilz describes in her chapter, most importantly, the contours of individual rights are indeterminate and uncertain, and unilateral efforts by some to impose their views about justice on others do not secure equal freedom, rather, they subordinate some to the judgment of others. The state overcomes this problem by establishing a general system of rights and duties that are enforced by an omnilateral, rather than a unilateral, will.

52

Jonathan Quong

On this view, the role of the state is to render our rights and duties determinate, thereby securing the conditions of mutual independence, where each person pursues his or her own plan of life, free from the wrongful interference of others. This picture of the state does not seem compatible with a view where political institutions are shaped to correspond to some people’s conceptions of the good or judgments about what has intrinsic value. If state power is used in this way, some people might be constrained from pursuing their plans, not on the basis that the constraints are required to secure equal freedom for all or mutual independence, but rather because doing so is required to ensure a better degree of correspondence between our shared political institutions and some people’s values or judgments about issues that are not matters of justice. Using the power of the state for this reason is not justifiable on the Kantian view, at least as I understand it. In Kantian terms, others cannot use or damage my means (those things that are rightfully mine) without my consent, merely so that they may pursue their own purposes. To constrain me in this way involves subordination and is a violation of the categorical imperative. The Kantian state must be neutral about the good life.2 Stilz evidently agrees with this conclusion—this explains why her conception of correspondence is restricted to judgments about the appropriate way to establish justice. And in distinguishing her view from Michael Walzer’s, she says, “I do not believe Kantian political autonomy requires us to favor institutions that reflect a majority’s religious or cultural values. On my view, correspondence is valuable when it comes to judgments about how best to secure the conditions of mutual independence.” But this leaves Stilz with a dilemma. On the one hand, she can insist that the relevant form of political autonomy is concerned only with achieving correspondence between political institutions and citizens’ shared judgments about how to achieve justice. But this disables her account from explaining intuitively wrongful instances of colonialism or annexation such as Just Annexation. On the other hand, Stilz could expand her conception of correspondence to include matters of religion, culture, and the good life, but doing so seems inconsistent with the Kantian foundations of her view.

In Defense of Functionalism

53

A Functionalist Argument Functionalist theories of political legitimacy are, recall, allegedly vulnerable to the objection; they are apparently unable to explain the wrongness of certain forms of colonialism and involuntary annexation.3 In this section I argue that functionalist theories are not, in fact, vulnerable to the objection. Contra Stilz, functionalist accounts can easily explain the wrongness in the relevant cases of colonialism and annexation without appeal to further notions such as political autonomy or nationality. Before proceeding with the argument, it’s worth emphasizing that the range of cases that allegedly pose a problem for functionalist views is very narrow. All real cases of wrongful colonialism and annexation of territory involve the unjust violation of individual rights or wrongfully threatening to violate individual rights.4 The wrongfulness of all these cases is not difficult for the functionalist to explain. It’s only a very unusual type of case—where a political authority somehow manages to seize control of another territory without, apparently, infringing any standard individual rights— that is our focus. I emphasize the unusualness of these cases not because I think there is something objectionable about considering unusual or fanciful cases in political philosophy. My point is rather that we must be very careful not to confuse our intuitions about colonialism and annexation as we know them with our judgments about the unusual cases under consideration. With that clarification out of the way, here is how functionalist theories can explain the wrongness of colonialism or annexation in the relevant range of cases: 1. Each person has a pro tanto claim of justice against having his or her citizenship or political status changed without consent, waiver, or forfeiture. 2. The cases of colonialism and involuntary annexation under consideration involve changing people’s citizenship or political status without consent, waiver, or forfeiture. Therefore: 3. The cases of colonialism and involuntary annexation under consideration are pro tanto unjust.

54

Jonathan Quong

Let’s call this the individual status argument. The main advantage of this argument is obvious. It shows how even the most apparently benign instances of colonialism and involuntary annexation infringe the rights of individuals, and thus it provides functionalist theories with a straightforward way of explaining why such practices are at least pro tanto unjust. Even if acts of annexation do not involve violence or threats of violence, they still violate individual rights on a large scale. Just as the functionalist can condemn acts of colonialism that involve infringing individuals’ rights against being killed, assaulted, or infringe individual rights to participate in democratic decision-making, the functionalist can also condemn acts of colonialism for infringing individual rights against having one’s citizenship or political status changed without consent. There is no need to appeal to more collective notions such as political autonomy or a shared will to explain these cases; the wrongness can be entirely explained by appeal to the infringement of individual rights. Another advantage of this argument is that there are only two premises, and the premises are hard to deny. The second premise seems clearly true—it’s just a description of what happens in the cases we are considering. And the first premise also seems clearly true. Imagine a rogue government official is somehow able to strip you of your American citizenship and grant you Australian citizenship instead. Even if this does not cause you any material harm—even if it doesn’t affect where you can live or work—it’s clear that the rogue official has infringed your rights. Moreover, the first premise gains additional plausibility from the fact that it doesn’t preclude the permissibility of changing people’s citizenship without their consent—the claim against such changes is only pro tanto and thus may be overridden if there are sufficiently compelling reasons for doing so, a point to which I’ll return below. The objection to functionalist theories rests on the assumption that there can be cases of colonialism or involuntary annexation where no individual rights are infringed. The individual status argument, if sound, shows this assumption to be false. Whenever a group of persons are subject to colonialism or involuntary annexation, their rights are infringed since their citizenship or political status is revoked or changed without their consent. Infringing people’s rights in this way is a serious injustice, and thus the

In Defense of Functionalism

55

functionalist can easily say that such acts of colonialism and annexation are pro tanto wrongful, and very likely impermissible unless there are very weighty reasons to perform them (e.g., averting a mass slaughter or preventing a society from sliding into a dangerous anarchy). The individual status argument provides a straightforward and compelling explanation of the cases under consideration. It also seems to me to have at least two advantages over Stilz’s account. First, unlike Stilz’s view, it has no trouble explaining a case like Just Annexation. In annexing country B, country A changes the citizenship of each member of B without consent and thus infringes an important right of each member of B—this is surely a central part of what makes the annexation wrongful. The individual status argument makes no appeal to the religious reasons why the members of B reject the annexation—it’s sufficient that the annexation involuntarily changes the political status of the B’s. Second, it provides a more compelling explanation for the importance of democratic procedures when re-drawing political boundaries. On Stilz’s view, democratic procedures are “neither necessary nor sufficient to realize self-determination. . . . So long as the cooperators in the state’s population share a political will that supports the state’s institutions and their shared will was formed under conditions of free deliberative reasoning, the state will be legitimate.” Democratic procedures can be a good means of ascertaining whether a political authority or set of political institutions corresponds with the shared will of the local population, but such procedures are ultimately inessential—what matters is the correspondence itself. On the individual status view, by contrast, democratic procedures play a more essential role. Whenever borders are re-drawn, this involves changing some people’s citizenship. Since each person has a pro tanto claim against having her citizenship unilaterally altered, justice condemns re-drawing political borders without seeking the consent of those involved. Unanimity about such decisions is, of course, extremely unlikely, but democratic supermajorities may represent a second-best solution in cases where there is significant support for re-drawing political boundaries. Democratic procedures at least give each person an opportunity to express their view regarding the proposed boundary changes,

56

Jonathan Quong

and if a sufficient number are opposed, no one’s political or legal status will be changed. Before considering some objections, it is worth clarifying one point. I’m not denying that there can be such a thing as a shared will or that there can be such a thing as collective political autonomy. These things may exist, and they may play a role in explaining the wrongness of acts of colonialism and involuntary annexation. I’m simply making the modest claim that we don’t need these notions to understand why acts of colonialism and annexation involve injustice even where violence is not used or threatened. Because we don’t need these collective ideas to explain the wrongness of colonialism and annexation, functionalist accounts of political legitimacy aren’t vulnerable to the objection. Since many believe the objection provides a powerful reason to reject, or at least supplement, functionalist theories of legitimacy, this conclusion is significant. Objections Let’s consider some objections to the individual status argument. Explaining All the Cases First, a skeptic may complain that the argument does not explain all the relevant cases. A colonial power, for example, might rule over the inhabitants of a territory without changing the citizenship of the local population. The individual status argument is apparently powerless to explain the wrongness of this form of colonialism (holding constant the assumption that the colonial power does not use any violence or threats of violence in imposing its rule). To assess whether this constitutes a genuine counterexample, however, the details of the case need to be specified. If the colonial power rules over the local population by denying the local population effective rights to vote and run for elected office, then the colonial power violates important political rights of the local population, and the functionalist view has no difficulty in explaining the wrongness of the case. If, on the other hand, the colonial power grants the local population the same effective rights to vote and run for office as the original citizens of the colonial power,

In Defense of Functionalism

57

then the colonial power has changed the citizenship or political status of the local population, regardless of whether this is officially acknowledged. Their political status has been effectively changed since they now have the legal status to vote and run for office in another country (the colonial power) and not their original country. This illustrates a more general point. There can be no counterexamples of the sort the skeptic imagines, since either the local population is denied political rights, in which case the functionalist view can easily condemn the colonialism, or else the members of the local population are granted equal political rights, in which case it’s clear that their political status or citizenship has been effectively changed. Proving Too Much Consider a different objection. Some may object that the normative premise of the argument proves too much. It apparently entails, for example, that if the residents of a territory participate in a free and fair election and vote to secede from the country to which they currently belong, the establishment of the new country constitutes a pro tanto injustice against each of the residents who did not wish to secede, since these dissenters will involuntarily have their citizenship changed. This conclusion, says the critic, is implausible: No claims of justice are infringed in a case like the one just described. Again, I think this objection can be defused once we think about various cases in greater detail. Suppose the majority of the members of territory B vote to secede from a larger country A, and a new country B is established. But the secession is acrimonious and country A refuses to allow any members of B who wish to remain citizens of A to move to A and retain their citizenship. Intuitively, a wrong is done to the dissenting members of B who wish to remain citizens of A. The normative premise of the individual status argument explains this intuition—the dissenting members of B each have a pro tanto claim against having their citizenship changed without consent. Now imagine a variant of the case where the secession is not acrimonious—country A passes legislation guaranteeing that any

58

Jonathan Quong

member of B who wishes to remain a citizen of A is entitled to do so, and both countries establish a fund to ease the transition costs for those members of B who wish to move to A in order to remain citizens of A. Under these conditions, no one’s citizenship has been changed without consent, waiver, or forfeiture. Any member of B who wishes to retain their citizenship in A is able to do so. It’s true that doing so may be costly and inconvenient, but the normative premise of the individual status argument doesn’t hold that a person has a claim of justice to costlessly retain her citizenship or political status. Sometimes, however, borders are re-drawn in ways that make it impossible for people to retain their existing citizenship. Suppose the members of country A decide, in a free and fair election, to dissolve the country into four distinct countries, B, C, D, and E. Those who voted against the dissolution—those who wish to remain citizens of country A—cannot retain their citizenship in A since the country will cease to exist. In a case like this, I don’t think it is any embarrassment to the individual status argument to conclude that the dissenters have a pro tanto claim against the result. The dissolution of country A may be justified, all things considered. But changing a person’s citizenship against her will is pro tanto unjust, and this fact should weigh heavily when we consider whether the dissolution of a country is justifiable. In sum, the normative premise of the individual status argument does not prove too much. When we consider different examples where people choose to re-draw existing borders via free and fair elections, the normative premise yields intuitive results. Those who dissent from what the majority decides should, whenever possible, be given a fair opportunity to retain their existing citizenship. When they are denied this opportunity, this is intuitively wrongful, and the normative premise of our argument explains this fact. Privileging the Status Quo The individual status argument might seem objectionable to some for a different reason: because it privileges the status quo. It does so by granting persons a right against having their citizenship or political status changed. But most people’s citizenship or political

In Defense of Functionalism

59

status is something determined by birth, and thus not something chosen. Since the status quo is—for most people—arbitrary in this sense, why should there be a right against changing it? When the objection is stated in this general way, it doesn’t seem difficult to rebut. Many features of persons are unchosen, and yet people have rights against having these features changed without their consent. Most people do not experience their gender or sexual orientation as something they choose, and yet it would clearly constitute a serious wrong to change someone’s gender or sexual orientation without consent.5 So the objection cannot succeed when presented in such a general way. But the objection can be made more plausible. Gender and sexual orientation are central to many people’s sense of self and plans of life, and thus changing these features of a person constitutes a very serious interference with an individual’s freedom or autonomy to shape their identity and plan of life. But when we turn to the issue of citizenship or political status, things may seem different. Consider the perspective of an individual person whose citizenship is changed without consent in a case like Just Annexation. This person will, by hypothesis, be a member of a just state that protects all her important rights and freedoms, and thus she will be given all the freedom to shape her identity and life that any person can justly demand. What, then, is the basis for her alleged right against having her citizenship involuntarily changed? Suppose we answer in the same way we did when considering gender and sexual orientation. We say that citizenship is central to many people’s sense of self and plans of life, and thus changing a person’s citizenship constitutes a very serious interference with that person’s freedom or autonomy to shape their identity and plan of life. This answer might seem inadequate if we focus only on the individual. Why is being a citizen of country B so important to shaping someone’s identity or plans of life if we stipulate that the person can continue to live and work in the same place, remain friends with the same people, and retain all of her civil and political liberties? To provide a satisfactory answer to this question, it might seem that we must appeal to something like collective political autonomy. An individual’s current citizenship matters primarily because it protects her ability to participate in an existing form of collective

60

Jonathan Quong

political autonomy. If changing a person’s citizenship without consent is wrong, it is for this reason. But if this explains why people have a claim against involuntary changes to their citizenship, then the individual status argument is not an alternative to Stilz’s account—it rather presupposes something like her account. I don’t think this is correct. Although many people may value their citizenship in part because they value participating in a form of collective political autonomy, this type of autonomy isn’t necessary to ground a claim against involuntary changes to one’s citizenship or political status. Consider the following sort of case. Country A involuntarily annexes a small territory from country B. The people who live in this territory are disaffected members of B—they don’t endorse many of the political institutions and policies in country B. These disaffected members also have no plans to try and reform B’s institutions—they know they are too small in number for this to ever succeed. In Stilz’s terms, there is a lack of correspondence between this small group of disaffected citizens and country B’s political institutions. Country A nevertheless infringes the claims of each of these people when it involuntarily changes their citizenship. If this is correct, then the mooted objection to the individual status account fails. Positing a claim against involuntary changes to one’s citizenship or political status need not presuppose the importance of collective political autonomy. But what grounds the claim against involuntary changes to citizenship or political status if not the appeal to collective autonomy? I suspect the answer has to do with power and equality. To change a person’s political status without consent is to exercise a kind of power over them. Absent a very compelling justification, it is wrong to exercise power over people in this way because it’s inconsistent with a certain form of equality. But accepting this positive suggestion isn’t essential to rebut the objection. We only need to see that collective political autonomy isn’t necessary to ground the right that the individual status argument posits. Consent in Functionalist Clothing Let’s consider one final objection. The individual status argument effectively requires that the local population must accept or consent to acts of annexation in order for those acts to avoid being

In Defense of Functionalism

61

pro tanto unjust. A skeptic might protest that the individual status argument thus helps itself to a consent-based view of political legitimacy. If true, this shows that functionalist theories are vulnerable to the original objection—they cannot evade the objection without help from one of the competing theories of legitimacy. This worry is perhaps best expressed in the form of dilemma. If state A cannot annex the members of B without their consent, then why do functionalist theories allow state A legitimate authority over the citizens of A despite the fact that most of those citizens never consented to state A’s authority? The functionalist thus faces a dilemma: A state can either gain legitimate authority over persons without their consent or it cannot. If it can, then functionalism cannot help itself to the individual status argument on pain of incoherence. If it cannot, then the functionalist account collapses into a partly consent-based view.6 This objection does not succeed for at least two reasons. First, the individual status argument does not require that consent is necessary for political legitimacy. It makes the weaker claim that changing people’s citizenship or political status without consent is pro tanto unjust. Second, it’s a mistake to assume that the functionalist view is a purely ahistorical conception of political legitimacy. The objection depends on the assumption that it doesn’t matter how an alleged authority came to be in a position of de facto control over a territory—all that matters is whether, here and now, the alleged authority satisfies the condition of performing various morally mandated functions in a reasonably just manner. But this is not an appealing version of the view, and not one that many functionalists defend. The best version of functionalism is, in part, historical. A state cannot become a legitimate authority over a territory by seizing control of the territory via systematic rights violations at t1, and then subsequently establishing reasonably just conditions at t2. What matters is not simply how a state currently operates—it also matters how it achieved its position of de facto authority. Of course, the incorporation of historical conditions into the functionalist view poses some well-known problems regarding the supersession of historical injustice.7 Intuitively, recent injustices seem to undermine a state’s claim to legitimate authority in a way that injustices in the more distant past do not. The challenge for

62

Jonathan Quong

functionalist theorists is to provide a principled way of explaining this intuitive difference. I believe that a principled explanation is possible, though I can’t defend that claim here.8 Moreover, a related problem arises for the consent theorist. Since every current state arose via injustice, the consent theorist must either provide her own principled explanation with regard to the supersession of historical injustice, or else she must concede that no state in human history will ever be legitimate given our existing history of injustice. Some philosophical anarchists may be happy to impale themselves on the latter horn of the dilemma, but to me this seems a much higher cost than the requirement to provide a principled approach to the supersession of historical injustice. In sum, so long as the functionalist view can be partly historical, then there is nothing puzzling or problematic in claiming that a state may be legitimate with regard to its current citizens even though those citizens did not consent to its authority, while insisting that the same state cannot annex further territory without being authorized by the inhabitants of that territory. Conclusion Some believe that functionalist theories of political legitimacy cannot explain the intuitive wrongness of colonialism or annexation where the colonizing state does not commit or threaten any violence, nor violate any other political or civil rights of the occupied persons, and subsequently establishes reasonably just conditions in the annexed territory. I advanced two main arguments with regard to this objection. First, I suggested that Stilz’s proposed solution to the objection is vulnerable to a serious dilemma: It can only explain the wrongness of certain forms of colonialism by appeal to a non-Kantian view of autonomy. Second, I offered an alternative response to the objection: the individual status argument. On this view, the intuitively troubling cases of colonialism and annexation can be easily explained as infringements of persons’ claims not to have their citizenship or political status changed without consent. In closing, it’s worth returning to a point made earlier. Every real-world instance of wrongful colonialism and territorial annexation has been accompanied by unjust violence, coercion, and other rights violations. The wrongness of all these cases of colonialism

In Defense of Functionalism

63

and annexation is overdetermined and easy to explain. Some people, however, believe that there is something distinctively wrong about colonialism—a wrongness that persists even when we imagine a case where there is no violence or threats to basic rights. For some, including Stilz, the distinctive wrongness has to do with thwarting people’s collective political autonomy. I don’t wish to deny that this often occurs—that colonialism is partly wrong for this reason. I’ve only tried to show that even a resolutely individualist picture of justice can explain the wrongness of colonialism and involuntary annexation. It’s pro tanto unjust to change people’s citizenship or political status without their consent. That’s enough for functionalist theories of political legitimacy to avoid the objection.9 Notes 1. It might seem that there is a straightforward explanation of the wrongness in this case: Country A thwarts the political autonomy of the members of country B because the annexation is involuntary and done without the consent of the members of B. But Stilz cannot help herself to this straightforward explanation. She is explicit in presenting her view as an alternative to voluntarist accounts of self-determination; for Stilz, consent is not necessary for political legitimacy. 2. For a defense of this conclusion see Japa Pallikkathayil, “Neither Perfectionism nor Political Liberalism,” Philosophy & Public Affairs 44 (2016): 171–196. 3. Stilz distinguishes between two types of functionalist theory: a maximizing version and a threshold version. The argument developed in this section can be deployed by either variant. 4. This point is emphasized in Laura Valentini, “On the Distinctive Procedural Wrong of Colonialism,” Philosophy & Public Affairs 43 (2015): 312– 331. 5. The wrong to the person is not reducible to a violation of bodily integrity. 6. Valentini advances a closely related objection against one interpretation of Lea Ypi’s account of the distinctive procedural wrongness of colonialism. See Valentini, “On the Distinctive Procedural Wrong of Colonialism,” 316–318. 7. For a forceful development of these concerns, see A. John Simmons, Boundaries of Authority (Oxford: Oxford University Press, 2016), especially ch. 7.

64

Jonathan Quong

8. The best explanation, I believe, will be grounded in a plurality of factors, including, among others: whether the current descendants of previous victims of injustice remain disadvantaged in terms of access to resources relative to others, whether the state’s constitution or basic political institutions have undergone significant reform, and whether appropriate acts of compensation and apology have ever been made by state officials. 9. For helpful comments and discussion, I’m very grateful to Annie Stilz and Rebecca Stone.

3 IS POLITICAL LEGITIMACY WORTH PROMOTING? AMANDA R. GREENE

All governments fear a loss of political legitimacy. This fear suggests that they consider political legitimacy to be something valuable— worth holding onto if they have it or restoring if it has been lost. But when there is a loss of political legitimacy, what exactly is lost? Is it necessarily something valuable? This chapter seeks to answer that question by developing and defending a new framework for political legitimacy. I am proposing a new framework partly in response to the following puzzle. Assessments of political legitimacy are increasingly common in current affairs, and they are often highly consequential. In these debates, everyone assumes that insofar as something lacks legitimacy, that’s a bad thing. The expression “that’s illegitimate” is treated as self-evidently negative by those of all political stripes, in all contexts. This way of talking makes “legitimacy” unusual among normative political concepts. Other political ideals such as democracy and autonomy tend to attract both fans and critics. An expression like “that’s undemocratic” is the sort of complaint that is appropriate only in certain political contexts, e.g., a policy debate but not a criminal trial. Not so legitimacy. The lack of political legitimacy is regarded universally as a defect. At the same time, legitimacy is assumed to be attainable, rather than an unattainable regulative ideal. Things that claim political legitimacy for themselves are not regarded as claiming what is obviously impossible. We could not make sense of how frequently people claim political legitimacy for themselves, and expect to be 65

66

Amanda R. Greene

believed, unless we consider legitimacy to be attainable. In sum, the lack of political legitimacy is universally seen as a defect that is fixable and worth fixing. However, most philosophical accounts of political legitimacy treat it as a regulative ideal. Two prominent approaches—voluntarism and contractualism—advance a standard of legitimacy that no society has ever met or could ever hope to meet. If political legitimacy is a utopian ideal, how do we make sense of the lack of legitimacy as a defect that is widely regarded as evidently fixable? To approach the puzzle from another angle, we can ask why we, as philosophers, insist on a standard of legitimacy that is so out of touch? I think it is because we fear that we might have trouble explaining the moral significance of an ideal that is more down-to-earth and attainable. In what follows I take the first steps in vindicating the everyday meaning of political legitimacy, while at the same time showing why legitimacy tracks something of genuine value in political life. Policing the Boundaries of Legitimacy Before I begin, I must set the stage by explaining briefly how I view the current landscape. Debates about political legitimacy rely on a shared understanding of what it is and why it matters, but as a matter of fact, a common definition eludes us.1 In my view, legal and political philosophers have not yet offered a clear enough definition of political legitimacy, as such. There are discussions of political legitimacy in the debates about political obligation and legitimate authority, but they do not take political legitimacy as the target of analysis.2 Some theorists have proposed conditions for legitimacy in a liberal democracy, without showing exactly why liberalism and democracy are themselves required for legitimacy.3 Others have discussed legitimacy in the context of human rights or republicanism, but they focus on the necessary conditions for legitimacy, leaving open the question of what would be sufficient.4 To the best of my knowledge, no contemporary political philosopher has offered a comprehensive theory of political legitimacy.5 As a result, we lack the philosophical development of conceptual tools that are adequate to its frequent use.6 In my view, the relative lack of progress on theorizing about political legitimacy can be largely explained by twentieth-century

Is Political Legitimacy Worth Promoting?

67

disciplinary specialization. The term is now used to mean quite different things by lawyers, philosophers, and social scientists. The sociological approach, exemplified by Max Weber, seeks to explain patterns of obedience in terms of beliefs about legitimacy.7 The approach of moral philosophers, represented by Rawls and Habermas, seeks to provide normative justification for liberal democracy through elaborating an ethics of citizenship in conditions of disagreement.8 The approach in jurisprudence, growing out of Hart’s conceptualization of a legal system, seeks to account for the authority of law.9 What unites the three approaches is that they all study the same objects: social organizations that persist over time through upholding structures of power and authority. But their approaches to explanation and justification differ greatly, leading to divergent usage of the term “legitimacy.” This balkanization of academic disciplines has led to the acceptance of a distinction between sociological, descriptive, or perceived legitimacy, on the one hand, and normative or liberal democratic legitimacy, on the other hand. This distinction is not only accepted by all parties; it is strictly enforced and policed, because it is viewed as a disciplinary boundary.10 For instance David Beetham, the most prominent political scientist writing on legitimacy, says, “I am concerned with legitimacy as a problem for social science rather than for philosophy [because] the two activities have a different purpose and a different logic.”11 Unfortunately, accepting this dichotomy as axiomatic has prevented theorists from seeing the way in which descriptively defined concepts can themselves have normative significance. In the case of concepts like poverty and cultural diversity, for instance, the definition corresponds to empirically verifiable facts in the world, but those facts themselves represent normatively important features of the political landscape.12 No one would claim that just because cultural diversity is empirically defined and measured it can’t have normative significance. And yet, when we insist on entirely separating normative and descriptive legitimacy, we rule out this theoretical possibility from the outset. Philosophers, for their part, tend to subsume the descriptive enterprise under the normative one.13 On the other side, social scientists incorporate the analysis of norms into their overarching descriptive enterprise, carefully stopping short of normative claims.14 Due to these disciplinary differences,

68

Amanda R. Greene

theorists often talk past one another.15 Within the context of this chapter, I am not interested in arguing that any of today’s disciplinary approaches are wrong, or even that the divide between them is wrong. Rather, I am interested in what might be missed by simply accepting this fragmentation of the inquiry. My account of legitimacy combines elements from all three approaches I have mentioned, and I hope to show that it is fruitful to bridge these divisions. It matters very much what we take to be the starting point of the inquiry, but it is difficult to define the subject matter of political legitimacy while bearing in mind these divergent methodologies. In my opinion, the typical starting points in normative theorizing about legitimacy—namely, authority, power, and coercion—are inadequate. Both power and authority are necessary for thinking about political legitimacy, in my view, but neither is adequate on its own. On the one hand, taking authority by itself as the starting point neglects those aspects of political power that don’t operate as law. For instance, Raz’s distinction between de facto authority and legitimate authority renders the analysis entirely in terms of obedience to legal directives.16 But political power goes far beyond obedience to the law. Even when we aren’t being asked to obey, we can ask whether the state’s exercise of power is legitimate, including the various actions that it takes, as well as the values that it expresses and promotes.17 On the other hand, taking power by itself as the starting point obscures the distinctiveness of the political. For example, Pettit’s definition of domination as subjection to an arbitrary will places the social and the political on the same plane.18 Pettit and other neo-republicans consider it to be a virtue of the account that subjection to the will of one’s husband or employer is analogous to the loss of freedom under a tyrannical state. But this unification of all subordination relations makes it difficult to see what is distinctively bad about an illegitimate political order. One might think that an easy starting point is to define legitimacy as the right to rule, but this doesn’t work insofar as the terms “right” and “rule” remain as opaque as legitimacy. Often the right to rule is glossed as a demand for the justification of coercive enforcement in conditions of reasonable disagreement about justice. But this construal assumes that what matters most is reciprocal justification, which unduly favors a contractualist approach

Is Political Legitimacy Worth Promoting?

69

to justifying coercion. Less obvious but more important is the fact that such an approach presumes that there is a distinction between licit and illicit violence that precedes particular social and political circumstances.19 So long as this presumption goes undefended, it simply restates the questions we hoped to answer with a theory of political legitimacy.20 In fact, I don’t think coercion is necessary to trigger questions of political legitimacy, though power and authority are normally involved. Therefore, I will assume that when we ask about political legitimacy, we are seeking to evaluate the relations of power and authority that are expressed and exercised in a political order (at this point, I deliberately leave the mode of evaluation unspecified). The term “political order” is meant to apply to a broad array of forms of political organization, including societies governed by customary law, all of which I will refer to as regimes.21 In light of this very broad definition of the subject matter, I want to propose a new way of thinking about political legitimacy, one that makes space for legitimacy to have distinctive meaning and value. The Plan The discussion begins with an outline of my view, as set forth in the next section, “Political Legitimacy as ‘Quality Assent.’” My proposal is that a regime is legitimate insofar as it achieves quality assent to being ruled among its subjects. Quality assent to rule, on my account, involves the willing acceptance of one’s status as a subject, based on an acknowledgment that one’s regime is valuable as a political order. This judgment, in turn, relies on the essential claim of a political order, which is to deliver benefits to subjects through the exercise of power and authority. Next, in the section titled “Significance of Political Legitimacy,” I argue that legitimacy, as I have defined it, represents something of moral significance. Far from being merely descriptive, achieving quality assent realizes a moral ideal of voluntary subjection to rule. This ideal is realized when social order is achieved on the basis of a shared perspective regarding the values that justify the exercise of power and authority, thereby securing political stability without widespread alienation. I argue that the connection with a moral ideal of voluntary rule shows why political legitimacy is valuable, though its value is dependent on certain conditions. In the final section,

70

Amanda R. Greene

“Comparative Advantages,” I dispel some apparent objections and highlight the advantages of understanding legitimacy this way. Not only does my account of legitimacy give us a mode of evaluation that is distinct from liberal democratic morality, it also shows us why legitimacy ought to be promoted for its own sake. Political Legitimacy as “Quality Assent” To introduce my view, I begin by considering the revolutionary, someone who completely rejects their current political order and seeks to radically change it. I would describe this person as an unwilling subject, someone who disavows being ruled to such a degree that they are willing to engage in disruptive and costly resistance (if it stands a chance of being effective) in order to fundamentally change the constitutional order. The contrary case is someone who is willing to remain a subject of their political order, all things considered. To get some purchase on this idea of a willing subject, I propose a notion of assent to being ruled. This assent is present when (i) a subject forms an overall subjective assessment of governance, and (ii) on that basis, accepts her political order. These assessments can be based on a variety of views about the fundamental aims of a political order. The judgments that are involved rely on subjective valuations of apparent goods that have been achieved by the regime, at either the individual or collective level. A positive governance assessment, therefore, is the judgment by a particular subject of some success in delivering goods through rule. Subjects may properly regard their regime as beneficial in a variety of different ways, according to their views on the nature and purpose of government. Thus, the generic notion of a governance assessment allows for idiosyncratic judgments by individuals about the goods government ought to serve. All that is required is that a subject takes her political order to be valuable as a political order. This requirement, in turn, presupposes that there is a formal, generic way of making sense of political orders as valuable. Accordingly, I suggest that regimes tacitly claim something for themselves, and it is in light of that claim that we make sense of them. Namely, they claim to exercise power and authority in order to provide some goods that are connected to the interests and

Is Political Legitimacy Worth Promoting?

71

aims of those it directs. No doubt the claim is implicit and subject to a wide variety of interpretations by those making and hearing it. The regimes may not succeed at what they claim, or even try very hard, but they make this kind of claim in virtue of ruling. Insofar as a system of organized power presents itself as a system of rule, it claims to benefit subjects through the exercise of power. This claim corresponds to the raison d’être, its overarching justifying aim, so I call it the essential claim of rule.22 Assent to being ruled is not just a matter of judgment. It also involves a voluntary element of endorsement, i.e., “I willingly accept my subjection.” As I understand this acceptance, it requires that I judge that my subjection has a point to it, a point that corresponds to some values that I can recognize, affirm, or endorse. It is not a judgment that submission is good for its own sake, but rather, that submission is worthwhile in light of some further goods. In other words, my acceptance is grounded in the following judgment: I am willing to continue being a subject of this particular political order because it provides some goods, or advances some values, through ruling. I may not affirm every aspect of my political order or be fully reconciled to it, but on some level, I accept its validity. I acknowledge and accept, even if it is only minimally, my own political subjection. This acceptance cannot simply be equated with obedient behavior, as some non-accepting subjects may comply with the law, and some disobedient subjects may willingly accept their status as a subject. But normally, the disposition to comply and readily cooperate with exercises of power and authority is a good indicator of this acceptance. As Beetham notes, “[The] degree of legitimacy is to be assessed by the extent to which those subject to it acknowledge it as rightful. And this can in principle be empirically ascertained through attitudinal and behavioral indicators which are available in the public domain, although the precise choice, design, and weighting of these is always open to disagreement and improvement in practice.”23 Hence, behavior can be taken as an imperfect indicator of acceptance. Quality Assent Having now defined this notion of “assent to being ruled,” here is my proposal, in unrefined form: A political order is legitimate

72

Amanda R. Greene

insofar as its subjects willingly assent to being ruled. Now comes the refinement: Not all assent to rule contributes to legitimacy. Some of it is disqualified, because it obviously clashes with the essential claim of rule. What does it mean to clash obviously? Though the essential claim is capacious, i.e., designed to include all the goods that regimes may claim to provide, there is one that is undeniable: the provision of basic security for all subjects. This is because the provision of basic security—understood as the protection from violence—is presupposed by any other claim that a political order could possibly make about the goods it exists in order to provide.24 Though regimes may fail to, or not even attempt to, actually provide for all subjects’ basic security, they nevertheless make this minimal claim as a presupposition of their essential claim. Hence, I call this claim the minimal claim of government (MCG). Consequently, the assessment by subjects of the regime as a system of rule must tacitly acknowledge that it claims, at the very least, to provide basic security. If a subject’s assessment of a regime is based on a denial of this minimal claim, then it clashes with the essential claim so obviously that it would no longer be recognition of a political order as a political order. To make things more precise, let’s presume that the judgment that underlies quality assent has this abstract form: “I see value in being under this system of rule [while tacitly acknowledging this system as such a thing that aims at least to provide basic security for all subjects].” The corresponding assent is then disqualified if the judgment that a political order is valuable lacks that tacit acknowledgment, either implicitly or explicitly. It would lack it explicitly if the basis of a subject’s assent is that the regime does not aim to protect the basic security of all subjects. This subject might be a dominator, someone who believes that the regime should neglect others’ basic security—or worse, values the regime precisely because it does so. Or, this subject might be an underling, someone who believes that they themselves are not worthy of protection. Both of these examples are ways of assenting to rule while explicitly denying the minimal claim. But it is also possible for the assent to implicitly lack an acknowledgment of the minimal claim. It would lack it implicitly if the subject’s assent is robustly independent of available evidence about whether it provides for some subjects’ security. For, a robust insensitivity indicates that one considers these facts to be irrelevant to one’s acceptance of a political order.

Is Political Legitimacy Worth Promoting?

73

Thus, I identify the following as a “quality filter”: Quality assent is assent to rule that is based on governance assessments that do not obviously clash, either implicitly or explicitly, with the minimal claim of government. The quality filter is crafted to exclude those whose acceptance of a political order is based, directly or indirectly, on the judgment that some subjects are to be denied basic security. The assent in those cases would not be quality assent because it would not count as recognizing a political order as a political order. It is possible to define the quality filter in a way that is more substantive, raising the bar on what would count as quality assent to a higher standard. As I explain later, I set a low bar in order to avoid embedding liberal assumptions, but the abstract idea of quality assent is compatible with imposing a quality filter that is more epistemically or morally demanding. All assent to rule that satisfies the quality filter contributes to legitimacy, in my view. Someone might then ask, when assent to being ruled is based on fear, is it included in quality assent? That depends. When the fear arises from believing that your regime would harm you if you didn’t acquiesce, then it would disqualify the assent since this judgment conflicts with MCG. But acquiescence in order to avoid an external threat does not clash with the MCG (assuming it is freely given). So the presence of fear and intimidation sometimes disqualifies the assent, but not always. This feature enables the view to explain why regimes that rule on the basis of terror, such as a tyranny or some kinds of mafia, cannot attain legitimacy even when the acquiescence to those regimes is voluntary. Thus, my refined proposal is that a political order is legitimate insofar as it attains quality assent to being ruled among its subjects. Legitimacy is achieved, then, when a political order governs in such a way as to be widely regarded as valuable by its subjects. The upshot is that subjects have a special standing, or kind of authority, to select their own terms when assessing their government, but not if those terms obviously conflict with the essential claim of a political order. The Political Subject On my view, legitimacy rests on the assent of all and only the subjects. Who counts as a subject? I doubt this question can be definitively settled, but it will help to indicate roughly what I mean. For

74

Amanda R. Greene

one thing, being the subject of a regime is different from being under the power of a regime. For example, the US government exercises power over the inhabitants of Mexico, but it does not rule the inhabitants of Mexico. This implies that “having power over” is not sufficient for subject status. On the other hand, subject status is not exhausted by a mere claim of authority, because an established pattern of obedience is also required. On my view, subjection to rule occurs whenever de facto authority or power is exercised effectively over a group of individuals, combined with a claim to benefit them through that exercise. This conjunction is able to demarcate being a subject of a political order from other relations, such as subjection of conquered or enslaved peoples. Roughly speaking, in order for an individual S to be a subject of regime P, there must be some degree of acknowledgment on both sides. That is, subject S considers himself to be (is disposed to believe and behave as though he is) under the de facto authority of regime P. For its part, regime P considers individual S to be under their de facto authority. This means there is some minimal recognition of S’s standing within the legal system of Regime P (including recognition through customary law). Individual S is acknowledged as a person who is entitled to make complaints to the officials regarding his or her basic security. These avenues of protection and redress must be more than mere formalities; they must be moderately effective for the regime to count as acknowledging their legal standing. Otherwise, a failure to provide basic security would not be a failure of a publicly acknowledged aim. These conditions for subjecthood must remain somewhat vague, because a determination of which individuals count as subjects is based on an array of facts having to do with the patterns of obedience in a group, and subjects’ views and expectations of their fellow subjects. So the basis for determination is a highly diffuse set of social facts, something that is difficult for just one person to manipulate. Normally the magistrates of the regime are not able to unilaterally decide who its subjects are, lest they seek to secure legitimacy by intentional exclusion, but they nevertheless have a great deal of influence over who counts as a subject.25 While the boundaries of becoming or ceasing to be a subject are not always clear, nevertheless we can identify and distinguish the endpoints of the process for individuals in transition, in both directions. We

Is Political Legitimacy Worth Promoting?

75

can say that there was a time when individuals have come to count as subjects, and a time when they have ceased to count as subjects, and this intuitive distinction is sufficient for my purposes. Hobbesian Security and Weberian Acceptance Thus far I have outlined a view of political legitimacy in terms of the quality assent to being ruled on the part of subjects. I used the essential claim of rule to identify what counts as non-quality assent, thereby creating a “quality filter.” Quality assent includes all those who recognize a political order as valuable, as long as their judgment does not obviously clash with the essential claim of rule. In contrast, non-quality assent includes those whose acceptance of a political order is based, even implicitly, on the judgment that some subjects are to be denied minimal security. Therefore, the account of political legitimacy that I am proposing has affinities with Hobbes and Weber, and so I want to clarify how I am combining elements of those theories. Solving the problem of Hobbesian order is clearly an ingredient in political legitimacy, because legitimacy is not possible while there is open conflict and threat of violence—something has gone terribly wrong if civil war breaks out.26 While some measure of peaceful order is necessary for legitimacy, it is not right to think of it merely as a “precondition,” as this implies that it is prior to and separate from the essential aims of political order. But this is not plausible. Whatever the story is about why political order is valuable in the first place, the avoidance of civil war is part of that story, simply because stable peace and order are required for any other valuable aim of politics to even make sense. Thus, I tip my hat to Hobbes for insisting that peaceful order is a proper aim of political order, not merely a precondition. At the same time, I tip my hat to Weber, who argued that legitimacy arises from the inculcation of beliefs held by a group about the valid sources of authority.27 The emphasis on acceptance of a system of command, where this acceptance is sustained by a belief about valid authority, is vital for my view. It connects with two deliberate choices I have made in theorizing about legitimacy. I treat legitimacy as a system-level property that pertains to the whole political order, as opposed to a property that can vary by

76

Amanda R. Greene

individual subject. I think legitimacy comes in degrees—that is to say, it is scalar rather than binary. Second, I have chosen to use the term assent, rather than consent, to avoid giving the impression that my view of legitimacy involves the authorization of a regime by its subjects, in such a way that authority arises through a deliberate contractual arrangement.28 As I have developed the idea, quality assent stands for a combination of a volitional state (acceptance) and a belief state (judgment). A decent proxy for this combination of willing and believing is a set of behavioral dispositions—to obey, first and foremost, but also to evince general support for the order upheld by the regime. Overall, quality assent is an acknowledgment of a regime as to-be-obeyed because it is worth obeying, where this is normally backed up by corresponding actions and dispositions. Accordingly, the view owes as much to Weber as to Hobbes. Significance of Political Legitimacy Now, someone might ask: If we achieve this thing that I call legitimacy, what exactly have we achieved, and how important is it that we have achieved it? In this section I will show why political legitimacy, understood as quality assent, has distinctive value and ought to be promoted for its own sake.29 It is essential to my account of political legitimacy that quality assent has dependent value, but the nature of this dependence requires some elaboration. The basic idea is that attaining legitimacy in a particular political order is valuable only when the political order is valuable. In this sense, the value of legitimacy is contingent on particular political circumstances, because any given political order may fail to be valuable. However, this contingency by itself does not show that legitimacy isn’t extremely important in the pantheon of human values. In fact, many goods that we aim to realize have only conditional value. For example, equality is not valuable if we are all equal slaves. Solidarity is not valuable if it is based on hatred. In spite of there being conditions, there are still some important things we can say about why equality and solidarity are valuable. So, too, with legitimacy. Besides this simple conditionality, there is a deeper kind of dependency at play. The value of legitimacy depends in part on the significance of the item whose legitimacy is in question. When a political

Is Political Legitimacy Worth Promoting?

77

order is delivering the important human good of basic security, and when subjects recognize it, the recognition is itself valuable. Though the recognitional value is distinct, it is nevertheless tied to the value of what is being recognized. Thus, it makes sense to say that insofar as we value a political order, we should value that it be legitimate. This idea should be contrasted with the notion that quality assent is valuable because it enables the effective functioning of government. That may well be true, but it wouldn’t exhaust the value of quality assent, in my view, because the value of quality assent is not entirely derived from what is assented to. Rather, quality assent has some non-derivative, non-instrumental value. In this sense, it is like loyalty to one’s friends or commitment to a cause. The value of loyalty or commitment depends in certain ways on its intentional object, without being entirely reducible to the value of the object. These considerations lead me to claim that quality assent has “dependent, non-instrumental value.”30 This terminology may sound abstract, but I am drawing on an idea that is simple and intuitive. Consider the mundane context of a patient interacting with his physician.31 The doctor makes an implicit claim, in virtue of her role, to instruct the patient on the basis of expertise about his health. If the doctor is minimally successful in fulfilling this claim, i.e., she has considerable medical knowledge and the ability to apply it through directives, then there is value in the patient acknowledging that the implicit claim is fulfilled and then relating to the doctor accordingly. In other words, when the doctor’s authority over the patient has some value, then the recognition and corresponding acceptance of that authority, by the subordinate, has further value. Likewise, consider the case of a judge deciding a civil dispute between two parties. The judge makes an implicit claim, in virtue of her role, to resolve the dispute in a fair and impartial way. If the judge is minimally successful in fulfilling this claim, then there is value in the disputants acknowledging that the implicit claim is fulfilled. This corresponds to the ancient dictum: Justice must be done, and it must be seen to be done. The disputants accept the verdict under the rubric of “justice having been done.” The dictum implies that the value of seeing that justice has been done is not reducible to the value of justice. Although, it is not entirely independent of justice.

78

Amanda R. Greene

Quality assent to one’s political order is like these examples. Its value is dependent, but not wholly derivative or instrumental. It rests on a recognition whose value is not reducible to the political order, but neither is it completely independent of it. The upshot is that when the political order is minimally valuable for subjects as political order, then quality assent to it has non-instrumental value. In particular, I argue that its value corresponds to three interrelated goods: avoiding political alienation, establishing political stability on a durable basis, and fostering alignment on the values that justify the exercise of political power. Non-Alienation and Stability The value of quality assent should first be observed at the individual level. Quality assent represents the attainment of something objectively valuable: the avoidance of alienation from one’s own political order. This alienation arises because one cannot authentically affirm the regime to which one is subject as in any sense worthwhile, i.e., as providing at least some benefits that one values subjectively. This sort of alienation is objectively bad for the individual. Consider again the rebel or the revolutionary, whose alienation is extreme. Even if they find solidarity or community in other arenas, they disavow the supreme ruling organization in their territory, the one with final say. Thus, when they reject their political order, it is a pro tanto bad state of affairs, as well as being bad for them. I’m relying on an intuitive contrast between willing and unwilling subjection in order to mark out a morally significant ideal of voluntary rule. I claim that each person has a fundamental interest in attaining voluntary rule for themselves, no matter what other benefits the political order may be delivering. Hence, I call my view of political legitimacy the voluntary rule conception.32 Another reason quality assent is valuable is that it secures political stability. Philosophers have been too quick to dismiss the significance of stability as an end in itself.33 While it is true that stability may be valuable only when certain conditions are met, it cannot be inferred that its value is only instrumental. Stability is integrally related to securing justice and the common welfare. The easiest way to see this is to imagine circumstances with a high degree of political instability, thereby presenting an unpredictable threat

Is Political Legitimacy Worth Promoting?

79

of disorder, disruption, and violence. This sort of civil unrest can reduce well-being directly when it occurs, and indirectly whenever it looms as a threat. Thus, political stability is an integral component of the social conditions that, taken altogether, count as realizing the goods of political association, including peace and justice. Quality assent makes a quantitative and qualitative difference to stability. For one thing, it delivers stability on a more reliable and durable basis. Weber himself declares that quality assent (what he calls “legitimate domination”) produces the most stable rule, while “purely material interests and calculations of advantages . . . result . . . in a relatively unstable situation.”34 David Beetham draws on the Weberian idea a century later to emphasize the connection between legitimacy and avoiding violence. He says, “[An understanding of legitimacy] helps explain the erosion of power relations, and those dramatic breaches of social and political order that occur as riots, revolts, and revolutions  .  .  . it is only when legitimacy is absent that we can fully appreciate its significance where it is present, and where it is so often taken for granted.”35 The social psychologist Tom Tyler, having defined legitimacy as “a belief about the moral right of [authorities] to possess and exercise power and influence,” thinks that legitimacy helps stabilize the criminal justice system.36 Tyler argues that when a belief in the moral right of authorities is present among subjects, the use of coercion and force (including punishment) will be accepted as valid. Subjects will consider the system’s rules and norms to be valid and appropriate, thereby motivating them to self-regulate and cooperate with law enforcement.37 It even appears that in some contexts, young men become less likely to pursue a life of crime.38 These findings suggest that, when subjects accept their ruling order as valid, regulation of behavior can be achieved with correspondingly less coercion and intimidation. Quality assent also makes a qualitative difference to political stability. Edmund Burke speaks of a society in which the political order has so much popular allegiance that it makes “power gentle and obedience liberal.”39 For Burke, this “mixed system of opinion and sentiment” delivers political stability of a more admirable sort than can be had through brute force alone. The pro-social behavior is maintained not through fear or even blind habit, but through sentimental allegiance to the social order that is being

80

Amanda R. Greene

preserved by the regime. This allows subjects to view their regime from a perspective that is oriented toward values—as a social scientist recently put it, it allows citizens to “relate to the powerful in moral rather than self-interested terms.”40 Generally speaking, it is better that obedience be cheerful and generous rather than grudging and mistrustful. In other words, if there must be power, it is better that it be gentle rather than brutal. Thus, stability based on quality assent can help to peacefully preserve the political and constitutional culture. Political Alignment When political stability has been secured through widespread and sustained quality assent to rule among subjects, it represents rough convergence on the values that justify the fundamental norms constraining the exercise of power. I call this political alignment on the proper use of political authority and power, and it is significant for several reasons. Quality assent to rule requires that there be some value system through which the regime’s use of power and authority makes sense to the individual. It indicates at least some alignment between what an individual values and what goods are promoted by the political order to which he is subject (though it may also promote other goods, ones valued by other subjects). When this assent is widespread, it represents some degree of alignment between what is proposed by a regime and what is accepted by subjects about the aims and functions of the political order. When there is a congruence between a regime’s exercise of power and the attitudes of those subject to that power, then there is likely to be some nexus of values that most subjects can endorse as justifying the use of political power, which usually involves claims of authority and threats of violence. When this nexus is stable, the political order is supported by a shared sense that the value system(s) served by the ruling organization are worthwhile. Even if the coherence of values is incomplete, the rough alignment shows that the ends and aims of government are the object of broad support. In other words, there is a kind of shared perspective on the values that justify exercising power and authority in the name of the whole community. This allows subjects to sustain an attachment and allegiance to the fundamental

Is Political Legitimacy Worth Promoting?

81

norms related to the exercise of power, even if they disapprove of particular policies or laws. This, in turn, exerts a form of pressure on the government to keep its exercise of power aligned with the value system(s) endorsed by subjects, which in turn keeps tyranny in check. Because there is willing deference to the guiding norms of the community, this alignment brings an otherwise compulsory form of organization closer to a voluntary association.41 While the alignment improves the vertical relationship between rulers and ruled, it also improves the horizontal social relations among subjects. It is good for me if I am subject to a system of political subjection in which my fellow citizens are willing subjects as well. Conversely, the clear absence of voluntary rule—involuntary subjection— instantiates a kind of disvalue, for an individual and for a political community. Therefore, political alignment allows a citizen’s vertical and horizontal relations to approximate more closely the moral ideal of a voluntary association—a moral ideal that animates many approaches to democracy and liberalism, as it happens. Thus, the significance of political alignment can be seen from the first-person perspective of a subject as well as the third-person perspective of an outsider viewing a political order. Sustaining a political order on the basis of this sort of alignment represents a morally significant achievement for individuals as well as groups.42 When it exists, we ought to promote its conservation. Where it doesn’t exist, we ought to foster and cultivate it. This means that legitimacy is the sort of value that should be promoted by political orders, taking its rightful place alongside other goods that we promote, like justice, liberty, and welfare. Accepting for the Right Reasons Philosophers who think about endorsement of hierarchy often have the following thought: Either there is value in accepting authority relations when it is for the right reasons, or, the way to understand the value of the acceptance is in instrumental terms, e.g., it leads to cheaper compliance, more utility, and so on. But those are not the only options, in my view. In cases where one party is directing another, it is valuable that the director relate to the directed party on terms that inspire acceptance of their power

82

Amanda R. Greene

and authority. This is true even when the subordinate party doesn’t have a full grasp of the reasons behind it, and so cannot be accepting it “for the right reasons” in that sense. Hence, there is a kind of value that is not reducible to the downstream consequences, but which doesn’t require the beliefs that produce the acceptance to be entirely true or perfectly correct. Still, someone may insist that there is simply no value in the acceptance if the judgment on which it is based is incorrect. There is a sense in which my view accepts that critique, but another sense in which it doesn’t. For one thing, notice that the notion of correct acceptance is ambiguous. When we say, “Sonia’s acceptance of her physical therapist’s right to be in charge of her rehabilitation is correct,” this could mean two different things. It could mean that it is correct for Sonia to accept it, or it could mean that the beliefs on the basis of which Sonia accepts it are correct. To see how these can come apart, imagine that Sonia has a capable physical therapist who instructs her to carry out a therapy routine that is effective but tremendously painful. Imagine that the routine is so painful that Sonia does not believe it is good for her, but she is charmed, cajoled, or coaxed by the therapist into going along with it—say, for instance, that the therapist is particularly charismatic and has a commanding presence. Her acceptance of his authority can be correct, even if her acceptance is not grounded in correct beliefs about the therapist’s authority or her reasons for obeying. Willingly subjecting herself to his direction is valuable for Sonia, and thus correct in a sense, even when it isn’t based on believing in his medical expertise (though the correctness of the acceptance is conditional on his having medical expertise).43 Only in this sense does the value of acceptance depend on correctness.44 Still, someone might be tempted to ratchet up the epistemic conditions in order to guard against cases of false consciousness. They might identify two salient defects in the beliefs underlying quality assent, both of which might seem to threaten to cancel out the value of the acceptance. One is where the content of the beliefs is false; the other is where the process of belief formation involves manipulation or deception. While it is tempting to add these as conditions on the value of quality assent, that would be a mistake. Both of these defects might turn out to make the acceptance incorrect and thereby lacking in value, but neither necessarily makes that

Is Political Legitimacy Worth Promoting?

83

the case. If they did necessarily—that is, if these defects were taken to always cancel out the value of acceptance—then you would be excluding some cases where the acceptance would be correct and valuable, in the sense illustrated above. Acceptance can be valuable even when it doesn’t emerge from a faultless apprehension of the relevant facts or have immaculate epistemic credentials. To insist otherwise is to prize having correct beliefs at the expense of everything else that might matter. Legitimacy without Justice Likewise, my view deliberately holds open the possibility that there may be quality assent to rule even when the regime is not fully just or rights-respecting. This returns us to the question of what conditions must be met in order for quality assent to be valuable, an issue that I postponed while I elaborated the value in terms of voluntary rule. Consider a mundane analogy again, that of a teacher and a student in the classroom. The teacher makes an implicit claim to educate the student. If the teacher is minimally successful in fulfilling this claim, then there is value in the student acknowledging that the implicit claim is fulfilled and accepting the teacher’s supervision over them. To insert myself into the scenario: Provided that my teaching is minimally valuable as teaching, it is a further good that my students recognize that I have some expertise and authority. I may not have exactly the authority that they think that I have, but it is still valuable that they recognize me as having some kind of authority and accept me as their teacher on that basis. Students who are entirely new to philosophy can only have a faint idea of what grounds my authority to teach, necessarily.45 No doubt there is some level of incompetent, malevolent, or harmful teaching that would cancel out the value of the teaching. But above this threshold, the teaching has value, and so the assent to the teaching has value too. The same holds for doctors and patients, and parents and their children. Someone who is inclined to deny that quality assent has value in these contexts may just be disagreeing about the conditions under which it is valuable. For if I were to increase the threshold sufficiently high, I suspect that this person would accept that the absence of quality assent represents a distinct loss of value. If so,

84

Amanda R. Greene

this person should accept that, like the teacher and doctor, there is value in a particular regime garnering quality assent, as long as the regime is above some threshold of value. Failing to provide for basic security, in my view, is what places it below this threshold, because the regime would fail to fulfill the minimal claim of a political order, thereby losing its claim to be at all valuable. Nevertheless, there are some who would be willing to accept the dependent value of legitimacy but insist that the threshold should be much higher, e.g., attaining justice. But it is too demanding to say that the value of legitimacy depends on whether justice has been achieved. An analogy would be to maintain that it is only good for a person to love someone who is always honest with them. It is probably better for someone to love an honest person, but that doesn’t mean it can’t be good for them to love someone who bends the truth. In other words, normally a life without love is worse than one with love, even in the absence of perfect honesty. It is difficult to uphold such a demanding view about the value of love without making its realization unattainable. As honesty and love are two distinct values in a human life, so justice and legitimacy are two distinct values in a political community. Legitimacy is worth preserving even without a full measure of justice, just as it is worth preserving love even when there has been a lapse in honesty. As I mentioned above, political ideals such as equality, democracy, and solidarity all work this way. For instance, while the ideal of equality does not have unconditional moral value, it is something to be aimed at and preserved, even while we pursue other ends at the same time. In a similar way, I would say that legitimacy has noninstrumental though conditional value. It is important enough to be its own aim, an end pursued by rulers and subjects alike, not merely as a means to justice. Comparative Advantages What are the advantages of this account when compared to other accounts of political legitimacy? In order to identify the comparative advantages, it will be helpful to revisit and clarify the distinctive features of my account. First, my account treats legitimacy as a system-level property, as opposed to saying that the very same regime is legitimate for one of its subjects but not for another.46

Is Political Legitimacy Worth Promoting?

85

Reflection on the role that legitimacy judgments play in our actual political deliberation shows us that there are certain advantages that come with treating legitimacy as a system-level property. For example, if a decision to support rebels in a civil war depends on whether the current regime is legitimate, then we are interested in a property of the regime, not a property that varies according to each individual subject.47 Another way of taking a system-level approach is to introduce a singular hypothetical perspective, but this imputation of unanimity erases important facts about the differing levels of quality assent across the population. My account, on the other hand, incorporates this individual variation, thereby avoiding the pressure to collectivize while retaining the systemlevel feature.48 This brings me to the second feature that I want to highlight. My account is summative, in the sense that it treats the system-level legitimacy of a regime as an aggregative function of the individual assessments of governance by each subject. No individual’s quality assent is necessary for the regime to be legitimate, but each person’s quality assent factors into the overall determination of legitimacy. The assent and non-assent of each subject are aggregated on an equal basis in order to reflect the moral ideal of voluntary subjection to rule, and its fundamental importance for each subject. The summation allows for the existence of quality assent to be a matter of contingent interests and values at the level of the individual. This approach to legitimacy judgments isn’t purely relativist because, for any given regime at a given time, there is an objective fact of the matter what degree of quality assent has been achieved. Nevertheless, the view retains a sensitivity to historical context, since quality assent depends on the actual beliefs of the regime’s subjects. Treating each person’s assent as relevant but not decisive for legitimacy allows my view to better address a tension that troubles other accounts, namely, the tension between the group and the individual. Do the views of outsiders count in the summation? This point relates to a distinction that is typically drawn between internal and external legitimacy. Sometimes the recognition of outsiders, such as the “international community,” matters for the effective functioning of a regime. However, the lack of internal recognition seems to be a greater threat to a regime’s legitimacy than

86

Amanda R. Greene

lacking external recognition.49 This suggests, correctly in my view, that the two audiences are not on a par. I relate this observation to the essential claim of rule, because if you were to assume that the international community didn’t exist, then it would still be possible to make sense of a political order as benefiting its subjects. But the same is not true if you were to assume that subjects didn’t exist. Thus, I am inclined to say that internal legitimacy is paradigmatic of the concept, and what we call external legitimacy matters only indirectly.50 The third feature of my approach that I want to highlight is that legitimacy is a matter of degree. Given that I understand legitimacy in scalar terms, one might worry that we lack the ability to simply call a particular regime illegitimate. But this is too quick. We can judge that a piece of fruit is unripe, even though we recognize that ripeness comes in degrees. In a similar way, a scalar concept of legitimacy need not preclude simple attributions of illegitimacy. I suspect, however, that any remaining unease about treating legitimacy as a scalar concept reflects a deeper worry. The worry is that legitimacy should properly be understood as a side constraint, in such a way that it can’t be tallied up as a quantifiable good that might, in principle, be balanced against other goods. But I think this worry is not as serious as it might first appear. Indeed, it brings to the foreground the most distinctive feature of my account. I think legitimacy should be understood as a good to be promoted, rather than a source of categorical prohibitions.51 There are both theoretical and practical advantages that come with embracing this feature in our thinking about legitimacy, and I address these in the following sections. Liberal Democratic Morality The first theoretical advantage of this feature—treating legitimacy as a good to be promoted that is distinct from other political goods—is that it clears up the role of liberal democratic morality in legitimacy assessments. There is a family of views of political legitimacy that take liberal democratic presuppositions about permissible coercion to be axiomatic; Bernard Williams has recently criticized these under the heading “political moralism.”52 One difficulty with such views is that they build substantive claims about

Is Political Legitimacy Worth Promoting?

87

what justifies the authority of the state into their conception of permissible coercion, in such a way that it is conceptually impossible for wrongful state coercion to occur in a legitimate regime. But this way of conceptualizing things makes it harder to recognize the dimensions of political illegitimacy that are distinct from mere wrongful coercion. The same is true for views of legitimacy based on liberal rights. If we define political legitimacy in such a way that the legitimate regime by definition doesn’t violate its subjects’ human rights, then, again, we obscure the dimensions of political legitimacy that aren’t included in a mere inventory of rights. In both cases, I suggest, these moralistic accounts of legitimacy muddy the waters. Building in liberal democratic commitments as axioms is both unnecessary and unhelpful in theorizing about legitimacy: unnecessary, because it is possible to show how achieving political legitimacy is valuable without these presuppositions, and unhelpful, because building them in clouds the relationship between legitimacy and liberal democratic values. It is entirely understandable that those with substantive liberal democratic commitments would want to see respect for liberal rights and democratic requirements as being necessary conditions of political legitimacy. But it is a mistake to think that one needs to build those substantive commitments into one’s notion of legitimacy in order to remain faithful to them. In certain contexts, it is natural to make assessments about the legitimacy of a regime that are, prima facie, orthogonal to judgments about whether and how that regime respects the requirements of liberal and democratic principles. Here an account of legitimacy like mine has an advantage, since it enables us to make sense of the kinds of assessments we are compelled to make in those contexts. And nothing essential is lost, for the liberal democrat, in adopting such a conception of legitimacy, since the only theoretical concession that this requires is that one regard one’s various evaluations of a regime—as being illegitimate, illiberal, or undemocratic—as complementary, rather than coextensive, evaluations. Certainly, in practice, this shift means that the liberal democrat cannot frame his democracy or rights-based objections to illiberal regimes as if those objections are about legitimacy. But those framings would, in any case, prematurely treat it as a settled question as to whether the requirements of legitimacy are coextensive with the requirements of

88

Amanda R. Greene

liberal democracy.53 My account can still explain why democracy and liberalism contribute to legitimacy, without building them into the concept of legitimacy.54 While it avoids the difficulties of moralism, my account is also able to avoid some serious problems with a realist theory of political legitimacy.55 Unlike realists, I don’t describe the value of legitimacy in terms of the peaceful maintenance of order and security. Hence, a stable and secure Hobbesian state can still be illegitimate if it fails to win subjects’ quality assent. Thus, the standard of quality assent can allow that subjects are being ruled illegitimately without slipping into the implication that they are no better than slaves or internal enemies. Because the standard of quality assent can distinguish between political order and legitimate political order, it is able to explain why a stable dictatorship that rules through fear is not a legitimate political order. In general, this is an explanatory burden that is difficult for realists to discharge. Distinct Basis for Criticism The standard of legitimacy in my account—namely, the achievement of quality assent—serves as a ready basis for criticism. No regime would deny that it has failed in some respect if it fails to attain quality assent. If we adopt quality assent as the standard of legitimacy, then regimes that lack it would be criticized for lacking something that they themselves acknowledge as worth having, that they should have.56 We can criticize them with a standard that they could not intelligibly disavow. This is not the case when we criticize a state for being illiberal or undemocratic, and so these standards provide less solid grounds for criticism. Furthermore, because the epistemic standards do not require the beliefs to be faultless, the standard is more attainable. Since it is something that can be attained, as opposed to a utopian regulative ideal, it removes a potential source of resistance to the critique. Moreover, an appeal to liberal democratic standards for political legitimacy can appear parochial, and for good historical reasons. There are many examples of liberal democratic ideals being used to legitimize neo-imperial projects.57 Unlike the aims involved in the minimal claim of rule, liberal democratic aims are not ones that all political regimes need to acknowledge insofar as they claim

Is Political Legitimacy Worth Promoting?

89

to rule. Thus, my account of legitimacy provides a basis for critique that is far less tainted with neo-imperialistic connotations. Still, it might be thought that quality assent is too morally thin. For example, there are various moral considerations that aren’t included explicitly. What about a state that behaves immorally, either to insiders or outsiders? These are valid concerns, but immoral treatment is under-described. In some sense of “immoral treatment,” all states are guilty as charged, e.g., unjust taxation schemes, failure to pay reparations, exploitative trade deals with other nations, etc. If these behaviors compromised legitimacy, then no state would be legitimate, and the standard would simply duplicate political morality in general. Since legitimacy is sometimes used to overcome intractable differences regarding what political morality requires, it seems better to retain other moral registers for these sorts of evaluations, such as whether a state is fair, egalitarian, or fully just.58 The standard of political legitimacy would have little point if it weren’t distinct from other modes of evaluation.59 By keeping legitimacy distinct from other modes of evaluation, we can better see what is lost in the absence of legitimacy. As I have defined it, a lack of legitimacy is not the same thing as a lack of justice or a lack of rights. In order for political legitimacy to remain a distinct mode of evaluation that corresponds to a distinct moral defect in a state, it cannot include all the things that we hope for, such as rights, equality, and democracy. In any case, we already have standards of moral evaluation for those things. What my account of legitimacy can show, instead, is how those things can diverge from political legitimacy. For instance, a regime may make gains in terms of justice but lose legitimacy along the way (as occurred with the European Union, arguably). Thus, the fact that quality assent corresponds to something unique and distinct counts in favor of using it as a standard of legitimacy. Only when it is conceptually independent can it show us what is at stake when we face trade-offs between multiple political values. Promoting Legitimacy The ability to treat legitimacy as a political good that is distinct from liberal democratic morality has theoretical advantages, but it

90

Amanda R. Greene

has practical ones as well. First, it enables policy-makers to aim at legitimacy as an end in itself. Since policy-makers need to be able to reckon with trade-offs, it allows them to treat the securing of legitimacy as one aim among others in a deliberation about what to do all things considered, whether it is a question of domestic health-care reform or foreign trade deals. My account of political legitimacy allows political actors to see legitimacy as a property of political orders in its own right, rather than treating legitimacy as a downstream by-product of justice. In order to spell this out, we need to refer back to the points made in the earlier section on dependent value, about why legitimacy is valuable in a non-instrumental way. My claim was that the value of legitimacy is grounded in a combination of social goods— non-alienation, stability, and political alignment—with which the realization of legitimacy is positively correlated. These are goods that a regime has reason to realize independently of any further benefits their realization might have, e.g., with respect to justice. Inasmuch as that’s the case, a regime will sometimes have reasons to act so as to promote its legitimacy, even where this conflicts with reasons derived from its other valuable aims. But equally, a regime will sometimes have reasons to prioritize its other valuable aims, even where this conflicts with its reasons to promote legitimacy. Legitimacy is one source of proper aims among others for any political regime. Even John Locke thinks that it is a distinct shortcoming of government if it acts in such a way that it loses the support of the people. He says, “When the governors have brought it to this pass, to be generally suspected of their people, . . . [this is] the most dangerous state which they can possibly put themselves in; wherein they are the less to be pitied, because it is so easy to be avoided . . . if he really means the good of his people, and the preservation of them.”60 Thus, even Locke acknowledges that something like quality assent is a proper aim of government, something they can be criticized for failing to attain even if they have acted justly. What I am trying to capture in these remarks is the kind of realpolitik that vulnerable regimes routinely have to contend with, where they are trying to realize the goods that political authority necessarily aims at realizing (e.g., related to the welfare and interests of the population), while simultaneously trying to strengthen

Is Political Legitimacy Worth Promoting?

91

their own claim to being the actor that oversees the realization of these goods. Williams describes these concerns in terms of the notion of political survival.61 If a moderate party has to choose between reducing immigration and staying in power, or increasing immigration and being replaced by a right-wing nationalist party, then my view shows that the former option—acting to maintain quality assent, while compromising other values—is not necessarily an illicit response to these challenges of governance. Those who understand legitimacy as an all-or-nothing side constraint, by contrast, can’t make sense of this dilemma as one that involves genuinely conflicting ends. Instead, they can only represent the standards of legitimacy that political regimes answer to as trumping or being coextensive with any other valid demands. Obligation and Rebellion I’ve been describing some of the comparative advantages of my account, compared to rival accounts of legitimacy. But one might think that my account has a decisive disadvantage compared to its rivals. In short, we often want to know what difference legitimacy makes. Does it tell us whether we have an obligation to obey the law, a duty not to rebel, or a right to intervene in sovereign states? One might think that my account doesn’t give us the resources to answer these sorts of questions. These are legitimate concerns, but they contain assumptions that I want to challenge. Suppose we think that legitimacy straightforwardly corresponds with a duty to obey. One way to capture this kind of view is simply to define legitimacy as whatever makes it the case that there is (domestically) a duty to obey and (internationally) a duty not to interfere. But doing so risks putting the cart before the horse. Deliberations about whether to rebel or intervene require us to first judge whether a regime is illegitimate. Consider an analogy. Suppose that a medical establishment advises that a patient undergo weight loss surgery if he is obese, but then simply defines “obese” as that weight above which one should undergo weight loss surgery. This would overly functionalize the definition of “obese,” reducing its meaning to its role in a decision. Similarly, some theorists, e.g., Allen Buchanan, have an overly functionalized concept of legitimacy, reducing its meaning

92

Amanda R. Greene

to its function as a decision criterion.62 When international lawyers have a debate about whether intervention is justified in Syria, for example, one of the questions they ask themselves is whether the Assad regime is legitimate. And they ask this question in a way that suggests that they are trying to latch onto a feature of the world that is analytically distinct from the question of what is to be done. The way that their question bears on their all-things-considered decision would make no sense if legitimacy just meant “such as to not be interfered with.” On my view, judgments about legitimacy do correspond to a feature of the world that we can first investigate, and then use in deliberation. Saying that the questions are distinct doesn’t mean that legitimacy is irrelevant for questions of whether to rebel or intervene. It just means that it doesn’t settle those questions by itself, by definition. Legitimacy normally contributes something to the reasons we have to obey the law, but it is a factor that hangs in the balance alongside other factors, in such a way that it can either strengthen our duty to obey the law, or be outweighed by factors that count against obeying the law.63 In other words, it is not merely equivalent to those questions. In fact, by not treating the verdicts on legitimacy as equivalent to verdicts on the duty to obey, we can resolve a paradox about the value of legitimacy. Mathew Coakley argues that legitimacy is the sort of thing that is capable of mattering only in those cases when it shouldn’t matter.64 That is to say, legitimacy could only come to matter in those cases where we lack interest- or rights-based reasons to obey the law. But in those cases, we simply lack sufficient reason to obey the law. And, he argues, in those cases we would be using the notion of legitimacy to “tip the scales” in circumstances where we shouldn’t tip the scales. I find Coakley’s argument persuasive, but what it demonstrates is not that legitimacy is normatively vacuous, but that we are wrong to seek a theory that equates legitimacy with the duty to obey. My view of legitimacy is able to avoid this paradox. Conclusion I have elaborated and defended a new theory of political legitimacy based on quality assent. According to my account, a regime is legitimate insofar as its subjects exhibit quality assent to being

Is Political Legitimacy Worth Promoting?

93

ruled. Assent to being ruled is a judgment-based acceptance by an individual subject that is consistent with the essential claim of rule, which is to provide benefits through the exercise of power and authority. Assent is quality assent when it is consistent with the minimal presupposition of this essential claim, i.e., that regimes provide for the basic security of their subjects. Therefore, I argued, regimes gain legitimacy by delivering benefits through ruling in such a way as to be recognized by their subjects as valuable instances of political order. This means that my view combines an emphasis on peaceful order, drawn from Hobbes, with an emphasis on subjective beliefs, drawn from Weber. I have argued that legitimacy, understood in these terms, is morally valuable. It corresponds to an ideal of voluntary rule, wherein each subject willingly accepts their political order in light of some values that they endorse. I argued that the realization of voluntary rule corresponds to three social goods. First, it correlates with the individual good of avoiding political alienation, such as that experienced by a rebel or revolutionary. Second, it corresponds to durable political stability, where widespread obedience is sustained without resort to fear or intimidation. And third, it corresponds to political alignment, a situation in which subjects converge broadly on the norms and values that govern the exercise of power, and their allegiance to them forms a kind of collective commitment to upholding them. The realization of voluntary rule has non-instrumental, though dependent, value. The value does not require the attainment of justice, but only basic security. Furthermore, the value does not depend on the beliefs that generate the acceptance being correct. So quality assent need not have the most pristine epistemic pedigree in order be valuable. I have argued that when political legitimacy is understood this way, it has significant advantages over rival accounts. It avoids embedding liberal democratic morality in the concept of legitimacy, a feature that leaves other normative standards of legitimacy open to objection on grounds of being parochial or ideological. It furnishes an undeniable basis of criticism, since no state can deny that it has failed insofar as it lacks quality assent. It also allows us to see legitimacy as an end worth promoting for its own sake, thereby showing how it could be weighed against other political goods that regimes seek to realize. I argued that allowing for various defects

94

Amanda R. Greene

in subjects’ beliefs permits legitimacy to be an attainable standard, thereby strengthening its role as a basis for universal critique. Finally, I argued that it helps us resolve a paradox in normative theorizing, which is that if legitimacy were to be equated with the duty to obey or right to rebel, then it wouldn’t be possible to see why it mattered. The advantages that I’ve explained and summarized here arise because I regard legitimacy as something attainable, yet normatively distinct and valuable. This makes my view better at explaining the puzzle with which I began: Why do we commonly regard a lack of legitimacy as a defect that ought to be fixed and is fixable? My account takes a more everyday view of what legitimacy requires: It bases the standard for legitimacy on what is essential to and undeniable about politics. It admits that politics involves subjection, rather than seeking to re-describe it in terms of nonsubjection—self-rule cannot be the aim, even in theory. At the same time, my conception better navigates the tension between the group and the individual that confronts any account of legitimacy. On my view, there is no individual demand for or right to justification of the state that, when satisfied, leads directly to legitimacy. Rather, the claim is that it is good for individuals to be willing subjects, and it is bad for them to be unwilling subjects. This ideal of voluntary rule, therefore, allows us to see that political legitimacy is a good to be promoted—alongside justice, but distinct from it. Acknowledgments For helpful feedback and discussion, I would like to thank Kristen Bell, Joshua Cohen, Tom Dannenbaum, David Estlund, Gordon Finlayson, Chandran Kukathas, R. J. Leland, Eliot Michaelson, Enzo Rossi, Andrea Sangiovanni, Debra Satz, Melissa Schwartzberg, Scott Shapiro, Robert Mark Simpson, Zofia Stemplowska, Anna Stilz, Adam Swift, Laura Valentini, Daniel Viehoff, Leif Wenar, Han van Wietmarschen, Allen Wood, Lea Ypi, and audiences at Chicago, KCL, LSE, Milan, Northwestern, NYU, Oxford, Sussex, UNC, Warwick, Yale, and York.

Is Political Legitimacy Worth Promoting?

95

Notes 1. For a discussion of the difficulty of defining political legitimacy, see section 1 of Fabienne Peter, “Political Legitimacy,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, Summer 2016 (Metaphysics Research Lab, Stanford University, 2016), https://plato.stanford.edu. 2. On political obligation, see A. John Simmons, Justification and Legitimacy: Essays on Rights and Obligations (Cambridge: Cambridge University Press, 2000); Margaret Gilbert, A Theory of Political Obligation: Membership, Commitment, and the Bonds of Society (Oxford: Oxford University Press, 2006); George Klosko, Political Obligations (Oxford: Oxford University Press, 2005). On legitimate authority, see Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986); Joseph Raz, “The Problem of Authority: Revisiting the Service Conception,” Minnesota Law Review 90 (2006): 1003–1044; Leslie Green, The Authority of the State (Oxford: Clarendon Press, 1990); Stephen Perry, “Political Authority and Political Obligation,” Oxford Studies in Philosophy of Law 2 (2013): 1–74; John Tasioulas, “The Legitimacy of International Law,” in The Philosophy of International Law, ed. Samantha Besson and John Tasioulas (Oxford: Oxford University Press, 2010), 97–116. 3. See Joshua Cohen, “Deliberation and Democratic Legitimacy,” in The Good Polity: Normative Analysis of the State, ed. Alan P. Hamlin and Philip H. Pettit (Oxford: Basil Blackwell, 1989), 17–34; John Rawls, Justice as Fairness: A Restatement, ed. Erin Kelly (Cambridge, MA: Belknap Press of Harvard University Press, 2001); David Estlund, Democratic Authority: A Philosophical Framework (Princeton, NJ: Princeton University Press, 2007); Thomas Christiano, The Constitution of Equality: Democratic Authority and Its Limits (Oxford: Oxford University Press, 2008). 4. For an argument that human rights are a necessary condition for legitimacy, see Allen Buchanan, The Heart of Human Rights (New York: Oxford University Press, 2013). For an argument that republican nondomination is necessary for legitimacy, see Philip Pettit, On the People’s Terms: A Republican Theory and Model of Democracy (Cambridge: Cambridge University Press, 2012). 5. One exception may be the revival of political realism, though their allegiance to the method of critical theory hampers the construction of a full-blown theory of legitimacy. Raymond Geuss, The Idea of a Critical Theory: Habermas and the Frankfurt School (Cambridge: Cambridge University Press, 1982); Raymond Geuss, Philosophy and Real Politics (Princeton, NJ: Princeton University Press, 2008); Bernard Williams, In the Beginning Was the Deed: Realism and Moralism in Political Argument, ed. Geoffrey Hawthorn (Princeton, NJ: Princeton University Press, 2005); Enzo Rossi and Matt

96

Amanda R. Greene

Sleat, “Realism in Normative Political Theory,” Philosophy Compass 9, no. 10 (October 2014): 689–701. For an analysis of the limitations of a realist theory of political legitimacy, see Amanda R. Greene, “The Promise of Political Realism” (Rochester, NY, 2018), https://papers.ssrn.com. 6. Awareness of this gap has been growing. For instance, Jane Mansbridge argues that we need an account that specifies the sufficient conditions, because we lack the levels of legitimate coercion that are needed to solve major global problems like climate change. Jane Mansbridge et al., “Listening to One’s Constituents? Now, There’s an Idea” (Brian Barry Memorial Lecture, London School of Economics and Political Science, May 15, 2017), http://www.lse.ac.uk/. 7. Max Weber, Economy and Society: An Outline of Interpretive Sociology (Berkeley: University of California Press, 1978); David Beetham, The Legitimation of Power (Atlantic Highlands, NJ: Humanities Press International, 1991); Tom R. Tyler, Why People Obey the Law (Princeton, NJ: Princeton University Press, 2006). 8. John Rawls, Political Liberalism, with a New Introduction and the “Reply to Habermas” (New York: Columbia University Press, 1996); Jürgen Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy (Cambridge, MA: MIT Press, 1996). 9. H.L.A. Hart, The Concept of Law (Oxford: Clarendon Press, 1961); Raz, The Morality of Freedom. 10. Those who work on normative legitimacy need not exclude its descriptive sense from their theory, but in practice they usually do. For endorsements of the strict distinction, see among others Allen Buchanan and Robert O Keohane, “The Legitimacy of Global Governance Institutions,” Ethics & International Affairs 20, no. 4 (December 2006): 405– 437; Estlund, Democratic Authority; David Copp, “The Idea of a Legitimate State,” Philosophy & Public Affairs 28, no. 1 (1999): 3–45; Rawls, Justice as Fairness; Tamsin Shaw, “Max Weber on Democracy: Can the People Have Political Power in Modern States?,” Constellations 15, no. 1 (2008): 33–45; Fabienne Peter, Democratic Legitimacy (New York: Routledge, 2009). Legal theorists reproduce the divide within the discipline by distinguishing between analytic and normative jurisprudence, though this distinction is contested in H.L.A. Hart, “Positivism and the Separation of Law and Morals,” Harvard Law Review 71, no. 4 (1958): 593–629; Ronald Dworkin, Law’s Empire (Cambridge, MA: Belknap Press, 1986). 11. Beetham, The Legitimation of Power, 7. 12. Sabina Alkire and James Foster, “Counting and Multidimensional Poverty Measurement,” Journal of Public Economics 95, no. 7 (August 2011): 476–487; James D. Fearon, “Ethnic and Cultural Diversity by Country,” Journal of Economic Growth 8, no. 2 (June 2003): 195–222.

Is Political Legitimacy Worth Promoting?

97

13. Peter, Democratic Legitimacy; Estlund, Democratic Authority. 14. Beetham, The Legitimation of Power; Margaret Levi, Audrey Sacks, and Tom R. Tyler, “Conceptualizing Legitimacy, Measuring Legitimating Beliefs,” American Behavioral Scientist 53, no. 3 (November 1, 2009): 354–375; Bruce Gilley, The Right to Rule: How States Win and Lose Legitimacy (New York: Columbia University Press, 2009). 15. From time to time the sociological concerns break into the normative discussion, e.g., debates about feasibility. See David Estlund, “Utopophobia,” Philosophy & Public Affairs 42, no. 2 (March 2014): 113–134, and non-ideal theory, Charles W. Mills, “‘Ideal Theory’ as Ideology,” Hypatia 20, no. 3 (2005): 165–184; A. John Simmons, “Ideal and Nonideal Theory,” Philosophy & Public Affairs 38, no. 1 (January 2010): 5–36. 16. Raz, The Morality of Freedom; Raz, “The Problem of Authority.” 17. Green, The Authority of the State. 18. Pettit, On the People’s Terms. 19. Walter Benjamin, “Critique of Violence,” in Reflections (New York: Schocken Books, 2007), 277–301. 20. That this presumption stands in need of defense becomes especially apparent in recent libertarian approaches to the legitimacy of the state. Michael Huemer, The Problem of Political Authority: An Examination of the Right to Coerce and the Duty to Obey (Basingstoke: Palgrave Macmillan, 2012); Jason Brennan, “When May We Kill Government Agents? In Defense of Moral Parity,” Social Philosophy and Policy 32, no. 2 (2016): 40–61. 21. The term regime comes from the Aristotelian definition of politeia: “an arrangement of a city with respect to its various offices, particularly the one that is supreme over all matters,” in Aristotle, Politics, trans. C.D.C. Reeve (Indianapolis, IN: Hackett, 1997), 1278b10–11. In jurisprudential terms, political order corresponds to a legal system, defined as a union of primary and secondary rules (see Hart, The Concept of Law). Accordingly, a political order consists of abstract institutional features (what a Rawlsian would call constitutional essentials) as well as concrete instantiations of those features (e.g., how the ruling party or government is currently exercising power via those institutional features). 22. The idea that political orders make an essential claim is related to the idea that law makes a moral claim for itself. Though this idea goes back to Plato’s observation that law aims at moral correctness (Minos 315– 317), it continues to be defended today. Plato, Complete Works, ed. John Cooper and D. S. Hutchinson (Indianapolis, IN: Hackett, 2005), 1310– 1313; Raz, The Morality of Freedom; John Gardner, “How Law Claims, What Law Claims,” (Rochester, NY, 2008), https://papers.ssrn.com; Scott J. Shapiro, Legality (Cambridge, MA: Belknap Press, 2011). 23. David Beetham, The Legitimation of Power, 265.

98

Amanda R. Greene

24. I deliberately avoid the term private violence because many regimes have treated private violence as licit, i.e., authorized de facto if not de jure (e.g., domestic violence, corporal punishment). 25. In this sense my account differs from the vague proposal by Schmitt that the capacity to effectively determine who counts as a member of a community is definitive of political power. Carl Schmitt, The Concept of the Political: Expanded Edition (Chicago: University of Chicago Press, 2007). 26. Here I deliberately say civil war, because I think a warrior society that is engaged in regular military activity vis-à-vis outsiders (e.g., Sparta) can attain political legitimacy. Eliminating the threat of civil violence is also what social scientists take to be relevant for attaining legitimacy, in both established and emerging states. Kristin M. Bakke et al., “Convincing State-Builders? Disaggregating Internal Legitimacy in Abkhazia,” International Studies Quarterly 58, no. 3 (September 2014): 591–607; Kristin M. Bakke, Andrew M. Linke, John O’Loughlin, and Gerard Toal, “Dynamics of State-Building after War: External-Internal Relations in Eurasian de Facto States,” Political Geography 63 (March 2018): 159–173; Douglass C. North, John Joseph Wallis, and Barry R. Weingast, Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History (Cambridge: Cambridge University Press, 2009). 27. I defend this reading of Weber in Amanda R. Greene, “Legitimacy without Liberalism: A Defense of Max Weber’s Standard of Political Legitimacy,” Analyse & Kritik 39, no. 2 (2017): 295–324. 28. In the past I defended a view of legitimacy based on a form of tacit consent. Though the view remains broadly the same, I have come to think that this term is misleading. See Amanda R. Greene, “Consent and Political Legitimacy,” in Oxford Studies in Political Philosophy, Volume 2, ed. David Sobel, Peter Vallentyne, and Steven Wall (Oxford: Oxford University Press, 2016), 71–97). 29. One can accept that quality assent has value even if one is reluctant to assign it the name “political legitimacy.” 30. For a discussion of dependent non-instrumental valuation, see Martin E. Sandbu, “Valuing Processes,” Economics and Philosophy 23, no. 2 (2007): 205–235; Joseph Raz, “Facing Up: A Reply,” Southern California Law Review 62 (1988): 1228. There is a related debate in ethics about the relation between two distinctions: final/instrumental goods and conditional/ unconditional goods: Rusty Jones, “The Real Challenge of the Republic” (presentation, Ancient Greek and Roman Philosophy Workshop, University of Chicago, May 6, 2015); Christine M. Korsgaard, “Two Distinctions in Goodness,” Philosophical Review 92, no. 2 (1983): 169–195. 31. This example is drawn from Plato’s Laws, where the main speaker compares doctors who address slaves with doctors who address free men

Is Political Legitimacy Worth Promoting?

99

(Laws, 720). Plato, Complete Works, 1406. He analogizes willing acceptance of laws with the free person’s willing obedience of a doctor. For discussions of the implication for Plato’s late view of political authority, see Amanda R. Greene, “Legitimacy and Democracy: A Platonic Defense of Voluntary Rule” (PhD thesis, Stanford University, 2014). 32. The view was first outlined as the “sovereignty conception” in Greene, “Consent and Political Legitimacy.” 33. For instance, Rawls treats stability as a problem of moral psychology that bears on justice only indirectly. He is concerned about individuals’ beliefs and their consequences for the willingness to comply, and he treats this merely as a precondition rather than as an ingredient of justice. John Rawls, A Theory of Justice (Cambridge, MA: Belknap Press of Harvard University Press, 1971), 336–337; Rawls, Justice as Fairness, 185–186. 34. Weber, Economy and Society, 213. 35. Beetham, The Legitimation of Power, 6. 36. Tom R. Tyler and Jonathan Jackson, “Future Challenges in the Study of Legitimacy and Criminal Justice” (Rochester, NY, 2013), https:// papers.ssrn.com, 13. 37. Tyler, Why People Obey the Law. 38. J. Jackson et al., “Monopolizing Force? Police Legitimacy and Public Attitudes Toward the Acceptability of Violence,” Psychology, Public Policy, and Law 19, no. 4 (November 2013): 479–497. 39. Edmund Burke, Reflections on the Revolution in France, ed. J.G.A. Pocock (Indianapolis, IN: Hackett, 1987), 67. 40. Gilley, The Right to Rule, 141. 41. Though I lack space to defend it here, I also see this alignment as corresponding to nineteenth-century ideals of recognition or subjective freedom. 42. This element of the view is the chief point of contrast with political theories that emphasize values like solidarity and reconciliation (e.g., communitarianism). 43. I take this to be little more than an elaboration of one way in which legitimate authority can be established through satisfaction of the “normal justification condition” that is outlined in Raz’s service conception. Raz, “Problem of Authority,” 1014. 44. I thank David Estlund for pressing me to clarify this point. 45. I regard this point as akin to the problem of proleptic reasons, as outlined recently by Agnes Callard. Taking the example of the music appreciation class, she argues that the new student cannot grasp the intrinsic value of music, but they nevertheless have a reason for enrolling that is related in a nuanced way to the intrinsic value of music. Agnes Callard,

100

Amanda R. Greene

“Proleptic Reasons,” in Oxford Studies in Metaethics 11, ed. Russ-Shafer Landau (Oxford: Oxford University Press, 2016), 129–155. 46. Simmons, Justification and Legitimacy. 47. For a rich discussion of how our theorizing ought to be responsive to the functional role of legitimacy assessments, see ch. 5 of Buchanan, The Heart of Human Rights. 48. For further elaboration of this advantage, see Greene, “Consent and Political Legitimacy,” 91. 49. Ilaria Cozzaglio, “Beliefs in Legitimacy and the Normative Role of Coherence” (presentation, Graduate Conference in Political and Legal Theory, University of Warwick, 2017). 50. However, this indirect relation might turn out to be very important. It might be that a new government within a state requires external resources and favor in order to govern effectively, and so external recognition becomes an essential precondition for securing internal legitimacy. Bakke et al., “Dynamics of State-Building after War.” 51. I thank R. J. Leland for pressing me to clarify this point. 52. Williams, In the Beginning Was the Deed. 53. I thank Robert Mark Simpson for pressing me to clarify this point. 54. For an argument that understanding legitimacy in terms of voluntary rule provides the most satisfactory explanation of the legitimacy of democracy, see Greene, “Legitimacy and Democracy: A Platonic Defense of Voluntary Rule.” 55. Williams, In the Beginning Was the Deed. 56. This idea is connected to Socratic cross-examination. Jonathan Lear argues that Socrates repeatedly shows his interlocutors that they are failing to live up to the essential implicit standards of being a human being, where these include epistemic and moral norms. The failure to live up to an ideal that is implicit in one’s self-presentation is a source of shame and internal dissonance. Jonathan Lear, A Case for Irony (Cambridge, MA: Harvard University Press, 2011). 57. Human rights discourse is accused of being an instrument of global hegemony, often with a sound historical basis. Samuel Moyn argues, for example, that the rise of human rights talk in international affairs coincided with America’s liberal internationalist approach to foreign policy, which was clearly to the economic advantage of the United States. Samuel Moyn, The Last Utopia: Human Rights in History (Cambridge, MA: Belknap Press, 2010). 58. Peter, “Political Legitimacy.” 59. This leaves room for Ronald Dworkin’s proposal that the “internal legitimacy” of a state depends on its behaving morally in the international arena, since the state’s claim of authority over its citizens depends partly

Is Political Legitimacy Worth Promoting?

101

on fulfilling its duty to uphold a system of international law. Ronald Dworkin, “A New Philosophy for International Law,” Philosophy & Public Affairs 41, no. 1 (January 2013): 2–30. 60. John Locke, Second Treatise of Government, ed. C. B. Macpherson (Indianapolis, IN: Hackett, 1980), 106. 61. Williams, In the Beginning Was the Deed, 13. 62. A functionalist account of legitimacy has been defended by Buchanan in the past, although his views on legitimacy have evolved significantly. In his latest published work, he proposes that the exchange of legitimacy judgments allows for successful coordination on minimally justified institutions, i.e., the “meta-coordination view.” In my view, this is still too functionalist—it defines the criteria for the concept in terms of one role it plays in coordinating joint action, as seen from the outside. Buchanan, The Heart of Human Rights. 63. For example, Thomas Shelby has argued that failures of justice for some subjects can affect their political obligations. Tommie Shelby, “Justice, Deviance, and the Dark Ghetto,” Philosophy & Public Affairs 35, no. 2 (April 1, 2007): 126–160. 64. Mathew Coakley, “On the Value of Political Legitimacy,” Politics, Philosophy & Economics 10, no. 4 (November 2011): 345–369.

4 THE SOVEREIGN AND THE REPUBLIC A REPUBLICAN VIEW OF POLITICAL OBLIGATION EKOW N. YANKAH

How a sovereign can legitimately impose obligations on individuals seems not only a core question but perhaps the first question of political philosophy. The legitimate authority to govern is naturally prior to inspecting the various goals at which one can aim while governing. The question is all the more pressing given that questions of legal authority do not stay abstract long; legal authorities define rights and duties, imply the obligation to obey the law as well as the power to coerce others into obedience. A political theory without a convincing answer as to how the power of the sovereign can be justified (even if it answers in the negative) hardly seems to qualify as a political theory. In outlining the answer to political questions, whether the nature of legal equality, the justification of property holdings or punishment, that an answer serves as the basis of legally enforceable rights lends these questions their particularly political nature. The legitimization of sovereignty is particularly vexing because we live in an avowedly liberal age. In a very real sense the sovereign casts a shadow over every political question. Liberalism, in its many forms, is tied together by a core principle that secures pride of place to individual autonomy or freedom. Because individual autonomy is central to liberal theories, there is a natural tension between the liberal starting point and what most take to 102

The Sovereign and the Republic

103

be the obvious need to justify some form of governance. Save the anarchists among us, few can be persuaded by a theory that does not generate some view of justifiable government. Yet the answers at which liberal theories inevitably arrive, I will argue, are either implausible to most or involve a bit of philosophical alchemy in which the distinctly autonomy-loving and freedom-preserving core of liberalism is substituted for something very different. This philosophical transformation travels under well-known labels; it hides under the “veils of ignorance,” within hypothetical contracts that have never been made, or in the shape of combined wills that spring forth instantaneously. The reason liberal theories either fail to convince or twist themselves into unrecognizable knots is, I suggest, they start with the wrong view of political obligation. From the point of view of liberalism, human beings are defined first and foremost by their autonomy or freedom-preserving nature. Kant, the foundational thinker of much of modern liberalism, premised his legal philosophy on the fact that human beings uniquely have rational wills that can govern their brute desires in complex ways.1 If each person possesses a rational will that cried out for respect, no one person can unilaterally justify interfering with another person’s freedom. In the words of Arthur Ripstein, the core Kantian impulse is that you be the master of your own life and slave to none.2 On this view, the human condition, at least in terms of political obligation, consists of independent rational beings (wills) jealously guarding their boundaries against encroachment. From this view, the anxiety of policing the justification of the state’s intrusive use of power makes perfect sense. It is only familiarity that has led to such widespread acceptance of what is actually a very strange picture as a plausible basis of political justification. Nowhere in social science, much less our everyday lives, does this image accord with our most natural intuitions. Human beings, to foreshadow the argument, do not guard their independence above all. Rather, human beings are deeply social creatures; not only do people seek to live together but for most, deprivation of other persons constitutes tragic deformation in their development or a form of near torture. That this is true in our social lives, I will argue, has important implications in our political lives.

104

Ekow N. Yankah

In what follows, I argue for an ironically radical view of political sovereignty. The proposed view is radical because it displaces the liberal independence that is nearly hegemonic in political theory today. It is ironically so because it is based on a most natural and intuitive picture of human relationships. Indeed, the republican justification of political sovereignty is so natural that I will point out that many opposing (liberal) views, either explicitly or implicitly, rely on republican intuitions to legitimize sovereignty. Rather than start with liberal independence followed by some fancy footwork to arrive at thinly disguised republican theories, we should start with a full-blooded republican theory that highlights the civic bonds and duties that constitute political obligation and illustrate most clearly that sovereignty is only justified when it is mutually constituted by our shared civic bonds, rights, and duties. One last bit of throat clearing. The limitation of state power is of such particular importance to the modern mind that even political theories whose roots are set in very different soil have slowly evolved to center on individual liberty. Indeed, to nominate a republican theory of political obligation in the modern philosophical landscape risks misunderstanding from some and causes others to question whether one is offering a genuine alternative. Republicanism has come to be associated with the “civic republicanism” of Quinton Skinner, which made central the freedom of the individual from arbitrary state interference.3 Its most powerful modern proponent, Philip Pettit, champions a republicanism based on the value of “non-domination,” that is, the right to be free of the threat of arbitrary interference. For Pettit, what is crucial is that legal interference under roughly just conditions does not constitute arbitrary interference and thus is not illegitimate.4 Though both theories share commonalities, Skinner and Pettit’s Machiavellian or “Roman” republicanism not only secures but makes central the individual’s freedom from government action, downplaying the civic bonds that I wish to highlight. In seeking a political theory premised on the more accurate and intuitive view of human interdependence rather than independence, a natural place to look is to classical republican views of political obligation and civic bonds. The republican view I propose is explicitly premised on an Aristotelian or Athenian theory of civic duty that makes interconnectedness central to political

The Sovereign and the Republic

105

justification. In fact, there is something like a paradox in examining Aristotelian republicanism to build a theory of sovereignty. Despite the fact that our shared civic duty is the basis for a republican theory of sovereignty, Aristotle’s work does not explicitly address the question of the state’s power over the individual, at least not as it is expressed in modern terms.5 As we shall see, for Aristotle the idea of people being bound in a political community was so natural that there was no need to take up the question in it contemporary form. Thus, constructing a full model of Aristotelian sovereignty will require inspecting the underlying premises of ancient Greek political philosophy as well as inspecting his views on how civic bonds bind and particularly where they fail. Despite being grounded in Aristotelian tradition, the project ahead is not primarily one of exegesis; the claims are meant to be live ones.6 Though I will rely on Aristotle for the scaffolding, the goal is to extract a persuasive view of reciprocal civic bonds binding citizens to each other and flowing back and forth between citizen and state. Not surprisingly, the goal is to render a picture of political obligation that sheds light on contemporary notions of political obligation and legitimacy and gives purchase on today’s pressing questions. Social and Civic Beginnings The question of sovereignty appears vexing to modern liberal political theories because they are fundamentally premised on the image of atomistic individuals who, afraid of crashing into each other, jealously guard their independence from invasion. Making freedom from another the central justification of political theory means that when the time comes to propose some form of justified government, such theories must somehow transform the right to coercively impose legal duties into a species of individual freedom, for example through hypothetical consent. Twisting oneself into knots is the inevitable end of a theory that starts from an unnatural beginning. Rather than begin from the antiseptic liberal starting point, let us begin, with Aristotle, by examining human beings as they actually are. Aristotle, like Kant, describes humans as unique in our highly developed rational capacity. But rather than join Kant in viewing our rational

106

Ekow N. Yankah

capacities as an austere freedom-seeking will, Aristotle starts with the perfectly natural notion that human beings are deeply social animals.7 In the normal and overwhelming set of cases, we seek each other out. This perfectly obvious observation is in stark contrast with the “state of nature” from which liberal theories often expend great energy to escape.8 Aristotle explicitly eschews any suggestion of the state of nature, hypothetical or otherwise, a theoretical commitment that turns out to be meaningful. Aristotle instead argues that even pre-political man was driven to find a mate, couple, and form a family.9 Forming a family is only the earliest manifestation of man’s social instinct. Small families are naturally drawn to live together.10 In the first place, this is because living together is a matter of material necessity; households and small villages can more easily secure the necessities of human life—food, protection from enemies and the natural elements, and so on. As such communities develop, another natural element of human nature pronounces itself. Human beings are not simply satisfied to live or even to live comfortably. Once the material necessities of life are secure, our rational nature is driven to find ways of living well. Again, this observation is so simple as to hide its import. Lions, fed after a hunt on the Serengeti, strike a beautiful pose of power and contentment. They will remain so until their hunger moves them again. While people may replicate this life for stretches at a time, enjoying a beach vacation, human lives do not resemble that of a lion. Secure against hunger and pain, human beings begin to pursue and construct more complex visions of lives worth living. Even when we are tempted to scoff at the inveterate television watcher, a moment’s reflection reveals the truth of this proposition. One need not pursue the life of the proverbial Oxford don to demonstrate the natural inclination to build a life that reflects values. The job one chooses, the hobbies and friends, the vacation for which one saves, the city to which one moves for its action and noise, or the city to which one refuses to move because of its action and noise; these and a thousand other decisions so subtle we hardly notice them, are the stuff through which we pursue a vision of good life. Our ability to pursue a good life is so dependent on our living together that Aristotle argues that there is a real sense in which

The Sovereign and the Republic

107

reaching our full humanity turns on our bonding.11 The claim is not a metaphysical one, of the nature that can be read into Rousseau. It is rather a simple observation that many of our most elevated capacities cannot be developed in the absence of a flourishing community that has secured the necessary material benefits for its members. If one were only to encounter a single person or a small band struggling daily to survive—think of the heartbreaking pictures of war-torn or drought-ridden nations—one could hardly imagine the heights to which human beings can develop. Our museums, jazz clubs, chess tournaments, and haute cuisine are all possible just because we bond together not only to live but to live well. Having highlighted that part of the natural expression of human rationality is to seek out an intelligible version of the good life, another feature becomes clear. Human beings seek out the good life together. In Aristotle’s words, we are not only social animals in that we seek each other out but we are political animals in that we largely define a good life in common.12 We do not merely live beside each other, we live together. To return to the animal analogy, human beings do not simply seek to live around each other, as cows do, but find the good life in shared and interwoven projects.13 In noticing this, we realize that our deep desire to live together is not based simply on maximizing food, shelter, protection, utility, or any other such good.14 We rely on each other not just for material well-being but for social and emotional support.15 Shifting one’s attention from the merely instrumental benefits of shared living to the richer goods it provides highlights another compelling truth about human beings: They are not simply rational agents, they are moral agents with thick and rich ends, both individual and shared, that extend beyond a crass view of singular interests.16 Unlike the atomistic view that undergirds modern liberalism, Aristotelian republicanism makes central that we are interconnected in complex ways that ground our communal and political identities.17 Put so plainly, this claim seems a bit of abstract armchair sociology. It is worth noting, however, that this view is truer to our life experiences and accords with the social science literature as it turns its attention to our natural interconnectedness. Our social and civic interconnectedness dissolves problems

108

Ekow N. Yankah

of legitimacy that are baffling from the point of view of the committed liberal. Take the stock liberal argument, perhaps made most famous by H.L.A. Hart but presaged by Locke and endorsed by Rawls, that justification for coercive imposition of laws in a liberal state can be premised on a certain sense of “fair play.”18 On this view, the benefits each individual gains because others submit to a system of laws justifies coercive enforcement of those laws (and presumably promotion through taxation) in order to prevent the dangers of free riders. As Nozick points out, however, fair play arguments need to explain why the mere fact that one’s neighbors begin a volunteer system of free public radio broadcasts does not obligate you to participate in the scheme (even if you enjoy the music very much).19 Put more accurately, no system that elevates autonomy or freedom as a singular political justification can adequately justify forcing one to participate based simply on an argument from fair play.20 But a step back from Nozick’s example allows us to ask how convincingly it captures our purported natural freedom. If everyone in a neighborhood spruced up their home, the natural reaction in most would not be to wonder if neighbors would be justified in forcing one to do so as well. Rather, the natural inclination would be to do so as well; to avoid the embarrassment of being the “rundown” house on the block.21 Indeed, I suspect that for most people embedded in a community bonded enough to organize Nozick’s public music organization, one would be more likely to feel pain or anxiety in not being included than fear of being forced to participate. Obviously, such charming hypotheticals do not resolve the question of whether it is justified to pressure the holdout or the recalcitrant, but noting how different our natural reactions would be from those idealized when we make autonomy our primary motivation begins to reorient our view of political justification. Reorienting our initial instincts allows for a more persuasive foundational view not just of social interaction but of political justification, to which we will turn shortly. The liberal and the libertarian are, to be sure, different political creatures. Nonetheless, because individual freedom is the central human characteristic to be preserved, both will render fair play an implausible ground for political obligation. Shifting our gaze from the preservation of freedom highlights the idea that the impulse to

The Sovereign and the Republic

109

join with one’s community to pursue goods that are jointly defined as valuable is entirely natural. Social scientists are increasingly focusing on the ways people are naturally driven to pursue various goods when they are informed that others within their community value and pursue those same goods. For example, merely informing people that the majority of people in their community do not litter, take wood from national parks, or recycle results in people conforming their behavior to the community norm.22 This effect holds true even in cases of important and otherwise recalcitrant problems. Informing college students that the majority of their peer group do not engage in binge drinking reduces binge drinking among those who do.23 Remarkably, informing people of the percentage of fellow citizens who pay their full taxes on time increases compliance with tax laws.24 Whereas Nozick’s example misleads one’s intuitions by choosing a charming but innocuous hypothetical, what our experiences and social science reveal is that in matters small and large, we are naturally driven to define and pursue our values in common. Even this underestimates the claim. It is not simply that humans are susceptible to being persuaded by others to pursue certain goods. It is that people define that which is good by reference to each other. In the most intimate of situations, this is perfectly obvious; it is perfectly natural for a father to respond to the question of how he is doing by relating the latest triumph or mishap that has befallen his daughter.25 The well-constructed person experiences the well-being of those near and dear to her as part and parcel of her well-being.26 Though not as intimate, this feeling extends to those with whom we share a civic community.27 How well our world is going turns in part on how well our world is going for our fellow citizens. Seeing crippling poverty anywhere is distressing but seeing it in one’s community touches one in a different way. Visiting another country that seems to have scored better in taking care of its poor fills one with the thought, “why can’t we do this back home?” And this is true even if one’s wealth and gated community means poverty does not affect one directly. In fact, what one notices is that even if you are wealthy, the poverty or suffering of one’s fellow citizens does affect you directly. The good for our polity is part and parcel of the good for us; we share in its faults and its excellences. We cheer for our teams

110

Ekow N. Yankah

during the Olympics and are saddened and angered when we learn of some injustice done by our government in a way that is different than if the same tragic event was committed in someone else’s name.28 One need only think of the many ways in which our projects are only recognizable insofar as they are part of ongoing projects that will benefit our community long after ourselves. Sam Scheffler illustrates this elegantly, demonstrating the ways in which we naturally make sense of our lives by orientating ourselves around traditions, both individual and communal, and pursuing projects that extend beyond our own futures by benefiting our communities.29 Scheffler calls to mind the scientist working on a cure for cancer. That she cannot hope to cure cancer in her lifetime would not undermine the value of her research.30 Her project is valuable as a contribution to curing cancer even in the case where no benefits will appear during her lifetime. Learning that her community will disappear twenty-five years after her death, however, would drain her life’s work of value, again, despite that in either case, she will have died. One need only imagine that the scientist is working on a disease that particularly affects her community or a caretaker of some important cultural marker and the intuition grows stronger.31 The conclusion draws attention to the fact that much of what gives our lives value is the way in which our projects contribute to others generally and our community particularly, regardless of whether we will be otherwise affected or even be alive to realize the benefits.32 Here again, we begin to see the distinctiveness of the Aristotelian political view and its interconnection with views of human nature. The view that one’s well-being is not just affected by the well-being of one’s community but in many ways constituted by the well-being of one’s community seems obvious enough. Thus, it is important to note that one’s well-being is in turn often constituted by a particular political community with particular views of what constitutes a good life. Morality, on this view, loses an important form of liberal neutrality; that is, the content of morality, in some large range of questions, will not be independent from where and when I learned morality.33 This is not to acquiesce to radical relativism; Aristotle reminds that both ethical obligations must be justified by their role in promoting human flourishing and political obligations by their securing political equality in service of the

The Sovereign and the Republic

111

common good. Our worst worries about community-dependent morality come from moral codes that are impossible to describe as meeting either requirement. All of this is simply a way of elaborating what we already know. But it is important to reconnect these ideas to their distinctive political picture. An important part of how we define what is valuable is defined together as a community. Further, important parts of what we define as valuable are intertwined with the good of others, those near and dear to us first but including our wider community. We deliberately seek to share our lives and grow together as communities to pursue a more robust vision of the good life, form friendships and partnerships, and care for emotional and social needs. It is this interconnectedness that gives credence to Ross’s colorful translation of Aristotle, “Man is born for citizenship.”34 Of course there are some rare individuals who wish to spend the vast majority of their lives in complete solitude. But these individuals attract our attention precisely because they are rare, and when taken to its limits such behavior can be evidence of deep emotional trauma or mental illness. These observations do not pretend that there are no holdouts in important social situations. Clearly, there are those who hold out from neighborhood projects and from much more important civic projects as well. In particular, as visions of the good life become increasingly pluralistic or political demands become increasingly suffocating, it will be less likely one can assume the kind of compliance indicated, a point to which we will return. Still, it is obvious that, on the whole, human beings seek out and flourish as part of social and political communities.35 Of course, there is nothing in the Kantian or liberal picture that need deny that one would feel some amount of social anxiety if excluded from neighborhood organizations. That is because, at least on the Kantian view, liberal autonomy does not state a thesis about human nature or psychology. Rather, it makes a claim about moral ontology; that is, that human beings are morally selfsufficient. Because persons have innate freedom or independent moral worth, their rights do not derive from others or their community. This view allows that social interdependencies may be driven by rational self-interest just so long as interactions are on terms consistent with equal moral worth.36

112

Ekow N. Yankah

The Aristotelian picture on offer, however, not only shifts our attention but invites us to question why one ought accept the Kantian picture. If the Kantian or liberal story does not exclude human sociability, its focus on innate rights naturally leads to an insistence on our rights as against each other and the state. Highlighting the enduring feature of human nature that we seek to work together not only for our material well-being but to both define and pursue a vision of the good in common not only illuminates a psychological feature about persons but places it as a foundation of political justification. Reorienting ourselves to this more natural view of sociability as political justification allows us to understand the much broader goals that may justify state (coercive) organization. Nor does doing so surrender the important Kantian claims that we must interact on conditions of equal moral worth. Rather, this view highlights that there are ways of securing and interacting as moral equals, even in public, that do not depend on solicitous focus on our rights as against each other. That human beings pursue lives in common is the trait that leads Aristotle to describe human beings as “political animals.” It is man’s political nature, our natural need and deep desire to live together in pursuit of a shared common good, that replaces liberal independence as the primary political justification in a republican theory of political obligation.37 Political obligation and in turn legal justification are embedded in the social project of living together as a civic community. Indeed, even the idea that one can determine the shape of rights, beyond perhaps the most basic rights to bodily integrity, outside human communities is rather speculative and abstract.38 A well-formed individual will understand that laws are one of the most significant ways in which we organize our joint pursuits, including the basic pursuit of living peacefully and flourishing as a community; that laws represent the agreed upon boundaries of acceptable behavior.39 Law provides normative guidance not only in the particular actions but in “setting parameters within which individuals carry on together their social interactions, pursue their individual and common projects, and the means for repairing the relationships when things go wrong.”40 Thus, republican political obligation is largely defined as establishing the respect that is owed to each individual as a citizen of their polity as well as the duties each in turn owes to their polity in light of our shared civic project.

The Sovereign and the Republic

113

The Republic and the Common Good We have gone a long way to discover what I trust are fairly intuitive conclusions. In this case, however, I hope the protracted explication to arrive at intuitive conclusions can be viewed as a strength of a republican theory. Seeing that freedom and independence are implausible as a primarily or uniquely justifying political justification, as is asserted by liberal theories, we can start again by building a more organic theory that places our civic nature front and center. Once we see that we are compelled by nature to live together and that being part of a civic society is good for persons, then it is easy to see that participating fully in that society must be part of that good.41 Focusing on our innate sociability provides an argument for more than an attractive or ethical life but places civic interconnectedness at the foundation of political life. Displacing competing versions of autonomy, even those that pay some heed to equality, as the central political justification dissolves persistent contemporary tensions and renders our joint political projects sensible. One important note bears immediate attention before we embark fully on describing republican citizenship. Though I use Aristotle as inspiration, it is clear that important updates are necessary. In particular, Aristotle held ugly views on who could qualify as a citizen, excluding women, barbarians, “natural slaves,” and the vast majority of workers.42 Aristotle’s stated reason is that, for one reason or another, such persons did not share fully in the rational capacities required for citizenship.43 In the case of workers, Aristotle is unclear whether their rational capacity is entirely absent or if the nature of manual labor in ancient Greece, which typically meant nearly unceasing toil and little chance for education, would prevent one from being prepared to participate in the duties of an engaged citizen.44 In the case of women and slaves, the ostensible reason was that both constitutively lacked such capacity in full.45 Even here, Aristotle hesitates. Well aware of the debate even in his day of the evils of slavery, Aristotle admits a distinction between those who are slaves by nature and those who are unjustly enslaved as a consequence of poor circumstances, such as a military defeat.46 Such hedging and a justification that ranks among Aristotle’s worst passages47 in the Politics may reveal his fundamental

114

Ekow N. Yankah

dislike of slavery, a begrudging acceptance of the institution as an economic and political necessity of his time.48 In any case, this is not the place to be an Aristotle apologist, and I hope to take on these faults more fully in future work. To state the obvious: We can set aside Aristotle’s ugly omissions of these groups, as we have no reason to believe that there are groups that systematically lack the rational capacity to participate in citizenship.49 What matters, then, is inspecting the vision Aristotle has for those considered citizens for its attractiveness. It is again worth remembering that it is the core vision of civic obligations and duties as reciprocal rather than an exegesis of the history of political thought that drives this project. Highlighting that we are political animals—that we wish to live together in pursuit of common civic projects—puts the question of political legitimacy in a dramatically different light than modern liberal conversations. It is no longer a question of how to justify the right of a person imposing their coercive will on another; the common political philosophy 101 question, “how are the orders of the state different than those of a gunman,” is shown to be a false analogy. Liberal theories ask how state power can be justified given that persons cannot avoid crashing into each other. On such a view, the state is reduced to just another someone, albeit a powerful someone. A republican justification takes a dramatically different view. On a republican view, state power is no longer viewed as a necessary evil, an imposition on liberty that can only be justified by the greater liberty it secures in exchange, as proposed by Hobbes and Locke. Rather, we discover that sovereignty is fundamentally a question of how we are to govern ourselves, together. Both because we must and we wish to live together in various common pursuits, freedom must be found not in justifying state power as though each individual had a right as against all the world to be left alone. Rather, freedom is justified within a conception that takes into account that we live together in an ongoing common pursuit. To use Pettit’s words in earlier work, republican freedom is the freedom of the city, not the freedom of the heath.50 Ensuring one’s freedom if one is committed to living with others means being able to live openly among others without fear; that the laws protect the basic goods of bodily integrity and other minimal goods without which living together is impossible.51 But

The Sovereign and the Republic

115

just as important, seeing that our political bonds are social in nature highlights that one can only experience oneself as free to the extent one is assured that he is living on equal terms with every other person, accorded the same respect by one’s community as every other person.52 One’s sense of freedom is ultimately social and turns on knowing you have a certain standing relative to others. To not be treated as an equal is to know that one is not fully sharing in the same civic project. Thus, equality before the law is essential to the justification of republican political obligation.53 But even so said would understate the matter; unlike many liberal theories that can only view law as an ever-present danger that needs to be justified, a republican theory understands that as much as equality before the law is essential to legal justification, law can reciprocally be constitutive of equality.54 Franchise and the Vote The combination of these two features, living freely and without fear among your fellows and the widespread knowledge that the law both treats and establishes equality in civic matters, leads naturally to a third feature. If our rational nature is expressed in our coming together to live as a community as equals, then part of that rational nature must be in having a voice in the affairs, structuring, and governance of one’s community.55 This idea comes naturally enough when we consider why having the right to vote is important for reasons far beyond its instrumental benefits. The right to vote is an important constituent of political membership because it is a way of conveying equal respect for the views and goals of each.56 Today, the term franchise is used solely to invoke the right to vote, but the republican view I am suggesting harkens back to the original and much richer meaning of the word franchise.57 Democratic voting procedures are one, but only one, way of reflecting a shared voice that is critical to a just polity. Franchise is the sense that one is a complete and equal member of the polity with all its attendant rights and responsibilities, to use the modern idiom.58 For Aristotle, a critical part of participating in one’s community was sharing in governance, particularly taking a turn as a public official and participating in public deliberation.59

116

Ekow N. Yankah

Understanding franchise as the basis of a deep claim about the nature of citizenship contextualizes voting as an intrinsic part of one’s equal standing as a citizen rather than justified by some guarantee of epistemic superiority or bald procedural justice. As Melissa Schwartzberg has argued, voting is just one component in the deeper claim that citizenship is fundamentally defined by equal capacity for deliberation in core civic matters.60 Indeed, focusing solely on voting falls into the modern habit of imagining citizens as merely choosing who should rule over them rather than as active citizens engaged in self-rule. Voting is best seen as of an inextricable piece with the attendant opportunities and responsibilities of democratic deliberation.61 This obviously includes the ability and opportunity to hold political office. Likewise, the permission and obligation to serve as a juror reflect the position of citizens as equals insofar as civic deliberation is concerned. Franchise does not turn on a naïve view that all in fact equally possess intellectual or ethical virtues. Instead it reflects the idea that when it comes to certain civic capacities, the normal citizen, be she a day laborer, nurse, or rocket scientist, has an equal voice as a citizen.62 As a practical matter, the range of political judgments necessary are within the reach of ordinary citizens. To be sure, a well-functioning polity is obligated both to educate its members adequately to execute their duties as citizens and to form good laws and institutions that will frame decisions such that they are within the grasp of ordinary judgment.63 But just as important as the idea that citizens are equally capable of deliberation under basically just and achievable conditions is the idea that citizenship, in turn, helps constitute a kind of natural equality. This republican view brings a distinctly different tenor to the question of governance than the contemporary positions on legitimation of political power. From the republican view, a state must work to promote franchise, i.e., the active participation of all in sharing the benefits and burdens of governance.64 Of course, in the vast majority of unglamorous offices, these benefits and burdens will be nearly identical; serving in them will be a way to make one’s voice in your polity heard as well as contribute to your share of its upkeep. Even today, in a world of professionalized public administration and a population too large to make each taking a turn in public office practicable, we recognize the importance of

The Sovereign and the Republic

117

having access to run for office as an important constituent of both our social basis for self-respect and our place in our political community.65 Simultaneously, we recognize that holding many “privileged” offices—head of a neighborhood association or ParentTeacher Associations—are a part of one’s duty in maintaining the well-being of one’s (political) community. Thus, the republican view of governance is not one that first asks of the justification for the state ruling over the populous but a view of “the free ruling over the free.” The conditions necessary to promote a republican view of political participation are no less than the conditions of citizenship in a polity. It is under these conditions that one can both express and fulfill our drive to pursue a life in common while being assured that our voice will be recognized in our civic community. The republican view of political obligation, however, is not merely procedural. While ensuring all have a voice in the polity will go a long way toward promoting just laws, a republican polity does not blithely assume that any particular procedure, democratic or otherwise, can by itself guarantee a just legal system. The conditions and procedures of civic governance are constitutive of republican citizenship because they are critical in fostering civic virtue.66 Still, they do not in and of themselves secure political justification. A polity justified by its pursuit of the common good is one that is sensitive to the shortcomings of relying solely on procedural mechanisms such as popular voting. As Waldron highlights, citizens often continue to feel themselves bound together through contested political issues because the rule of law assures they may continue to contest certain decisions or that their preferences will carry the day on other equally important decisions.67 It is equally important to ensure that the common good is not exhausted by a commitment to voting or any other procedure alone. Obviously, the problem of persistent minorities pressing recognizably legitimate claims that lose on referendum must be addressed if claims of political obligation are to be justifiable, particularly to that minority. A political system premised on the republican value of franchise—an equal voice in pursuit of the civic common good— requires addressing the interests of persistent minorities to illustrate their concerns are truly understood to be part of the

118

Ekow N. Yankah

common good. Such instances broaden fundamental guarantees of equal voice in order to strengthen the bonds between dominant majorities and persistent minorities. These are not cases of giving fundamental voting rights to a racial minority; that “baseline” case is one of guaranteeing a political voice in the first instance. One should also notice more subtle cases that are nevertheless important markers of civic concern; the local pool installing special hours to accommodate the special privacy requests of an ethnic population or the closing off of certain streets at night to accommodate religious observations. That these examples are quotidian is precisely the point; they are the everyday ways we recognize that ensuring all see themselves reflected in the pursuit of the common good is not merely defined by voting numbers.68 Franchise and the Common Good We noted that the willingness to share in the burdens of communal governing is a critical feature of civic virtue. Aristotle described the unjust person as grasping, motivated by wanting more than his fair share of benefits and less than his fair share of burdens.69 Thus, when the polity is successful in cultivating civic virtue, citizens will internalize the need to care and contribute to the public and common good ahead of the interests of the individual or his own small interest group.70 The well-constructed citizen realizes that kin, tribe, race, political party, and lobby group are owed an important share of political concern, but no more than a share. Thus, the polity, its laws and political institutions, are justified to the extent it promotes the good of the entire political community ahead of the interests of any particular person or interest group.71 “[T]rue forms of government [ ] are those in which the one, or the few, or the many, govern with a view to the common interest; but governments which rule with a view to the private interest, whether of the one, or of the few, or of the many are perversions.”72 This picture does not rest on some naïve or utopian view of citizens who lack recognizably human ambitions or preferences for their own projects. It is all too easy to dismiss such republican ambitions as belonging to a romanticized past, whether it be a Rockwellian America or the near mythic time of the American founding; some time, it is imagined, when people of political

The Sovereign and the Republic

119

virtue were in abundance. Likewise, the danger in invoking Athenian political thought is its easy dismissal as applicable only in an ancient and unrecognizable world. It is too easy to idealize Athens, a small, homogeneous city-state where familiarity and consensus about the common good made possible Aristotle’s ideal of ruling and being ruled in turn. This view leaves few lessons for our contemporary diverse and cacophonous political world, with its sharp disagreement about the common good. But such easy dismissal depends on both unsupportable romanticization of the past and unnecessary pessimism about our political future. I do not hold any idealized view that past Americans were by natural constitution selfless and more reliably orientated toward the shared common good. The long and continuing fights, from the beginning of our nation’s history, through an almost unimaginably bloody civil war, generations of racial oppression, the studied subjugation of women, and the current fight for equal rights by gay Americans and others should quickly disabuse us of such fantasy. Nor should we imagine that either the people or political challenges testing the civic commitments of ancient Athens were unrecognizable to our modern mind. Ancient Athens was a large, thriving and, most important, diverse political culture, hosting persons unknown to each other from disparate homes and backgrounds, speaking different languages, advocating and contesting the vision of the common good.73 No less than today, there were political battles over who truly counted as part of the polity, who held genuinely Athenian values, and who truly cared for the common good.74 The point, for our purposes, is that the success of this republican theory does not rely on an unrealistic view of selfless citizens who ignore their every need to pursue the good of their fellow citizens.75 As historian Josiah Ober puts it: The “standard Greek polis” is often taken, by liberals and communitarians alike, as a very thick community, deeply grounded in tradition and situated commitments . . . Athens [was] actually considerably less tradition-bound than is often supposed [ ]. Athens allowed, I will argue, many (though not all) of its people to authorize their own histories, define for themselves a variety of goods,

120

Ekow N. Yankah

and it provided many of them with resources adequate to pursuing a variety of diverse ends. And so Athenian political coherence remained thin enough to be consistent with the demands of political theorizing that must engage with the real condition of multicultural modernity.76

Indeed, this republican view turns on the explicit recognition that people are status-oriented animals and thus make critical space in securing their status as civic equals. No polity could remain healthy if the civic duties it imposes are so suffocating they leave no space for the individual projects of its citizens, including the ability to prefer those near and dear.77 Thus, the Aristotelian vision rejects Plato’s arguments that the personal projects of individual citizens ought to be subsumed by that of the polity.78 Nothing in a republican view imagines citizens foreclosed from pressing for or advocating for their own interests, particularly pressing or insufficiently noticed ones. After all, the ability to favor those dear to you, to pursue personal projects, and to give time and energy to those projects most meaningful to each individual are also parts of civic flourishing. It does seem perfectly natural, however, that the well-composed citizen will frame her projects in a way that is at least compatible with and more often supports some portion of the good of her civic community.79 Aristotle’s famous example invokes sailors on a ship. Though each has his own project and duties, they cumulatively contribute to the well-being of the ship. Because our concepts of justice are themselves deeply influenced by the standards of our political community, a citizen often will not even be able to formulate visions of success or claims of justice without incorporating a shared concept of the common good.80 Whether one counts her career as a true success, for most, turns on her civic community’s judgment on whether she was a pharmaceutical titan or a black market drug dealer. The information I feel you must fairly disclose to me upon the sale of a house and thus the complaints I can level at you, for example, will be partly constituted by where our society has drawn the line between required disclosure and buyer beware.81 The intuitive analogy is of a group of roommates. In a basically functioning home, even the way one makes demands seeks to be

The Sovereign and the Republic

121

plausibly compatible with the common good of the house. The announcement that you do not want to do the dishes when it is your turn because you do not like doing dishes will be met with confusion and puzzlement; ultimately it will be rejected as an argument at all. Just as individual citizens will frame their projects to be compatible with the common good, so too must laws, taken individually, and political and legal institutions taken collectively. This is particularly clear when civic agents, e.g., office holders, seek personal gain for themselves, their friends and family, or special-interest groups. Just as individuals are unjustified in twisting civic power for their own benefit, Aristotle saw that factions doing so were equally unjustified. This danger could exist in virtue of unchecked populism, where the poor seek only to siphon money from the rich. More likely, of course, is a world in which the rich and powerful use the law to exploit the poor. In either case, our civic bonds are ruptured when law no longer aims at the common good.82 In justifying law by pursuit of the common good, the republican view of political justification frees itself from much contemporary political thought where law is only justified by some catalog of rights or hypothetically consented-to liberties. As we shall see, the republican view presents a distinct picture from modern liberal views, whether based on Kantian freedom, Nozickian libertarianism, or Rawlsian liberalism. But the republican view expressed here stands in sharpest contrast with what passes as the modern enlightened view that all politics is interest-group warfare and rent seeking. It remains equally dangerous when we collectively acquiesce to “enlightened politics,” harnessing crass interest-group politics to counteract the same, even with the best of intentions.83 In dire situations, this postmodern Machiavellian view of balancing interest groups may well be the best that can be done to secure the common good, but it is a shadow of institutions (and a public ethos) that are constructed, in the main, to focus on the common good.84 It is not surprising that even the most disingenuous attempts to serve special interests are so often veneered with arguments of how they will serve the common good. No matter how we may rationalize the politics as power-play view, those who must sell us these ideas know that something inside us deeply understands that civic institutions are only legitimate when they seek the common good.

122

Ekow N. Yankah Republican Sovereignty and Citizenship

Our examples so far have straddled the personal and political, but each has suggested how the shape of the justifications we rely upon in communal activities extends into our political lives as well. Since an Aristotelian republicanism is premised on our political nature, described as our drive to live in common as active participants in our civic governance, and is justified by political and legal institutions that aim at the common good, its justification of political sovereignty is dramatically different from much contemporary treatment. Fundamentally, republicans are committed to building a civic polity in which the most natural view of the state is not of something outside of ourselves, requiring justification that is analogous to a gunman. The change of tone can feel so dramatic that it is as though the question of the political sovereign has disappeared. The feeling that a political theory assumes away one of the central questions of contemporary political theory raises the suspicion that the theory is fundamentally flawed. This reaction is a symptom of well-worn liberal paths of thinking rather than any defect in republicanism. I hope to illustrate that the way a republican theory resolves the related questions of what obliges a citizen to obey the law and what permits others to impose legal obligations on the unwilling presents an attractive picture of sovereignty. Indeed, we shall see that in examining sovereignty, republicanism captures much of the intuitive in liberalism while discarding many of the least plausible features. To go further, we eventually notice that liberal theories borrow much of their plausibility by relying on deeply republican intuitions. Law-Abidingness A republican theory makes political and legal obligation internal to the theory itself. Because law is one of the focal ways by which we coordinate our pursuit of the common good and resolve conflicts arising from our individual projects, even when law is imperfect, a citizen with civic virtue will adopt a supportive and deferential attitude toward much of law.85 The rule of law, when oriented toward the common good, will secure great communal goods. Those who cast aside the laws of their community, in some sense, reveal an

The Sovereign and the Republic

123

arrogant or oligarchic character.86 An attitude of law-abidingness will be embedded in the same civic virtue that promotes the common good and each citizen’s part within it.87 Law-abidingness is something broader than the well-explored question in jurisprudence of whether there is a prima facie obligation to obey the law. Law-abidingness describes a dispositional civic virtue, a willingness to generally follow roughly just laws even when they seem to clash with one’s most immediate interests. It is captured in Rawls’s sense of reasonableness and described in Aristotle’s virtue of justice. Law-abidingness does not mean that individuals believe they are obligated to follow every law, nor is it defeated by a singular instance of law breaking. At the same time, law-abidingness should not be understood as a slavish or passive acceptance of any legal standard.88 As we noted, Aristotle is insistent that in a functioning policy, each must rule and be ruled in turn. To this he adds that each must be able to rule and be ruled well.89 A citizen must be sensitive to the law’s function as supporting a civic project that promotes the flourishing of its members.90 Thus, when the laws become directly hostile to human flourishing, as in Nazi Germany or the antebellum South, they will be recognized by the well-constructed citizen as impossible to obey or internalize.91 So understood, because law-abidingness is part and parcel of supporting your civic community, it includes the impulse not only to support your community’s laws but to improve them, particularly in ways that steer them away from instrumental factionalism and aim them toward the common good.92 Indeed, improving the laws of one’s community while displaying an attitude of respect toward the project of law-making—think of civil rights marchers in the American South, simultaneously breaking unjust laws while showing respect for law itself—is often a way of exhibiting the greatest civic virtue.93 At its core, however, republicanism may be said to take a more optimistic view of a citizen’s obligation to follow the law than what is most often communicated by liberal anxiety. In the normal case, a citizen does not take her individual good to be in conflict with the good for the many.94 A republican theory is embedded in the idea that the good for the many is typically part of the good for the individual. Thus, one’s sense of justice will be part of one’s

124

Ekow N. Yankah

reasoning, guiding consideration about how one’s plans fit into or at least should not come into conflict with the good for one’s civic community.95 Political and Legal Obligations Of course, a natural sense of law-abidingness will not settle the question of sovereignty, for that would assume the question away. While the question of sovereignty is related to the obligation to obey the law, it also fundamentally focuses on the question of the legitimate right to rule.96 The republican conception of the power to impose legal and political obligations, as well as to coerce or punish those who resist or break legal norms, follows from the preceding discussion. On a republican view, criminal punishment, for example, is not solely the state’s enforcement of pre-social rights or justified by retributivist imperatives. Rather, legal and political power is embedded in the very social project of our living together as a civic community. Rights and society, it may be said, go together, and our rights against one another cannot be abstracted from the civil society we share. Even the definition of the crimes typically depends on social decisions about the boundaries of rights and obligations.97 Whereas many liberals see sovereignty as stemming, first and foremost, from the ability to enforce preexisting legal rights, the republican view understands the duties of law as reciprocally held between citizens and community. This does not mean the republican denies that persons also possess certain inalienable rights; there are things one cannot morally do to anyone regardless of whether one shares a civic community. Nor are crimes outlawed only because they hurt the metaphorical public good. As Duff has persuasively argued, the visceral and most important wrong we understand when contemplating many crimes is the harm and the violence done to the victim.98 In the case of the most obvious laws, both concern for individual harms and our concern for a shared civic project will produce the same requirements. In any community that claims to be dedicated to the project of flourishing as a community, the most basic human interests—protection from murder, rape, or assault, kidnapping, and some level of protection of property—will require protection, for human beings cannot otherwise plausibly flourish.99

The Sovereign and the Republic

125

Outside of core protections, however, we begin to see more clearly the persuasiveness of the republican justification for political and legal power. Regardless of where one falls, from the largely lonesome frontiersman to the urban bon vivant, we pursue life in common; none can survive without the ability to coordinate both actions and expectations with others to define the parameters of acceptable behavior.100 Freedom certainly belongs to all, but whether I may enter this building, whether this area is a public park or private property, whether five parts per million of this heavy metal is your poisoning my community or an acceptable industrial by-product, these and countless other acts cannot be determined without reference to socially determined expectations.101 This is equally true in many other fields, from tort liability to repair the inevitable accidental injury to contract law, which allows us to organize our lives and commercial pursuits in increasingly complicated ways. While criminal law doctrines rightfully attract our attention, it may be more commonplace legal arenas that highlight the core workings of a republican sovereign. To make more politically salient our earlier example of shared household duties, consider zoning laws. Zoning laws do not plausibly instantiate intrinsic property rights of particular persons or preserve autonomy. Rather, zoning laws are a perfectly ordinary instance of using laws to coordinate and enforce disparate individual actions to preserve common goods. Recognizing the ways in which safety can be enhanced or disease avoided leads to numerous fire and safety codes. While some efforts will concern reconciling noncompatible claims of rights, as much or more will concern making competing uses not just compatible but harmonious or enjoyable. Zoning laws pertain to not only potentially offensive industrial use but to establishing recreational parks and historic districts preserving past civic characteristics. Further, because at a certain size and level of complexity, living together becomes unmanageable if not collectively organized, local authorities are empowered to authoritatively shape the city, separating industry from residential and so on. Of course, in many instances, there will be no definitive best choice but that an authoritative choice is made will itself be a great boon.

126

Ekow N. Yankah

Voice and the Common Civic Good That there is rarely a unique or dominant choice is another reason zoning regulations nicely capture our republican intuitions. Freeing ourselves from the claim that laws can only be justified by the protection of rights recognizes that there is a remarkably reasonable range in our shared conception of a flourishing polity. Should an area be solely residential, industrial, or mixed use? Does having a neighborhood become a beacon of nightlife add to a thriving town or is this neighborhood oversaturated with bars? These and countless similar decisions shape the lives available to all citizens and will ultimately constitute the civic good in which we share. Precisely because there are countless decisions, small and large, required in any zoning plan, there will be great contestation over many decisions. (Anyone who has attended a community board meeting on granting a bar license or land variance instantly recognizes that even in perfectly good faith there will be heated disagreement over how a neighborhood should develop.) That the choices will be contested does not (necessarily) undermine the civic goods captured; under conditions designed to resist corruption and capture, contestation is precisely the instantiation of a shared voice in the common good. Of course, it takes little searching to notice that zoning laws, like any legal arena, are too often subject to manipulation by the powerful for their own interest. Such abuse is so common that the phrase NIMBY—Not In My Back Yard—is a commonly known description of residents avoiding their share of unpleasant but necessary infrastructure burdens. The point, however, is that the potential for abuse cannot be avoided by retreating to a view of law as solely rights-regarding but only by building political institutions and a public ethos that reinforces the shared benefits and burdens that constitute the common good. To the extent that some parts of law, whether it be zoning laws, tax laws, and particularly criminal law, lose touch with their civic justification, a republican theory warns to search for factional abuse and provides an argument for modesty and legal curtailment.102 Franchise prompts a polity to structure its institutions such that citizens naturally support or behave compatibly with the shared civic good rather than antiseptically focus law merely on individual rights or allow law to imperialistically overrun its

The Sovereign and the Republic

127

bounds and serve their merely personal values. Indeed, that legal and political power is only justified where necessary to permit and improve our civic functioning represents an important limiting feature that prevents the state from running wild to coerce any particular image of personal virtue.103 Justifying political obligation by premising it on our need to come together to pursue a common good may strike many as banal. First, the common good may be insufficiently fine-grained or simply too empty a concept to do any work. History is full of political actors arguing that their abuse of power is in the service of the common good; Aristotle’s thinking on natural slaves is an immediately relevant example.104 Focusing only on laws that result from abusive interest groups leads us to give up on the idea that the common good can do work. Remember that the republican vision of political obligation focuses on the reciprocal duties between citizens and their state and makes central that all should be active participants, to the extent possible, in the governance of their community. A world in which there are increasing avenues of citizen involvement, from community boards that have actual power in legislating in their neighborhood; police-citizen boards that are structured to and empowered to effect change; changes in campaign finance including public matching funds that open the political process and limits on donations from any one source; shared military service; and perhaps even a legal requirement to vote, gives real teeth to the idea that the common good will not merely be a veneer from behind which the powerful oppress.105 It is also true, as Aristotle realized, that building a civic community wherein citizens were prepared to contribute and sacrifice for the common good required real investments in the education and support of citizens from an early age; the state’s job would be in a real way to invest in creating citizens.106 These conditions require public investment and a change in the current view of interest-group-only politics, but such changes are not unimaginable. In a state that promoted citizenship, having a voice in governance to pursue a common good would not be an empty promise. The natural complement to having a voice would be the ability of citizens to contest public policy and legal decisions.107 The republican state would both encourage and be legitimized by the ability of citizens to contest decisions in a wide variety of ways. An example of this lies in our not too distant

128

Ekow N. Yankah

past in the town hall meetings advocated by champions of participatory democracy and today in the citizen review boards of many major cities that must often approve civic matters ranging from bar licenses to disruptive commercial endeavors (such as filming a movie). To paraphrase Philip Pettit in describing a different but here sympathetic republican theory, republicanism is not legitimated in light of hypothetical consent, it is legitimated in the presence of real contestation.108 Such contestation, while having real cost in the ability to govern cheaply and quickly, has immeasurable benefits in ensuring that the pursuit of the common good is real. Legal Obligation I have argued that a certain sense of law-abidingness is natural to a basically well-ordered polity; it is so deeply interwoven, we often fail to notice that the overwhelming majority not only plan their actions to be compatible with the law but internalize legal standards as part of how actions and life plans are valued. Nonetheless, a republican theory owes an explanation of the reach of legal authority for those who breach justified laws. I have already indicated the obvious starting point; where certain acts make living together as civic equals impossible—most apparent in malum in se crimes—criminal law appropriately prohibits and punishes. Further, where state authority is legitimated by all having shared a voice in its creation, the conditions of proper legal authority, including rights of enforcement, are met. Describing these features as constituting legitimate authority implies the power of the state to coerce recalcitrant citizens into compliance. Returning to the unglamorous example of zoning laws highlights how distinct from purely rights-loving theories a full-throated republican theory is and yet illustrates that such enforcement often feels so unremarkable as to go unnoticed. In a well-ordered polity, zoning laws are a way of both managing the competing demands of development and shaping a vision of shared public space for the common good. Undoubtedly, competing visions and interests will result in such laws being highly contested. Institutions and procedures will have to be constructed to avoid capture by special interests and ensure all have a voice

The Sovereign and the Republic

129

in shaping such zoning, including paying special attention to the claims of persistent minorities. It is of course essential to note that punishment should naturally be a last resort; many projects can be accomplished by positive intervention by the state. Tax benefits or direct state projects, for example, may shape the character of a neighborhood. Further, there will be sections of The City where regulation is not considered necessary, where the unruliness of the market is tolerated or even welcomed. Still, after the above conditions are met, justified enforcement flows naturally. Once a section of a city has prohibited industrial use or, more subtly, multifamily dwellings, violators will be rightfully sanctioned.109 Without the ability to sanction the recalcitrant, the very project of shaping the polity for the common good would fail. Once all have had a say in a matter where a decision is needed to secure a civic good, it will be natural to secure that good with enforcement. Notice, nowhere in the above discussion have I mentioned inherent property rights or justified state authority by reference to individual autonomy or freedom. The breadth of our zoning examples illustrated why “rights” are insufficient to capture the various civic goods we pursue in common. This republican view justifies legal enforcement by the power needed to pursue civic goods rather than any commitment to theories of property rights often endorsed by competing liberal theories. This is not to assert that many, perhaps most, theories of property cannot accommodate these perfectly ordinary views. Rather, it is to point out that they do so by curtailing any stringent view on rights with republican limits, a point to which we will immediately turn.110 The Limits of Authority It should be noted that while franchise as a political value establishes a different justification for enforceable rights than theories premised on innate rights, autonomy, or freedom, this does not mean that state coercive power is imagined to be limitless or free to pursue any chosen good. Recall the basis of state power is tied to the pursuit of the civic common good. Zoning is a telling example precisely because it is focused in the first instance on the necessary shaping of civic, rather than purely personal, goods.

130

Ekow N. Yankah

Even in ancient considerations, without our contemporary views of liberal rights, republican justifications linked the appropriate span of government power as limited by appropriate civic projects. In contemplating the way political offices and political power in particular should be distributed, Aristotle argues that state goods and benefits can only be aimed at promoting those elements that enter into “the composition of the state”; in other words, projects that constitute the ways in which we live together in a civic polity.111 Thus, political benefits could not be distributed on the basis of excellences that were outside the civic good.112 More important, civic power was abused if it was aimed at imposing individual virtues outside of civic necessity. This limitation distinguishes republican justifications from purely communitarian justifications, as I understand them, and is directly relevant to some of the most troubling contemporary uses of coercive law to impose punishment based on a majority’s view of what is personally good or virtuous for an individual.113 Liberalism and Republicanism Not only is a republican vision of governing plausible but, as I have argued elsewhere, I suspect much of what makes liberal theories attractive is actually hidden in republican intuitions.114 Take Kantian forms of liberalism, surely one of the dominant modern strains. Kant premised legal power on the need to guard one’s innate right to freedom from imposition of any one person’s unilateral will onto another.115 In order to do so, Kant points to the need to ensure that all coercive laws are the product of an “omnilateral will.”116 But the mysterious “omnilateral will” turns out to be exactly the submitting of individual decisions about justice to collective political decisions that are taken in the name of all. Indeed, as long as an act is taken for public rather than private purposes, we are told that the act is the product of the omnilateral will and thus legitimately coerced. That is to say, despite the singular focus on innate freedom with which we begin, Kant’s liberalism turns out to be a disguised and much less attractive form of republicanism. It is disguised because it pretends to be justified only by its protection of individual freedom without making explicit that the philosophical alchemy that turns the unilateral will into the

The Sovereign and the Republic

131

omnilateral will is thoroughly republican in nature. It is much less attractive because it is solely justified by the protection of individual freedom. Lacking a fuller republican picture, it must twist itself into knots to justify anything like a state that the vast majority of us find attractive. Natural parts of the common good, on Kant’s conception, must be excluded when not justified in the name of freedom alone. I have focused on zoning because it is an unnoticed yet ubiquitous way in which we marshal the law to serve the common good, but there are countless examples of everyday legal and political power aimed at our civic flourishing rather than the protection of individual rights or external freedom. It remains unclear how those who support a thin Kantian view of the state can legitimate a state that supports the arts, builds public parks, pursues remedial racial or ethnic policies, funds purely scientific endeavors or a national health-care system, and so on in the name of freedom alone.117 Last, lacking the republican highlighting of active civic participation, Kant holds the view that once the sovereign has ruled, the citizen has no right to disobey any law or, outside of petition, modify the legal system.118 Modern variations of liberal theories likewise fall short, excluding important civic goals many find naturally encompassed in our civic project. A brief example, to be further explicated in later work, may highlight why liberal theories are implausible, or at least incomplete, without republican reinforcements. On the one hand, recall that Kantian liberalism authorizes all laws passed by a legislature (the omnilateral will) just insofar as they are justified by protecting freedom.119 Thus, a Kantian legislature, it would seem, is permitted to pursue public health initiatives insofar as public health risks threaten freedom; building sewers and preventing epidemics and the like would be permitted, and one might generously include treatment of serious pain and debilitating conditions.120 Further efforts, say, a concerted government scheme to lower prices and assure better health outcomes to more citizens at lower cost, however, are justified not by defending freedom but rather by aiming at the common good. Even the rich resources of Rawlsian liberalism make no room for civic goals outside of the (admittedly important) impulse to benefit the poorest among us. Rawlsian liberals certainly have

132

Ekow N. Yankah

richer resources to meet these challenges. Though great detail will have to await further explication, Rawls is committed to principles of justice that first secure basic primary goods (including basic bodily integrity). One might think that national health-care schemes are captured by or at least touched upon by this principle. Moreover, Rawls’s second principle of justice commits one to political institutions that serve the interests of the least advantaged among us before all others. This concern with distributive justice certainly mandates political institutions to control costs and ensure better health outcomes. Yet it is not difficult to imagine serious medical and social goods that are not solely captured by our concerns over distributive justice. There are times and places where new but dangerously unhealthy habits are adopted by those not among the poorest. A growing middle class in a developing nation may begin mimicking Western fast food habits or recreational drug use. Perhaps it becomes fashionable and status conferring for those who can afford it to use baby formula despite the best evidence suggesting breast-feeding. American examples surrounding healthy eating, smoking, and the like come immediately to mind. The current debates about the wisdom of government interjection over health care and “nanny states’” regulation of health incentives (consider reduced health-care costs for gym memberships and limiting the size of movie theater sodas) may be a fair public debate. But that the dominant liberal theories of our time mark such efforts as unjustifiable or prohibited sets a high bar to clear if such liberal theories are to be convincing. The point, of course, is that many, perhaps most, of the important tasks to which we legitimately set government cannot be justified solely as protecting freedom, preserving primary goods, or securing distributive justice. The government may seek to reinforce important issues of public health simply because they are in the interest of a flourishing civic polity. This worry is not one of high-minded academic debate. In the past few years the United States has struggled mightily to join the rest of the industrialized world in expanding government guarantees of health insurance. This vast public debate in America has been misshapen to a remarkable extent because the American public has so widely and deeply internalized the language of rights that the entire debate often ran aground on the question, “what gives the government

The Sovereign and the Republic

133

the right to force me to pay for someone else’s illnesses.” Even setting aside obvious prudential arguments about the uncertainty of who gets sick when, it is striking that for many, the only acceptable replies had to take the form of justifying conflicting rights. The obvious justification that some measure of security against illness is fundamental to a flourishing society was almost completely stifled as a reply. Indeed, current liberal theories struggle to justify something as simple as posters that encourage replacing soda with water. This is not even to mention the intuition that it is nontrivial that we collectively support, at least modestly, excellence in arts, natural beauty, and architecture among other cultural positives, simply because they represent civic excellences. That so much of what we take to be critical to a plausible, attractive, and justified state cannot be reconciled with liberal freedom or autonomy shows both that republicanism represents a distinct view of legitimate governance and that liberalism is often made plausible only by internalizing republican intuitions. Legal Conclusions I have tried here to build a compound picture of legitimate sovereignty; political obligation as law-abidingness and political authority as civic duty. Combined, the republican picture I presented may be thought of as “franchise.”121 Republican franchise, I have argued, stems from our most basic nature as social and political beings that both need and desire to live in common. It is an attempt to step back from Rawlsian and Kantian liberalism that has to either imagine that all political institutions come from the imagined consent of asocial persons or that the only justified state is one that extends as far as necessary to protect individual freedom. Republicanism is distinct in casting off some of liberalism’s less plausible restrictions. I have already argued that it is only by focusing on the common good rather than Kantian freedom that one can justify funding some form of national health care to alleviate ever increasing costs of medical care for the general public. Likewise, funding for public arts and public parks, in correct and modest proportion, are natural parts of the common good that cannot be easily described as freedom enhancing as required by Kantian liberals. Take two further examples. First, the absolute freedom at the

134

Ekow N. Yankah

heart of Kantian liberalism now famously justifies the exceptional position in Anglo-American law that one need not rescue another citizen from grave, even fatal harm, even where that rescue can be done with no risk or little cost to oneself.122 This unattractive position is naturally resolved under a republican view of citizenship that prizes the common good and would find it easy to balance the costs to citizens in cases of great danger and costless rescue.123 Likewise with the Rawlsian difference principle. While the spirit of Rawls’s difference principle is certainly compatible with the shared concern of republican citizenship, its lexical nature is unnecessarily inflexible for the kind of republicanism seen here. That one could never allow a material benefit to any portion of society if it did not advance the worst off in society would prohibit massive gains to many or even most if wealth could not be reliably transferred to the poorest. A theory that focuses on the common good allows one to prioritize wealth creation that alleviates the burdens of the poorest, pay due regard to inequality yet not flatly prohibit contributions to a flourishing society even where it may be only partially spread. Notice also that a republican justification makes sense of why a state may be justified in pursuing ameliorative polices of racial justice. Imagine a city has decided to rectify racial underrepresentation among its firefighters.124 In this city, though a past history of discrimination had rendered the fire department disproportionately hostile to racial minorities, imagine, counter to the most similar actual cases, that the current practices evidenced no contemporary discrimination. Nonetheless, the effects of past generations’ discrimination have meant few minorities joined the fire department. Perhaps citizens are vastly more likely to apply to the fire department if a parent was a firefighter. In any case, notice that a municipality may have reasons to expend energy and resources to diversify its fire department despite the fact that the reasons for the disproportionate makeup of the fire department was not the result of anyone within two generations having had their freedom curtailed. A city may believe that certain important positions of civic trust, care, or enforcement, from fire departments, police officers, and even municipal workers generally, should if possible symbolize civic equality quite apart from important questions of preserving individual freedom.

The Sovereign and the Republic

135

If franchise illustrates the legitimate reach of state power beyond that based solely on rights violations, it also illuminates important conclusions about the limits of the state’s justified claims of legal authority. Earlier we saw Aristotle taking part in public deliberations about civic matters to be constitutive of citizenship. Interestingly, his reasons were not instrumental—at least not as ordinarily understood. The motivating claim did not depend on deliberations by the many being, for example, epistemically superior. Rather, his suggestion is that not giving groups of people a voice in civic matters inevitably leads to resentment and ultimately instability of the state.125 But as Schwartzberg points out, Aristotle does not reason purely as a matter of numbers; the idea is not merely that giving people a civic voice is placating.126 Instead, having a civic voice is a distinctive way in which a polity ensures each citizen can achieve his potential as a political animal. This reading indicates an importantly non-instrumental value in securing civic equality, but Aristotle’s arguments highlight an even darker facet that refocuses us on the questions of legitimate political authority. The more worrying claim is not merely that robbing a person or a group of their civic voice leads to the perception of unfairness. Were that the case, the only question would be one of containment—the necessary amount of force or placation to quell unrest. Rather, Aristotle claims that a polity where some have no voice is “full of enemies.”127 This is not simply a descriptive claim. It is a normative claim that must be read in contrast to his understanding of civic bonds as a form of friendship. Aristotle understands that those denied their civic voice or measure of equality rightfully understand themselves to not be full citizens.128 Put plainly, where a polity does not ensure its citizens (or groups of citizens) their civic equality, those citizens are to that extent no longer obligated to view the state as making legitimate claims of authority. Polities that deny citizens or groups civic equality declare bit by bit that “those people” are not truly part of the polity; that they are being ruled over rather than engaged in a common project. Members of such oppressed communities may insist that their civic status be recognized premised on underlying civic contributions or historical bonds.129 But equally, they may reject the state’s claim that they are obligated by its laws and recognize their relationship to the state is “catch as catch can.”

136

Ekow N. Yankah

This claim may be sound in political theory but is anything but abstract. The belief held by many communities of color in the United States that police represent little more than an oppressive occupation force reflects this intuition without philosophical varnish. Nor is such disaffection limited to complaints about policing; it is the product of understanding that the law, in countless large and small ways, has been historically bent to the disadvantage and exclusion of racial minorities. Highlighting that those without a civic voice rightfully resent the state is to see how people can recognize themselves as politically alienated, understand themselves to be metaphorically at war with their own state and reject, to greater or lesser degree, the state’s claims of legitimate obligation. Reading republican justification this way makes clear what is risked by a polity that fails to extend a rich franchise to all. Still, republican citizenship should not be understood as a hodgepodge of popular left-leaning politics. Indeed, one of the features of republicanism most relevant to questions of sovereignty has unsettling, yet I believe plausible, implications. Because republican citizenship is premised on our civic project, it prizes civic priority. Put another way, it will often be the case that threats that are aimed at the civic polity or betrayals of the civic trust justify responses that are greater than “ordinary crime.” In the ordinary case, this is easy enough to understand. In ancient Athens, civic authorities had to undergo a public hearing before leaving office to ensure they had not abused their office for personal gain.130 This ancient ritual echoes our sensitivities today when we hear of public corruption by a person or administration. More troubling, given our current struggles, a polity’s right to defend itself has ominous implications for combating terrorism. Terrorism, whether by foreign enemies or home-grown and selfdeclared enemies of our own government, is distinct precisely because it takes as its aim not personal gain but the weakening or destruction of our ability to govern ourselves and continue our civic project. While I, like many Americans, condemn many of the steps taken in the “war against terror,” from acts that must count as torture to the extradition of persons to strip them of American laws, I have also struggled with my belief that the same laws that govern the prosecution of ordinary crime do not necessarily govern the prevention and prosecution of terrorist attacks.131

The Sovereign and the Republic

137

Most troubling, it seems to me that some of the ordinary rights that a citizen relies upon may be curtailed in the case of imminent danger to the very integrity of our civic union.132 To be sure, I believe such instances to be much rarer than is the habit to claim, and even under dire circumstances a polity must be sure not to lose its commitments to the principles by which it defines itself.133 But I cannot shake the idea that under circumstances where the very project of governing is attacked, our notion of individual rights must be thoughtfully balanced. Part and parcel of citizenship is understanding that there are times when your individual rights are subsumed by the needs of your civic community.134 We understand this intuitively in the case of conscription; where the entire nation’s ability to continue as a civic community is threatened, those who are able have the obligation to set aside their current projects and indeed risk their very lives to protect their country.135 Thus, I find the recent controversial use of a drone to kill an American citizen actively engaged in violent terrorism to be a complex question of tactical wisdom but philosophically justifiable. I hasten to add that a philosophy that authorizes exceptional measures, when strictly required to protect the civic polity as a whole, should not be mistaken for a polity that justifies any action under the guise of patriotism. Whether or not some of the less attractive measures in our current “war on terror” are philosophically permissible certainly does not answer whether they are wise. Moreover, the same republican justification that highlights the special duties owed to one’s fellow citizens is premised on sharing a voice in civic governance. The ongoing political machinations that have occurred outside of public view, with little or no discussion and, just as disturbingly, with little oversight by elected representatives, surely cannot be convenient with a republican justification. That the political procedures that resulted in a variety of repellant legal practices were not subjected to republican scrutiny is no trivial procedural objection; submitting controversial political practices to public scrutiny both importantly alters the practices a nation is willing to employ and simultaneously requires a public to discuss and endorse the set of values for which the polity stands. Still, some will rightfully note that a political theory that premises our political rights on our civic bonds rather than individual freedom or autonomy will be more

138

Ekow N. Yankah

liable to curtail what modern sensibilities view as rights against the government, especially in the extraordinary cases such as terrorism or threats to the polity as a whole. This is an ultimately troubling conclusion; an awkward result of a theory that has as its first impulse viewing the power of the state as less external and ominous. The natural downside to understanding that we have a special relationship to those within our civic community is to make sense of the natural intuition that we have dramatically different (and fewer) duties to those outside our civic community. Most starkly, this view recognizes that those who attack our civic bonds as such represent a unique danger to the project of our civic governance through which we express our most basic political nature. Making sense of this intuition, one hopes, will ultimately spur us to engage in serious conversation about the appropriate responses to attacks on our ability to govern and police the dangerous excess of its manifestations rather than hide this impulse in the dark.136 This troubling but I argue plausible conclusion is the result of seeing the question of legitimate sovereignty as a question not of how to justify the imposition of something alien and threatening but of how we are bonded together. Notes 1. Immanuel Kant, Metaphysics of Morals, trans. Mary Gregor (Cambridge: Cambridge University Press, 1996), 13–14. 2. Arthur Ripstein, Force and Freedom: Kant’s Legal and Political Philosophy (Cambridge, MA: Harvard University Press, 2009). 3. Quentin Skinner, The Foundations of Modern Political Thought, Volume I (Cambridge: Cambridge University Press, 1978); “The Idea of Negative Liberty: Philosophical and Historical Perspectives,” in Philosophy in History, ed. Richard Rorty, J. B. Schneewind, and Quentin Skinner (Cambridge: Cambridge University Press, 1984), 193–221. 4. Philip Pettit, Republicanism: A Theory of Freedom and Government (Oxford: Oxford University Press, 1997). 5. Richard Kraut, Aristotle: Political Philosophy (New York: Oxford University Press, 2002), 380–381; Jonathan Barnes, “Aristotle and Political Liberty,” in Aristotle’s Politics: Critical Essays, ed. Richard Kraut and Steven Skultety (Lanham, MD: Rowman & Littlefield, 2005), 187. 6. Paul Gowder, “Democracy, Solidarity, and the Rule of Law: Lessons from Athens,” Buffalo Law Review 62, no. 1 (2014): 2–4.

The Sovereign and the Republic

139

7. Aristotle, “Politica (Politics),” in The Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 1941), I.2. 8. Kraut, Aristotle, 245–246. 9. Ibid., 240–243. 10. Ibid., 242–249. 11. Ibid. 12. Ibid., 245, 249–250. 13. Aristotle, “Ethica Nicomachea (Nicomachean Ethics),” in The Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 1941): VIII.12; “Politics,” I.2.1252b29–30; Kraut, Aristotle, 249–250. 14. Kraut, Aristotle, 243, 247–248. 15. John Cooper, “Political Animals and Civic Friendship,” in Aristotle’s Politics: Critical Essays, ed. Richard Kraut and Steven Skultety (Lanham, MD: Rowman & Littlefield, 2005), 65–70, 74–75. 16. Roberto Gargarella, “Tough on Punishment: Criminal Justice, Deliberation, and Legal Alienation,” in Legal Republicanism: National and International Perspectives, ed. Samantha Besson and José Luis Martí (Oxford: Oxford University Press, 2009), 167–186. 17. Aristotle, “Nicomachean Ethics,” VIII.12.1162a16–19; “Politics,” I.2.1252b29–30; Kraut, Aristotle, 257. 18. Herbert Lionel Adolphus Hart, “Are There Any Natural Rights,” in Theories of Rights, ed. Jeremy Waldron (Oxford: Oxford University Press, 1984). John Locke famously argued that simply by living in a state and thus “accepting” the benefits it confers implies one’s consent to its laws— Two Treatises of Government, ed. Peter Laslett (Cambridge: Cambridge University Press, 1960), 2.120. Rawls endorses this principle in rough form in “Legal Obligation and the Duty of Fair Play,” in Collected Papers (Cambridge, MA: Harvard University Press, 1999), 117–130, though his views add complexities, to which we will return, in A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), 342–343. 19. Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 93–94. 20. Ripstein, Force and Freedom, 187–190. 21. Myles Burnyeat, “Aristotle on Learning to Be Good,” in Essays on Aristotle’s Ethics, ed. Amélie Rorty (Berkeley: University of California Press, 1980), 71–73, 78–79. 22. Robert B. Cialdini, Carl A. Kallgreen, and Raymond R. Reno, “A Focus Theory of Normative Conduct: Recycling the Concept of Norms to Reduce Littering in Public Places,” Journal of Personality and Social Psychology 58, no. 6 (1990): 1015–1026. 23. H. Wesley Perkins and Alan D. Berkowitz, “Perceiving the Community Norms of Alcohol Use among Students: Some Research Implications

140

Ekow N. Yankah

for Campus Alcohol Education Programming,” International Journal of the Addictions 21, no. 9–10 (1986): 961–976. 24. Steve Martin, “98% of HBR Readers Love This Article,” Harvard Business Review, October 2012. 25. Cooper, “Political Animals and Civic Friendship,” 75–76. 26. Ibid., 72–76. 27. Aristotle, “Nicomachean Ethics,” I.7.1097b11; “Politics,” I.2.1253a2–3. 28. Cooper, “Political Animals and Civic Friendship,” 73–73, 79–80. 29. Samuel Scheffler, Equality and Tradition: Questions of Value in Moral and Political Theory (Oxford: Oxford University Press, 2010), 284–294. 30. Samuel Scheffler, Death and the Afterlife (New York: Oxford University Press, 2016). 31. Anthony Appiah, The Ethics of Identity (Princeton, NJ: Princeton University Press, 2003), 114–120, 130–141. 32. Scheffler, Death and the Afterlife, 1–27. 33. Alasdair MacIntyre, Is Patriotism a Virtue? (Lawrence: University of Kansas Press, 1984), 7–9. 34. Aristotle, “Nicomachean Ethics,” I.7.1097b11; “Politics,” I.2.1253a2–3. 35. Cooper, “Political Animals and Civic Friendship,” 65–70, 74–75. 36. I am grateful to Alan Brudner for incisive critique here and in many other places. 37. Philip Pettit, “The Freedom of the City: A Republican Ideal,” in The Good Polity, ed. Alan P. Hamlin and Philip Pettit (Oxford: Blackwell, 1989), 163. 38. Nicola Lacey, State Punishment: Political Principles and Community Values (London: Routledge, 1994). As Nicola Lacey astutely notes: the conception of an a- or pre-social human being makes no sense. What individual human beings perceive as the proper bounds of autonomy around themselves, what they regard as just distributions, how they regard their relations with each other and a thousand other questions central to political philosophy, are ones which we simply cannot imagine being answered outside of some specific social and institutional context. (171) 39. David Keyt, “Aristotle and Anarchism,” in Aristotle’s Politics: Critical Essays, ed. Richard Kraut and Steven Skultety (Lanham, MD: Rowman & Littlefield, 2005), 207. 40. Gerald J. Postema, “Law’s Ethos: Reflections on the Public Practice of Illegality,” Boston University Law Review 90, no. 4 (2010): 1856. 41. Kraut, Aristotle, 252; Ekow N. Yankah, “Legal Vices and Civic Virtue: Vice Crimes, Republicanism and the Corruption of Lawfulness,” Criminal Law and Philosophy 7, no. 1 (2013): 71.

The Sovereign and the Republic

141

42. Aristotle, “Politics,” VII.8.1328a33–36, III.9.1280a30–34; Melissa Schwartzberg, “Aristotle and the Judgment of the Many: Equality, not Collective Quality,” Journal of Politics 78, no. 3 (2016): 734–736. 43. Schwartzberg, “Aristotle and the Judgment of the Many,” 734–736. 44. Kraut, Aristotle, 249–252, 279–305; Dorothea Frede, “Citizenship in Aristotle’s Politics,” in Aristotle’s Politics: Critical Essays, ed. Richard Kraut and Steven Skultety (Lanham, MD: Rowman & Littlefield, 2005), 177–178. 45. Aristotle, “Politics,” I.4.1253b32, I.13.1254b22–1260a12; Kraut, Aristotle, 249, 285–287. 46. Aristotle, “Politics,” I.3.1353b20–23, I.6.1244a4–5; Keyt, “Aristotle and Anarchism,” 208–209, 212; Kraut, Aristotle, 277–303. 47. Bernard Williams, Shame and Necessity (Berkeley: University of California Press, 1993), 117. 48. Frede, “Citizenship in Aristotle’s Politics,” 177–78; Keyt, “Aristotle and Anarchism,” 216. 49. Frede, “Citizenship in Aristotle’s Politics,” 172,176–178. 50. Pettit, “Freedom of the City,” 159; Pettit, Republicanism, 66–67. 51. Ekow N. Yankah, “Republican Responsibility in Criminal Law,” Criminal Law and Philosophy 9, no. 3 (2015): 457. 52. Pettit, “Freedom of the City,” 149–150. 53. Gowder, “Democracy, Solidarity, and the Rule of Law,” 21–41. 54. Ibid., 14–15; Gregory Vlastos, “Isonomia,” American Journal of Philology 74, no. 4 (1953): 337. In ancient Athens, isonomia represented the idea that in a well-ordered society, by requiring all to be subject to and restrained by the same law, legal equality secured political equality, allowing the poor and powerless to hold the rich and powerful accountable on the same terms. 55. Aristotle, “Politics,” III.4; Kraut, Aristotle, 251–252; Cooper, “Political Animals and Civic Friendship,” 69. 56. Kraut, Aristotle, 362; Jean Hampton, “Punishment, Feminism, and Political Identity: A Case Study in the Expressive Meaning of the Law,” Canadian Journal of Law & Jurisprudence 11, no. 1 (1998): 23. 57. Pettit, “Freedom of the City,” 141–142. 58. Vlastos, “Isonomia,” 354–357. 59. Aristotle, “Politics,” III.4.1276b36–1277b34. 60. Schwartzberg, “Aristotle and the Judgment of the Many,” 736–737, 741–742. 61. Ibid., 742. 62. Ibid., 736. 63. There is no reason to read a view that promotes republican governance as incompatible with the recognition of the need for technical expertise. It is sufficient that a polity will try to maximize as much as prac-

142

Ekow N. Yankah

ticably possible the opportunity for widespread voice, contestation, and shared governance. Ibid., 740–741; Josiah Ober, “Democracy’s Wisdom: An Aristotelian Middle Way for Collective Judgment,” American Political Science Review 107, no. 1 (2013): 827. 64. Pettit, “Freedom of the City,” 145; Yankah, “Legal Vices and Civic Virtues,” 70–74. 65. Barnes, “Aristotle and Political Liberty,” 189–190. 66. Yankah, “Legal Vices and Civic Virtue,” 72–74. 67. Jeremy Waldron, “Why Law—Efficacy, Freedom, or Fidelity?,” Law and Philosophy 13, no. 3 (1994): 259. 68. It also bears noticing that these quiet ways in which we naturally rely on the common good arguably overlook the Rawlsian injunction that basic political institutions be neutral regarding competing conceptions of the good. The Rawlsian account and its claim that neutrality is demanded only of the basic structure is sufficiently nuanced that pursuing the point more fully will have to wait for a future occasion. 69. Kraut, Aristotle, 102, 265–272. 70. Josiah Ober, Athenian Legacies: Essays on the Politics of Going on Together (Princeton, NJ: Princeton University Press, 2005), 25. 71. Aristotle, “Politics,” III.6.1279a17–21, III.6.1284b4–19. 72. Ibid., III.7.1279b28–32. 73. Ober, Athenian Legacies, 1–4, 12–14, 24. 74. Ibid., 11–15. 75. Ibid., 24–25. 76. Ibid. 77. Yankah, “Legal Vices and Civic Virtues,” 73–74; Aristotle, “Nicomachean Ethics,” VIII.9. 78. Aristotle, “Politics,” II.2–5; Schwartzberg, “Aristotle and the Judgment of the Many,” 735. 79. Kraut, Aristotle, 271–275. 80. Yankah, “Republican Responsibility in Criminal Law,” 463–464. 81. Burnyeat, “Aristotle on Learning to Be Good,” 71–73, 78–79; William A. Edmundson, “The Virtue of Law-Abidance,” Philosophers’ Imprint 6, no. 4 (2006): 16. 82. Aristotle, “Politics,” IV.1.1295b2–1296b2; Yankah, “Legal Vices and Civic Virtue,” 75–76; Gowder, “Democracy, Solidarity, and the Rule of Law,” 14. 83. Heather Gerken, “Getting from Here to There in Redistricting Reform,” Duke Journal of Constitutional Law & Public Policy 5, no. 1. (2010): 1–15. 84. Niccolò Machiavelli, The Discourses on Livy, trans. Luigi Ricci, ed. Eric R. Vincent (New York: Random House, 1950), ch. 4; Maurizio Viroli,

The Sovereign and the Republic

143

“Machiavelli and the Republican Idea of Politics,” in Machiavelli and Republicanism, ed. Gisela Bok, Quentin Skinner, and Maurizio Viroli (Cambridge: Cambridge University Press, 1990), 143–172. 85. Aristotle, “Nicomachean Ethics,” V.1.1129b24–25. 86. Gowder, “Democracy, Solidarity, and the Rule of Law,” 9–10. 87. Kraut, Aristotle, 117–118, 271; Edmundson, “The Virtue of LawAbidance,” 1, 11–12, 17. 88. Yankah, “Legal Vices and Civic Virtue,” 67; Jill Frank, A Democracy of Distinction: Aristotle and the Work of Politics (Chicago: University of Chicago Press, 2005), 122. 89. Aristotle, “Politics,” III.4.1277a25–7, III.4.b14–17; Kraut, Aristotle, 367–369. 90. Lawrence B. Solum, “Natural Justice,” American Journal of Jurisprudence 51 (2006): 104–105. 91. Ibid.; Aristotle, “Nicomachean Ethics,” III.1.1110a4–27; III.4.1277a27; Kraut, Aristotle, 114–118. 92. Kraut, Aristotle, 370–371. 93. Ibid.; Edmundson, “The Virtue of Law-Abidance,” 1, 6–7, 13–16. 94. Kraut, Aristotle, 271–272. 95. Ibid., 270–271, 380. 96. Stephen Perry, “Political Authority and Political Obligation,” in Oxford Studies in Philosophy of Law: Volume 2, ed. Leslie Green and Brian Leiter (Oxford: Oxford University Press, 2013), 1–75. 97. As Lacey puts it in State Punishment, 171. 98. Antony Duff, Answering for Crime: Responsibility and Liability in the Criminal Law (Oxford: Hart, 2007), 141. 99. Yankah, “Republican Responsibility in Criminal Law,” 463–464; John Braithwaite and Philip Pettit, Not Just Deserts: A Republican Theory of Criminal Justice (Oxford: Clarendon Press, 1990), 63–64, 69. 100. N.R.E. Fisher, Hybris: A Study in the Values of Honour and Shame in Ancient Greece (Warminster: Aris and Phillips, 1992), 1; Yankah, “Republican Responsibility in Criminal Law,” 464. 101. John Finnis, Natural Law and Natural Rights (Oxford: Oxford University Press, 1980), 232. 102. Yankah, “Legal Vices and Civic Virtue.” 103. Aristotle, “Politics,” III.12.1282b23–1283a16; Kraut, Aristotle, 121, 265–267, 275, 378–379. 104. Aristotle, “Politics,” I.6, III.6. 105. Pettit, Republicanism, 173. 106. Aristotle, “Nicomachean Ethics,” X.9.1179b32–1180a4, XX; Kraut, Aristotle, 359–361, 365–366. 107. Pettit, Republicanism, 196–200.

144

Ekow N. Yankah

108. Ibid. 109. The choice of multifamily dwellings is not accidental. Because multifamily dwellings tend to allow larger numbers of people to live in a (desirable) area, they can often have a democratizing effect. Further, in America, minority and immigrant communities and other culturally distinct groups have often preferred multifamily dwellings due to either economic necessity or cultural preference. Thus, the prohibition of multifamily dwellings (or multiple families in a shared dwelling) has often been seen as a proxy for excluding groups. It is precisely this kind of negotiation, between the desire of some to preserve neighborhoods of certain characteristics, the balancing of wealth versus access, but most particularly, the use of law for illicit factional abuse, that is at the heart of both cultivating a shared voice in the community and sensitivity to persistent minorities. 110. As mentioned earlier, such limits are so natural as to often escape notice. Only those tempted by libertarianism, for example, are struck by Nozick’s famous example rejecting the obligations imposed by a neighborhood music association. Indeed, neighborhood associations provide for neighborhood festivals, regulate music, and control noise all the time. 111. Aristotle, “Politics,” III.12.1282b23–1283a16. 112. For Aristotle, this was true of the most important forms of excellence, be it wealth, birth status, and even that great Greek good of beauty. 113. See the tension between the majority and Justice Scalia’s dissent in Lawrence v. Texas, 539 U.S. 558 (2003). 114. Ekow N. Yankah, “Crime, Freedom and Civic Bonds: Arthur Ripstein’s Force and Freedom: Kant’s Legal and Political Philosophy,” Criminal Law and Philosophy 6, no. 2 (2012): 255–272. 115. Immanuel Kant, Groundwork of the Metaphysics of Morals, trans. Mary Gregor (Cambridge: Cambridge University Press, 1997), 6:256— 6:257; Ripstein, Force and Freedom, 164–169, 190–195. 116. Kant, Groundwork, 6:355; Ripstein, Force and Freedom, 54–55, 190– 195. 117. Yankah, “Crime, Freedom and Civic Bonds,” 266; Victor Tadros, “Independence Without Interests?” Oxford Journal of Legal Studies 31, no. 1 (March 2011): 193–213. But see Arthur Ripstein, “Private Authority and the Role of Rights: A Reply,” Jerusalem Review of Legal Studies 14, no. 1 (December 2016): 64–86. It should be noted that there are competing pictures of Kant’s political project that make much more central Kant’s goal that the law should accord with the laws that would obtain in a Kingdom of Ends, championed by thinkers such as Barbara Herman and Seana Shiffren. Also see Barbara Herman, Moral Literacy (Cambridge, MA: Harvard University Press, 2007).

The Sovereign and the Republic

145

118. Jeremy Waldron, “Kant’s Legal Positivism,” Harvard Law Review 109, no. 7 (May 1996): 1535–1566. 119. Ripstein, Force and Freedom, 192–193, 208, 211, 218. 120. Though from a purely Kantian perspective, I am torn as to whether debilitating conditions that arise from natural causes can be counted as impositions on one’s “freedom.” Remember that Kant believes that freedom is imposed upon by others and is strict in describing one’s needs outside of poverty to be mere wishes from the view of others. 121. Pettit, “Freedom of the City,” 142. 122. Tadros, “Independence without Interests?.” 123. Yankah, “Crime, Freedom and Civic Bonds,” 263. 124. Ricci v. DeStefano, 557 U.S. 557 (2009). 125. Schwartzberg, “Aristotle and the Judgment of the Many,” 738. 126. Ibid., 742. 127. Aristotle, “Politics,” III.11.1281b28–30. 128. Schwartzberg, “Aristotle and the Judgment of the Many,” 738. 129. Here I am thinking of a case similar to the establishment of the American Colonization Society which in the early 1800s proposed to send free blacks back to Africa, arguing that they faced better chances for success there than in America. Despite living in a society in which their skin color was a sign of social inferiority and bondage, many prominent African Americans fiercely rejected the idea. Fredrick Douglass wrote movingly of the idea that despite its staggering evil, America was the only land nearly any black American had ever known and that the blood of the slave had built America. Thus, it was better for those able to stay and demand their proper place as citizens. In future work, I hope to explore this pressing question of how, despite a polity’s failure to grant a person or group rights as citizens, they may demand entrance as citizens. 130. Gowder, “Democracy, Solidarity, and the Rule of Law,” 13. 131. Ekow N. Yankah, “When Justice Can’t Be Done: The Obligation to Govern and Rights in the State of Terror,” Law and Philosophy 31, no. 6 (2012): 643–672. 132. Ibid., 665; François Tanguay-Renaud, “Making Sense of ‘Public’ Emergencies,” Philosophy of Management 8, no. 2 (2009): 36–41. 133. Yankah, “When Justice Can’t Be Done,” 668. 134. Kraut, Aristotle, 265–267. 135. Ibid., 269–270. 136. Yankah, “When Justice Can’t Be Done,” 668–672.

5 POLITICAL LEGITIMACY UNDER EPISTEMIC CONSTRAINTS WHY PUBLIC REASONS MATTER FABIENNE PETER

Political decisions regulate many aspects of our lives. For example, they might concern the legal permissibility of abortions, influence the extent of income and wealth inequalities, or determine membership in international or global institutions. Such political decisions are legitimate if either the decisions themselves, or the political institutions through which they were made, are justified.1 What does the justification of political decisions, whether it is substantive or procedural, demand? Philosophers generally understand justification in terms of reasons. A belief is justified if there are sufficient or decisive reasons to support that belief. An action is justified if there are sufficient or decisive reasons that support that action. On this view, which I will adopt here, a political decision is justified if there are sufficient or decisive reasons in support of it. But what sort of reasons can justify political decisions? In political life, we often disagree about the answer to that question. Sometimes, these disagreements have their origin primarily in divergent moral or religious beliefs. We might disagree, for example, about the extent to which the government should protect individual liberty relative to other values and thus disagree about the legitimate extent of government interference with individual choices. Other disagreements have their origins primarily in divergent beliefs about non-moral facts. Examples 147

148

Fabienne Peter

include disagreements about climate change policies that derive from conflicting assessments of the effectiveness of carbon taxation or disagreements about the desirability of a minimum wage policy that are based in conflicting assessments about the effects of such policies on unemployment. These disagreements, too, might yield incompatible assessments of whether a government policy is justified. This fact of disagreement about the justification of political decisions needs to be distinguished from the normative question of what grounds the justification of political decisions. It is this normative question, a question about the sort of reasons that have the power to justify political decisions, that is at the heart of the debate on public reason and that is the focus of this chapter. According to defenders of a Public Reason Conception of political legitimacy, the reasons that can justify political decisions are those that all can accept or that no one can reasonably reject, under some suitable qualification of what counts as an agreement or as a reasonable rejection.2 On this conception, at least some (but typically not all) of our disagreements are normatively significant in the sense that they have the power to undermine the justification of political decisions. John Rawls, for example, took some moral and religious reasons to be subject to normatively significant disagreements and argued that those reasons cannot justify political decisions.3 Because the reasons in the justifying set are those that are not subject to a normatively significant disagreement, I want to characterize the reasons that can justify political decisions on the Public Reason Conception as agreement-dependent. On the opposing conception of political legitimacy, which I call the Objective Reason Conception, the justification of political decisions is agreement-independent. According to defenders of this conception, normative practical reasons are, or are given by, objective facts, and whether political decisions are justified depends on how well they are supported by those reasons.4 It is an implication of this view that our disagreements are not normatively significant; disagreements do not have an impact on practical justification and, as a result, do not have an impact on the justification of political decisions. A related way of understanding the contrast is this. According to the Public Reason Conception of political legitimacy, political

Political Legitimacy Under Epistemic Constraints

149

decisions must be justified to the citizens. Disagreements are normatively significant when they imply a failure to justify a particular decision to the citizens. According to the Objective Reason Conception, by contrast, practical justification, in general, is not justification to, but justification simpliciter. Whether a political decision is justified depends on objective reasons, not on whether there is some sort of agreement. My aim in this chapter is to argue against the Objective Reason Conception of political legitimacy and, in this way, provide support for the claim that public reasons matter for political legitimacy. My starting point, however, is to grant that there are objective reasons and to allow that they can affect the justification of political decisions. I will show that granting this premise to the defenders of the Objective Reason Conception does not mean that the justification of political decisions is necessarily in terms of those objective reasons. Agreement can still matter. The main problem with the Objective Reason Conception of political legitimacy arises from its stance on disagreements. According to this conception, as I mentioned, disagreements do not affect the justification of political decisions; disagreements are normatively neutral. Disagreements such as the ones I described at the beginning are the result of mistaken beliefs about reasonrelevant facts—normative or non-normative—and such mistakes, it is thought, do not undermine the justification of political decisions. I will argue that defenders of a Public Reason Conception of political legitimacy are right to emphasize that not all disagreements are mere mistakes; some disagreements have normative implications. When they do, political legitimacy depends on overcoming those disagreements. Of course, a defender of a Public Reason Conception does not need to claim—and typically does not claim—that all disagreements undermine political legitimacy. Some disagreements are not normatively significant because they are the result of mistakes or an unwillingness to consider relevant evidence or arguments. A successful defense of the claim that public reasons matter for political legitimacy needs to answer the question of which disagreements are normatively significant. I will say a lot more below on what I mean by normatively significant disagreements. For the time being, let me just note that I propose to answer this question

150

Fabienne Peter

not on the basis of some moral or political values, as has been common in the literature on public reason.5 Instead, I will start from the epistemology of practical reasoning. I will provide an account of practical reasoning that is compatible with the Objective Reason Conception and thus refrain from loading the dice against that conception from the start. I will then use this account to identify a category of normatively significant disagreements, however, and show how such disagreements affect practical justification, at least in the context of political legitimacy. This will allow me to pinpoint where the Objective Reason Conception goes wrong and to build my argument for the significance of public reasons for political legitimacy. The chapter is organized as follows. I will start with a more exact portrayal of what is at stake in the debate between defenders of the Objective Reason Conception and defenders of the Public Reason Conception. I will then introduce my account of the epistemology of practical reasoning and present a first argument against the Objective Reason Conception—the argument from permissible practical reasoning. Next, I develop the idea of normatively significant disagreements and show how it figures in a second argument against the Objective Reason View. In the section that follows, on Political Legitimacy, I will present my third and final argument against the Objective Reason View. This argument demonstrates that public reasons matter for political legitimacy, at least under some epistemic circumstances. I will end with a brief discussion of some objections that one might raise against the view I defend in this chapter. Two Conceptions of Political Legitimacy It is time to characterize the Objective Reason Conception (OR) of political legitimacy with greater precision. I understand it as committed to the following two claims. The first claim is this: OR: There are normative practical reasons that are, or are given by, attitude-independent facts; they are objective reasons.

To accept OR is to accept that at least some normative practical reasons have their source of normativity in attitude-independent

Political Legitimacy Under Epistemic Constraints

151

facts. Such reasons do not derive from beliefs we have about what is desirable, for example, or from what we happen to desire. There are two ways in which one can spell out OR. The strong version says that all normative practical reasons are, or are given by, attitude-independent facts. Many have endorsed this claim—it has become a very popular view in recent meta-normative theory.6 OR can also be interpreted more weakly, however. On this interpretation, the claim is only that some normative practical reasons (but not necessarily all) are objective reasons. I accept the weak interpretation of OR, and it is an upshot of the argument I present in this chapter, that we should reject the strong interpretation, at least for the political context.7 While the first claim is about normative practical reasons, the second claim is about how these reasons bear on the justification of political decisions. ORJ: The justification of political decisions is necessarily in terms of objective reasons.

ORJ states that the only reasons that matter for political legitimacy are those given by attitude-independent facts.8 Joseph Raz and David Enoch are among the philosophers who have most prominently advocated versions of the Objective Reason Conception. The Normal Justification Thesis that is at the core of the conception of political legitimacy that Raz advocates states that the legitimacy of political decisions depends on whether those decisions allow the citizens to better conform with the reasons that apply to them anyway, independently of the decisions of the political authority.9 The reasons in question are fact-given reasons; they are not reasons that citizens agree on or endorse. Enoch, similarly, draws a distinction between the facts that are reason-giving and people’s beliefs about those facts. He argues that defenders of a Public Reason Conception are wrong to focus on people’s perspectives; what justifies political decisions is not what people believe about those decisions, but whether or not the facts support those decisions.10 My main target in this chapter is ORJ, the claim that political justification is necessarily in terms of objective reasons. As mentioned above, I grant that there are objective reasons, and I accept

152

Fabienne Peter

the weak version of OR. But I will argue that objective reasons are not the only ones that can justify political decisions. My argument against ORJ hinges on the role of practical reasoning in political decision-making and in political justification. Practical reasoning is reasoning about what we should do and political decision-making relies on it. Our practical reasoning, in general, is not determined by the facts, however, but by our perspective on the facts. As a result, what we have objective reason to do and what we believe we should do might come apart. The following example illustrates the point. Suppose you consider offering your friend a drink. You believe—and have every reason to believe—that the glass in front of you contains gin. You decide that you should offer your friend the drink. Unfortunately for you and your friend, the glass does not contain gin. It contains petrol. If all normative practical reasons are given by attitudeindependent facts, as the strong version of OR claims, then you have no normative reason to offer your friend the drink. If only reasons given by attitude-independent facts determine practical justification, then you are not justified in offering your friend the drink. Yet, there seems to be nothing wrong with your reasoning. When we reason practically, we necessarily reason on the basis of what we believe about the practical circumstances we find ourselves in. Even defenders of the strong version of OR accept that this is a feature of our practical reasoning, of course. They deal with the issue in the following way. In addition to objective reasons, there are subjective, or apparent, reasons.11 Subjective reasons reflect our perspective on the situation—as given by our beliefs and/or our evidence. And as the gin/petrol example shows, what we have objective reason to do and what we have subjective reason to do can come apart. Defenders of the strong version of OR interpret subjective practical reasons in relation to our beliefs and/or our evidence.12 Those who reject the strong version of OR—either in favor of a hybrid view or because they reject OR altogether—often characterize subjective reasons in other ways.13 Although this is an important topic, I cannot fully discuss it here. It is also somewhat tangential to the argument I develop here, as my argument does not depend on a particular characterization of subjective reasons. For

Political Legitimacy Under Epistemic Constraints

153

the purposes of this chapter, I shall define subjective practical reasons as reasons that are or involve beliefs about the relevant practical circumstances. This is a fairly loose characterization that leaves open a number of questions—whether the relevant beliefs need to be justified or not; what, if anything, justifies those beliefs, for example. What interests me here above all is the question of what sort of reasons can ground practical justification in the context of political legitimacy—objective reasons or subjective reasons? Consider the following example from the political context. Suppose there is a policy A which would increase the health status of the poor, and which should be implemented for that reason. A defender of the Objective Reason Conception would argue that a political decision to implement policy A is justified on the basis of the fact-given reason that it increases the health status of the poor. But in our practical reasoning about whether or not we should implement policy A, we might form divergent beliefs about the empirical facts— whether or not policy A does, in fact, increase the health status of the poor—and about the normative facts—whether the fact that a policy increases the health status of the poor is a reason to implement it. From the perspective of our subjective reasons, the case for implementing policy A might not be clear-cut. Do subjective reasons play a role in the justification of political decisions? Defenders of the Public Reason Conception of political legitimacy answer positively. Different interpretations of this view work with different interpretations of subjective reasons. What they all have in common, however, is the claim that political decisions must be justified on the basis of subjective reasons—whether they are substantive reasons or procedural reasons.14 In my understanding, the Public Reason Conception is thus committed to the following claim: PRJ: The justification of political decisions is necessarily in terms of an agreement grounded in subjective reasons.

Of course, PRJ should not be understood to mean that the relevant agreement that can justify a political decision must include any possible set of subjective reasons. Defenders of a Public Reason Conception typically allow for some restrictions in the admissible

154

Fabienne Peter

set. Some interpret the relevant restriction in terms of reasonableness, which is a restriction on public deliberation.15 In this interpretation, subjective reasons that do not meet the reasonableness test, however it is specified, do not undermine the justification of political decisions. Only disagreements that survive the reasonableness test can undermine the justification of political decisions. Others formulate the restriction in terms of individual deliberation. Gaus, for example, argues that only good reasoning gives rise to what I call subjective reasons here.16 Note also that my characterization of the Public Reason Conception does not distinguish between the so-called consensus accounts and convergence accounts. According to consensus accounts, agreement-based justification of political decisions is in terms of a single, shared set of reasons. According to convergence accounts, it is sufficient that an agreement that is relevant for the justification of political decisions obtains at the level of the decisions themselves; it does not have to obtain at the level of the reasons that justify a particular decision. Such an agreement might thus be supported by diverse reasons. My characterization does not distinguish between the two accounts because it leaves open what exactly the relationship is between agreement and the reasons that justify a political decision.17 Whatever one might want to say about which subjective reasons are relevant for the justification of political decisions, the important point to note is that it is with regard to the normative significance of subjective reasons that the Public Reason Conception and the Objective Reason Conception come apart. Defenders of the Objective Reason Conception reject the normative significance of subjective reasons. They draw a distinction between practical reasoning and practical justification and maintain that even if practical reasoning is in terms of subjective reasons, practical justification is still necessarily in terms of objective reasons. We can make progress in this debate if we examine the epistemology of practical reasoning, or so I will argue in this chapter. Even if we grant that there are objective reasons and that they sometimes matter for the justification of political decisions, the epistemic circumstances that characterize the political justification are often such that we lack a good grip on what they are. And when objective reasons are epistemically out of reach, they

Political Legitimacy Under Epistemic Constraints

155

are not available for the justification of political decisions. If there is any justification for political decisions in those circumstances, it will depend on what is accessible in valid practical reasoning and, therefore, on subjective reasons. ORJ should thus be rejected and be replaced by a conception of what is required for the justification of political decisions that recognizes that subjective reasons play an important role in political justification, at least in some—epistemic—circumstances.18 Epistemic Constraints on Practical Reasoning What do I mean by practical reasoning? Let me start with a comment on the contrast with theoretical reasoning. Theoretical reasoning answers a question about what should be believed. Practical reasoning, by contrast, answers a question about what should be done. But it would be wrong to think that there is no overlap between theoretical and practical reasoning. Practical reasoning depends on our beliefs about non-normative facts and, if one accepts at least a weak version of OR, as I do, it also depends on our beliefs in relation to normative facts—facts that are or give rise to normative practical reasons. Because of that fact-dependence, we can ask about the epistemic constraints that apply to practical reasoning. Some have argued, building on Timothy Williamson’s work on the knowledge norm for assertion, that the relevant constraint is knowledge.19 The knowledge norm for assertion is this: KA: It is permissible for you to assert that p iff you know that p.

For example, suppose a lost tourist asks you about the way to his hotel. It is permissible for you to assert “the hotel is just down the road” if (and only if) you know that the hotel is just down the road, but not if you do not and just want to be seen as friendly. Williamson explains that “asserting that P without knowing that P is doing something without having the authority to do it, like giving someone a command without having the authority to do so.”20 Assertions are, of course, a type of action and some have thus generalized Williamson’s account to reasons for action and to

156

Fabienne Peter

practical reasoning.21 Applied to practical reasoning, our topic here, the Knowledge Norm says this: KPr: Your use of p as a premise in practical reasoning is permissible iff you know that p.

We can apply the knowledge norm both to ordinary practical reasoning and to normative practical reasoning. Using the example from above again, the knowledge norm says that your reasoning that a policy A should be implemented because it increases the health status of the poor is a permissible instance of practical reasoning just in case you know that the policy increases the health status of the poor (non-normative) and that it does so is a reason to implement the policy (normative).22 Is the Knowledge Norm a plausible restriction on practical reasoning? There are two main objections that aim to show that it is not.23 The first is that there are many cases in which knowledge of p does not seem necessary for permissible practical reasoning in p-related contexts. The second is that there are cases in which knowledge does not appear to be sufficient. Although I find both objections convincing and relevant to the topic under discussion, I will only be able to focus on the necessity objection here. However, in the final section of this chapter, I will briefly touch on the sufficiency objection as well. The gin/petrol case I mentioned earlier can serve to illustrate the necessity objection. Suppose you are justified to believe that the glass contains gin, not petrol—all the other glasses on the table contain gin, you have had a drink yourself earlier from this table, you have seen a waiter fill the glasses with a bottle of gin in his hand, and so on. It seems permissible for you in this case to use the premise “this is gin” when deliberating about whether to offer your friend the drink. As it happens, the glass contains petrol, a fact you do not know. So, the knowledge norm is violated, but there does not seem to be anything wrong with your practical reasoning. We will also accept your excuse that you did not know this fact when it turns out that your action made your friend very ill. This example seems to show that it can be appropriate to invoke p in your p-dependent practical reasoning and choice, even if p is false. Knowledge, which I am assuming is factive, is thus not necessary.

Political Legitimacy Under Epistemic Constraints

157

The right response to the necessity objection is to weaken the knowledge norm and replace knowledge by justified belief.24 The Justification Norm for permissible practical reasoning says this: JPr: Your use of p as a premise in practical reasoning is permissible iff you are justified to believe that p.

In the gin/petrol case, JPr says that if you are justified to believe that the glass contains gin, then your use of that premise in practical reasoning about whether to offer your friend the drink is permissible; knowledge is not required. In the political case, JPr says that if you are justified to believe that a policy is feasible, for example, then your use of that premise in practical reasoning about which policy options to consider is permissible; knowledge is not required. If this is correct and we should replace KPr by JPr, we can use this to formulate a first argument against the Objective Reason Conception of political legitimacy. The argument goes like this. The Objective Reason Conception is only plausible if practical reasoning is plausibly restricted by KPr, at least in the normative context. This is so because of the facticity of knowledge. To see this, assume p refers to a fact that gives rise to an objective reason. Only if you know that p do objective reason (given by p) and subjective reason—the reason that determines your practical reasoning, given by your belief that p—necessarily coincide. But we saw that Kpr is not plausible and that we thus cannot assume that objective reasons and subjective reasons necessarily coincide. The more plausible norm, JPr, implies that knowledge is not necessary for permissible practical reasoning. This means that there might be a gap between what you permissibly reason you should do and what you should do in the objective reason sense. Ah, you might say, but this argument is flawed. JPr only shows that permissible practical reasoning depends on subjective reasons, not that practical justification does. But the Objective Reason Conception is only concerned with practical justification, not with permissible practical reasoning. The whole point of the gin/petrol case is to show that I might be permissibly reasoning that this is gin and, on that basis, decide to offer the drink to my friend. But to say that I am reasoning permissibly is only to say that what I am

158

Fabienne Peter

doing is rational, that I can be excused for my action, and so on. It is not to say that I am justified in doing so in the full-blooded sense—that I have a normative reason to do so. Even if we accept that practical reasoning is necessarily in terms of subjective reasons, we do not have to accept that subjective reasons carry any independent normativity. And you are right of course; the argument I have presented is flawed in this way. A successful argument against the Objective Reason Conception needs to show that subjective reasons not only determine permissible practical reasoning, but that they can have normative import in their own right and that ORJ is thus false. My next argument against ORJ is an argument from disagreement. Normatively Significant Disagreements Characterized quite generally, disagreements involve attitudes that are not only incompatible but that cannot be jointly held.25 Disagreements between doxastic attitudes are a case in point. If you express a belief that p (e.g., the policy is feasible) and I express a belief that not-p (the policy is not feasible), we have a disagreement, as the two attitudes cannot be jointly held: Either the policy is feasible or it is not. But there are also non-doxastic disagreements. An interesting case is disagreement about taste. Can we have disagreements about taste? Say you maintain “haggis is tasty” and I maintain “haggis is not tasty.” Clearly, the two attitudes are incompatible. But there does not seem to be anything particularly problematic about that. It certainly does not follow that one of us has made a mistake. So while the two attitudes are incompatible, it appears that they can be jointly held—by different individuals, at least. If that is the relevant test, we do not have a disagreement. But if we apply a more demanding test and specify that the relevant attitudes that constitute a disagreement are attitudes that one individual could not jointly hold, then we have a disagreement about taste.26 The topic of this chapter is practical disagreements— disagreements about what we should do. Practical disagreements can result from attitudes such as desires, intentions to act, beliefs relating to practical reasons, and the like. If the relevant attitudes are desires, the practical disagreement might look like a

Political Legitimacy Under Epistemic Constraints

159

disagreement about tastes—if there is a disagreement at all, it is faultless and typically unproblematic. But clearly there are constellations of practical attitudes that constitute a disagreement. Consider the following attitudes. You hold “we should do x”; I hold that “we should do not-x.” Our incompatible attitudes may still be faultless, i.e., not involve a mistake on either side but simply arise from different desires. But it is a clear case of a practical disagreement as our incompatible attitudes toward what we should do cannot be jointly realized and hence jointly held by us. Given that we can have practical disagreements, there is a further distinction that is relevant for my purposes here, the distinction between fact-dependent and fact-independent disagreements. In a doxastic disagreement, the disagreement is about which proposition is made true by the facts: Is the policy feasible or not? Such a disagreement is obviously a fact-dependent disagreement. In a disagreement about taste, if it occurs at all, there is probably no fact of the matter. If there is not, then it is not a fact-dependent disagreement. Practical disagreements can be of either kind. I will focus here on fact-dependent practical disagreements, and the relevant facts might be either nonnormative or normative.27 To use my political example again, you and I might disagree about whether or not policy A increases the health status of the poor and/or about whether or not the fact that A increases the health status of the poor justifies implementing A. Returning now to practical reasoning, accepting JPr has the following implication: Permissible practical reasoning may result in fact-dependent disagreements. This is an important difference between KPr and JPr. If knowledge sets the constraint for permissible practical reasoning, then we cannot have disagreements that are compatible with permissible practical reasoning. If we find ourselves in a disagreement about p, at least one of us will have reasoned impermissibly. But if JPr is the correct norm for permissible practical reasoning, then it is possible that one of us permissibly concludes that p and the other permissibly concludes that not-p.28 Suppose we both agree that policy A is more desirable than policy B. But, to use my non-normative example again, suppose that while you justifiably believe that A is feasible, I justifiably believe that it is not feasible. We then find ourselves in a disagreement

160

Fabienne Peter

about whether to choose A or B that results from our permissible reasoning. I want to claim that fact-dependent practical disagreements that result from permissible reasoning are normatively significant, at least in the context that is the focus of my chapter here—the justification of political decisions. They are normatively significant in the sense that they have implications for practical justification. To see the plausibility of this, consider the contrast between a disagreement that results from impermissible reasoning and one that results from permissible reasoning. If I have simply not bothered to take into account evidence for the feasibility of A although it would have been easy for me to so, for example, then our disagreement is of a different kind than if my reasoning in favor of B is permissible. If I am to blame, epistemically speaking, for my false belief that A is not feasible, then it does not follow from our disagreement that choosing A is not justified. Choosing A might still be justified, even if I believe that we should choose B. But if neither of us is reasoning impermissibly, the situation is different. Practical disagreements that result from permissible reasoning can affect justification in the following way. If you justifiably believe p and I justifiably believe not-p, a p-dependent policy choice is either not justified to me or it is not justified to you. In the policy example, if you permissibly reason that A is feasible and that we ought thus to do A, and I permissibly reason that A is not feasible and that we thus ought to choose B, choosing A is not justified to me and choosing B is not justified to you. The same line of thought can also be developed for disagreements that involve normative beliefs. We can use this point to formulate a second argument against the Objective Reason Conception of political legitimacy—the argument from normatively significant disagreements. Recall that ORJ says that the justification of political decisions necessarily depends on objective reasons. This implies that disagreements about normative practical reasons are normatively insignificant. But I have just shown that disagreements about which objective reasons apply that result from our permissible practical reasoning can undermine practical justification, including the justification of political decisions. Such disagreements are thus normatively

Political Legitimacy Under Epistemic Constraints

161

significant, and subjective reasons may affect what counts as a justified decision. Therefore, ORJ is false. Ah, you might say now, but this argument is flawed too. To show that ORJ is false, I would have to show that objective reasons are not necessary for the justification of political decisions. But I have not yet shown that. I have only shown that there might be a gap between justification simpliciter and justification to—justification in terms of objective reasons and justification in terms of subjective reasons. And you are right again. My belief that A is not feasible might be mistaken, whereas you correctly hold that A is feasible. Choosing A might then still be justified, even if it is not justified to me. That is all that ORJ claims. I have thus not yet shown that ORJ is false, let alone that we should adopt some version of the Public Reason Conception of political legitimacy. While more work is needed to show that ORJ is false, the second argument establishes something. It cannot establish that objective reasons are not necessary for the justification of political decisions, but it does show that the justification of political decisions may be epistemically underdetermined when permissible practical reasoning leads to disagreements about which policy should be chosen and neither side can justifiably claim the epistemic high ground. If nobody knows—in a factive sense—the reason-relevant facts, then ORJ does not have a grip on the justification problem. Permissible practical reasoning may lead to a gap between objective and subjective reasons, and this gap implies that practical justification in terms of objective reasons is epistemically underdetermined. This does not yet show that agreement at the level of subjective reasons is required for the justification of political decisions, as the Public Reason Conception claims, but it does show that disagreements that result from permissible practical reasoning can undermine the applicability of ORJ. Political Legitimacy and Disagreements To show that ORJ is false, I have to show that objective reasons are not necessary for the justification of political decisions, at least in some contexts. I will do so by showing that there are epistemic circumstances in which the justification of political decisions requires

162

Fabienne Peter

an agreement involving subjective reasons. I will not, however, argue for the Public Reason Conception of political legitimacy as it is normally understood. Recall that I characterized this conception in the following way: PRJ: The justification of political decisions is necessarily in terms of an agreement grounded in subjective reasons.

What I will defend in this section is a weaker version of this conception. The weaker conception does not claim that political legitimacy always requires an agreement involving subjective reasons. It only claims that an agreement involving subjective reasons is sometimes necessary for political legitimacy, the idea being that it is necessary when the justification of political decisions in terms of objective reasons fails because of epistemic limitations. Specifically, what I will defend is the following claim: PRJ*: In some epistemic contexts, the justification of political decisions is necessarily in terms of an agreement grounded in subjective reasons.

To show that ORJ, the claim that political justification is necessarily in terms of objective reasons, is false and that PRJ* is correct, I need to make a stronger case than I have so far for the normative significance of agreement-based justification of political decisions. As a start, consider again the non-normative example that I have already used a few times. You justifiably believe p—that policy A is feasible—and permissibly support policy A on that basis. I justifiably believe not-p and permissibly support policy B on that basis. Suppose that p is true and A is the correct choice. For the time being, assume that you do not know that, however.29 What are the implications of this case for the political legitimacy of choosing A? First, note that finding ourselves in such a disagreement does not imply that choosing A cannot be justified to me. How can I permissibly reason that choosing A is justified? After all, I just said that the disagreement we find ourselves in is normatively significant in the sense that it results from the permissible reasoning of each of us. True. But that was our separate reasoning. After establishing the disagreement, we might turn to collective

Political Legitimacy Under Epistemic Constraints

163

deliberation to try to resolve it. I see two main ways this might happen (there might be others).30 First, you might testify that p and this might give me epistemic reason to adjust my belief that p in response to your testimony. Such reasons could include that you have a reputation for making reliable judgments in the p area, that you are in a more privileged epistemic position with regard to p than I am, or that the methods that you used to form your belief that p are generally trustworthy. If I have such epistemic reasons to accept your testimony, then I cannot justifiably hold on to my original belief. I might have sufficient reason to come around and form a belief that p. Alternatively, choosing A may become justified to me if you can explain to me why p. For example, you might be able to show me evidence that demonstrates conclusively that p. Again, if the force of epistemic reasons is on your side, you may be able to change my permissible reasoning, in which case I cannot justifiably stick to my original belief that not-p. While it is thus possible that A—the correct choice, I have assumed—can be justified to me, it might not be possible to do so. The problem arises because justifiably believing—or even knowing—that p neither entails being a good testifier nor that one is able to explain why p. Whether one is a good testifier depends on a range of factors that are only indirectly, or not all, related to p. And “why p” is a different proposition than “p.” So none of your reasons for believing that p (or even knowing that p) may be reasons that you could mobilize to explain to me “why p.” If neither condition is met, then I have no way of forming a justified belief that p and hence to conclude that we should choose A. Therefore, our disagreement that results from permissible reasoning persists and, I want to claim, remains normatively significant. I have assumed that we are in a symmetrical epistemic situation—we have both reasoned permissibly and although your belief that A is feasible happens to be correct and my belief that A is not feasible is not correct, you do not know that and you have not been able to give me good epistemic reasons for believing otherwise. Given that we both lack access to the relevant facts, there is no epistemic basis for privileging your belief and dismissing mine and thus for insisting that choosing A is justified. Under those circumstances,

164

Fabienne Peter

the fact that A is not justified to me undermines the political legitimacy of choosing A. Briefly consider what happens if we modify the example and assume that you do not only have a justified belief that p, but you know that p. In the example I have been using, suppose you know that A is feasible. Is that sufficient for the legitimacy of a political decision in favor of A? Although I cannot fully argue it here, I am inclined to say that, at least in a political context, knowledge without an ability to testify or to explain is not sufficient for practical justification. This point is based on the sufficiency objection to the knowledge norm of practical reasoning that I mentioned earlier. This objection says that sometimes even knowledge-level justification is not sufficient for permissible practical reasoning.31 The objection can be illustrated with the following example. Suppose a surgeon has been informed by her extremely competent and reliable team that the tumor she is about to remove is in the left kidney. On the basis of this testimony by her team, the surgeon knows that the tumor is in the left kidney. But we would probably consider her irresponsible—engaging in impermissible practical reasoning—if she did not double-check the evidence herself before operating. Similar circumstances regularly obtain in the political context, and what creates the extra pressure on justification is the interpersonal constellation. Consider the health example again and suppose that you know, not only that policy A will increase the health status of the poor, but also that this empirical fact favors implementing the policy. Suppose that I justifiably believe that it will not increase the health status of the poor or that, while it does, this is not a reason to implement it (because of countervailing reasons, for example that the policy is too expensive). I want to claim, deploying the sufficiency objection, that for me to be required to defer to your claim that we should implement A, it is not enough that you know what you do about A. Additional conditions have to be met, minimally conditions relating to your ability to testify or to explain to me that the policy has the empirical and/or normative properties that you claim. This sketch of an argument will not convince the hard-nosed objectivist about normative reasons. He will deny the sufficiency

Political Legitimacy Under Epistemic Constraints

165

objection and insist that knowledge of the objective reasons that apply is sufficient for practical justification, including the justification of political decisions. This, then, is a loose end I cannot argumentatively tie up here. Let me bracket this issue and continue with the assumption I made above, that justification in terms of objective reasons is not epistemically accessible to either of us: Neither of us has knowledge-level justification for the beliefs that drive our practical disagreement. This assumption echoes the epistemic skepticism that Rawls expressed in his essay on “The Independence of Moral Theory.”32 I agree with Rawls that our knowledge of normative facts is not as robust as our knowledge of non-normative facts. With science, we have a fairly reliable, though of course not infallible, route to factual knowledge. In the context of normative claims, we lack a similarly reliable methodology. Especially if one endorses some version of non-naturalism, there thus remains an epistemological challenge: How do we gain knowledge of normative facts? The proposals that defenders of OR have offered for how to deal with the epistemological challenge remain controversial.33 In addition, in policy evaluations that usually depend on a complex mix of normative and non-normative assessments, knowledge also very quickly comes under pressure. I thus do not think that a theory of practical justification that is suitable for the political context can be plausibly restricted to instances where we know all the relevant facts, normative and non-normative. My third and final argument against the Objective Reason Conception of political legitimacy is therefore this. If justification in terms of objective reasons is not epistemically accessible, then what we can justify to each other is all the justification we can get. Under those epistemic circumstances, if there is to be any justification for political decisions, it has to be justification in the sense of justification to, not of justification simpliciter. Disagreements that result from permissible practical reasoning can undermine political legitimacy. Conversely, a political decision is only legitimate, in those epistemic circumstances, if the decision can be justified to each of us. In those circumstances, agreement-based justification has to take the place of justification in terms of objective reasons. Let me illustrate the point once gain in terms of my example of a disagreement about non-normative facts. For A to be the

166

Fabienne Peter

legitimate choice under the circumstances I have assumed, choosing A must be justifiable to me. If our normatively significant disagreement persists and A cannot be justified to me, then choosing A is not legitimate. On the other hand, if choosing B cannot be justified to you, then choosing B is not legitimate under the circumstances I have assumed. The only policy choices that are legitimate under those circumstances are those that can be justified to both of us, i.e., those that are supported by an agreement grounded in subjective reasons. Granted, the choices we make in this fashion may not be justified in terms of objective reasons, were they accessible. But there is no other route to justification than via (some account of) subjective reasons. Under those epistemic circumstances, ORJ is thus false and PRJ* is correct. Objections Let me end this chapter by briefly considering some objections to the argument I have presented. A first objection is political. One might object that a conception of political legitimacy that requires agreement is too demanding. Such a condition of political legitimacy will identify most policy choices as illegitimate, as indeed it did in my example of a choice between policy A and policy B. In reply, I want to say something about how I understand the need to reach agreement through permissible practical reasoning. My examples throughout have highlighted the substantive level—which policy should we choose? I did this to keep things simple. I do not mean to imply that political legitimacy always requires agreement at the substantive level. I share the worry articulated in the objection that this would result in an overly minimal set of justified policies. The agreement required by PRJ*, by the weak version of the Public Reason Conception that I endorse, may be at the procedural level. An agreement at the procedural level is an agreement about how we should make decisions regarding controversial policy choices. In many democratic societies, for example, the view that democratic procedures help resolve controversial policy issues is still very widely held. Nothing in what I have said above precludes developing the Public Reason Conception of political legitimacy in such a procedural direction.

Political Legitimacy Under Epistemic Constraints

167

My argument can thus support the claim that in the face of normatively significant disagreements about policy issues, political legitimacy requires a democratic decision.34 The second objection I want to consider is conceptual. One might object that the epistemology of practical reasoning is the wrong starting point for a defense of a Public Reason Conception because the justification of political decisions—whether substantive or procedural—is a practical problem, not a theoretical problem. The problem is what we have reason to do, not what we have reason to believe. In reply, note, first, that the main focus of the epistemology of practical reasoning is not on what one should believe. Instead, it is on what appropriate practical reasoning requires. And the thought is that appropriate practical reasoning cannot ignore the question of what one should believe about alternative courses of action. There are, of course, practical contexts where what one should believe is largely irrelevant. For example, when you are deliberating about which ice cream flavor to buy, what you should believe about those flavors is probably not very important. What matters is which flavor you like best. I am happy to grant that even some political decisions may be of this kind. But most of our political decisions, and many of our private decisions, are not of this kind, or so I want to claim. When you are deliberating about which charity to give to, what you should believe about those charities is relevant. Your deliberation is criticizable if you rely on beliefs you should not hold. Likewise, your argument that we should not choose policy A because A is not feasible is not defensible if you should believe that A is feasible. And, finally, your argument that we should not choose policy A because the health status of the poor is not a political concern is not defensible if you should believe that the health status of the poor is a political concern. A third objection grants that the epistemology of practical reasoning is relevant for the justification of political decisions, but highlights that philosophers are just as likely to disagree about the correct epistemological theory as they are about the correct moral or political defense of a Public Reason Conception of political legitimacy. My approach thus does not offer progress. My argument is indeed vulnerable to this objection because I have made controversial claims about the epistemology of practical reasoning.

168

Fabienne Peter

I have claimed, for example, that the appropriate norm for practical reasoning is justified belief, not knowledge. If this premise is false, my argument is unsound. In reply to this objection, let me emphasize that my aim in this chapter has been somewhat narrower than the objection assumes. My aim has been to show that even if one accepts that there are objective reasons and that they matter for the justification of political legitimacy, it does not follow that they are the only reasons that matter. It does not follow because political justification often faces significant epistemic constraints, and these constraints can imply that no justification in terms of objective reasons is available. The only justification that is available in those circumstances is a justification based on public reasons. Some epistemologies of practical reasoning—strongly externalist ones—will indeed be incompatible with this line of argument. But it is at least conceivable that epistemologies other than the one I have adopted here will support an argument roughly along those lines. And that will be sufficient for my main claim to go through. In addition, exploring the epistemological assumptions supporting different conceptions of political legitimacy is, I believe, illuminating in its own right. A final objection to the view that I have developed in this chapter concerns its implications for the Objective Reason Conception. A defender of an Objective Reason Conception can grant my argument but maintain that a weaker version of ORJ can still be upheld. I have argued that we should reject the following claim: ORJ: The justification of political decisions is necessarily in terms of objective reasons.

The objection states that one could weaken ORJ in analogy to how I have weakened the Public Reason Conception. The weaker version of ORJ, call it ORJ*, states this: ORJ*: In some epistemic contexts, the justification of political decisions is necessarily in terms of objective reasons.

I agree that my argument is compatible with ORJ*. What is more, I would want to endorse ORJ*. If we have sufficiently robust knowledge of what political decision is objectively correct, I agree

Political Legitimacy Under Epistemic Constraints

169

that political legitimacy is undermined if the wrong decision is made. To see the plausibility of this, suppose a political regime makes decisions that are obviously morally impermissible. For example, suppose a regime issues laws that violate essential human rights of some minorities without any redeeming justification. Such decisions do, I want to maintain, lack legitimacy because they disregard moral knowledge that is sufficiently robust. While I accept ORJ*, it is important to recognize the limitations of building a conception of political legitimacy on ORJ*. My argument in this chapter is motivated by the question of what legitimacy requires when we lack sufficiently robust knowledge of what the correct political decision is. I believe this to be the normal case in politics. I have no disagreement with a defender of an Objective Reason Conception who accepts that if the epistemic circumstances limit reliance on objective reasons, political legitimacy will require that political decisions are justified in terms of public reasons. My disagreement is primarily with those defenders of an Objective Reason Conception who subscribe to ORJ (the strong version) and who thus claim that public reasons are irrelevant for the justification of political decisions. And I also want to press upon on all defenders of an Objective Reason Conception that the question of what legitimacy requires when epistemic circumstances limit reliance on objective reasons is in need of an answer. Acknowledgments I am most grateful for the detailed and insightful written comments on a previous version of this chapter that I have received from Jennifer Rubenstein and Micah Schwartzman. I have also greatly benefited from the comments of other participants at the NOMOS conference at the March 2017 Central APA meeting in Kansas City and from the comments I have received on previous versions of this chapter, including at the workshop on Factual Disagreements and Political Legitimacy at the University of Copenhagen, the New Directions in Public Reason conference at the University of Birmingham, the Pavia graduate conference in political philosophy, the workshop on Religion and Public Justification at University College London, the Philosophy, Politics, and Economics conference in honor of Luc Bovens at LSE, the

170

Fabienne Peter

Political Epistemology conference in London, and at research seminars at the University of Leiden, the ANU, and the University of Melbourne. I particularly want to thank Elizabeth Anderson, Ian Carter, Emanuela Ceva, Tony Coady, David Estlund, Elizabeth Erdenberg, James Fleming, Gerald Gaus, Dan Halliday, Klemens Kappel, Jack Knight, Cécile Laborde, Christian List, Steve Macedo, Jonathan Quong, Kristen Rundle, Debra Satz, Melissa Schwartzberg, Kai Spiekermann, Steve Wall, Han van  Wietmarschen, Jo Wolff, Andrew Valls, Daniel Viehoff, and Alex Voorhoeve, among many others. Notes 1. The concept of political legitimacy has been understood in many different ways, of course. I give an overview in Fabienne Peter, “Political Legitimacy,” Stanford Encyclopedia of Philosophy (Summer 2017 Edition), ed. Edward N. Zalta, https://plato.stanford.edu. To sidestep possible controversies on this issue, I take an ecumenical approach and understand political legitimacy in relation to the justification of political decisions. In addition, I do not, for the most part, distinguish between substantive and procedural conceptions of political legitimacy. A substantive conception targets the justification of political decisions directly; a procedural conception targets the justification of decision-making institutions. I thus intend to cover both when I speak of the justification of political decisions. 2. Note that my focus in this chapter is on conceptions of legitimacy, not on conceptions of justice. It is not my aim here to defend a public reason conception of justice. 3. See John Rawls, Political Liberalism (New York: Columbia University Press), and “Reply to Habermas,” Journal of Philosophy 92, no. 3 (1995): 132–180. 4. See, for example, Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986); David Enoch, “Authority and ReasonGiving,” Philosophy and Phenomenological Research 89, no. 2 (2014): 296–332; and Enoch, “Against Public Reason,” Oxford Studies in Political Philosophy, Volume 1 (Oxford: Oxford University Press, 2015), 112–142. 5. See, for example, Charles Larmore, The Autonomy of Morality (Cambridge: Cambridge University Press, 2008), and Jonathan Quong, Liberalism without Perfectionism (Oxford: Oxford University Press, 2011). 6. See Derek Parfit, On What Matters, Volumes 1 and 2 (Oxford: Oxford University Press, 2011), and Thomas Scanlon, Being Realistic about Reasons (Oxford: Oxford University Press, 2014), among others.

Political Legitimacy Under Epistemic Constraints

171

7. If one understands objective reasons as a subset of all normative practical reasons, one is committed to a hybrid view of practical normativity. Such views are far less popular than the monist views. Ruth Chang, “Grounding Practical Normativity: Going Hybrid,” Philosophical Studies 164, no. 1 (2013): 163–187 defends a hybrid view. I defend a hybrid view on the grounds of political legitimacy in Fabienne Peter, The Grounds of Legitimacy, manuscript. 8. If the strong version of OR were true, ORJ would directly fall out of that—if all normative practical reasons are objective, there are no other reasons that could play a role in the justification of political decisions. But ORJ can be made compatible with the weak version of OR as well, and the combined claim is that even if there are normative practical reasons that are not objective reasons, only objective reasons can justify political decisions. An additional argument would be necessary to show why only those reasons can justify political decisions. 9. Raz, The Morality of Freedom. 10. Enoch, “Against Public Reason,” 130ff. 11. Parfit, On What Matters. 12. This is a very loose characterization, compatible with both the understanding that a subjective reason is a belief that one has an objective reason and with the counterfactual understanding that a subjective reason is a belief that, if true, would be an objective reason. On this topic, see also Daniel Whiting, “Keep Things in Perspective: Reasons, Rationality, and the A Priori,” Journal of Ethics and Social Philosophy 8 (2014): 1–22. 13. They define subjective reasons as dependent on attitudes such as desires, for example, or on procedural factors such as endorsement. 14. To keep the presentation simple, I will mainly focus on a substantive interpretation of public reasons, even though this is not actually the interpretation that I believe we should adopt—see my “Epistemic Foundations of Political Liberalism,” Journal of Moral Philosophy 10, no. 5 (2013): 598–620. I will return to this issue in the last section of this chapter. 15. Quong, Liberalism without Perfectionism. 16. Gerald Gaus, The Order of Public Reason (Cambridge: Cambridge University Press, 2011), 246. 17. For a discussion of the two accounts, see Kevin Vallier, “Convergence and Consensus in Public Reason,” Public Affairs Quarterly 25 no. 4 (2011): 261–280. A related way to put the point is that my characterization of the Public Reason Conception leaves open whether the aggregation of individual judgments about which a political decision is justified should be premise-based or conclusion-based. According to the consensus conception, it is the former; according to the convergence conception, it is the latter. See Luc Bovens and Wlodek Rabinowicz, “Democracy and

172

Fabienne Peter

Argument—Tracking Truth in Complex Social Decisions,” in Deliberation and Decision, ed. Anne van Aaken, Christian List, and Christoph Lütge (Aldershot: Ashgate, 2003), 143–157; and Christian List, “The Discursive Dilemma and Public Reason,” Ethics 116 no. 2 (2006): 362–402. 18. As I will explain below, I will not defend PRJ, but a weaker principle instead. 19. See Timothy Williamson, Knowledge and Its Limits (Oxford: Oxford University Press, 2000), 257. 20. Williamson originally only formulated the necessity claim: It is permissible for you to assert that p only if you know that p. The tendency has since been to focus on the defense of the stronger, biconditional, claim KA, which combines the necessity claim with a sufficiency claim: knowing that p is sufficient for permissibly asserting that p. 21. On this, see, among others, Jessica Brown, “Knowledge and Practical Reason,” Philosophy Compass 3 no. 6 (2008): 1135–1152; Jason Stanley and John Hawthorne, “Knowledge and Action,” Journal of Philosophy105, no. 10 (2008): 571–590; and Mikkel Gerken, “Warrant and Action,” Synthese 178, no. 3 (2011): 529–547. 22. In what follows, I will use the feasibility example more than normative examples, as it is simpler. 23. See, e.g., Jennifer Lackey, “Norms of Assertion,” Noûs 41 (2007): 595–626, and Brown, “Knowledge and Practical Reason.” 24. See also Lackey, “Norms of Assertion,” and Ralph Wedgwood, “Contextualism About Justified Belief,” Philosophers’ Imprint 8, no. 9 (2008): 1– 20. 25. See James Dreier, “Relativism (and Expressivism) and the Problem of Disagreement,” Philosophical Perspectives 23, no. 1 (2009): 79–110. 26. Conflicting perceptions of color give rise to a related example, taken from a recent Internet hype. Individually, we cannot both hold that a particular dress looks white/gold and that it looks blue/black. But different individuals might well see color differently. On the more demanding test, if two individuals see the dress as having a different color, they have a disagreement about color. On the less demanding test, they do not have a disagreement. 27. One might object at this stage that this move implies that my argument only works for those interpretations of the Public Reason Conception that accommodate fact-dependent normative disagreements. I understand Gaus’s interpretation of the Public Reason Conception that he puts forward in Gerald Gaus, The Order of Public Reason, for example, to be in terms of disagreements that are not fact-dependent. According to Gaus, a normative order is justified if it is supported by everyone’s preferences, suitably understood, for rules of social morality. But those who

Political Legitimacy Under Epistemic Constraints

173

defend a version of the Public Reason Conception of this kind could accept a hypothetical version of the argument that I intend to present: ORJ is false even if one accepts the weak version of OR. This hypothetical version leaves open whether one should accept OR or whether one should construe normative practical reasons in the political context on different premises altogether. 28. This holds at least as long as epistemic justification is non-factive; see Clayton Littlejohn, Justification and the Truth-Connection (Cambridge: Cambridge University Press, 2012) for an opposing view. 29. You might object that if you have a justified belief that p and p is true, you have knowledge. But just think of Gettier-style examples. P might be true even though the reasons for which you happen to believe p have nothing to do with the facts that make p true, so you do not know p. 30. A related question, one that I cannot address here, is whether we might have reasons to adjust our beliefs in response to the fact of disagreement. A version of this question is the focus of the literature on the epistemology of disagreement. I have discussed the implications of the epistemology of disagreement for political legitimacy elsewhere, in my “Epistemic Foundations of Political Liberalism” and in “The Epistemic Circumstances of Democracy,” in The Epistemic Life of Groups, ed. Michael Brady and Miranda Fricker (Oxford: Oxford University Press, 2016), 133– 149. 31. For example, Brown, “Knowledge and Practical Reason.” 32. John Rawls, “The Independence of Moral Theory,” Proceedings and Addresses of the American Philosophical Association (1974): 5–22. 33. A prominent suggestion in this regard is that the epistemological challenge can be met via a reflective equilibrium approach. I believe that this strategy fails for reasons similar to those articulated by Thomas Kelly and Sarah McGrath, “Is Reflective Equilibrium Enough?” Philosophical Perspectives 24, no. 1 (2010): 325–359. 34. In Democratic Legitimacy (New York: Routledge, 2009), I defend a procedural interpretation of public reason, and in “The Epistemic Circumstances of Democracy,” in The Epistemic Life of Groups, ed. Michael Brady and Miranda Fricker (Oxford: Oxford University Press, 2016), 133– 149, I provide a negative epistemic argument for democracy.

6 LEGITIMACY AS A RIGHT TO ERR DANIEL VIEHOFF

The legitimacy of political and legal institutions—by which I mean, their normative legitimacy, rather than their descriptive or legal legitimacy—has been analyzed in a number of different ways in recent philosophical discussions.1 One influential approach focuses on these institutions’ claim that they have the authority to make morally binding, or perhaps coercively enforceable, decisions for their subjects, and takes legitimate institutions to be those that have the authority they claim to possess.2 Another asks whether an institution’s exercise of power (coercive or other) over its subjects makes the relationship between institution and subject morally problematic, and takes legitimate institutions to be those whose relationships to their subjects are morally acceptable.3 This chapter proposes a third approach which, though it shares features with both of the more prominent views just mentioned, is importantly different from each.4 Legitimacy, on the proposal I sketch here, is centrally a right to err, or to make mistakes that set back interests of others that are ordinarily protected by rights. This may sound similar to a point commonly emphasized by those who focus on an institution’s authority to impose binding or enforceable duties: The moral power to make the decision, they say, is effective even where the decision to impose the duty rests on a mistake. But legitimate institutions have a right to err that goes in important ways beyond authority so understood. In particular, reflection on our moral and legal practices reveals that institutions wielding legitimate authority are not merely empowered to impose duties, or lay down enforceable rules, even when the decision to do so is mistaken. They are also insulated from compensatory demands for 174

Legitimacy as a Right to Err

175

making substantively unjust decisions, at least as long as these decisions are made in the right way (where making decisions in this way is in turn a condition for the legitimacy of the institution’s exercise of its power). More specifically, I suggest that legitimate institutions have a distinctive liberty right to set back others’ interests that other agents normally lack; and these institutions’ subjects in turn lack certain permissions to avoid, or redirect, the costs of the institutions’ mistakes in ways that would be permitted if similar costs were imposed by agents who lack legitimacy. Legitimate institutions have this special liberty right because, and insofar as, they act for their subject (in a specific sense) and do so only for the subject’s sake. As a matter of fairness, (some of) the costs of the institutions’ actions ought to be borne by the subjects for whom they are undertaken, rather than by the institutions (or the officials acting on their behalf). In turn, where an institution fails to act for the subject in the relevant way, that institution (and those acting for it) may have to bear the costs of its errors, which the subject may thus be morally permitted to redirect by suitable acts of resistance. Legitimacy and Error That there is a significant link between legitimacy and error is suggested by reflection on the moral rights of officials and others (such as parents or teachers) who exercise what we think of as legitimate power. To highlight this link, I discuss in this section a number of cases in which someone exercising power has a right to err of the sort I suggest is central to legitimacy. I begin with the relatively straightforward case of a judge adjudicating a civil suit, before I discuss the somewhat less perspicuous case of legislative authority. (The points I make here in principle also extend to executive action; but since executive action also gives rises to special problems, I won’t discuss it in detail in this chapter.) Let me also flag that the following discussion is meant to capture what I take to be quite general moral features of the kinds of institutions I am describing. Each legal system will have its own way of institutionalizing some, or all, of these features, and reflection on how certain familiar legal systems have institutionalized them may occasionally be helpful for sharpening one’s understanding of the general moral phenomenon. Yet my concern is not with these

176

Daniel Viehoff

(often quite parochial, and highly context-dependent) ways in which the general moral norms are legally implemented (or not, for that matter), and so I will say nothing about corresponding doctrinal matters. Imagine Judge finds against Defendant in a tort suit, and holds that Defendant owes Plaintiff some amount (say, $1,000) in compensatory damages. Often enough, this is sufficient to motivate the parties to pay up (assuming they can). If they don’t, there is an elaborate system of enforcement procedures that ultimately lead to an executive official (a sheriff, for example) taking possession of Defendant’s property so as to ensure the judgment is satisfied. Let me begin with two features of this case that are familiar enough from philosophical discussion. First, there are right or wrong answers to the questions that Judge must answer in adjudicating the matter: Did Defendant indeed do what would make her legally liable? (Was she acting negligently, or otherwise tortuously? Did her tortuous action cause harm to Plaintiff?) And what remedy is Plaintiff entitled to? (How great a harm did Plaintiff suffer?) Etc. Second, legal institutions (and those speaking for them) take it for granted that Judge’s legitimate exercise of her adjudicatory power is not conditional on the correctness of her answer to these questions, or of the decision she ultimately makes (though it may be conditional on various other considerations). Her decision stands even if she made a substantive mistake, and Plaintiff ought not to have been awarded $1,000 in damages. This is in principle true whether the mistake was one of law or of fact.5 We must not be misled here by observations about the grounds on which decisions by a trial court may be appealed in familiar legal systems. For if the court of final appeals ends up siding with the trial court, the appeal court’s decision too is considered binding (or otherwise normatively effective) even if it is mistaken. Once a final court decision to that effect has been made, Defendant is obligated to pay $1,000, even if the decision was in fact substantively incorrect; and if Defendant does not pay, then various officials may be permitted to take steps, including coercive measures, to ensure Plaintiff receives the awarded damages, even if this means that Defendant ends up losing $1,000 (or more) that, but for Judge’s mistaken decision, she would be entitled to.

Legitimacy as a Right to Err

177

All of this, I said, is assumed by our legal system and the officials speaking for it. But it is not only officials who take this view. Empirical research into citizens’ attitudes toward legal-political institutions indicates that people generally grant such institutions significant leeway for error when they see these institutions as legitimate.6 (Indeed, it is this willingness of ordinary citizens to go along with decisions even if they disagree with them on the substance that makes what we may call “empirical legitimacy”—the citizens’ belief that their governing institutions possess the moral right or status we associate with legitimate authority—so valuable to legal and political institutions: Empirical legitimacy enables them to pursue certain ends even in the face of substantive disagreement about the end’s choice-worthiness.) In line with many philosophers, I will call the court’s capacity to create duties or permissions at will, simply by making and communicating a decision to create them, its “normative power.” A similar power is also claimed by various other institutions, such as legislatures and administrative agencies. There are a number of interesting questions that can be asked about the normative power claimed, or held, by political and legal institutions. (Is it a normative power to impose duties? Or to create permissions for officials to enforce the institution’s decision? What, if anything, distinguishes this power from other ways of bringing it about that such duties, or permissions, exist? And so on.7) But these questions will not be the focus of the following discussion. For, though their claim to possess such normative power is indeed important for understanding central legal and political institutions that claim legitimacy, such institutions may possess this normative power and yet lack legitimacy as a right to err. And conversely, an institution or agent could in principle have a right to err, without having or exercising such normative power. Finally, leaving room for an understanding of legitimacy as a right to err is important for grasping central features of our legal and political practices. Or so I want to suggest now. Bearers of legitimate authority have, I think further reflection reveals, a right to err that is not exhausted by the normative power to impose duties (or create permissions to enforce) just mentioned. Consider again a court (including a court of appeals) that makes a mistaken decision. Grant that Judge’s verdict puts Defendant

178

Daniel Viehoff

under a moral duty to pay, or that, in case of non-payment, some official may permissibly enforce the judgment against Defendant, by executing against Defendant’s wages, bank accounts, or personal property. Judge, in other words, makes a decision with the intended effect that $1,000 of Defendant’s money goes to Plaintiff, money which Defendant in fact ought not to be required to pay to Plaintiff (though Judge believes otherwise). In many circumstances, if A gives an order, or makes a decision, with the intention to bring it about that another person B loses control over $1,000, money to which B is in fact entitled (or, if one cares to put the matter in a more neutral fashion, “money which B ought not to lose control over in this way”), then B would have a compensatory claim against A, even if A in good faith believed that B lacked the entitlement. Imagine you don’t know that A is a judge. And imagine you hear that A orders B to hand over to C some property to which B would otherwise be morally and legally entitled, and that if B doesn’t do so, D will come and make B do it against her will. A normal reaction would be to think that A is somehow on the hook, morally (and, one might hope, legally), for the loss in property that B suffers as a result of A’s order. And yet Judge incurs no compensatory liability for mistaken decisions—decisions that require Defendant to pay money to Plaintiff though Defendant ought not to have been required to do so—if those decisions are made in the right way, pursuant to Judge’s legitimate authority over the parties. Though Judge made a mistake and is in a straightforward sense responsible for Defendant’s loss, Defendant has no moral right that Judge compensate her for the loss she suffered. Nor does Defendant in this case have a compensatory claim against the community in whose name the court speaks.8 To be clear, all of this is true only if the court’s decision satisfied certain important procedural requirements (about which more later). But, crucially for our purposes, it is true even if, though the procedural requirements were satisfied, the court’s decision was substantively unjust. The feature just highlighted is not peculiar to judicial authority. Consider next legislative decision-making. Imagine, for instance, that a legislature, in the exercise of its legitimate authority, passes a law that imposes a tax on the sale of particular goods. As a result, various people’s interests are set back: Some buyers end up still buying the good, but pay more than they otherwise would. Some

Legitimacy as a Right to Err

179

buyers end up not buying the good, though they would have benefited from buying it at the price it would have cost sans tax. And some sellers or producers end up selling less, and earning less, because the would-be buyers in the previous category abstain from purchasing the good. All of this may be justifiable: The benefits of raising taxes in this way may be great enough, and the distribution of burdens fair enough, to make adopting this tax objectively justified. (In saying this, I am assuming that, at least within certain limits, there is a right or wrong answer to the question whether the tax scheme the legislature adopts is fair, just, efficient, and thus ought, or ought not, to be adopted.) But at least sometimes legislatures get the matter wrong: Taxes are imposed that create greater burdens than benefits, or distribute the burdens and benefits unfairly among the members of the community. Just as in the case of the court, so in the case of the legislature, the decision to impose a duty to pay taxes, or to create permissions to coercively enforce tax rules, is often normatively effective even if it is mistaken. And, importantly for the purposes of this chapter, as long as the legislature adheres to certain procedural standards, the legislature’s mistake does not give rise to compensatory claims by those on whom it inflicts losses: Even if the tax rules it adopts are in fact unfair, imposing an undue burden on some citizens, the legislators making the decision are not required to somehow make whole those who are unfairly deprived of some benefits, or unfairly suffer particular burdens.9 And, I want to suggest once again, neither do the losers have straightforward claims to compensation against the community in whose name the legislature speaks. What makes it especially easy to overlook this feature of legitimate legislative authority is that the legislature is under a moral duty to consider the distributive effects of its policies on the community as a whole, and to take into account these considerations in making policy choices. Whereas a court (for good reason) usually loses jurisdiction over the parties once the dispute has been disposed of, the legislature’s authority over citizens is ongoing. A citizen can thus sensibly raise a complaint against a particular policy’s distributive effects and demand that future policies make up for it; and if the legislature recognizes these distributive effects and yet fails—now knowingly—to consider them when making future policy, then the legislature has indeed failed to exercise its law-making

180

Daniel Viehoff

power legitimately. But, crucially, this does not mean that the legislature committed a compensable wrong by making the mistake in the first place. (Nor, for that matter, would it commit such a wrong by erroneously but in good faith concluding that the citizen’s complaint about the past distributive effects is groundless.) It merely made a mistake that should figure in its future decisions about how to fairly distribute the burdens and benefits of collective policies among the citizens over time; but if it fails to correct its past mistake, and yet satisfies the various procedural conditions that make its decisions legitimate, then the decision it makes does not entitle the victim to compensation. It is tempting but misguided to think of this denial of a right to compensation as a merely institutional (or legal) claim: “From the point of view of the institution, which fails to recognize that it made a mistake, the citizen is not entitled to compensation. But this does not bear on the moral issue of whether the citizen is in fact owed compensation.” If this were all that denying the right to compensation amounted to, then the victim would be morally (though not institutionally) permitted to take steps to seek compensation from the institution and those acting for it. But then there would be nothing distinctive about legitimacy, for legitimate as well as illegitimate decision-makers purport to get the matter right and will adhere to their own assessment of the situation. If the institution is indeed legitimate, then it is not just the case that the institution claims to be getting the matter right. It is also the case that, when its legitimate decision is nonetheless wrong, the victim may not seek to unilaterally recover the cost of the mistake from the institution or its agents, or otherwise see to it that they fall on them rather than her. In practice, this means, for example, that the victim may not—morally, and not just legally—pursue selfhelp, either by evading certain costly demands that the institution makes on it or by redirecting the costs of the institution’s mistakes on officials by, e.g., resisting them when they come to enforce the legitimate decisions of the legislature or court. Authority and Legitimacy Someone may accept that those exercising authority do indeed escape compensatory obligations for the costs their legitimate

Legitimacy as a Right to Err

181

yet mistaken decisions create, but deny that this requires us to go beyond a conception of legitimate authority as a normative power. Protection against such liability, they may say, is already built into the very idea that the institution has the normative power to change the subject’s rights and duties. In this section I argue that assimilating the right to err to authority understood as a normative power would be a mistake. Someone may possess the kind of normative power associated with authority and yet lack the right to err that is central to legitimate institutions (and vice versa). As a slogan: Justified authority, the normative power to change another’s situation at will, does not amount to legitimate authority, a fullblown right to err when exercising one’s normative power.10 The conceptual separability of a normative power and a right to err is obvious enough. One can easily imagine a contractual relation between A and B in which B promises to obey A’s directives on certain matters in exchange for A’s promise to compensate B for any costs that such obedience imposes on her. So A has normative power over B, and yet has no right not to compensate B for the costs that exercising his normative power imposes on her. Whether this conceptual possibility is realized depends, however, on how the authority relation is justified. Someone therefore may suggest that, though a conceptual gap between justified and legitimate authority exists, practically they go hand in hand, because standard justifications of political and legal authority establish both an institution’s authority and its full-blown right to err. But this too is mistaken, as reflection on an influential account of political authority reveals. On what is perhaps the most common view among those who do not insist on consent as a necessary condition of legitimate authority, bearers of political authority have the capacity to impose duties on others at will because their having and exercising such power enables the resolution of certain morally pressing problems, most obviously perhaps collective problems that require large-scale coordination among the members of our community.11 (Such arguments could underpin views on which political authority involves a normative power to impose duties on the subjects, as well as those on which it involves a power to create permissions for—and perhaps simultaneously duties on—officials to enforce certain rules against the subjects. For simplicity’s sake I focus here on the former.) Crucially, however, establishing

182

Daniel Viehoff

that a decision-maker has the relevant power grounded in the importance of coordination is not sufficient to establish that the decision-maker is insulated from bearing the costs her directives impose on the subjects. On a coordination-based view, what justifies the decisionmaker’s power over the subjects is the need to resolve an urgent problem that requires coordination; such coordination can only, or best, be achieved if the subjects obey the decision-maker’s directives; and the problem is urgent enough that the subjects are morally required to obey the directives despite the costs that this requirement imposes on them. But that the urgency of the problem is sufficient to justify imposing the cost of obedience on the subjects does not establish that this cost should be left lying with them if it could instead be redistributed. And whether it should be redistributed depends crucially on the origins of the problem that needs solving. If the decision-maker is also culpably responsible for creating the problem in the first place, then the decisionmaker may both possess the power that solves the problem and bear liability for the costs that solving the problem imposes on others. This would, crucially, include the costs that the subjects incur when they obey the decision-maker’s directives. To make this more concrete, consider an example: A throws child C into the pond, and E, F, and G are bystanders. Realizing that his crime has been observed, A seeks to mitigate future punishment by now trying to rescue C. But rescuing C in fact requires cooperation among a number of people, who must act in a highly coordinated fashion. For some reason or another, among those present, only A knows what it takes to save C, and the only way in which E, F, and G can do their part in saving C is by following A’s directives about what to do. Now if A directs E to do x, and F to do x*, and G to do x**, so that they jointly bring about C’s rescue; if following A’s directives is the only way of rescuing C open to E, F, and G; and if following A’s directives isn’t overly burdensome on them; then E, F, and G have duties to follow A’s rescue-related directives. Crucially, E, F, and G may have to follow A’s directives even if A makes a mistake and gives them directives that are distinctly suboptimal—directives, for instance, that make saving C harder, and more costly, for E, F, and G—as long as A’s not giving any directives at all, or E, F, or G’s not following A’s directives,

Legitimacy as a Right to Err

183

would be worse still. But then A is very plausibly thought to have authority—the power to impose duties on the subjects at will— over E, F, and G.12 And yet A is clearly under an obligation to compensate E, F, and G for the costs they incur when discharging their duty to rescue C, because A is responsible for creating the need to save C in the first place. And if A makes a mistake that increases the cost of rescuing C, then that additional cost also falls on A. So even though he has justified authority over others, he pays for his own mistakes when exercising his normative power. This example highlights that even if authority is justified by appeal to coordination, a right to err of the sort held by fully legitimate authorities can be absent. But the example may seem farfetched and of limited importance for thinking about political and legal institutions: Even bad legal and political authorities, one may think, aren’t usually responsible for the creation of the very problem they are now tasked with solving. I am somewhat less sanguine that political and legal institutions are not complicit in the creation of some of the problems that they are also empowered to solve. But setting that aside, it is important to recognize that the right to err is also absent in other scenarios—including, crucially, scenarios where the issue is not that the decision-maker culpably created the problem that needs solving, but that the decisionmaker gives a directive that is mistaken in a particular way. The kind of mistake at issue in these cases does, I want to suggest, very plausibly arise in legal and political institutions on a regular basis; and so, such institutions too may possess authority and yet lack a right to err. Consider the following example: C has once again fallen into the water, though this time the fault is C’s own (rather than A’s). The only way to save C is once again for A to give binding directives to E, F, and G. If A directs them to do x, x*, and x**, then E, F, and G will each bear some cost c, and A will bear some significantly smaller cost c−1. (For instance, each of E, F, and G have to wade into the water up to their thighs, and will ruin their shoes and trousers. A will have to get into the water just enough to ruin his shoes.) If A instead directs them to do y, y*, and y**, then E, F, and G bear a greater cost c+1, and A will bear no cost at all. (Each of E, F, and G now have to get into the water up to their neck, thus ruining all their clothes; but A need not get into the water all.)

184

Daniel Viehoff

The likelihood of saving the child, C, is the same in either case. Thus, a fair distribution of benefits and burdens would require the adoption of policy x, x*, and x**. Imagine all of this is known to A. Yet to avoid cost c−1 to himself, A chooses the second strategy, and directs E, F, and G to do y, y*, and y**, at cost c+1 to each of them. For the sake of saving C, they must obey A’s directive. And yet they should not have to bear the cost of A’s choice to pursue this policy, a choice A made for the sake of minimizing the costs to himself even if this meant imposing a greater, and unfair, cost on others. So, A ought to compensate them for his unjust directives. (And even though A would often be entitled to compensation from C for the costs associated with saving C from the emergency C culpably created, A would have no such entitlement here, at least with regard to the excess cost he created beyond what he himself reasonably recognized was necessary to save C.) Matters would be quite different if A had directed E, F, and G to adopt the more costly course of action (y, y*, y**) because A reasonably but mistakenly believed that this course would be safer for C. In that case, A would have made an error—objectively, it would have been better if A had chosen the less costly, and equally safe, course of action (x, x*, x**). And yet A would not be responsible for bearing the cost of his good-faith mistake (which would instead properly fall on C, who culpably created the problem that A tried to help solve). So whether A has normative power over another even where A’s directives are mistaken is separable, not just conceptually, but even in a wide range of practically relevant cases, from the further question whether A must bear the costs of any mistakes he makes when giving directives to his subjects. The examples discussed here also shed light on why this is so: The considerations that determine whether A has authority diverge in important respects from those that bear on whether or not A must bear the costs his mistakes create, and thus whether A has the full-blown right to err that we attribute to legitimate authorities. The answer to the first question—whether A has justified authority over B—depends on how well (or badly) B would do, with regard to the urgent moral problem at hand, if A’s directives were absent or were not obeyed by B. By contrast, whether A

Legitimacy as a Right to Err

185

must bear the costs of B’s following these directives—whether A has a full-blown right to err, or (for short) whether A’s possession and exercise of authority are legitimate—depends on other considerations, including who could reasonably have avoided, and been expected to avoid, making choices that created the costs in the first place, and what a fair distribution of these costs consequently amounts to. It is, finally, worth pointing out that sometimes political institutions exercise power, not by giving authoritative directives, nor even by making coercive threats, but by directly changing the world around us in certain ways, or by providing services—other than the service of giving useful directives—to us. For instance, they build roads for us, or set up rescue services. These exercises of power too involve claims to legitimacy and may be covered by a right to err. Still, given the centrality of authority relations to our legal and political institutions, it is on power-as-authority that I focus throughout this chapter. Giving Content to the Right to Err A legitimate authority, it is sometimes said, has a right to rule.13 Philosophical discussions of this right to rule often focus on the authority’s normative power: whether it is the power to impose duties, or just that to create permissions to enforce; whether the duties created are owed to the decision-maker, if they are owed to anyone; etc.14 But the discussion up to now suggests that, even if we focus on institutions that act by making purportedly binding or enforceable decisions, the right to rule they claim to possess is not exhausted by a normative power. It also includes a right to err when exercising this power. But what does this right to err precisely consist in? What (to put the matter in the language of legal philosophers) is the proper Hohfeldian analysis of the moral relation (or relations) that make up this right? On one interpretation, the right to err is, first and foremost, a claim right that others not interfere with the authority’s making and implementing its decisions even if these decisions erroneously burden the subject. This could, among other things, include a claim right against being forced to compensate the victim of one’s mistake, or even against being made to account for

186

Daniel Viehoff

one’s mistake before another authority, such as a court. (In practice, this would then usually be institutionalized as an immunity to another’s legal claims. But the immunity would derive from the underlying claim right, rather than being self-standing.) On another interpretation, the legitimate authority’s right to err is, centrally, a liberty right (or what is sometimes also called a Hohfeldian privilege, or moral permission) that others lack. This liberty right would in turn explain why a legitimate authority has various claim rights against outside interference. Agents are ordinarily protected (at least within certain limits) by claim rights against certain forms of interference with their actions, though that protection falls away (frequently, if not always) when they violate moral requirements that apply to them. Because a bearer of legitimate power may permissibly act in ways that would be morally impermissible if done by others, certain actions she undertakes are also protected by claim rights that would not protect others undertaking the same actions. The legitimate authority’s special claim rights thus derive from special liberty rights, rather than being explanatorily independent or self-standing. I think this second understanding of the right to err is preferable. The first misleadingly assimilates a legitimate authority’s right to err to what Jeremy Waldron famously called a “right to do wrong”: a claim right against outside interference with one’s action even where that action violates a moral duty one has.15 We have, for instance, a right to do wrong (in Waldron’s sense) where our intentional or negligent promulgation of racist stereotypes is protected by our freedom of speech: we have a moral duty not to intentionally or negligently promulgate such stereotypes, and may be blamed and criticized for it, and yet may have a right that others not interfere with our speech so as to enforce our compliance with the duty. But the right to err held by those exercising legitimate power is not best understood as a “right to do wrong” of this sort. The errors that the right to err covers are, even though morally regrettable, not ones for which we can morally criticize or blame the person making them. If they were acting legitimately, then the judge and legislators discussed earlier did what they could reasonably be expected, or asked, to do. Indeed, they often did what they were required to do. And yet their doing so ultimately led to a morally suboptimal

Legitimacy as a Right to Err

187

outcome. They committed, I think we should say, an error that set back another’s interest-protected-by-a-right, and yet they did not wrong the victim, because the right that ordinarily protects the victim’s interest was not a right held against judges or legislators when they are engaged in a legitimate exercise of their power. (Our moral response to the error is similar, in some respects, to how we treat moral wrongdoing for which the agent has an excuse. But we must not overemphasize this similarity. Most important, excuses do not usually insulate us from compensatory obligations: I may be excused, or even justified, in taking your car without your consent if I need it to drive my wife to the hospital in an emergency. Yet I still owe you compensation for my non-consensual use of your property. So the sole purpose of highlighting the similarity to excuses is to resist the assimilation of a legitimate authority’s right to err to the “right to do wrong” Waldron analyses.) Focusing now on the right to err understood in this second way, what more can we say about it? Two points are worth highlighting. First, we can easily explain the legitimate authority’s insulation from compensatory demands for (some of) its mistakes: Bearers of legitimate power have an underlying liberty right which undercuts such compensatory claims, rather than a self-standing claim right against being forced to compensate victims who nonetheless retain a moral right to compensation. (This is not to deny that there are also cases where the latter is the more plausible explanation. Certain instances of sovereign immunity are best understood as involving, not the genuine absence of an obligation to compensate the victims of one’s errors, but instead a claim right against enforcement, or immunity against legal proceedings, grounded in institutional considerations about the separation of powers. Immunity so justified seems to me to be sufficiently different from the “right to err” that I set it aside here.) Second, we can say more about the precise content of the liberty right. In a related context, Thomas Christiano has suggested that legitimate authorities have a liberty right to “make decisions as [the authority] sees fit.”16 Yet this does not seem to capture how we think about standard instances of political and legal authority. Just consider again a court exercising its legitimate authority over the parties before it. Clearly there are limits to the considerations on the basis of which the judge may make a decision. And though

188

Daniel Viehoff

matters are once again less perspicuous in the case of legislative authority, the same is true there. Where, for instance, a democratic legislature fails to even treat as a reason for or against a policy its effects on the interests of certain of its subjects, or fails to make an adequate effort to get these reasons right, and its erroneous policy decision harms these subjects, the decision lacks moral insulation from these subjects’ compensatory demands. A legitimate authority’s right to err is, I want to suggest, best understood as constituted by the absence of a duty to get the matter right, but the presence of a duty to take due care to get the matter right, and to make the decision based on the right considerations. Consequently, a bearer of legitimate authority is, in an important sense, not less restricted than someone who acts without legitimacy; she is, rather, restricted in different ways, subject to different moral requirements. In particular, she is usually under various constraints about how to deliberate, and on the basis of what reasons to act, that do not apply to others; and it is the satisfaction of these constraints that insulates her from the alternative duties that those others are under. Whence the Right to Err? Can we say more about the bases of legitimacy understood as a right to err? How can we explain that someone has a liberty right to engage in pursuits even if she sets back interests of others ordinarily protected by (claim) rights, but is in turn under a deliberative constraint to act on the basis of certain considerations only? In this section I sketch what I think is the most plausible answer (without, however, being able to defend the proposal in anything like the detail that the issue deserves). I begin by saying something about the kinds of cases with which I have (mostly implicitly) contrasted the actions of legitimate authorities: cases, familiar from our everyday interactions with other private citizens, in which agents must generally bear the costs of their own mistakes. Only then will I turn to cases—if not less familiar, then at least much less discussed among philosophers—of agents who are insulated from such costs. Before I begin the discussion proper, let me note that I will say nothing here about consent, even though it is one basis on which an agent can avoid compensatory obligations for costs her

Legitimacy as a Right to Err

189

actions impose on another. Our concern is ultimately with explaining how legitimate political and legal institutions (and those acting for them) can have a right to err; and all the familiar arguments against consent-based justifications of political authority apply equally to justifications of a political institution’s right to err. On a familiar view of central moral norms governing our interactions (and, especially, our right to compensation where interactions go wrong), these norms strike a balance between the interests each agent has in pursuing her own projects and aims, and the interests others have in not being inhibited in their own, similar, pursuits. Put somewhat grandly, these norms enable us to live alongside each other as free and equal persons. Let me offer some examples to make this thought more concrete. As an embodied agent, most (if not all) of my pursuits require use of my body. So I have a corresponding interest in controlling my body and what happens to it. This interest explains why others must generally not touch or use my body without my consent. “Generally” only because the prohibition might be lifted where allowing another to touch or use my body without consent would impose a small burden on me, while disallowing it would impose a great burden on them. (If you are about to be crushed by a boulder, and the only way to escape it is to jump out of its path and onto my foot, you may do so even if without my consent.17) But even then, it will often be the case that, though what you did was justified, you must compensate me for the setback to my interests that I suffer as a result. (Your bruising my foot was justified, yet you still have to pay for my visit to the doctor.) Similarly, since the pursuit of our projects requires the use of external resources, my corresponding interest in using, and more generally controlling the use of, objects in the world explains why I have property rights; and others’ interests in pursuing their own projects explain why my property rights must be limited (including by their property rights). Normally others must not use my property without my consent; but sometimes they are permitted to do so (say, in case of an emergency). Yet when they do, they will usually have to compensate me for the use they made of what is properly mine. And (to mention one last example) my interest in pursuing my own projects, without being overly dependent on the projects of others, explains why I am permitted to take actions that impose

190

Daniel Viehoff

some not unreasonable risks on others. But their interests in not being overly dependent on my choices explain why I must not take unreasonable risks, and negligently impose harm on them. When someone does engage in action that violates these duties, and thus oversteps the boundaries that strike a fair balance between her interests and those of others, then she must compensate those she has wronged, and put them back in the position to which they were entitled as a matter of justice or fairness. Much more would have to be said to fill in this (very partial) picture of some of the central moral norms governing our interactions. Still, I think what I have said may suffice to motivate two thoughts. First, it is plausible that the norms this particular account explains normally regulate only what we do, not why (for what reason) we do it. This reflects the asymmetric interests that the different parties have with regard to an agent’s permission to act. The agent herself has an interest in freely pursuing her own aims and projects. These aims and projects provide specific reasons for the actions she undertakes, or give her actions their particular point or purpose. So it matters greatly to her on what reasons she may act, because limiting these reasons would limit the aims and projects she can pursue. Others, by contrast, have an interest, first and foremost, in being insulated from the effects of the agent’s actions on their own autonomous pursuits. But these effects depend on what the agent does, not on why she does it. So, others’ interests justify imposing restrictions on the agent’s freedom to engage in certain actions (those that would unduly interfere with or set back others’ endeavors), but do not justify imposing restrictions on the reasons for which she may act. Second, the account just sketched also suggests an answer to the question how the costs of mistakes—including some reasonable mistakes—are to be distributed. The rights others have against me insulate them from bearing the costs associated with the pursuit of my projects and aims. My making mistakes about the proper limits of my sphere of freedom is among the costs of these pursuits. As long as I make a mistake, but my mistake does not lead me to overstep the boundaries set by others’ rights, the cost of the mistake falls on me. (If I mistakenly think that the property line runs closer to my house than it does, or that an action is riskier than it in fact is, and I therefore abstain from a course of action that would have

Legitimacy as a Right to Err

191

been more valuable to me than the one I take instead, then this is a cost that I must bear.) And if my mistake leads to my overstepping these boundaries, and I thereby create costs on others, I owe them compensation. Why? Because the mistake is part of my pursuing my own projects, and the very point of the boundary drawn by others’ rights is that it insulates them from bearing (certain) costs associated with these pursuits. It is against the background of this (to philosophers I believe reasonably familiar) picture of central moral norms governing interactions and compensation that I want to explain the seemingly special case that has been the focus of this chapter: the case of an agent who, acting legitimately, is at liberty to engage in actions that set back interests of others ordinarily protected by rights, and thus does not owe compensation where her actions do set back these interests. Here is the basic idea: The picture just sketched assumed that each agent acts in the free pursuit of her own projects and aims, and must bear the burden of the mistakes she makes in that pursuit if she oversteps the boundaries set by others’ rights. But sometimes an agent acts, not in the free pursuit of her own projects and aims, but for another (in a sense yet to be clarified). And where the agent acts for another in this way, and does so for the other’s sake (in a sense that also requires further clarification), there the costs of (at least some of) her mistakes properly fall on the person for whom the agent acts, rather than on the agent who in fact commits the error. To make this thought intuitively plausible, consider a simple example. Imagine once again that C has fallen into the water (having ignored a warning sign against playing on the bank), and A is available to rescue C. To rescue C, A borrows B’s boat, which gets damaged in the process. It seems intuitively clear that C is morally required to bear the cost of the rescue mission: Even though A in fact damaged the boat, C must compensate B for it. (This will in practice often be institutionalized as a duty to indemnify A against B’s demand for compensation.) C must do so even if A could, in fact, have rescued C without using B’s boat, and even if the damage to the boat is due to a mistake A made (say, he didn’t see a buoy in the water, and collided with it, thereby denting the boat). Under what conditions would we want to say that A acts for C in cases like this, so that C is on the hook for what A did (including

192

Daniel Viehoff

A’s mistakes)?18 There are at least two such conditions. The first is a deliberative condition: A must be acting on the basis of certain reasons only—in the example just mentioned, reasons tied to the project of rescuing C. For, imagine that A borrows B’s boat, not to rescue C, but for his own pleasure. Yet once A is out on the water (and after he has collided with the buoy), he sees, and rescues, C. Even if taking B’s boat would have been necessary to rescue C, and so A ought to have borrowed the boat had he seen C’s predicament from the shore, the fact that A in fact took B’s boat for another reason (and dented it while pursuing the project associated with those reasons) is sufficient to put A, rather than C, under a duty to compensate B for the dent. What explains this judgment? I want to suggest that, where A deliberates in the right way, A acts for C, not just in the sense of benefiting C, but in the sense of adopting C’s practical position. A adopts C’s practical position by acting only on the reasons that C has19—in light of C’s aims and projects, and of the rights of others insofar as they limit their pursuit—and foregoing the freedom he, A, would ordinarily have to act on whatever reasons he sees fit in light of his own projects and aims. In practice, this means that A must, for instance, choose what course of action to take by considering only how different courses affect C’s projects and aims, not by how they affect A’s own. But why should the fact that A acts for C in this way suffice to make it incumbent upon C to bear the costs of A’s action? By itself, it does not. For the deliberative condition is only the first of two that must be met. There is also a second, justificatory, condition: A’s acting for C (in the way just described) imposes a duty on C to bear the costs of A’s so acting only if A’s so acting is itself in C’s interest even if it makes C liable for the costs of A’s so acting. In other words, whether A acts for C, constrained by C’s reasons, may be up to A. (It may just happen to be one of A’s projects.) But whether A’s acting for C is such that C must bear the costs of A’s action is not up to A, but depends on C’s interests only. It is only if A’s acting for C (even if this means C has to bear the correlative costs) can be justified by appeal to C’s interests alone, independently of any additional interest A has in doing what he is doing (or that anyone else has in A’s doing so), that A’s acting for C is compatible with C’s not being unfairly put on the hook for A’s projects or aims.

Legitimacy as a Right to Err

193

Notice that what matters here is whether C’s interest justifies A’s acting for her, not whether it imposes a duty on A to do so.20 C’s interest may not suffice to put A under a duty to rescue C, because A’s countervailing interests in not being subject to excessively costly requirements prevails.21 But that A is free not to act for C does not show that A is not justified in acting for C solely on the basis of C’s interests; and when A does, then A is indeed fully acting for C, with all of the consequences this has for the proper distribution of the benefits and burdens of A’s mistakes. To see this, imagine that the rescue mission is in fact dangerous: The water into which C has fallen is quite treacherous, so that the risk to any would-be rescuer is too great to make the rescue mandatory. Yet A, in an act of supererogation, goes ahead anyway, and saves C, though once again at the cost of denting B’s boat. It would seem odd to suggest that C would owe no compensation here, just because A had a choice, morally speaking, whether to rescue C, when he exercises that choice solely on the basis of the reason that it is in C’s interest to be rescued. If A’s action satisfies the two conditions, A has a right to err when rescuing C—A is at liberty to take actions (like use C’s property, touch C’s body, or even break C’s arm) that would ordinarily be impermissible, and thus does not owe compensation for them. Why? Because if A abandons his ordinary freedom to pursue his own projects and instead acts solely in the pursuit of C’s projects and aims, and his doing so is good for C (and that is why A does it), then it seems fair that C, for whose sake A’s action is undertaken and in whose interest A’s undertaking the action is, also bears the burdens associated with it. Put differently: Where the relevant conditions are satisfied, A’s action is counterfactually dependent on the benefits that it provides to C. It is if, and only if, A’s acting for C (that is, pursuing C’s project) is good for C, and A in fact acts for C (that is, makes his choices based solely on the basis of C’s reasons for action), that A’s acting for C also puts C on the hook for the potential downside of A’s so acting. And when A acts for C under these conditions, then (i) making A liable to bear the costs of his own mistakes threatens to put him on the hook for another’s projects and aims in a problematic fashion, whereas (ii) making C liable to bear the costs of A’s mistakes, though it makes C dependent on A’s actions, does not put C on the hook for

194

Daniel Viehoff

another’s projects and aims (since the aims and projects that A is pursuing are necessarily C’s, in the sense that A’s actions are subject to a deliberative condition tied to the reasons that derive from C’s projects and aims). This account also explains which of A’s mistakes C is indeed liable for, and which ones A must bear. C is liable for those of A’s mistakes that someone with A’s capacities would make when properly guided solely by C’s reasons. For instance, when deciding whether to spend more time on getting the matter right, or instead acting now based on more limited information, the costs and benefits of either option must be weighed from C’s point of view, in light of C’s projects and aims. And if A’s capacities are such that the likelihood of mistakes is so great that the likely costs of A’s acting for C exceed the benefits, then the justificatory condition is not satisfied, and A’s action cannot in fact be attributed to C. The general structure of the argument sketched here straightforwardly extends to institutions and their actions.22 They (and those acting for them) are at liberty to engage in certain otherwise wrongful actions vis-à-vis their subjects, and insulated from compensatory demands that would otherwise apply to agents committing such acts, if these institutions act for their subjects (the institutions satisfy certain deliberative constraints tied to the subject’s projects and aims) and if their acting for their subjects in this way is itself in the subjects’ interest (the institutions satisfy the justificatory constraint). That the legitimacy of legal and political institutions is structurally continuous with the distinctive rights of certain individuals— those suitably acting for another—may seem surprising. But I think it simply reflects the fundamental unity of the moral domain. There is, however, one important feature of legal and political institutions that makes the account offered here especially apposite, and a concern with legitimacy as a right to err especially pressing. Both authoritative directives and coercive threats—the two forms of power most central to legal and political institutions— involve one agent’s intentionally changing another’s reasons for action.23 What is more, these exercises of legal or political power are normally intended not just to provide a minor change in the subject’s reasons, but to practically settle how the subject ought to

Legitimacy as a Right to Err

195

act—to create a conclusive reason for a particular course of action. If A has such directive power over S, then A is enabled to effectively settle what S will do (at least if S is rational), and thus to direct S’s rational agency at will. And just as A’s touching S, or harming S, is acceptable only if A’s doing so is respectful of their asymmetrical relation to S’s body, so A’s having and exercising directive power over S is acceptable only if it respects their asymmetrical relation to S’s agency. S’s body is not simply available for use by others to pursue their own aims and projects; and neither is S’s agency. So if S has no duty to contribute to A’s projects and aims, or to bear any burden to enable their pursuit, then A’s exercising directive power over S in their pursuit is incompatible with S’s status as an independent moral agent.24 In turn, if A’s directive power over S is to be compatible with S’s moral independence, and the relation between A and S morally appropriate, then A’s having and exercising directive power over S must be suitably conditioned on S’s own projects and aims. As a slogan: When it comes to A’s power to direct S’s agency, A must have and exercise that power for S. And so we are also in a position to better understand the relation between the different conceptions of legitimacy that I sketched at the opening of this chapter, which focus, respectively, on the possession of a normative power, the existence of a morally appropriate relationship between ruler and subject, and an institution’s right to err. That an institution has authority (and, more generally, directive power) over subjects raises the question whether the institution stands in a morally appropriate relationship to the subjects, one that is respectful of the subjects’ status as independent moral agents. To ensure that its authority is compatible with the subjects’ moral independence, the institution must have and exercise its authority for, and for the sake of, the subjects. And if the institution acts in this way, its reasonable, good-faith errors need not, as a matter of fairness, be borne by the institution or its agents rather than by the subjects for whom the institution acts. Conclusion By way of conclusion, let me return to one of the examples from which this chapter began. Consider again a court adjudicating a

196

Daniel Viehoff

dispute. Such a court has directive power—the power, for instance, to impose on Defendant a duty to pay Plaintiff $1,000, even though Defendant would otherwise have been morally free to choose what to do with that money. Why does Judge not owe compensation to Defendant for the loss Defendant suffers as a result of the erroneous decision? Because Judge is, in the relevant sense, acting for Defendant: Defendant has reason, and perhaps an obligation, to resolve the dispute with Plaintiff in a certain way, peacefully and fairly. Judge helps Defendant—and Plaintiff—to achieve this, by adjudicating the dispute in light of the considerations that properly bear on the question at issue between them. And so Judge, acting on the basis of the appropriate considerations, acts for them, even where the ultimate decision is erroneous. The matter would be different, however, if Judge’s decisionmaking did not satisfy the deliberative or justificatory condition. At times this may effectively undermine Judge’s authority and leave Defendant without a duty to obey the verdict. But often it will not have this effect: Defendant will still be under a duty to obey Judge’s decision, because that is, say, the best way to effectively settle the dispute with Plaintiff in a way that is suitable to their relationship. In that case, Judge has the power to impose a duty on Defendant, yet lacks a liberty right to exercise that power even where doing so sets back the Defendant’s interest in keeping the $1,000, which is ordinarily protected by Defendant’s property rights. Defendant therefore might be morally permitted to seek to recover the loss from Judge—if not via the legal system, then by acts of self-help. What if Judge sends a sheriff to enforce the verdict? If the sheriff enforces the verdict conditional on its actually imposing a duty on Defendant, then sheriff might in fact be acting for Defendant in the relevant sense, and act legitimately even where the decision enforced was made illegitimately. But if the sheriff’s action is not suitably counterfactually dependent on Defendant’s duty, and instead simply depends on the sheriff’s beliefs about his own obligations, qua executive official, to obey Judge’s order, then the sheriff’s action would be illegitimate too. And if the enforcement action would unjustly burden Defendant (as is likely the case if the judgment it enforces is erroneous), then Defendant may be morally permitted to redirect the cost of the illegitimate enforcement toward the sheriff, and impose, for instance, defensive harm

Legitimacy as a Right to Err

197

on the sheriff to ensure that she is not deprived of what is, by right, her own. So the issue of legitimacy as a right to err also bears, in crucial ways, on our right to resist government actors who abuse their power, and treat it as if it were available to them for purposes other than the sole legitimate one, of serving those over whom the power is exercised. Acknowledgments I presented an early version of this chapter at a Pacific APA panel on “The obligation to obey the law.” I am grateful to Rob Hughes for organizing the panel, and to the audience and my co-panelists, David Lefkowitz and Margaret Gilbert, for helpful comments and questions. I am also grateful to Andrei Marmor and Emad Atiq for inviting me to present a more recent version to their legal philosophy seminar at Cornell; and to them and their students for insightful comments. For written comments on, or extended discussions of, this chapter, I owe special thanks to Stephen Galoob, Felix Koch, Melissa Schwartzberg, and Juri Viehoff. Notes 1. For a useful articulation of this distinction, see Richard Fallon, “Legitimacy and the Constitution,” Harvard Law Review 118, no. 6 (2005): 1787–1853. Notice too that I treat an institution’s legitimacy as conceptually, and practically, separate from its overall justifiability. For an influential discussion of this distinction with which I am in outline (though not in its details) sympathetic, see A. John Simmons, “Justification and Legitimacy,” Ethics 109, no. 4 (1999): 739–771. 2. For approaches fitting this characterization, see, e.g., Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986); Simmons, “Justification and Legitimacy”; Arthur Isak Applbaum, “Legitimacy without the Duty to Obey,” Philosophy and Public Affairs 38 (2010): 215–239. 3. For otherwise quite different theories that nonetheless all focus on the quality of the relationship between ruler and subject, see, e.g., John Rawls, Political Liberalism, Paperback ed. (New York: Columbia University Press, 1996); Niko Kolodny, “Political Rule and Its Discontents,” Oxford Studies in Political Philosophy 2 (2016): 35–68; Amanda Greene, “Consent and Political Legitimacy”; Daniel Viehoff, “On Authority, Legitimacy, and Service: Discussion of the Morality of Freedom,” Jerusalem Review of Legal

198

Daniel Viehoff

Studies 14, no. 1 (2016): 113–124; Viehoff, “The Truth in Political Instrumentalism,” Proceedings of the Aristotelian Society 117, no. 3 (2017): 273–295. Notice that asking about the moral acceptability of the relationship does not settle whether the institution is justifiable all things considered, and thus legitimacy so understood need not collapse into justifiability. 4. Let me flag that I do not take these approaches to be competitors so much as alternative, and in principle compatible, attempts to articulate normative claims central to familiar legal and political institutions and practices. So, in offering this third account of legitimacy, my aim is not to supplant but rather to supplement existing discussions of the normative features that are central to, and important for understanding the structure of, such institutions and practices. 5. Whether it also extends to cases where the applicable law fails to properly reflect the parties’ moral entitlements depends on whether that law is morally authoritative for the subjects. 6. Studies by Tom Tyler and others find that people who view a judicial institution as fair, and therefore legitimate, also accept its adjudicative decisions even if these decisions are (the subject believes) substantively mistaken or unjust, and harmful to the subject’s own interests. See, e.g., Tom R. Tyler, Why People Obey the Law (Princeton, NJ: Princeton University Press, 2006). 7. See, e.g., David Enoch, “Authority and Reason-Giving,” Philosophy and Phenomenological Research 89, no. 2 (2014): 296–332; Stephen Perry, “Political Authority and Political Obligation,” Oxford Studies in Philosophy of Law 2 (2013): 1–74. For Hohfeld’s original account of normative powers, and their relation to other normative incidents, see Wesley Newcomb Hohfeld, Fundamental Legal Conceptions as Applied in Judicial Reasoning, ed. David Campbell and Philip A. Thomas, Classical Jurisprudence Series (Aldershot, UK; Burlington, VT: Ashgate/Dartmouth, 2001). It is common in discussions of authority to speak of the subject’s duty to obey. As Perry points out, the proper counterpart of a Hohfeldian power is a Hohfeldian liability, rather than a duty. But since the duty to obey is a conditional duty (a duty to do x if directed to do x), it is easy enough to see how duties of obedience can often (though perhaps not always) be theorized as a matter of normative power and liability. Sometimes it is suggested that legitimate authority involves, not just such a duty, but a corresponding claim right to obedience. I think this is implausible as a general condition on authority (legitimate or other), and so just set it aside here, for the reasons discussed in Enoch. 8. I leave it open here whether Defendant has a moral claim against Plaintiff. On the one hand, I think it is clear that Defendant has a duty to pay up, and also a duty not to enforce whatever claim she may have

Legitimacy as a Right to Err

199

against Plaintiff. On the other hand, if Plaintiff comes to realize that Defendant did not in fact wrongfully harm him, then it is conceivable that Plaintiff has a moral duty to return the damages award to Defendant. 9. There is an ambiguity here worth flagging: I mean to say both that the legislators qua individuals do not owe such compensation, and that the legislators qua legislators—or, better, the legislature as a body—does not owe compensation, though it does (as the next paragraph explains) have an obligation to seek a fair distribution of benefits and burdens over time, and so must seek to include, among the considerations on which it makes future decisions, facts about past decisions and their distributive effects. 10. In fact, justified authority differs not just from legitimacy as a right to err, but also from the third sense of legitimacy mentioned earlier, according to which it is a matter of the moral appropriateness of the (power) relation between ruler and subject. For discussion, see Viehoff, “On Authority, Legitimacy, and Service.” 11. See, for example, Raz, The Morality of Freedom; Perry, “Political Authority and Political Obligation.” It also plays a central (though in some ways less direct) role in recent theories of democratic authority, including Thomas Christiano, “The Authority of Democracy,” Journal of Political Philosophy 12, no. 3 (2004): 266–290; David Estlund, Democratic Authority: A Philosophical Framework (Princeton, NJ: Princeton University Press, 2008); Daniel Viehoff, “Democratic Equality and Political Authority,” Philosophy and Public Affairs 42, no. 4 (2014): 337–375. 12. For a more detailed discussion of examples of this sort, and why we should still consider them instances of practical authority properly understood, see “On Authority, Legitimacy, and Service.” 13. See, e.g., Raz, The Morality of Freedom. 14. See, e.g., Enoch, “Authority and Reason-Giving.” 15. Jeremy Waldron, “A Right to Do Wrong,” Ethics 92, no. 1 (1981): 21–39. 16. More specifically, he says that a right to rule (as contrasted with a mere permission to coerce or a mere power to impose duties at will) “includes a liberty right on the part of the authority to make decisions as it sees fit [as well as] a power to impose duties on citizens.” Thomas Christiano, The Constitution of Equality: Democratic Authority and Its Limits (Oxford: Oxford University Press, 2008), 240–241. Christiano also suggests that a right to rule includes a claim right to obedience. As mentioned earlier, I simply set this issue aside here. 17. I borrow this example from Victor Tadros, “Independence without Interests?,” Oxford Journal of Legal Studies 31, no. 1 (2011): 200.

200

Daniel Viehoff

18. In answering this question I am, as I said earlier, setting aside cases where C has consented to A’s acting for C. 19. And, ordinarily, only on some of these: Legitimate power tends to be circumscribed by subject matter, for reasons that I cannot discuss here. 20. Someone might in fact suggest that C does not owe such compensation precisely because A had a duty to rescue C: The duty reflects the fact that A could be asked to bear a certain burden for the sake of rescuing C; and so A is not entitled to have the burden relieved by subsequent compensation. But, as discussed earlier, that A must bear the burden of rescue if the alternative is to impose a much greater burden on C (letting C drown) does not show that A must bear it if the alternative is to impose that very same burden on C (by requiring C to compensate A for his loss). Whether A has a duty to rescue, and whether A is entitled to compensation for the loss borne when fulfilling the duty, are separate questions. 21. It is worth mentioning here that, though we commonly speak of a “duty to rescue,” what we plausibly have in mind is usually a duty to try to rescue. This is easy enough to overlook because, in many of the cases commonly taken to give rise to a duty to rescue, what a successful rescue requires is sufficiently straightforward that there is no significant gap between (seriously) trying and succeeding. 22. This is not to say that working out the details of its application to institutions does not pose a number of challenges. To mention just one: Institutions serve many people simultaneously, and so one needs an account of what an agent ought to do when acting for several beneficiaries at once, especially when their interests aren’t neatly aligned. 23. This and the next paragraph draw on Daniel Viehoff, “The Truth in Political Instrumentalism,” Proceedings of the Aristotelian Society 117, no. 3 (2017): 285. 24. Notice that the problem just described arises even if the exercise of power imposes no further costs on the subject. Using me to pursue another’s projects (absent some of the special conditions discussed earlier) disrespects my status as an independent moral agent, quite apart from any more concrete harm I suffer as a result.

7 OFFICIAL INTENTIONS AND POLITICAL LEGITIMACY THE CASE OF THE TRAVEL BAN MICAH SCHWARTZMAN

Seven days after his inauguration, President Donald Trump signed an executive order suspending entry into the United States of citizens from seven countries with Muslim majorities.1 National security officials were caught off guard by the order, and its hasty implementation resulted in chaos and confusion at airports in the United States and around the world. The order was immediately challenged in federal courts on both statutory and constitutional grounds. A court in Washington State blocked its nationwide enforcement, and after a federal Court of Appeals upheld that decision, President Trump tweeted, “SEE YOU IN COURT, THE SECURITY OF OUR NATION IS AT STAKE!”2 Instead of challenging the appellate court’s decision, however, President Trump revoked his first executive order and issued a second one that largely resembled the first.3 This order, too, was purportedly justified on grounds of national security. And it, too, was immediately blocked by federal courts, which held that the second order, like the first, was motivated by hostility toward Muslims in violation of the Establishment Clause of the First Amendment, which prohibits religious discrimination.4 This time, President Trump appealed and was rebuked in the strongest terms. Writing for the Fourth Circuit Court of Appeals, Chief Judge Gregory declared that the order “drips with religious intolerance, animus, 201

202

Micah Schwartzman

and discrimination.”5 As evidence, the court cited numerous statements made by President Trump and his associates, including a “Statement on Preventing Muslim Immigration” that “call[ed] for a total and complete shutdown of Muslims entering the United States until our representatives can figure out what is going on.”6 The court concluded that “[t]here is simply too much evidence that [the second order] was motivated by religious animus for it to survive any measure of constitutional review.”7 Six months later, in an effort to remove the taint of animus, President Trump issued yet a third order, calling this one a “Proclamation.”8 This order indefinitely suspended entry of all immigrants from six of the original Muslim-majority countries, along with some citizens from Venezuela and North Korea. Despite the government’s claim that the Proclamation was justified by a comprehensive, multiagency security review, federal courts again rejected the order as motivated by hostility toward Muslims. Chief Judge Gregory was explicit about the court’s reasoning. “Our constitutional system creates a strong presumption of legitimacy for presidential action,” he wrote, “[b]ut the disposition of this case is compelled by . . . undisputed evidence that the President of the United States has openly and often expressed his desire to ban those of Islamic faith from entering the United States. The Proclamation is thus not only a likely Establishment Clause violation, but also strikes at the basic notion that the government may not act based on ‘religious animosity.’”9 At the time of this writing, the constitutionality of President Trump’s travel ban is pending before the US Supreme Court. But, however the Supreme Court ultimately decides the case, the decisions of the lower courts—which consistently opposed the ban— suggest a legal, and perhaps also a moral, view about the relationship between the president’s intentions and the legitimacy of his conduct. According to this view, official intentions are directly relevant to determinations of constitutional legitimacy. If public officials act with intentions that are prohibited under some constitutional provision, such as the Establishment Clause, courts understand themselves to have powerful reasons to invalidate the conduct in question. And, perhaps implicit within this view is a subtler normative one, according to which it makes sense, morally speaking, for constitutional provisions to prohibit conduct by officials who act from wrongful intentions or motivations.

Official Intentions and Political Legitimacy

203

My aim is to defend this normative view against an objection developed by some moral philosophers and legal scholars who are skeptical about the relevance of intentions in determining the moral permissibility of an agent’s actions.10 These critics argue that whether an agent is allowed to take some action depends not on the agent’s intentions or motivations but on whether the action is, in fact, justified under the circumstances. President Trump may have had bad intentions in implementing his travel ban, but if the ban was justified by considerations of national security, then it was legally and morally permissible, regardless of the president’s intentions. If this argument, which I will call the permissibility objection, is correct, it provides a basis for denying the relevance of intentions in evaluating whether actions taken by public officials are legitimate. This objection has been applied recently to challenge judicial review of legislative intent and other constitutional doctrines that rely on intent-based inquiries.11 The logic of the objection could also be extended to cover executive and administrative actions of the kind at issue in the travel ban litigation. In what follows, I reject the permissibility objection and defend the moral view, reflected in the decisions of the lower courts in the travel ban cases, that the intentions of public officials are relevant to the political legitimacy of their conduct. After setting out some preliminary observations about official intentions and clarifying the scope of my argument, I sketch several ways in which such intentions can feature in accounts of political legitimacy. I then rehearse the permissibility objection and offer some arguments against it. Finally, I respond to the problem of how to cure conduct that has been tainted by wrongful intentions. Defining Official Intentions Before turning to how intentions can be relevant to determinations of political legitimacy, let me offer some preliminary points about the nature of official intentions and the scope of my argument about them. My aim is to avoid some of the standard debates about the role of intentions, motivations, and purposes of public officials, especially in the exercise of judicial review. In many areas of constitutional law—including freedom of religion, freedom of speech, race and sex discrimination, regulation of reproductive

204

Micah Schwartzman

choice, and LGBT rights—courts engage in searching inquiries into the intentions of political actors. And for decades, this practice has been the subject of extensive debates among legal scholars.12 But much of the literature has focused on institutional questions about whether it makes sense to ascribe intentions to legislatures, and if so, how courts can determine those intentions.13 As much as possible, however, I want to leave aside such questions and focus instead on the moral issue of whether intentions, assuming they exist and are discernible, matter for evaluating the permissibility or legitimacy of official actions. So, to clarify the scope of my argument, I distinguish between three types of questions about the intentions of official actors— legislators, executives, judges, and perhaps citizens when they exercise political power (as when they vote directly in referendums)— which I will refer to as official intentions. First, there is an ontological question about whether such intentions exist, especially in cases involving collective political bodies. There is now a large and rich literature devoted to demonstrating the possibility of collective intentionality, including on a mass scale, that might account for groups like legislatures forming, maintaining, and acting upon shared intentions.14 But even if we are skeptical about group agency and collective intentionality and therefore reject attempts to show that legislatures have intentions, it is still possible to inquire about the intentions of individual political actors. In the travel ban cases, courts inquired into the intentions of President Trump, free from skepticism about group intentions. Dissenters in those cases raised objections to that inquiry, but no one argued that it was impossible for the president to have intentions (although some questioned whether courts could discern clear or stable ones).15 Thus, for present purposes, I set aside ontological questions that have dominated debates about legislative intention. Regardless of how those debates are resolved, in at least some cases, official intentions may exist and can be the subject of legal and moral inquiry. Another question I want to avoid is epistemic: Assuming that an agent can have intentions, how can we know them? What kind of evidence is sufficient to establish that an agent had a particular intention when taking some action? There is, again, extensive debate among legal scholars about how courts should approach these questions with respect to collective entities, especially

Official Intentions and Political Legitimacy

205

legislatures.16 But at least with respect to individuals, courts adjudicating criminal and tort law cases routinely inquire into the intentions of individual actors.17 And while there might be special considerations involving separation of powers or other legal concerns that apply to such inquiries in cases involving constitutional matters, as the lower courts in the travel ban cases noted, there is nothing epistemically distinctive about trying to discern the intentions of a particular public official for constitutional purposes.18 The third question, and the one I want to pursue here, is distinguished from ontological and epistemic concerns. Assuming that official intentions exist and that we can know them, are they relevant to whether officials are permitted to act in various ways? This question could be interpreted as a legal question or as a moral one or, depending on one’s jurisprudential views about the relation of law and morality, perhaps the two questions are inseparable. But again, I am not interested in begging jurisprudential questions. My focus is squarely on the moral question: Whatever their legal status, are official intentions morally relevant in determining whether actions taken by public officials are legitimate? This is a question to which legal scholars and political philosophers have paid relatively little attention. Indeed, to the best of my knowledge, we lack a general account—indeed any account—of the moral relationship between official intentions and political legitimacy.19 My last preliminary point concerns the concept of an intention. As I use it here, the concept refers to the reasons for which agents understand themselves to be acting.20 For any given action, we can ask a person: “Why did you take that action?” Someone who responds sincerely will state her intention by offering her reason (or reasons) for acting in the way that she did. In providing such reasons, a person might also be thought to describe her motivation, or what she takes to be the reason that caused her to act. Defining intention in terms of an agent’s subjective reasons for action brings the concept close to that of motivation. Perhaps that is why many courts and scholars do not distinguish sharply, if at all, between the concepts. Because I do not think anything in the moral arguments that follow turns on the distinction between an agent’s intentions or motivations—understood in terms of their subjective reasons for action—I shall also use the terms interchangeably.21

206

Micah Schwartzman How Intentions Figure in Theories of Political Legitimacy

In the travel ban cases, lower courts held that President Trump’s executive orders were illegitimate because they were based on intentions that violated the Establishment Clause of the First Amendment. The courts interpreted the Establishment Clause to forbid the government from acting on the basis of religious animus, or an intent to disfavor a particular religious group because of prejudice, hostility, or hatred of that group.22 Finding ample evidence that President Trump was motivated primarily by such animus, courts rejected the government’s assertions that the president’s actions were constitutionally legitimate. My question is this: Is it possible to give an account of the relationship between official intentions and political legitimacy that makes sense of this result? That is, can we give a moral justification for rejecting the legitimacy of an official action on the basis of the intention or motivation behind it? One way to approach this question would be to adopt a particular account of political legitimacy and show how that account supports evaluating the exercise of political power by taking into consideration the nature of official intentions. But this way of proceeding may narrow the field too quickly, leaving open the possibility of rejecting intent-based determinations because of disagreements about the underlying account of legitimacy. Instead, I want to sketch several ways in which political legitimacy might be related to official intentions. Political legitimacy is commonly understood to refer to the state’s right to rule by imposing moral duties and by exercising coercive power to enforce compliance with them.23 The traditional view is that a state’s right to rule creates correlative obligations on the part of subjects to obey the law.24 Some theories, however, reject this correlativity thesis and explicate the right to rule in terms of a moral power to impose duties without generating political obligations.25 Others argue that political legitimacy entails a right to non-interference with the state’s exercise of coercive power.26 But, regardless of how the right to rule is defined—whether in terms of claim-rights, powers, privileges, or some combination of Hohfeldian rights—an important feature of theories of political legitimacy is that they specify

Official Intentions and Political Legitimacy

207

the conditions under which it is morally permissible for the state to exercise coercive power. Such theories provide an account of when political officials are allowed to impose duties on others by enacting, enforcing, and interpreting the law. It is worth emphasizing that political legitimacy concerns the moral permissibility of official conduct. Many exercises of political power may be permissible even if they are mistaken. This point is often noted when distinguishing legitimacy from other aspects of political morality. Laws can be legitimate even if they are incorrect, unwise, or unjust.27 Of course, under most conceptions of legitimacy, the state must govern in a manner that is at least minimally or reasonably just. If that condition is satisfied, however, officials may have discretion to act in ways that are morally controversial and subject to reasonable disagreement. A theory of legitimacy can be thought of as setting a moral floor, the minimal conditions under which the state can assert the right to rule, with whatever moral content that right entails.28 Intentions might be relevant to determinations of political legitimacy in different ways. According to pure procedural theories, laws and other official acts are legitimate if they are the product of appropriate procedures. What makes procedures appropriate will depend, of course, on the values promoted by a particular theory of legitimacy. In democratic theory, some accounts of proceduralism are grounded in fairness and equality.29 Procedures may be fair because they provide each person with an equal share of political power or perhaps with an equal opportunity to influence democratic outcomes. Other accounts emphasize the deliberative or epistemic benefits of democratic procedures.30 But whatever the justification for pure proceduralism, the intentions of official actors will be relevant in determining their legitimacy only to the extent that the results of following the appropriate procedures make them relevant. Thus, if a constitutional or democratic procedure generates substantive legal requirements that turn on official intentions, then those requirements—which we can call intention-sensitive—are legitimate by virtue of having the proper procedural pedigree. For example, suppose political actors follow appropriate constitutional procedures in adopting legislation that prohibits religious discrimination by public officials. And suppose further that this legislation prohibits, or is properly interpreted as

208

Micah Schwartzman

prohibiting, official acts that are motivated by religious animus. In these circumstances, whether officials have acted with wrongful intentions would be relevant to determining the permissibility of their actions. Pure proceduralism makes intentions relevant only contingently. If political procedures are not used to prohibit wrongful intentions, then actions based on such intentions might be legitimate under such a theory. And even if procedures are used to make intentions relevant, proceduralism itself will not provide an account of why it makes sense, morally speaking, to do so. Of course, that might not be a deficiency according to procedural theories, which are often justified on the grounds that they avoid making substantive moral judgments of this kind. Some procedural accounts of legitimacy incorporate minimal substantive moral constraints on outcomes. Those constraints may, in turn, include limits on actions undertaken on the basis of wrongful intentions.31 For example, Simon May has recently advanced an account of democratic legitimacy that features two main principles: first, a procedural principle stating that “all citizens must have equal standing in fundamental political processes,” and, second, a substantive principle according to which “the government must treat citizens with equal concern and respect in the exercise of its authority.”32 Whereas the procedural principle is focused on the proper form of the political process—requiring equal voting, basic rights to freedom of speech and association, and some minimum economic capability to participate in the process—the substantive principle constrains policies adopted as the outcomes of that process. May gives his substantive principle an expressivist interpretation, such that it prohibits the government from adopting policies that convey a message of inferiority toward some citizens.33 The paradigmatic example is of an apartheid state, such as Jim Crow in the American South, which systematically violated the rights of African Americans and denigrated their status as equal citizens. According to May, an apartheid regime is illegitimate because it enacts policies that express views inconsistent with the substantive principle of equal concern and respect. But that substantive principle might also be interpreted in a way that emphasizes the relevance of intentions. One way to understand the requirement that the state treat citizens with equal concern would be to say that

Official Intentions and Political Legitimacy

209

citizens have a right to equal treatment that includes a prohibition on public officials subjecting them to policies based on racial animus. This principle need not compete with an expressivist view. The two might be complementary and reach similar results. The point here is only that a substantive principle of equal concern could be interpreted to limit procedural outcomes on the basis of impermissible intentions. And an intentionalist account, like an expressivist one, can be used to explain why certain policies, such as those constitutive of an apartheid regime, are not merely morally wrong but also politically illegitimate. A procedural view with minimal substantive constraints might make only a narrow range of intentions relevant to political legitimacy, perhaps excluding only the most egregious cases of intentional discrimination.34 But there are familiar conceptions of legitimacy that both incorporate a wider range of substantive moral values and that also exclude a broader set of reasons as possible justifications for the exercise of political power. The most prominent of these is the liberal principle of legitimacy, which in John Rawls’s formulation holds that “our exercise of political power is proper and hence justifiable only when it is exercised in accordance with a constitution the essentials of which all citizens may reasonably be expected to endorse in light of principles and ideals acceptable to them as reasonable and rational.”35 Because citizens reasonably disagree about broader religious and philosophical conceptions of the good, Rawls argues that constitutional essentials must be justified by sufficient public reasons.36 Reasons are public in the relevant sense when they draw on basic political values—including “the values of equal political and civil liberty; equality of opportunity; the values of social equality and economic reciprocity”37—and commonsense methods of reasoning.38 The upshot of all this, Rawls argues, is that the exercise of political power cannot be legitimate when it is supported only by nonpublic values and forms of reasoning, which are those based on particular conceptions of the good. Public officials must show that their actions are consistent with a constitution the essentials of which are justifiable on the basis of public reason.39 On its face, the liberal principle of legitimacy does not mention the intentions of public officials or make them relevant to the moral permissibility of official conduct. But intentions might enter into a Rawlsian account of legitimacy in at least two ways. First, as

210

Micah Schwartzman

under a procedural conception of legitimacy, intention-sensitive rules might be adopted pursuant to an otherwise publicly justified constitution. But second, and more fundamentally, intentions might be relevant because the moral principles and ideals that are acceptable to reasonable citizens, and which are used to structure constitutional essentials, may include limits on how public officials can be motivated. In this way, the liberal principle of legitimacy is similar to the substantive principle of equal concern mentioned above. But whereas the substantive principle might be construed narrowly to exclude only intentions that deny the basic equality of some citizens, the liberal principle of legitimacy might sweep more broadly, either by prohibiting public officials from exercising power with the aim of promoting a particular conception of the good or by requiring them to be motivated by public reasons.40 Having sketched several ways in which intentions might be relevant to political legitimacy, let me return to the question I asked earlier, which is whether it is possible to give an account of legitimacy that makes sense of lower court rulings in the travel ban litigation. The answer depends, of course, on the theory of legitimacy. For procedural accounts, the connection between the permissibility of official conduct and intentions may be entirely contingent. But if procedures are used to select intention-sensitive rules, then conduct that violates those rules may be politically illegitimate. For example, in the context of the travel ban cases, the Establishment Clause is a substantive rule adopted according to constitutional procedures, and if properly interpreted by authorized officials, including judges, to prohibit certain forms of intentional religious discrimination, then those actions are impermissible. By contrast, a substantive theory may incorporate intention-sensitive rules, such as the Establishment Clause, as moral restrictions on procedural outcomes. For a theory of this type, a rule like the Establishment Clause may be justified by a principle requiring the state to treat those subject to its control with equal concern and respect. And to the extent the president’s order amounts to intentional religious discrimination that fails to show such respect, those orders are politically illegitimate. Last, under the liberal principle of legitimacy, public officials may be required to act on the basis of public reasons. If their actions are not motivated by such reasons, they contravene the moral principles and ideals acceptable to

Official Intentions and Political Legitimacy

211

reasonable citizens when seen as free and equal members of a liberal democratic society. This conception of legitimacy can account for the travel ban decisions by saying that the president’s orders were based on intentions that violate the idea of public reason and that cannot be publicly justified. My point in describing and applying these various theories of legitimacy is not to defend one or another of them, but to suggest how intentions might matter, either contingently or intrinsically, in evaluating whether laws and other forms of official conduct are legitimate. Obviously much more argument would be needed to support a claim that a particular account of legitimacy is plausible, let alone correct, and that the relevant account properly incorporates intention-sensitive moral principles. But even this brief survey should be useful in showing how different theories of legitimacy might account for the presence of such principles and their role in evaluating the moral permissibility of official conduct. The Permissibility Objection Theories of political legitimacy that incorporate intention-sensitive rules make determinations about intentions relevant to whether official conduct is morally permissible. But in recent years, a number of philosophers have argued that intentions are never directly relevant in determining the permissibility of actions.41 If their contention is correct, then any theory of legitimacy that incorporates intention-sensitive rules is to that extent mistaken. I will call this the permissibility objection.42 In rehearsing this objection, it is important to distinguish two ways in which intentions might be relevant for determining permissibility. They can be either directly or indirectly relevant. Intentions are directly relevant when the intentions themselves matter for whether an action is permissible. And they are indirectly relevant when they matter because of their effects, or because they change the meaning of an action in a way that brings into play some further moral principle that bears on the permissibility of acts with such meaning.43 Some examples may help to mark this distinction between the direct and indirect relevance of intentions. In the travel ban cases, courts determined that the president’s intention to discriminate

212

Micah Schwartzman

on religious grounds made his orders impermissible. Courts treated the fact that the president acted with animus as bearing directly on whether his actions were legitimate. But the president’s intentions might also have mattered indirectly because of their “predictive significance,” which is the likelihood that an agent acting with certain intentions will produce a range of consequences.44 For example, courts might have been concerned that in acting on the basis of religious hostility, the president was likely to make mistakes in evaluating risks to national security. Stereotyping based on animus might corrupt threat assessments and lead to unnecessary rights infringements. To avoid those harms, courts might apply close scrutiny to actions based on wrongful intentions. They would be responding not to the direct relevance of the president’s intentions but to the risk that a president acting on wrongful intentions would produce consequences, such as rights violations, that are independently impermissible.45 The president’s intentions might also be indirectly relevant because, when publicly conveyed, they demean or denigrate religious minorities in ways that are impermissible.46 The idea here is that wrongful intentions themselves do not make the president’s order impermissible; rather, the expression of his intentions informs the social meaning of his actions. And if that social meaning is denigrating or demeaning to religious minorities, then there are again independent reasons—perhaps based on something like the substantive principle of equal respect mentioned above—for determining that conduct based on those intentions is prohibited.47 Given this distinction between direct and indirect relevance, it is important to note that the permissibility objection holds only that intentions are never directly relevant to permissibility. The objection does not deny that intentions might matter because those who act on them are more likely to cause harmful effects or because they will express their intentions in ways that are demeaning to others. The focus of the objection is on whether the intentions themselves make a difference in deciding whether some action is morally permissible. The main argument for the objection is that the direct relevance of intentions for moral permissibility cannot be maintained in reflective equilibrium.48 That is, testing intention-sensitive moral principles—or, more specifically, those that make intentions

Official Intentions and Political Legitimacy

213

directly relevant—against a range of intuitive judgments about particular cases shows that all such principles are incorrect and should be abandoned. As received, this claim is categorical. The objection states that intentions are never directly relevant, that there are no cases in which intentions matter in themselves for determining whether an action is permissible. The philosophical literature contains a wide range of cases testing the permissibility objection. The most familiar involve challenges to the Doctrine of Double Effect (DDE). Although there is no standard formulation of this doctrine, it is generally understood to hold that an action that harms others is permissible only if (1) an agent does not intend the harm, but rather merely foresees it, and (2) the positive effects of the action are proportional to its harmful or negative effects.49 Proponents of the permissibility objection have focused on DDE because it makes intentions directly relevant to determining whether a harmful action is permissible. If it can be shown that DDE is mistaken, then one powerful intention-sensitive moral principle must be abandoned. And if DDE falls, then other principles that are similarly intentionsensitive might also be rejected in a similar fashion. To see the force of this argumentative strategy, consider the primary examples used to challenge DDE, which at this point are nearly canonical.50 The first is: Tactical Bomber: A pilot bombs a military target with the intention of advancing the war effort. The pilot foresees but does not intend the deaths of some innocent civilians.

If the military target is sufficiently valuable and the foreseeable, but not intended, loss of civilian life is proportional, then the tactical bomber’s act satisfies DDE and is therefore permissible. But now consider: Terror Bomber: A pilot bombs a military target with the intention of terrorizing the civilian population. The pilot not only foresees but intends the deaths of some innocent civilians.

Whereas the tactical bomber’s actions are permissible under DDE, the terror bomber’s are forbidden, precisely because the

214

Micah Schwartzman

pilot intends, rather than merely foresees, harm to civilians. And because he violates DDE, his actions are morally impermissible. Proponents of the permissibility objection argue that the permissibility of the actions taken in these examples cannot turn directly on the intentions of the pilots. If the tactical bomber’s actions are permissible—if the pilot is morally allowed to fly the mission—and if the terror bomber performs exactly the same physical actions, then it is difficult to see why those actions are impermissible when undertaken by the terror bomber. Suppose the tactical and terror bombers are co-pilots flying a plane together on the same mission. All that differentiates their actions is their unspoken intentions. Critics of DDE want to ask: Can the intentions of the two pilots really make a difference in whether they are morally allowed to fly the mission? Those who argue for the irrelevance of intentions draw a sharp distinction between the moral justification of an act and whether an agent is motivated by that justification. The permissibility of an act depends on whether there are sufficient reasons for allowing it. For a given act, such reasons may exist, even if an agent does not act on the basis of them and, indeed, even if an agent is unaware of them. Similarly, there may be sufficient reasons to prohibit an action, regardless of whether an agent takes those reasons into account. As Scanlon writes, “[W]hat makes an action wrong is the consideration or considerations that count decisively against it, not the agent’s failure to give these considerations proper weight.”51 This distinction between the justification of an act and the motivation for it might ground an argument to defend President Trump’s travel ban.52 Suppose that a different president, one who had made no statements vilifying or demeaning Muslims, reviewed the national security threat assessments provided to the Trump administration and concluded that the suspensions were morally justified. And suppose that judgment was correct, that there were, in fact, sufficient moral reasons to suspend entry temporarily from those countries. Under those circumstances, according to this argument, the fact that President Trump acted on the basis of religious animus does not determine whether his order was permissible. Consider another example, offered by Richard Fallon, involving a religiously motivated legislature that passes a law to codify

Official Intentions and Political Legitimacy

215

the Sixth Commandment’s prohibition on murder.53 Even if the correct, or most reasonable, theory of legitimacy rejects religious justifications as sufficient reasons for exercising political power, there are independent public reasons to support a prohibition on murder, which is surely permissible and indeed even required. The legislature’s intentions to promote a particular religious view are thus irrelevant to the law’s moral (and legal) legitimacy. To be sure, for all these examples, defenders of the permissibility objection agree that intentions may be indirectly relevant to the permissibility of conduct.54 The terror bomber is likely to kill more civilians than the tactical bomber, which is a reason to prevent the terror bomber from flying missions. A president who hates religious minorities is more likely to take actions that harm them, which is a reason for courts to apply close scrutiny to his purported justifications.55 And a legislature with theocratic impulses may have a tendency to pass laws that are not morally justified, which is also a reason for applying heightened scrutiny to its decisions. But while all of these are reasons to be concerned about wrongful intentions, none of them shows that the agent’s intentions are directly relevant in the way that the objection denies. People who are badly motivated might have a propensity to act in ways that are morally unjustifiable, but that does not mean that they have done so for any particular action. Proponents of the objection will also agree that intentions are relevant to evaluating an agent’s moral character. The terror bomber is culpable for acting on a vicious intention. President Trump is liable to criticism for religious bigotry. And the religiously motivated legislature can be blamed for attempting to impose its religious views on others. The argument here is that moral disapproval of this kind is consistent with holding that the intentions of these agents are irrelevant to whether their actions are permissible. Bad people, motivated by bad reasons, can still do the right—or at least the permissible—thing. If the permissibility objection holds, then theories of political legitimacy that allow for intention-sensitive rules are, to that extent, mistaken. Under a procedural theory, the political process might be used to adopt rules that fail to track political morality. Substantive and liberal theories that incorporate claims about the direct relevance of intentions would be guilty of more immediate

216

Micah Schwartzman

moral error. Either way, the permissibility objection raises a challenge for accounts of legitimacy that make official intentions relevant in explaining why some exercises of political power are impermissible and therefore illegitimate. Denying the Permissibility Objection To the extent the permissibility objection is justified in reflective equilibrium, it is vulnerable to challenges based on cases in which intentions are intuitively relevant to determinations of permissibility. Drawing on a range of such cases—including familiar trolley problems and examples from criminal law, medical ethics, and just war doctrines—some philosophers have argued that the objection is fundamentally mistaken and cannot account for common moral intuitions.56 My own intuition is that examples like Tactical Bomber and Terror Bomber are difficult and provide some support for the claim that intentions are irrelevant or at least do not determine permissibility in some range of cases. Others may have similar reactions to national security claims in the travel ban cases, or in Fallon’s example of a religiously motivated law prohibiting murder. It is possible, however, that these cases can still be explained by an intention-sensitive view. The argument is that intentions are directly relevant and that they provide pro tanto reasons for prohibiting actions based on wrongful intentions. But in the cases surveyed here, those reasons are not sufficiently powerful or decisive given the nature of the harms that otherwise wrongful actions would prevent. If acting with bad intentions will contribute to ending a just war, or preserving national security against threats of terrorism, or allowing the state to punish murder, then perhaps it is permissible, all things considered, to allow the relevant actions despite the wrongful intentions of those who perform them. An important criticism of the permissibility objection is that the examples often used to motivate it have various confounding factors, including what Matthew Liao has called the “must act” factor.57 These are cases of moral urgency in which an agent is usually forced to confront a matter of life or death. In such cases, there is often only one permissible response, which is to save lives. The cases also typically require agents to make a single decision and to do so immediately and under pressure, without time for reflection

Official Intentions and Political Legitimacy

217

and deliberation. There are no chances to exercise discretion, to think about the reasons for an action, or to revisit it in the future. This criticism of the permissibility objection—that it relies too heavily on cases involving confounding factors—suggests that the objection might be open to counter-examples that lack such factors. In other work, I have argued that examples of employment discrimination might better track our intuitions about various types of official decisions. Such examples often involve actions by repeat players who have time to deliberate about their decisions and who exercise discretion in distributing benefits and burdens.58 Along similar lines, consider the following example: Prejudiced Diplomat: A diplomat is assigned to issue a certain number of visas. With one visa left to give, the diplomat must decide between two otherwise qualified applicants, one Christian and one Muslim. The diplomat has a negative view of Muslims and so gives the visa to the Christian applicant.

In this example, the diplomat has discretion to choose between the two applicants. There might be various legitimate reasons to prefer one over the other. Perhaps one has stronger in-country family connections, or has special skills, or presents a lower risk of imposing financial burdens on the state. If it turns out there are no legitimate reasons to prefer one applicant over the other, the diplomat might resort to chance—a lottery or flip of a coin—to reach a fair outcome. But whatever the available reasons (or lack thereof), the diplomat may not act on the basis of religious prejudice. The intention to discriminate on the basis of religion makes the diplomat’s decision impermissible. And if that is correct—if the diplomat’s intentions are relevant to whether it is permissible to give the Christian applicant a visa—then this case is a counterexample to the permissibility objection. Let me anticipate several responses to examples like Prejudiced Diplomat in which agents with discretion discriminate on wrongful grounds. The first is that the diplomat has role-based obligations that limit the reasons on which the diplomat is permitted to act.59 The problem with this response is that the diplomat’s role is not what drives the exclusion of religious prejudice as a ground for permissible decision-making. There is a more general moral

218

Micah Schwartzman

prohibition on acting for this reason that applies to the diplomat as it would to many other roles, both public and private. In an example involving an ordinary private employer deciding between two otherwise qualified job candidates, the same moral exclusion would hold. And if one thought that some employers are entitled to act on the basis of religious prejudice, then we might consider examples involving racial prejudice and reach the same result. Another, more direct, response is that intentions are not necessary to explain the wrongness of the diplomat’s discriminatory act. Instead, according to this argument, what makes the diplomat’s act wrong is the denial to the Muslim applicant of equal opportunity and access to important social goods.60 It does not matter whether that denial was intentional or not. Suppose a diplomat gives visas to Christian applicants in disproportionate numbers because of an unconscious bias against Muslims. The diplomat might act without any overt intention to discriminate, but Muslims would still be harmed by the denial of access to visas. And they could rightly complain that they have not received equal treatment, even if no one intended to harm them. The answer to this response is that even if intentions are not necessary to determine the impermissibility of acts, they can nevertheless be sufficient. The way to show this is to isolate the role of intentions by offering examples that strip away other wrongmaking properties. Adam Slavny and Tom Parr have pursued this strategy by offering cases of what they call “harmless discrimination.”61 Consider the following example: Visa Lottery: A diplomatic official is tasked with assigning visas to a limited number of otherwise qualified immigrants. Acting on the basis of hostility towards Muslims, the official rigs a lottery system with the aim of selecting a disproportionate number of nonMuslims. But when the lottery goes into effect, it fails to operate as the official planned and produces the same distribution of visas that would have resulted from a lottery that had not been rigged.62

The official’s action in this case is impermissible for the same reason as in Prejudiced Diplomat, namely, that the official acts on an intention that denies the equal moral worth of Muslims. In Visa Lottery, the official fails to harm anyone, but that does not make

Official Intentions and Political Legitimacy

219

the official’s action permissible. Here the official aims to produce a harm on the basis of a wrongful intention. And what explains the wrong of this attempt to discriminate is the nature of that underlying intention. Proponents of the permissibility objection might be tempted to explain away this example by appealing to the predictive significance of the official’s intention. What makes it wrong to rig the lottery, on this view, is that the official is likely to produce harmful effects. But this is not the most direct or straightforward explanation of why the official’s action is impermissible. To see this, consider this example given by Slavny and Parr: Racist Voter: A firm appoints new staff by taking a vote amongst existing members. One voter wants to reject a particular applicant because of her skin color. He knows that his vote will not affect the result because his co-workers have already agreed to appoint her. Nevertheless, he votes for rejecting the candidate.63

This is another example of harmless discrimination in which an agent nevertheless acts impermissibly. But unlike in Visa Lottery, this is not a case of attempt or mistake of fact. Here the racist voter knows that his action will have no harmful effect, and yet I think the proper response to this example is to say that the voter has acted impermissibly. The voter is not only morally culpable for acting on racist intentions. He can certainly be criticized for demonstrating bad moral character, but he also does something that he is not morally permitted to do, namely, cast a racist vote. And if that is right, the example shows that an agent’s intentions are directly relevant to the permissibility of an act, which means the permissibility objection is mistaken. A final point about Prejudiced Diplomat, Visa Lottery, and Racist Voter is that none of these examples implicate expressive harms. Such harms may be sufficient to make actions impermissible, but they are not necessary. These examples show that acts of discrimination (or attempted discrimination) can be wrong even if they do not convey messages that disparage or demean others. Agents may act impermissibly on the basis of prejudicial intentions without expressing those intentions publicly. It is possible to discriminate against others wrongfully even if no one else knows it.

220

Micah Schwartzman

Of course, in many cases of discrimination, the wrongness of the act will be overdetermined. As noted above, in the travel ban litigation, courts may have been concerned about the predictive significance of the president’s intentions. They could also be interpreted as attempting to prevent expressive harms resulting from the president’s public denigration of Muslims. But the courts’ decisions to block the president’s orders based on the wrongfulness of his intentions can also be taken at face value. Those decisions can be supported by accounts of political legitimacy that incorporate official intentions as directly relevant to the permissibility of the exercise of political power. Such accounts might be challenged on the basis of the permissibility objection. But that objection is not sustained in reflective equilibrium. And indeed, some of the most powerful counter-examples involve discrimination of the type at issue in the travel ban litigation. The permissibility objection thus seems weakest in precisely those cases where it is most likely to be invoked to explain away the relevance of official intentions. The Problem of Moral Taint Accounts of political legitimacy that incorporate intentionsensitive principles must address the problem of moral taint. If officials take some action that is impermissible solely because of the wrongfulness of their intentions, can they remedy the impermissibility of their action by repeating it on the basis of legitimate intentions? Or is their future conduct tainted by what they have previously done? And if so, how can they remove the taint? Such questions are sometimes posed in the form of an objection to intention-sensitive accounts. A famous example in American constitutional law comes from Palmer v. Thompson, which involved a decision by the city of Jackson, Mississippi, to close five public swimming pools rather than racially integrate them.64 In a much-criticized decision, the Supreme Court held that the city had not acted unconstitutionally by denying equal protection of the laws, even though there was clear evidence that the city’s decision was motivated by racial prejudice. As one constitutional scholar wrote shortly after the case was decided, “Almost everyone in Jackson, Mississippi, knew that the city closed its public swimming pools solely to avoid integration.”65 But the Court dismissed

Official Intentions and Political Legitimacy

221

what everyone knew, in part because of objections to relying on the motives of public officials. Writing for a bare majority, Justice Black argued that: [T]here is an element of futility in a judicial attempt to invalidate a law because of the bad motives of its supporters. If the law is struck down for this reason, rather than because of its facial content or effect, it would presumably be valid as soon as the legislature or relevant governing body repassed it for different reasons.66

Implicit within this argument from futility is an objection to moral taint. Suppose a legislature passes a law based on animus, and a court invalidates it. Now, at some point in the future, the legislature repasses the same law on the basis of legitimate reasons. A court might hold that the second law is tainted by the first and is for that reason illegitimate. And it might do the same for later iterations of the law. Technically, there would be nothing futile about this course of action. But, as the Palmer majority suggests, it would be absurd, or at least beyond the power of a court exercising constitutional review, to continue in this way perpetually. A court could not invalidate an otherwise permissible law that was motivated by legitimate reasons solely because, sometime in the past, the legislature had enacted an earlier version based on illicit motivations. In a concurring opinion, Justice Blackmun found “disturbing” the possibility that “the city was to be ‘locked in’ with its pools for an indefinite time in the future . . . just because at one time the pools of Jackson had been segregated.”67 Thus, in raising its futility objection, the Palmer Court implicitly rejected the idea that an earlier law could forever taint, and therefore make impermissible, later and otherwise justifiable reenactments. An initial response to this moral taint objection is that there is no reason to assume there will always be legitimate reasons to support reenacting a policy that was previously adopted on the basis of wrongful intentions.68 In Palmer, the city had no legitimate reasons for closing its pools, and it is not obvious that it would ever have had such reasons. As critics of Palmer have suggested, the Court should have required the city of Jackson to revisit its decision and show that it was, in fact, possible to adopt its policy without acting from racial hostility or the appeasement

222

Micah Schwartzman

of it.69 A remedy of that kind would have been neither futile nor absurd—indeed, far from it. When officials have acted from wrongful intentions, the predictive significance of their intentions may also provide reason to be suspicious about the permissibility of their future actions. As we have seen, wrongful intentions can be directly relevant to the impermissibility of conduct, but they can also be indirectly relevant in providing evidence about how officials are likely to act in the future. That is one reason why courts may apply a doctrine of taint. When courts have determined that officials were improperly motivated, they are justified in remaining skeptical about later conduct undertaken by the same officials.70 For example, in the travel ban cases, lower courts that invalidated President Trump’s first executive order determined that the president had acted on the basis of wrongful intentions. Those intentions were directly relevant to the illegitimacy of his actions. But when President Trump issued his second and third executive orders, his actions were rightfully tainted by evidence of past impropriety. Courts were suspicious that the religious animus that motivated the first order was still present in the adoption of the later orders. That suspicion provided grounds for applying heightened scrutiny to the reasons proffered by the government as legitimate justifications for its actions. When courts examined those reasons and found them wanting, that conclusion only reinforced their determination that the later orders were not merely tainted by prior illegitimate conduct but were themselves based on persistent religious animus.71 The intuition behind the moral taint objection is a powerful one. At some point, with changes of circumstance and enough passage of time, it must be possible for public officials to enact policies that are otherwise permissible, even if the same, or similar, policies had been previously enacted on the basis of wrongful intentions. But when officials have acted impermissibly and incurred moral taint, they carry a burden to show that their new actions are morally permissible. They can satisfy that burden by taking a number of steps. First, and most obviously, they can stop acting on wrongful intentions. As the Fourth Circuit observed in the travel ban litigation, “President Trump could have removed the taint of his prior troubling statements; for a start he could have ceased publicly disparaging Muslims.”72 Furthermore, officials

Official Intentions and Political Legitimacy

223

can recognize past moral wrongs, seek remedies for those disparaged and harmed, and issue appropriate apologies. In developing new policies, they can engage sincerely in public deliberation about permissible purposes and offer evidence to show that the policies in question serve those purposes. And they can administer those policies over time in a manner consistent with their stated purposes. In all these ways, officials acting in good faith can build credibility for their claims that the policies they have pursued are based on permissible reasons and no longer tainted by prior illegitimate exercises of political power. Conclusion When President Trump signed his first executive order suspending entry of foreign nationals from seven Muslim-majority countries, almost everyone knew that his purpose was to make good on his campaign promise to ban Muslims from traveling to the United States. But when courts held that the travel ban was unconstitutional because it was motivated by religious animus, they raised a difficult and important question about whether, and to what extent, the intentions of political officials are relevant to the legitimacy of their actions. I have suggested that various conceptions of political legitimacy—ranging from narrowly procedural to fully substantive—can incorporate intention-sensitive principles. And against the permissibility objection, I have argued that such principles can be sustained in reflective equilibrium. Theories of legitimacy that make official intentions directly relevant are not, for that reason, mistaken. And neither are judges who engage in motive inquiries and other forms of constitutional review that are morally justified by intention-sensitive accounts of political legitimacy. Such accounts rightly focus on how wrongful intentions can make actions impermissible, on how past wrongful conduct can taint future actions, and on how public officials can exercise power permissibly by recognizing past wrongs and acting credibly on the basis of legitimate reasons.

224

Micah Schwartzman Postscript

As this volume was in production, the US Supreme Court handed down its decision in Trump v. Hawaii, which reversed the lower courts and upheld the third iteration of the president’s travel ban against statutory and constitutional challenges.73 The Court, with a five-to-four majority, swept aside overwhelming evidence of President Trump’s hostility to Muslims. Instead, applying its most deferential standard of review, the Court held that “because there is persuasive evidence that the [Proclamation] has a legitimate grounding in national security concerns, quite apart from any religious hostility, we must accept that independent justification.”74 Rejecting this conclusion, four dissenters agreed with the lower courts that the Proclamation was motivated by religious animus in violation of the First Amendment. Justice Sotomayor wrote that even under rational basis review, evidence of the president’s prejudice was so powerful that “it simply cannot be said that the Proclamation has a legitimate basis.”75 In Trump v. Hawaii, the majority and the dissenters disagreed about the relevance of intentions in evaluating the actions of executive officials. Where the majority viewed evidence of animus as nearly irrelevant, at least in this context, the dissenters found that it undermined the government’s credibility in asserting a legitimate purpose. But the dissenters were on stronger ground here. For the reasons discussed above, including that moral taint should raise heightened suspicion of continuing impermissible conduct, intentions matter for the legitimacy of official actions, even and especially the president’s. Acknowledgments For helpful comments and discussion, I thank Kimberly Ferzan, Leslie Kendrick, James Nelson, Fred Schauer, Richard Schragger, and Nelson Tebbe. Notes 1. The countries were Iraq, Iran, Libya, Somalia, Sudan, Syria, and Yemen. See Executive Order 13,769, “Protecting the Nation from Foreign

Official Intentions and Political Legitimacy

225

Terrorist Entry into the United States,” 82 Fed. Reg. 8977 (January 27, 2017). 2. See Adam Litpak, “Court Refuses to Reinstate Travel Ban, Dealing Trump Another Legal Loss,” New York Times, February 9, 2017. 3. Executive Order 13,780, 82 Fed. Reg. 13,209 (March 6, 2017). 4. For the procedural history of the first and second orders, see International Refugee Assistance Project v. Trump, 857 F.3d 553, 572–75 (2017) [hereinafter IRAP I]. In the interest of full disclosure, I helped to organize and author an amicus brief arguing that the travel ban was motivated by animus and was therefore unconstitutional under the First and Fourteenth Amendments. See Brief of Constitutional Law Scholars as Amici Curiae in Support of Plaintiffs–Appellees and Affirmance, 4th Cir. No. 8:17-cv-00361 (April 19, 2017). 5. Ibid., 572. 6. Ibid., 594. 7. Ibid., 601n.22. 8. “Enhancing Vetting Capabilities and Processes for Detecting Attempted Entry into the United States by Terrorists or Other Public-Safety Threats,” 82 Fed. Reg. 45,161 (September 24, 2017). 9. International Refugee Assistance Project v. Trump, 833 F.3d 233 (4th Cir. 2018) [hereinafter IRAP II]. 10. See T. M. Scanlon, Moral Dimensions: Permissibility, Meaning, Blame (Cambridge, MA: Harvard University Press, 2008), 8–88; Deborah Hellman, When Is Discrimination Wrong? (Cambridge, MA: Harvard University Press, 2008), 138–168; F. M. Kamm, “Failures of Just War Theory: Terror, Harm, and Justice,” Ethics 114 (2004): 666; Judith Jarvis Thomson, “Physician-Assisted Suicide: Two Moral Arguments,” Ethics 109 (1999): 517. 11. See Richard Fallon, “Constitutionally Forbidden Legislative Intent,” 130 Harvard Law Review 523 (2016). For a response to Fallon and for some doubts about the applicability of the permissibility objection in various areas of constitutional law, see Michael C. Dorf, “Even a Dog: A Response to Professor Fallon,” 130 Harvard Law Review Forum 86 (2016). 12. See, e.g., John Hart Ely, “Legislative and Administrative Motivation in Constitutional Law,” 79 Yale Law Journal 1205 (1970); Paul Brest, “Palmer v. Thompson: An Approach to the Problem of Unconstitutional Legislative Motive,” 1971 Supreme Court Review 95; Lawrence H. Tribe, “The Mystery of Motive, Private and Public: Some Notes Inspired by the Problems of Hate Crime and Animal Sacrifice,” 1993 Supreme Court Review 1; Elena Kagan, “Private Speech, Public Purpose: The Role of Government Motive in First Amendment Doctrine,” 63 University of Chicago Law Review 413 (1996); Larry Alexander, “Rules, Rights, Options, and Time,” Legal Theory 6 (2000): 391–404.

226

Micah Schwartzman

13. See, e.g., Jeremy Waldron, Law and Disagreement (Oxford: Oxford University Press, 1999), 119–146; Antonin Scalia, A Matter of Interpretation (Princeton, NJ: Princeton University Press, 1997), 16–23; Richard Ekins, The Nature of Legislative Intent (Oxford: Oxford University Press, 2016). 14. See, e.g., Christian List and Philip Pettit, Group Agency: The Possibility, Design, and Status of Corporate Agents (Oxford: Oxford University Press, 2011); Scott Shapiro, “Massively Shared Agency,” in Rational and Social Agency, ed. Manuel Vargas and Gideon Yaffe (Oxford: Oxford University Press, 2014), 257–293. 15. See, e.g., IRAP I, 264 (Niemeyer, J., dissenting). 16. See, e.g., Larry Simon, “Racially Prejudiced Governmental Actions: A Motivation Theory of the Constitutional Ban Against Racial Discrimination,” 15 San Diego Law Review 1041 (1978): 1065–1127; Aziz Z. Huq, “What Is Discriminatory Intent,” 103 Cornell Law Review 1211(2018). 17. See Fallon, “Constitutionally Forbidden Legislative Intent,” 580. 18. IRAP I, 857 F.3d. at 600–602. 19. In moral philosophy, Scanlon has offered a general ethical account but not applied it to matters of political philosophy. Within legal philosophy, there are sophisticated accounts of criminal law and antidiscrimination law that give extensive consideration to the relevance of intentions. See, e.g., Victor Tadros, The Ends of Harm: The Moral Foundations of Criminal Law (Oxford: Oxford University Press, 2011), 139–166; Hellman, When Is Discrimination Wrong, 138–168. But I am not aware of any systematic treatment of this issue in political philosophy. Fallon has come the closest to providing such an account by following Scanlon in denying the relevance of legislative intentions to the moral permissibility of laws. Fallon, “Constitutionally Forbidden Legislative Intent,” 564. Fallon’s focus is on the justifiability of statutes, rather than on the legitimacy of official actions more generally. But his argument may point the way toward a more general skeptical account. 20. Here I follow Scanlon, Moral Dimensions, 10. 21. And here I follow Tadros, The Ends of Harm, 139, 148. 22. IRAP II, 27; see also 219–220 (Harris, J., concurring). 23. See, e.g., A. John Simmons, Justification and Legitimacy: Essays on Rights and Obligations (Cambridge: Cambridge University Press, 2001), 130; David Copp, “The Idea of a Legitimate State,” Philosophy & Public Affairs 28 (1999): 3–45; Richard Fallon, Law and Legitimacy in the Supreme Court (Cambridge, MA: Harvard University Press, 2018), 23. 24. A. John Simmons, Moral Principles and Political Obligations (Princeton, NJ: Princeton University Press, 1979), 195–200; Copp, “Idea of a Legitimate State,” 27–28. 25. Arthur Isak Applbaum, “Legitimacy without the Duty to Obey,” Philosophy & Public Affairs 38 (2010): 215–239.

Official Intentions and Political Legitimacy

227

26. N. P. Adams, “Institutional Legitimacy,” Journal of Political Philosophy 26 (2017): 84–102. 27. Simon Cabulea May, “Democratic Legitimacy, Legal Expressivism, and Religious Establishment,” Critical Review of International Social and Political Philosophy 15 (2012): 220. 28. For the distinction between ideal and minimal conceptions of legitimacy, see Fallon, Law and Legitimacy in the Supreme Court, 24–35. 29. See, e.g., Waldron, Law and Disagreement; Thomas Christiano, The Rule of the Many (Boulder, CO: Westview Press, 1996). 30. See, e.g., David Estlund, Democratic Authority (Princeton, NJ: Princeton University Press, 2008). For a survey of procedural theories, see Fabienne Peter, Democratic Legitimacy (New York: Routledge, 2009), 65–74. 31. In American constitutional law, the classic statement of this type of view is John Hart Ely, Democracy and Distrust: A Theory of Judicial Review (Cambridge, MA: Harvard University Press, 1980), 136–147; see also John Hart Ely, “The Centrality and Limits of Motivation Analysis,” San Diego Law Review 15 (1978): 1155–1161. 32. May, “Democratic Legitimacy,” 221. 33. Ibid., 223–224. 34. Again, see, e.g., Ely, Democracy and Distrust, ch. 6. 35. John Rawls, Political Liberalism (New York: Columbia University Press, 1996), 217. 36. See John Rawls, “The Idea of Public Reason Revisited,” in Collected Papers, ed. Samuel Freeman (1999), 573–615. See also Jonathan Quong, “On the Idea of Public Reason,” in A Companion to Rawls, ed. Jon Mandle and David A. Reidy (Chichester: Wiley Blackwell, 2014), 265–280. 37. Rawls, Political Liberalism, 224. 38. Ibid., 223–225. 39. Rawls, “Idea of Public Reason Revisited,” 576–577. 40. It is a difficult question whether a Rawlsian view of political legitimacy incorporates a requirement that public officials be motivated by public reasons. In the past, I have argued against a strong motivational requirement as unnecessary for liberal legitimacy and inconsistent with the assumption that reasonable citizens will be motivated in deeper ways by a diversity of religious and philosophical doctrines. See Micah Schwartzman, “The Sincerity of Public Reason,” Journal of Political Philosophy 19 (2011): 387–390; also see David Reidy, “Rawls’s Wide View of Public Reason: Not Wide Enough,” Res Publica 6 (2000): 57–63. But the idea of public reason does require that public officials be motivated to comply with its demands. This might yield a second-order motivational requirement that permits officials to act on nonpublic reasons, but only when they sincerely believe the policies they support are justified by sufficient public

228

Micah Schwartzman

reasons. For further discussion, see Micah Schwartzman, “Must Laws Be Motivated by Public Reasons?” in Courts and Public Reason, ed. Silje Langvatn et al. (Cambridge: Cambridge University Press, forthcoming); see also Wojciech Sardurski, “Searching for Illicit Motives: Constitutional Theory of Freedom of Speech, Equal Protection, and Separation of State and Religion,” Sydney Law School Research Paper No. 14/61. 41. See sources cited in note 10. 42. This part and the next draw on arguments developed in Schwartzman, “Must Laws Be Motivated by Public Reason,” parts 3–4. 43. Scanlon, Moral Dimensions, 12–13, 62–73. 44. Ibid., 30–32. 45. Fallon, “Constitutionally Forbidden Legislative Intent,” 577; see also Gordon G. Young, “Justifying Motive Analysis in Judicial Review,” 17 Williams & Mary Bill of Rights Journal (2008): 191–261. 46. For the idea of expressive harms, see Elizabeth S. Anderson and Richard Pildes, “Expressive Theories of Law: A General Restatement,” 148 University of Pennsylvania Law Review, 1503, 1527–31 (2000). 47. See Scanlon, Moral Dimension, 72; Hellman, When Is Discrimination Wrong, 34–41. 48. Scanlon has also argued that for an act to be morally permissible, an agent must be able to choose whether to undertake it. According to Scanlon, agents cannot choose their intentions, and so intentions cannot be directly relevant to moral permissibility. I shall leave aside this argument, which has received serious and, in my view, decisive criticisms. See Dana Kay Nelkin and Samuel C. Rickless, “Three Cheers for Double Effect,” Philosophy and Phenomenological Research 89 (2014): 141–143; Jeff McMahan, “Intentions, Permissibility, Terrorism, and War,” Philosophical Perspectives 23 (2009): 354–356; Halvard Lillehammer, “Scanlon on Intention and Permissibility,” Analysis 70 (2010): 580–583; Tadros, Ends of Harm, 145–146. On judgments about permissibility in reflective equilibrium, see Nelkin and Rickless, “Three Cheers,” 131; McMahan, “Intentions, Permissibility, Terrorism, and War,” 369–370. 49. For similar versions of DDE, Scanlon, Moral Dimensions, 1–2, 28– 29; Hellman, When Is Discrimination Wrong, 151–152; Thomson, “PhysicianAssisted Suicide,” 510. 50. For extensive discussion of these examples, see, e.g., Scanlon, Moral Dimensions, 28–32; Tadros, Ends of Harm, 148–149; Nelkin and Rickless, “Three Cheers,” 126; McMahan, “Intentions, Permissibility, Terrorism, and War,” 345–348. See also Warren S. Quinn, “Actions, Intentions, and Consequences: The Doctrine of Double Effect,” Philosophy & Public Affairs 18 (1989): 334–351; Matthew Liao, “Intentions and Moral Permissibility: The Case of Acting Permissibly with Bad Intentions,” Law & Philosophy 31

Official Intentions and Political Legitimacy

229

(2012): 704–706; Alec Walen, “The Doctrine of Illicit Intentions,” Philosophy & Public Affairs 34 (2006): 39–67. 51. Scanlon, Moral Dimensions, 22–23. 52. After announcing his first travel ban, President Trump issued a press release stating: “My policy is similar to what President Obama did in 2011 when he banned visas for refugees from Iraq for six months. The seven countries named in the Executive Order are the same countries previously identified by the Obama administration as sources of terror.” See Donald J. Trump, “Statement Regarding Recent Executive Order Concerning Extreme Vetting,” The White House (January 29, 2017). But see Jon Finer, “Sorry, Mr. President: The Obama Administration Did Nothing Similar to Your Immigration Ban,” Foreign Policy (January 30, 2017). 53. Fallon, “Constitutionally Forbidden Legislative Intent,” 569. 54. See, e.g., Scanlon, Moral Dimensions, 87. 55. Cf. Fallon, “Constitutionally Forbidden Legislative Intent,” 577. 56. McMahan, “Intentions, Permissibility, Terrorism, and War,” 347– 359; Tadros, Ends of Harm, 155–160; Nelkin and Rickless, “Three Cheers,” 141–147. 57. Liao, “Intentions and Moral Permissibility,” 712–717. 58. Schwartzman, “Must Laws Be Motivated by Public Reason.” See also Adam Slavny and Tom Parr, “Harmless Discrimination,” Legal Theory 15 (2016): 100–114. 59. This response is suggested by Scanlon, Moral Dimensions, 70. 60. See Scanlon, Moral Dimensions, 73; Hellman, When Is Discrimination Wrong, 141–156. 61. Slavny and Parr, “Harmless Discrimination,” 3. 62. This example is a variant of Deborah Hellman’s Orchestra Director. See Hellman, When Is Discrimination Wrong, 161. 63. Ibid., 107. 64. 403 U.S. 217 (1971). 65. Paul Brest, “Palmer v. Thomson: An Approach to the Problem of Unconstitutional Legislative Motive,” 1971 Supreme Court Review 95. 66. Palmer, 403 U.S. at 225. The Court also raised an epistemic objection to ascertaining the collective intentions of the Jackson city council. There was evidence of racial animus, the Court acknowledged, but there was also evidence that the city was motivated by safety and financial concerns. Ibid., 224–225. But this argument was both facile and implausible under the facts of Palmer. The city only had safety and financial concerns because of racial hostility toward integration, which the Court conceded would not “validate an otherwise impermissible state action.” Ibid., 226. See also Brest, Palmer, 130–133.

230

Micah Schwartzman

67. 403 U.S. at 230. In the travel ban cases, dissenting judges in the Fourth Circuit expressed a similar view: “It is undeniable that President Trump will continue to need to engage in foreign policy regarding majority-Muslim nations, including those designated in the Proclamation. Yet, the district court’s opinion presupposes that the Proclamation is tainted by prior campaign-trail statements and prior executive orders, clearly indicating that future actions might also be subject to the same challenges made today.” IRAP II, 265 (Niemeyer, J., dissenting). 68. See Ely, Democracy and Distrust, 138–139; Brest, Palmer, 125. 69. Brest, Palmer, 133. 70. Ibid., 126. 71. IRAP II, 51–52 (Gregory, J.) (“The contradiction between what the Proclamation says—that it merely reflects the results of a religion-neutral review—and what it does ‘raises serious doubts’ about the Proclamation’s proffered purpose, and undermines the Government’s argument that its multi-agency review cured any earlier impermissible religious purpose.”). 72. IRAP II, 49 (Gregory, J.). 73. 138 S. Ct. 2392 (2018). 74. Ibid., 2421. 75. Ibid., 2442 (Sotomayor, J., dissenting).

8 THE POLITICAL LEGITIMACY OF INTERNATIONAL NGOs JENNIFER C. RUBENSTEIN

Large-scale, Western-based, international non-governmental humanitarian and development organizations (INGOs) such as Oxfam, CARE, and Doctors Without Borders are not simply collections of well-intentioned do-gooders.1 Nor are they merely tools of the entities that fund them, or the governments of the countries where they work. They are political actors that wield considerable political power. For example, INGOs frequently participate in the decision-making processes of international organizations; they shape which issues get put on the international agenda and how those issues are framed; they help to discursively construct the social identities of the people they serve.2 More concretely, INGOs provide and control access to essential goods and services, such as education, medical care, water, food, sanitation, and information to large populations for years or decades. As one employee of the INGO Médecins Sans Frontières/Doctors Without Borders (MSF) recounted: in exceptional settings, such as refugee camps, governments often keep their distance, delegating the health administration of these populations to international agencies and NGOs. This gives MSF the advantage of “extra-territoriality.” It has the margin for manoeuvre it needs to take rapid control of all the stages in the response to an epidemic, including setting up and exploiting a surveillance system for the ongoing collection of health data, epidemic investigation (with diagnosis confirmation, when necessary using

231

232

Jennifer C. Rubenstein

biological tests) and the introduction of measures for reducing the number of infections and mortality (early detection and treatment of cases, isolation and immunization and vector control).3

As Anthony Donini writes, “the humanitarian organisations are also part of governance, if not of government. Major NGOs can be powerful vectors in shaping public opinion and government policy.”4 Over the past twenty-five years, INGOs, INGO scholars, and critics of INGOs have identified desiderata for INGOs that are responsive to the forms of power that INGOs exercise. Among other things, they have argued that INGOs should be “accountable,” “responsive,” “credible,” “representative,” and “transparent.”5 INGOs have sought to enact these desiderata in various ways. For example, they have created deliberative forums, feedback sessions, and complaints mechanisms; they have tried to establish ombudstype positions; they have promulgated new technical standards and altered their hiring and promotion procedures so that they are more fair to national-level staff.6 While the effectiveness of these efforts has been debated, the efforts themselves reflect widespread agreement that these desiderata are relevant to INGOs. As part of these discussions, INGOs, INGO scholars, and critics of INGOs have also taken up another concept, albeit less frequently: the concept of political legitimacy—by which I mean normative political legitimacy, understood, roughly, as the moral right to rule.7 This concept, as it has been developed in the political philosophy literature, has three dimensions: criteria for political legitimacy, such as, in the case of INGOs, accountability and good performance; the fact that these criteria represent minimum thresholds rather than ideals; and the thing that these criteria are minimum thresholds for: the moral right to rule.8 Most contemporary English-language discussions of INGO political legitimacy have focused narrowly on the first of these three dimensions: criteria that INGOs must meet to be normatively politically legitimate. They have said little or nothing about the political and ethical implications of those criteria being minimum thresholds for the moral right to rule. In contrast, the political philosophy literature on political legitimacy does explore all three dimensions of the concept of political legitimacy. It also addresses the question of whether the concept of political legitimacy can be

The Political Legitimacy of International NGOs

233

extended to non-state actors.9 However, it has mostly dealt with non-state actors as a group and/or non-state actors other than INGOs.10 As a result, the political philosophy literature on political legitimacy has also not elucidated the implications of the full concept of political legitimacy for INGOs. This chapter takes up this task. I argue that claims about INGO political legitimacy that bring the full “three-dimensional” concept to bear have the potential to contribute significantly to discussions of the roles, rights, and responsibilities of INGOs, aid recipients, and other subjects of INGO rule, in at least four ways. First, arguments about INGO political legitimacy flip the script of typical discussions of INGOs as charitable do-gooders: While discussions of INGOs as charitable do-gooders focus on how much good INGOs can or will do, discussions of INGO political legitimacy emphasize that they engage in a kind of rule and focus on whether they are morally justified in ruling. Second, claims that INGOs are politically legitimate (or not) construe aid recipients and other subjects of INGO rule not as passive victims, but rather as moral and political agents who exercise moral and political judgment and are capable of acting on principle, even if doing so goes against their immediate self-interest. Third, because claims about INGO political legitimacy involve a minimum threshold, they create loci for agreement among people who otherwise disagree about INGOs. Finally, because the INGO literature has focused heavily on articulating criteria of INGO political legitimacy while neglecting other dimensions of the concept (or so I argue), I do not here defend yet more criteria of INGO political legitimacy. Instead, I argue that democratic criteria of INGO political legitimacy—that is, criteria that hold INGOs to democratic norms, such as accountability to those affected—have the potential to work synergistically with other dimensions of the concept of political legitimacy to acknowledge and call forth the moral and political agency of aid recipients and other subjects of INGO rule. By elucidating these contributions that the concept of political legitimacy makes to debates about INGOs, I also clarify the distinctiveness and bite of the concept of political legitimacy more generally, especially in contexts of profound inequality, urgent need,

234

Jennifer C. Rubenstein

and dispersed political authority. While the concept of political legitimacy does not engage with all forms of power that INGOs exercise, it does encourage a direct confrontation with INGO rule. As such, it has some surprising and salutary implications, in terms of the questions that it raises, the political and moral subjectivities and forms of judgment that it calls forth, the types of collective action that it enables, and the way that it keeps focus on, without prejudging, the question of whether INGOs’ operations in any given context are morally justified. Before proceeding to the substance of the argument, I want to briefly note one complexity about terminology (which, like most terminological complexities, is about more than terminology). Many things can be judged legitimate or illegitimate, including children, excuses, questions, and reasons. In these contexts, “legitimate” means rightful, proper, or consistent with relevant rules or standards. My specific focus here is on political legitimacy, which pertains to entities that exercise, or decisions that involve, significant political power. INGO and INGO commentators often discuss INGO “legitimacy” without the modifier “political.” It does not necessarily follow, however, that the kind of legitimacy at issue in these contexts is non-political. INGOs frequently have a political interest in downplaying the political dimensions of their activities. For example, humanitarian INGOs often describe themselves as “neutral” and “impartial” because being perceived as apolitical sometimes enables them to gain access to civilian populations in conflict settings.11 INGOs that adopt a political strategy of neutrality have an incentive to avoid describing their legitimacy (or anything else about themselves) as “political.” More generally, INGOs that wish to characterize themselves as solely moral and not political actors have an incentive to reject the idea that their legitimacy has anything to do with the exercise of political power. This is why, in order to determine whether political legitimacy is at issue in a particular context, it is necessary to look at the activities and relationships involved and ascertain whether these include the exercise of significant power. If they do, then the kind of legitimacy that is needed is political legitimacy, regardless of how the situation is described by the actors involved.

The Political Legitimacy of International NGOs

235

The Conceptual Architecture of Normative Political Legitimacy Let us begin with a reasonably standard definition of normative political legitimacy in the political philosophy literature. According to this definition, political legitimacy is: the right to rule, understood to mean both that institutional agents are morally justified in making rules and attempting to secure compliance with them and that people subject to those rules have moral, content-independent reasons to follow them and/or to not interfere with others’ compliance with them.12

On this definition, political legitimacy can be seen as having a conceptual architecture consisting of three connected but distinct parts. (There are of course other ways to conceive of this architecture, but this one is useful for present purposes.) The first part is the criteria for political legitimacy. These criteria might include democratic procedures (e.g., minimally free and fair elections) or refer to other aspects of a regime (e.g., respect for basic human rights). In the context of INGOs, commonly cited criteria for INGO political legitimacy include being accountable, being transparent, and having close connections with “Southern” NGOs.13 The second part of the concept of political legitimacy is the idea that the criteria indicate not an ideal to be aspired to, but rather a minimum threshold or low bar. An entity that meets the minimum threshold is politically legitimate; an entity that fails to meet it is not. Political legitimacy is thus often contrasted with justice, which is a higher bar. There is widespread, though not universal, agreement about this aspect of political legitimacy in the political philosophy literature.14 The third part of the concept of political legitimacy is what the minimum threshold is a threshold for: the moral right to rule. One could imagine a minimum threshold for many things—say, for getting certified in ethical business practices by a third party, or receiving tax breaks. In the concept of political legitimacy, the minimum threshold is for the moral right to rule. As I discuss further below, while there is widespread agreement that the moral right to rule is

236

Jennifer C. Rubenstein

central to normative political legitimacy in the political philosophy literature, its implications are highly contested. To further elucidate these dimensions of political legitimacy, consider an analogy to doctors and their professional legitimacy. For doctors, the relevant criterion of professional legitimacy is (let us say, for the purposes of this example) medical expertise. The minimal threshold is the amount of expertise necessary to pass medical school and medical board exams. Rather than a moral right to rule, doctors have a professional (and legal) privilege (analogous to a right, for present purposes) to prescribe medicines and perform surgery. Thus, if we focus only on the criteria for being a doctor (medical expertise) and ignore what the criteria are criteria for (prescribing medicine), we bypass a host of questions about the substantive content of doctors’ privileges and responsibilities, as well as the implications of these privileges and responsibilities for patients. While there is of course a substantive connection between the criteria for being a doctor (passing an exam) and doctors’ privileges (prescribing medicine), they are not the same thing. Analogously, in the case of INGO political legitimacy, if we focus only on the criteria that INGOs must meet to be politically legitimate, we miss the implications, for INGOs and their subjects, of INGOs having, or lacking, the moral right to rule. Indeed, without the minimum threshold and moral right to rule aspects of political legitimacy, “criteria” of political legitimacy function more like desiderata for INGOs. In the doctor example, if medical expertise isn’t a minimum threshold that one must meet to exercise the privileges of being a doctor, then it is merely a desideratum. Likewise, accountability (for example) is often mentioned as a criterion for INGO political legitimacy. Within the context of the three-dimensional structure of political legitimacy described above, it can function as a minimum threshold for the moral right to rule. However, without this threshold, it functions more like a desideratum—something that is good for INGOs to have more of, but without teeth as a criterion for whether they are justified in ruling.

The Political Legitimacy of International NGOs

237

Is the Concept of Political Legitimacy Applicable to INGOs? There is widespread agreement among political philosophers that the concept of political legitimacy is applicable to large-scale nonstate actors, a category that is typically assumed to include largescale INGOs. As Hurrell and Macdonald argue: non-state actors such as multinational corporations (MNCs), NGOs, and hybrid institutional actors constituted by both state and nonstate actors, now wield many of these forms of decision-making power [i.e., forms “that have traditionally been subject to legitimating control”] (in fields of law-making, economic development, public service provision, and so on), alongside a plurality of states.15

Nonetheless, if political legitimacy is about the moral right to rule, one might wonder if humanitarian and development INGOs really engage in rule (what Buchanan and Keohane call “making rules” and “securing compliance”) in the relevant sense. I think that the definition of political legitimacy cited in the previous section can be applied to INGOs if we interpret “make rules” and “secure compliance” in specific ways. If “make rules” is taken to refer to supremacy in lawmaking—what Buchanan elsewhere calls “the (credible) attempt to achieve supremacy in the making, application, and enforcement of laws within a jurisdiction”—then few or no INGO activities can be deemed politically (il)legitimate, because INGOs hardly ever self-consciously (much less credibly) strive for lawmaking supremacy.16 A more useful interpretation of “make rules” for the context of INGOs would be something like “significant influence over the making of formal and informal rules and policies.” Likewise, “secure compliance” cannot refer only to enforcement via the threat or use of overt violent force by the police or military. While INGOs sometimes hire armed guards, they typically engage in the more indirect, diffuse, managerial, and discursive forms of power outlined at the outset of this chapter. Bringing these more expansive interpretations of “making rules” and “enforcing compliance” together, the aforementioned definition of political legitimacy is relevant to INGOs when they exercise significant political power in the making or enforcement

238

Jennifer C. Rubenstein

of rules and policies. This, then, is what I mean when I refer to INGO “rule.” While this formulation expands significantly beyond Buchanan and Keohane’s account of rule, it leaves the rest of their definition intact, in particular the idea that people subject to the rule of a politically legitimate entity “have moral, contentindependent reasons to follow them and/or to not interfere with others’ compliance with them.”17 A second potential difficulty with applying the concept of political legitimacy to INGOs involves the idea of a minimum threshold. Such a threshold might seem inappropriate for INGOs because (or insofar as) they operate in extreme and unpredictable situations. When life-saving aid can only be provided via morally compromised means or alongside significant unintended negative effects, it can be extremely difficult to identify “red lines.” This is why some aid workers argue that “everything is open to negotiation.”18 From another perspective, however, it is precisely in these sorts of situations where (higher) standards of justice are unlikely to be met that the lower bar of political legitimacy is most salient. Even if it is difficult to specify in advance substantive criteria of political legitimacy for INGOs (e.g., “do not pay bribes to local militia”), it might be easier to specify adequate procedures for making these judgments (e.g., “consult with local community about, carefully consider, and frequently revisit the decision to pay bribes”). INGO criteria of political legitimacy can also be counterfactual, e.g., it might be that for an INGO to be politically legitimate, it must not displace other entities that could do a better job or are normatively superior in some other way. In short, while specifying plausible minimum thresholds for INGOs’ moral right to rule is challenging, there is no reason to think it is impossible, or indeed significantly more difficult than establishing such thresholds for other types of actors, such as states. How the Concept of Political Legitimacy Is Deployed in the Context of INGOs While the concept of political legitimacy as a minimum threshold for the moral right to rule can plausibly be extended to INGOs, INGO scholars and commentators focus primarily on criteria for political legitimacy, and not on other aspects of the concept.

The Political Legitimacy of International NGOs

239

Sometimes this happens because the concept of political legitimacy is defined vaguely or not at all. Lister notes that “the term ‘legitimacy’ . . . is rarely defined[,]” and that “[l]iterature dealing with NGOs tends to be particularly weak in this area.”19 In other cases, political legitimacy is defined, and in a way that is roughly consistent with the definitions in the political philosophy literature, but the rest of the analysis doesn’t engage with the definition proffered. For example, in “Four Criteria of NGO Legitimacy,” Atack defines legitimacy as an answer to the question, “Who has the right to assert leadership, to organize people, and to allocate resources in the development enterprise?”20 He emphasizes that “[i]t is this concern with the right of different institutions (states or NGOs) to involve themselves in the development process that raises normative questions concerning the legitimacy of particular types of political and social action.”21 This emphasis on the right to exercise political power (to “assert leadership,” “organize people,” and “allocate resources”) is consistent with the political philosophical conception of political legitimacy. However, Atack does not discuss the implications of INGOs having or lacking this right; instead, the rest of the article elucidates four criteria that INGOs must meet in order to be normatively politically legitimate. Sometimes, discussions of political legitimacy in the INGO literature seem to focus on criteria for political legitimacy to the exclusion of other dimensions of the concept because they conflate normative political legitimacy with empirical political legitimacy. Normative political legitimacy (the type of political legitimacy I’ve been discussing so far) pertains to whether an entity, such as an INGO, is actually morally justified in ruling; empirical political legitimacy pertains to whether the members of a group or population that is being ruled believe that the entity ruling them is morally justified in so doing. The difference between normative and empirical political legitimacy can be understood, analytically, as the difference between a first-person and third-person perspective. When one is expressing one’s own views about whether some entity has the moral right to rule, one is typically invoking an account of normative political legitimacy, which includes both a first-order claim about the criteria that the entity must meet to have the moral right to rule and a second-order assessment of whether the entity meets these criteria. In contrast, when one is describing other people’s

240

Jennifer C. Rubenstein

views on these matters (typically, the views of a large group, constituency, or population being ruled), the subject shifts to empirical political legitimacy: what criteria members of that group believe an entity must meet to have the moral right to rule, and their assessment of whether the entity in question meets these criteria.22 These two conceptions of political legitimacy are frequently conflated in the INGO literature. For example, in an article about what he calls the “informal” legitimacy of MSF, Calain argues that “the kind of legitimacy which is frequently referred to in the humanitarian literature appeals more to feelings of ‘natural rights’ or ‘rightfulness of action.’”23 It is not clear, from this definition, whether the “feelings” in question are the author’s, in which case the subject is normative political legitimacy, or those of third parties, in which case the subject is empirical political legitimacy. Likewise, writing about INGOs, Michael Edwards defines legitimacy as “the right to be and do something in society—a sense that an organisation is lawful, admissible and justified in its chosen course of action.”24 While the first part of this definition (the “right”) refers to normative legitimacy, the second part (the “sense”) seems to refer to empirical legitimacy. Both Calain and Edwards put more emphasis on criteria for political legitimacy than they do on the implications of political legitimacy being a minimum threshold for the moral right to rule. One possible reason for this conflation of normative and empirical political legitimacy is that scholars don’t feel comfortable asserting criteria for INGO political legitimacy in a purely normative or prescriptive mode; they prefer to situate their normative claims within the context of a broader discussion of what others (INGOs themselves, donors, more rarely aid recipients) see as the sources of INGO political legitimacy. Another possibility is that INGO scholars and practitioners view empirical political legitimacy as itself a criterion for normative political legitimacy: They think that INGOs are only justified in ruling (and also, at a more practical level, are only able to rule) if the people they rule believe they are so justified. Whatever the reason(s) for it, one effect of this conflation of normative and empirical legitimacy seems to be to downplay questions that only arise in the context of normative political legitimacy, in particular, questions about the rights and duties of INGOs and the people they rule.

The Political Legitimacy of International NGOs

241

In “By What Authority? The Legitimacy and Accountability of Non-Governmental Organizations,” the prominent INGO scholarpractitioner Hugo Slim discusses INGO accountability and legitimacy at length. All of the sections about legitimacy either are called “sources of legitimacy” or they identify or discuss these sources. These sources of legitimacy are somewhat similar to criteria (they include things like credibility, good performance, and relationships, which are often mentioned as criteria for political legitimacy). However, the language of sources downplays the idea that there is a minimum threshold of these things that must be met—or that particular consequences follow when they are not met. Slim defines legitimacy as “the particular status with which an organisation is imbued and perceived at any given time that enables it to operate with the general consent of peoples, governments, companies and non-state groups around the world.”25 While this is a definition of empirical political legitimacy, Slim goes on to discuss both sources of (what I would call) empirical political legitimacy, such as “tangible support,” and sources of (what I would call) normative political legitimacy, such as consistency with international human rights law.26 By focusing on sources of empirical and normative political legitimacy, mingling these together, and construing them as sources rather than criteria, Slim’s discussion bypasses questions about the implications of political legitimacy as a minimum threshold for the moral right to rule. Here is one final example. In Evaluating Transnational NGOs: Legitimacy, Accountability, Representation, Steffek and Han use the language of sources and suggest that the subject of INGO legitimacy is exhausted by the question of what these sources are: In principle, organizational legitimacy can refer to the evaluation of almost all aspects of what an NGO is and does. We do not endeavor to suggest a neat definition of NGO legitimacy in this introductory chapter, but we contend that there are three types of issues that can structure an overview of the current debate. Slightly modifying the typology suggested by Collingwood and Logister (2005, p. 189) we distinguish the following potential sources of NGO legitimacy.27

While there are certainly exceptions, these examples are indicative of what I see as a broader tendency, in the INGO literature, to

242

Jennifer C. Rubenstein

focus on criteria for or sources of normative political legitimacy, while bypassing the implications of political legitimacy being a minimum threshold for the moral right to rule. More Politicized Invocations of Political Legitimacy in the INGO Context Like commentators writing about INGOs, INGOs themselves mention the term “legitimate” and its cognates much less frequently than they do the terms “accountable” and “transparent” (see fig. 8.1).28 When INGOs (or people speaking officially on their behalf) do discuss political legitimacy, they sometimes discuss the sources of INGOs’ normative political legitimacy.29 INGOs also sometimes invoke the conception of political legitimacy as a minimum threshold for the moral right to rule. However, this happens not (only) when INGOs are trying to argue that they are “over the bar” (as one might expect), but also, and often more forcefully, when they are trying to argue that they are under it. In particular, INGOs trying to pressure conventional governments to take on governance tasks that INGOs themselves do not want to take on sometimes argue that governments have normative political legitimacy that INGOs lack. For example, in a speech accepting the Nobel Peace Prize on behalf of MSF in 1999, then-president of MSF International, James Orbinski, argued: Ours is not to displace the responsibility of the state. Ours is not to allow a humanitarian alibi to mask the state responsibility to ensure justice and security. And ours is not to be co-managers of misery with the state. If civil society identifies a problem, it is not theirs to provide a solution, but it is theirs to expect that states will translate this into concrete and just solutions. Only the state has the legitimacy and power to do this.30

In this passage, Orbinski suggests that normative political legitimacy is necessary to address large-scale problems of justice and security, but that MSF does not have this legitimacy; it is therefore not right or proper for MSF to address these problems. MSF frequently finds that once it begins a project in a particular location in response to an acute crisis or need, it is drawn

Accountable Transparent Legitimate (agg) Legitimate Legitimacy

Oxfam

1198 694 108 83 25

Doctors-Borders CARE 419 104 96 63 31

151 76 41 19 22

CRS

114 33 11 10 1

World Vision Save the Children Total Mentions 21120 373 23375 89 301 1297 131 83 470 109 77 361 22 6 107

The Political Legitimacy of International NGOs 1000 900 800 700 600 500 400 300 200 100 0

Oxfam

Doctors-Borders Accountable

CARE Transparent

CRS

243

World Vision

Legitimate (agg)

Save the Children

Figure 8.1. INGO use of “legitimacy” and related terms.

and pressured into ongoing service provision to address more chronic issues. This happens in part because MSF perceives itself, and is perceived by others, as having stronger obligations to continue providing aid in particular places once it has begun doing so.31 To reduce these pressures, MSF seeks public recognition of the limits of its normative political legitimacy. It argues that there is a minimum threshold for the moral right to rule, which it does not meet. Because MSF doesn’t reach this threshold (according to MSF), it lacks the moral right to rule, and so cannot have a moral obligation to rule. For INGOs such as MSF, less normative political legitimacy can mean more organizational autonomy. INGOs invoke political legitimacy as the moral right to rule in another way. As INGOs are acutely aware, governments and other ruling regimes sometimes use their association with INGOs to boost their own empirical legitimacy. For example, “[i]n northwestern Pakistan, the army and international donors accord humanitarian aid a pivotal role in their ‘stabilisation’ strategy intended to establish the government’s legitimacy.”32 For INGOs that find themselves in a position of having to decide whether to allow themselves to serve this (empirical) legitimating function, questions about governments’ normative political legitimacy—about whether governments have the moral right to rule—are highly salient. This, then, is a second way in which a conception of normative political legitimacy as a minimum threshold for the moral

244

Jennifer C. Rubenstein

right to rule is salient for INGOs. However, it is applied to governments, not to INGOs themselves. In sum, commentary on INGO political legitimacy tends to focus on the criteria for or sources of political legitimacy. INGOs’ own discussions of political legitimacy sometimes engage more directly with the idea of political legitimacy as the moral right to rule, but often this is in a modality of retrenchment or denial— INGOs’ claim to lack political legitimacy—or it involves the political (il)legitimacy of governments, not INGOs themselves. The implications of INGOs being politically legitimate or illegitimate—of them having or lacking the moral right to rule—is largely evaded in these discussions. It is to these implications that I now turn. Political Legitimacy as a Minimum Threshold for the Moral Right to Rule for INGOs What comes into view when we conceive of INGO political legitimacy not only as criteria for political legitimacy, but rather as criteria that set out a minimum threshold for the moral right to rule? First, this focus flips the script of typical charity-based approaches to INGOs. These approaches assume that INGOs’ activities are good and beneficial: The more they do, the better. In contrast, a focus on INGOs’ political legitimacy foregrounds the question of whether INGOs meet minimum criteria to justify their rule. Rather than, “How might we persuade these generous INGOs to do more?, or even, “How might we tinker at the edges to make them even better,” the question becomes, “Do they meet minimum criteria to justifiably exercise political power?” This perspective not only pushes back against the assumption that because INGOs have good intentions they will necessarily have good effects; it also offers a kind of bulwark against the bias toward action that has often plagued INGOs, especially in emergency situations.33 Unlike more utilitarian approaches that also avoid this bias (i.e., approaches that ask whether “the benefits of action outweigh the costs”), a focus on normative political legitimacy keeps in view not only consequences for individual welfare (or consequences more generally), but also how political power is exercised. While the political legitimacy script is skeptical of INGO rule, it does not

The Political Legitimacy of International NGOs

245

simply reject it out of hand. It thus takes seriously the needs and claims of people that INGOs seek to assist. In addition to flipping the script of charity-based approaches, claims that INGOs are politically legitimate or illegitimate address the subjects of INGOs’ rule as moral agents—agents who make moral and political judgments, and who have, and are owed, moral and political obligations. This point holds in somewhat different ways for different accounts of political legitimacy. Most straightforwardly, if an INGO has the moral right to rule, then those it rules have good (though not necessarily undefeatable) moral reasons to cooperate with, or at least not stymie, it. As Buchanan and Keohane put it, if rules are politically legitimate, “people subject to those rules have moral, content-independent reasons to follow them.”34 When an entity has empirical political legitimacy—when those it rules believe that it has the moral right to rule—they are more likely to cooperate on moral grounds, even if they think that doing so is not in their self-interest.35 Correspondingly, the claim that an INGO is normatively politically legitimate is a claim that those it rules should cooperate with it on moral grounds, even if doing so is not in their self-interest. Thus, the claim that an INGO is politically legitimate or illegitimate invites its subjects to assess it on moral grounds and act accordingly. This invitation construes the subjects of INGO rule not as mere biological beings focused narrowly on their own survival (which is how they are sometimes portrayed to and perceived by outsiders), but rather as agents capable of making and acting on moral judgments. Some scholars argue that citizens have not merely strong moral reasons, but full-fledged moral obligations, to comply with their politically legitimate governments.36 For Buchanan, individuals owe this obligation not to their government, but rather to other individuals who are also subject to the government’s rule.37 To render this idea for the context of INGOs, we might say that individuals subject to an INGO’s rule have an obligation to other similarly situated individuals to cooperate with that INGO, if that INGO meets the criteria for political legitimacy.38 There are difficult questions here about which criteria for INGO political legitimacy, if any, could generate this obligation: For example, is good performance enough? Another interesting question to consider in this regard is what it would mean

246

Jennifer C. Rubenstein

for individuals to comply with this obligation given that there is frequently a whole complex of INGOs, NGOs, and other actors providing services in a given area, such that it can be difficult to figure out who is doing what, and who is responsible for which outcomes.39 The more important point for present purposes, however, is that these lines of argument treat subjects of INGO rule as active agents who make moral demands on, and have moral obligations to, others. Regardless of whether individuals have obligations to comply with politically legitimate INGOs, they are generally not obliged to comply with INGOs that are not politically legitimate. Indeed, individuals might have political responsibilities to each other to resist the rule of politically illegitimate INGOs. For example, if an INGO is politically illegitimate because it acts unjustly or is ineffective, then resistance to it might be justified. This is an important practical issue in the context of INGOs, because the communities where INGOs try to work sometimes reject them. They do this for a host of reasons, including the belief that an INGO provides poorquality services, suspicion that it has an ulterior political or religious agenda, and revulsion at the idea of accepting services from an organization that also assists their enemies.40 Common forms of resistance to INGOs include speaking out against them, refusing to use their services, not warning INGO workers about security threats, and shunning INGOs and/or their employees (e.g., refusing to sell them things or rent houses to them). The idea that individuals subject to an INGO’s rule might have a moral permission, or even a moral or political obligation, to resist it also opens up space to construe those individuals as moral and political agents. Just as conceptualizing social unrest as a protest rather than a riot brings its political and moral dimensions into view and construes the individuals involved as political actors rather than criminals or misbehaving children, describing an INGO as politically illegitimate brings the political and moral dimensions of its subjects’ non-cooperation into view.41 Some scholars of political legitimacy argue that political legitimacy alone does not generate an obligation to comply: An entity can be morally justified in ruling while those whom it rules can be morally justified in not cooperating.42 This formulation makes sense of the idea that an INGO can live up to some reasonable set

The Political Legitimacy of International NGOs

247

of criteria for normative political legitimacy, but that the individuals whom it rules are still justified in not cooperating, because they want, and think they can achieve, more for themselves than what compliance with a politically legitimate INGO will provide. As Eyal Weizman writes: It might be that only with the ultimate refusal of aid at a time of their choice—with the rejection of the very apparatus that sometimes keeps them in good health, and sometimes operates to manage their exclusion—with refugees constructing their own spaces, self-governing, posing demands and acting upon them—that the potentiality of their political life will actualize.43

We might therefore say that the duty to comply with INGOs arises under conditions that are more demanding than mere political legitimacy. Alternatively, we might say that individuals are always obliged to comply with a politically legitimate INGO, but that INGOs lose political legitimacy when other actors, such as domestic governments, are able to perform the same functions better. Either way, claims about INGO political legitimacy open up discussion of principled non-compliance and resistance by the subjects of INGO rule in a way that claims about what INGOs should do, do not.44 The three dimensions of political legitimacy—criteria, minimum threshold, and moral right to rule—cannot be cleanly separated. However, the moral right to rule dimension is a major part of why claims about INGO normative political legitimacy treat the subjects of INGO rule as moral and political agents. I turn now to some other implications of political legitimacy being a minimum threshold. Some of these turn more on the minimal part, and some turn more on the threshold part. One obvious implication of political legitimacy being a minimum threshold is that discussion of INGO political legitimacy does not address what INGOs should do ideally, but rather what they must do at a minimum (to be justified in ruling). For this reason, claims about INGO political legitimacy can be a locus of agreement among people who otherwise disagree. A group of people can agree that an INGO is politically legitimate, and so has a presumptive right to rule, while disagreeing about what it ought to do

248

Jennifer C. Rubenstein

or strive for ideally. Likewise, a group can agree that an INGO is politically illegitimate and so should not wield power, even if they disagree about why it is illegitimate, or how far below the bar it is. Claims of INGOs’ (il)legitimacy can therefore be a focal point for coalition-building and collective action. Of course, people can also agree that an INGO is unaccountable and cite different reasons for this belief, but because the concept of normative political legitimacy is substantively wide open—it can be based on any criteria—allegations of political (il)legitimacy can be a point of agreement among people with a wider range of views than allegations of unaccountability. An implication of political legitimacy being a minimum threshold is that claims that INGOs are illegitimate or legitimate have high stakes: Accusing an INGO of being politically illegitimate is tantamount to saying that it should either (a) become politically legitimate or (b) stop exercising political power. Thus, governments take advantage of the high stakes of political legitimacy claims by arguing both that INGOs are not politically legitimate, so that they can be excluded from decision-making forums, and that they are (democratically) politically legitimate, so that consulting with them “counts” as consulting with marginalized groups.45 Claims that INGOs are unaccountable (for example) don’t work this way. The claim that an INGO is unaccountable conveys only that it is unaccountable—and perhaps implies that it should be more accountable. It doesn’t tell us that because an INGO is insufficiently accountable, it must give up power. Of course, INGOs are sometimes said to be illegitimate because they are insufficiently accountable. In these cases, accountability is invoked as a criterion of political legitimacy: To be justified in ruling, an INGO must achieve a minimum threshold of accountability. But in this scenario, accountability is not doing the work in making the situation high stakes. Indeed, accountability could be replaced by any other criterion or set of criteria for INGO political legitimacy, such as transparency, compliance with the law, or effectiveness. What is doing the work, instead, is the idea of political legitimacy as a minimum threshold. Because the stakes of claiming that an INGO is politically legitimate—or illegitimate—are high, it matters that the minimum threshold is located in the right place. If overly demanding criteria

The Political Legitimacy of International NGOs

249

of political legitimacy are applied to INGOs, they will be wrongfully (or counterproductively) excluded from rule—which could be very bad for the individuals and groups who would otherwise benefit from their services. On the other hand, if too-weak criteria are applied, harmful or incompetent INGOs will soldier on, armed with something like a seal of approval (or at least acceptance). This means that the power to authoritatively identify criteria for INGO political legitimacy, and adjudicate whether a given INGO has met those criteria in a given case, is potentially quite significant. Not surprisingly then, there are sometimes political battles over who will wield these forms of power. As Collingwood writes: the powerful will always (attempt to) legitimise their actions with reference to a whole range of rules and principles (such as moral or “civilising” values, national interests, the interpretation of international law, and so forth) and the underdog will always claim that such behaviour is “illegitimate” according to a set of competing standards.46

While the normative implications of political legitimacy’s high stakes are therefore open-ended, the concept is, again, more distinctive than it is often taken to be, and has more teeth than it is often taken to have, in the INGO literature. The Relationship Between Political Legitimacy as the Moral Right to Rule and Democratic Criteria of Political Legitimacy I have argued that discussions of INGO political legitimacy have focused too much on the content of criteria for political legitimacy rather than other dimensions of the concept. That said, the criteria are of course still very important. One important question about the criteria pertains to the relative weight that should be put on democratic versus non-democratic criteria for INGO political legitimacy. Democratic criteria are responsive to INGOs as wielders of political power. They tell us that for INGOs’ exercise of power to be justified, it must be constrained, directed, informed, or shaped, directly or indirectly, by those who they rule. This might happen through mechanisms of before-the-fact

250

Jennifer C. Rubenstein

selection, after-the-fact accountability, and/or ongoing deliberation and power-sharing. Examples of criteria of democratic political legitimacy for INGOs include “the notion of legitimisation via ‘giving voice’ to or empowering abused or marginalised groups,” “representativeness,” “downward accountability,” empowerment, democratic membership structures that extend internationally, and “experience in, and links with, the South.”47 In contrast, nondemocratic criteria for INGO political legitimacy are responsive to other aspects of their role and function, for example that they are charged with preventing early death and/or providing accurate scientific information. They include things like having relevant technical expertise and a good track record of effectiveness.48 I believe that, when it comes to the political legitimacy of INGOs, there should be a strong presumption in favor of democratic criteria, both because of the intrinsic value of people having control over their lives, and because of the effects of such arrangements. Various objections can be lodged against this view, including that the people who INGOs rule might not necessarily want them to spend time and resources being deliberative, accountable, or responsive. I will not address these here. Rather, I want to point out that because the concept of political legitimacy as the moral right to rule treats the subjects of INGO rule as moral and political agents, it has the potential to work synergistically with democratic criteria for INGO political legitimacy to further empower the subjects of INGOs’ rule. If democratic criteria are brought together with a conception of political legitimacy as a minimum threshold for the moral right to rule, then the question of whether INGOs have the moral right to rule—a question that already engages with the subjects of INGO rule as moral and political agents—is decided by those subjects. This is especially the case if the criteria are not only substantively democratic, but were themselves selected through a democratic procedure. If political legitimacy understood as the moral right to rule is based on non-democratic criteria, such as technical proficiency or effectiveness, then the subjects of INGOs’ rule are still engaged as moral and political agents capable of making moral judgments and acting on them, but the bases on which they can make these judgments are limited and don’t include the degree to which INGOs share power or consult with them. Conversely, if

The Political Legitimacy of International NGOs

251

democratic criteria of assessment are in place but there is no sense of political legitimacy as the moral right to rule, then the criteria function more as desiderata—nice things for INGOs to have, to be, or to do, but without any real urgency or bite. In addition to providing a focal point for an overlapping consensus about INGOs, a full concept of political legitimacy that included democratic criteria would center that consensus around the views and priorities of those subject to INGO rule. Finally, the presence of democratic criteria of INGO political legitimacy not only would make debates about INGO political legitimacy high stakes, but the high-stakes question at issue would be whether INGOs should stop ruling because they failed to comply with democratic norms. To summarize, I have not provided an argument in favor of democratic criteria for INGO political legitimacy here. Rather, I have shown that the concept of political legitimacy as a minimum threshold for the moral right to rule can function synergistically with such criteria. Conclusion Critics sometimes argue that the literature on political legitimacy in political philosophy is excessively arid, abstract, and disconnected from real-world politics. Regardless of whether this is a fair critique overall, it is misleading, or at least dramatically incomplete, in the context of INGOs. Incorporating the full concept of political legitimacy as a minimum threshold for the moral right to rule from the political philosophy literature into the debates about INGO political legitimacy has the potential to (re)politicize those debates in surprising and salutary ways, especially if the criteria for political legitimacy attached to this concept include democratic criteria. The upshot of this argument is not a policy prescription. I am not claiming that it would be better, all things considered, if INGOs and their critics and commentators adopted the concept of political legitimacy as a minimum threshold for the moral right to rule as it has been developed in the political philosophy literature. Whether they should or should not do this, all things considered, is a matter of political strategy and contingent political

252

Jennifer C. Rubenstein

circumstances. Indeed, as the histories of terms such as “participatory,” “grassroots,” and “empowerment” teach us, seemingly promising terms and ideas are frequently co-opted by powerful actors and deployed in ways that contravene their original meanings.49 Rather than offer prescriptive advice about terminology, I have tried to show that the concept of political legitimacy as it has been developed and debated in the political philosophy literature has very specific substantive content and normative bite. Bringing this to bear on the context of INGOs raises new and important questions, calls into view overlooked subjectivities, enables coalitions among otherwise-dispersed groups, and raises the stakes of debates whose normative upshot is otherwise muddled or unclear. It also illuminates the discursive and political potentialities of the very concept of normative political legitimacy itself. Acknowledgments My thanks to Jennifer Petersen, Allison Pugh, Melissa Schwartzberg, Micah Schwartzman, and Denise Walsh for very helpful comments on an earlier draft of this chapter. I thank Matt Frierdich for comments and for help with the analysis shown in figure 8.1. Notes 1. This chapter focuses on a small subset of INGOs: large-scale, Western-based humanitarian and development INGOs, such as Oxfam, CARE, and Doctors Without Borders (henceforth “INGOs”). I concentrate on these INGOs because they are the focal point of the debates about normative desiderata for INGOs and INGO political legitimacy that I will be discussing. While the arguments presented here are relevant to any INGOs that engage in rule, I particularly have in mind the small number of very large, mostly multi-mandate INGOs—what Michael Barnett calls the “gang of six”—which receive the largest amount of donations and work in the largest number of countries. Michael Barnett, Empire of Humanity: A History of Humanitarianism (Ithaca, NY: Cornell University Press, 2013), 142. 2. Jens Steffek, “Introduction: Transnational NGOs and Legitimacy, Accountability, Representation,” in Evaluating NGOs: Legitimacy, Accountability, Representation, ed. Jens Steffek and Kristina Hahn (New York: Palgrave, 2010), 3–7; Vivien Collingwood, “Non-governmental Organizations, Pow-

The Political Legitimacy of International NGOs

253

er, and Legitimacy in International Society,” Review of International Studies 32 (2006): 439–454; Michael Edwards, NGO Rights and Responsibilities: A New Deal for Global Governance (Foreign Policy Center, 2000). On aid’s unintended negative effects, see, for example, Fiona Terry, Condemned to Repeat: The Paradox of Humanitarian Action (Ithaca, NY: Cornell University Press, 2002); Mary Anderson, Do No Harm: How Aid Can Support Peace—or War (Boulder, CO: Lynne Rienner, 1999). David Kennedy, The Dark Side of Virtue: Reassessing International Humanitarianism (Princeton, NJ: Princeton University Press, 2005). Jennifer Rubenstein, Between Samaritans and States: The Political Ethics of Humanitarian INGOs (Oxford: Oxford University Press, 2015); Vivien Collingwood and Louis Logister, “State of the Art: Addressing the INGO ‘Legitimacy Deficit,’” Political Studies Review 3 (2005): 175–192; Claire Magone, Michaël Neuman, and Fabrice Weissman, Humanitarian Negotiations Revealed: The MSF Experience (New York: Columbia University Press, 2011). 3. Claire Magone, “Nigeria: Public (Health) Relations,” in Humanitarian Negotiations Revealed, 130. 4. Anthony Donini, “The Far Side: The Meta Functions of Humanitarianism in a Globalised World,” Disasters 34 (2010): S220–S23 (citations omitted). See also Magone et al., Humanitarian Negotiations Revealed, 110. These arguments also apply to large-scale development INGOs, and the development activities of INGOs that have multiple mandates. 5. Sarah Lister, “NGO Legitimacy: Technical Issue or Social Construct?,” Critique of Anthropology 23 (2003): 175–192; Alan Hudson, “Making the Connection: Legitimacy Claims, Legitimacy Chains and Northern NGOs’ International Advocacy,” in New Roles and Relevance: Development NGOs and the Challenge of Change, ed. David Lewis and Tina Wallace (West Hartford, CT: Kumarian Press, 2000); David Hulme and Michael Edwards, Non-governmental Organization—Performance and Accountability: Beyond the Magic Bullet (West Hartford, CT: Kumarian Press, 1995). 6. See, for example, “Accountability to Affected Populations: Mapping of Current Initiatives, Interagency Projects and Key Reports—2016,” IASC Task Team on Accountability to Affected Populations and Protection from Sexual Exploitation and Abuse,” available online at https://www.chsalliance.org, as well as I. McWha-Hermann, J. Jandric, S. Wakefield, S. C. Carr, C. Grund, and M. Moutou, “Project FAIR: Exploring Practical Pathways for Reward Fairness in International NGOs” (Edinburgh, UK: University of Edinburgh, 2017), available online at http://project-fair.org/. 7. Throughout, I will be concerned with Western liberal democratic notions of political legitimacy. While it would be interesting and informative to do so, I do not grapple here with other conceptions of political legitimacy or its analogues.

254

Jennifer C. Rubenstein

8. As I discuss below, I use the term “rule” to refer to INGOs exercising significant political power. 9. For example, Allen Buchanan and Robert O. Keohane, “The Legitimacy of Global Governance Institutions,” Ethics and International Affairs 20 (December 2006): 405–437; Thomas Christiano, “Is Democratic Legitimacy Possible for International Institutions?” in Global Democracy: Normative and Empirical Perspectives, ed. Daniele Archibugi, Mathias KoenigArchibugi, and Raffaele Marchetti (Cambridge: Cambridge University Press, 2012); Andrew Hurrell and Terry Macdonald, “Global Public Power: The Subject of Principles of Global Political Legitimacy,” Critical Review of International Social and Political Philosophy 15 (2012): 553–571, DOI: 10.1080/13698230.2012.727305. 10. E.g., Hurrell and Macdonald, “Global Public Power”; Buchanan and Keohane, “The Legitimacy of Global Governance Institutions.” 11. Barnett, Empire of Humanity. This idea has, however, come under some pressure. See Magone et al., Humanitarian Negotiations Revealed, and Caroline Abu-Sada, In the Eyes of Other: How People in Humanitarian Crises Perceive Humanitarian Aid (United States: Doctors Without Borders, 2012). Some INGOs also have an economic incentive to downplay their political dimensions, because they package their activities and present themselves to donors as agents of charity, care, and concern that are above or outside politics. 12. Buchanan  and Keohane, “The Legitimacy of Global Governance Institutions”; also Fabienne Peter, “Political Legitimacy,” Stanford Encyclopedia of Philosophy (Summer 2017 edition), ed. Edward N. Zalta, available online at https://plato.stanford.edu. 13. Hudson, “Making the Connection”; Collingwood, “Nongovernmental Organizations, Power, and Legitimacy.” 14. Peter, “Political Legitimacy.” See also Richard Fallon, Law and Legitimacy in the Supreme Court (Cambridge, MA: Harvard University Press, 2018), 24–26. 15. Hurrell and Macdonald, “Global Public Power,” citations omitted. For a more skeptical view, Collingwood, “Non-governmental Organizations, Power, and Legitimacy.” 16. Allen Buchanan, Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law (Oxford: Oxford University Press, 2003), 145. 17. Buchanan  and Keohane, “The Legitimacy of Global Governance Institutions”; also Peter, “Political Legitimacy.” 18. Magone et al., Humanitarian Negotiations Revealed, 5. 19. Lister, “NGO Legitimacy.” 20. Iain Atack, “Four Criteria of NGO Legitimacy,” World Development 27 (1999): 855–864, quoting M. Bratton, “The Politics of NGO-Government Relations in Africa,” World Development 17 (1989): 569–587.

The Political Legitimacy of International NGOs

255

21. Atack, “Four Criteria.” 22. Cord Schmelzle and Eric Stollenwerk, “Governance Effectiveness and Legitimacy in Areas of Limited Statehood: Virtuous or Vicious Circle?” Journal of Intervention and Statebuilding 12, no. 4 (2018): 449–467; Jennifer Rubenstein, “The Virtuous Circle Argument, Political Judgment, and Citizens’ Political Resistance, Journal of Intervention and Statebuilding 12, no. 4 (2018): 584–602. 23. Calain, “In Search of the ‘New Informal Legitimacy.’” 24. Edwards, NGO Rights and Responsibilities, 20. 25. Hugo Slim, “By What Authority? The Legitimacy and Accountability of Non- governmental Organisations. International Meeting on Global Trends and Human Rights,” Geneva, January 10–12, 2002. International Council of Human Rights Policy, available online at http://www.ichrp.org. 26. Slim, “By What Authority?” 27. Steffek and Hahn, Evaluating Transnational NGO Legitimacy, 9–10 (emphasis mine). 28. In a search of the websites of the six largest INGOs in May 2018, all mentioned the term “accountable” far more often than “legitimate” and its cognates. Five out of six mentioned “transparent” more than “legitimate.” Removing references to legitimacy that are not about the organization itself from the analysis does not significantly change this outcome. 29. Hudson, “Making the Connection.” 30. James Orbinski, “Nobel Prize Speech,” online at https://www.nobelprize.org. Emphasis mine. 31. Lisa Fuller, “Justified Commitments? Considering Resource Allocation and Fairness in Médecins sans Frontières-Holland,” Developing World Bioethics 6 (2006): 59–70. 32. Magone et al., Humanitarian Negotiations Revealed, 70. 33. See Rony Brauman’s argument that “abstention is not necessarily an abdication,” cited in Terry, Condemned to Repeat, 25. 34. Buchanan  and Keohane, “The Legitimacy of Global Governance Institutions”; Christiano, “Is Democratic Legitimacy Possible for International Institutions?” 35. Schmelzle and Stollenwerk, “Virtuous or Vicious Circle?” 36. Peter, “Political Legitimacy.” See also Hurrell and Macdonald, “Global Public Power.” 37. Buchanan, Justice, Legitimacy, and Self-Determination, ch. 5. Buchanan argues that this obligation stems from the government’s political authority, not only its political legitimacy. 38. Abu-Sada, In the Eyes of Others. George Klosko and Andrew Gates, “Refugees and Liberal Theories of Political Obligation” (unpublished manuscript).

256

Jennifer C. Rubenstein

39. Note that this is a different and narrower question than the question of when individuals are obliged to cooperate with INGOs. Some answers to the latter question, for example involving fair play or humanitarianism, have nothing to do with INGOs’ moral right to rule. See Klosko and Gates, “Refugees.” 40. Abu-Sada, In the Eyes of Others; Magone et al., Humanitarian Negotiations Revealed. 41. Rubenstein, “Virtuous Circle Argument.” 42. Buchanan, Justice, Legitimacy, and Self-Determination, ch. 5; Peter, “Political Legitimacy.” 43. Eyal Weizman, The Least of All Possible Evils: Humanitarian Violence from Arendt to Gaza (London: Verso, 2011). 44. Rubenstein, Between Samaritans and States, ch. 3. 45. Peter Niggli and André Rothenbühler, “Do NGOs Have a Problem with Legitimacy?” (2003), available online at https://www.globalpolicy. org. 46. Collingwood, “Non-governmental Organizations, Power, and Legitimacy.” 47. Magone et al., Humanitarian Negotiations Revealed, 5; Collingwood, “Non-governmental Organizations, Power, and Legitimacy”; Lister, “NGO Legitimacy”; Atack, “Four Criteria.” 48. Slim, “By What Authority?”; Hudson, “Making the Connection”; Collingwood, “Non-governmental organizations, Power, and Legitimacy”; List, “NGO Legitimacy.” 49. Andrea Cornwall and Deborah Eade, eds., Deconstructing Development Discourse: Buzzwords and Fuzzwords (Warwickshire: Practical Action Publishing and Oxfam GB, 2010).

PART III

BELIEFS ABOUT LEGITIMACY AND COMPLIANCE

9 EVALUATING CONSENSUAL MODELS OF GOVERNANCE LEGITIMACY-BASED LAW TOM R. TYLER

What is the most desirable form of connection between legal authorities and the people in the communities they regulate? This question can be addressed through a comparison of two models of authority: coercive and consensual. The coercive model is instrumentally based. It suggests that legal authorities influence public behavior by providing benefits or by threatening sanctions. An example of a frequently promised benefit is safety and security through crime control. A threatened sanction is a threatened or delivered punishment for rule breaking. Through the combination of promising to provide safety or other valuable services and through influencing behavior by the threat or use of sanctioning those who do not follow the law, legal authorities seek to motivate the public behaviors needed from members of the community to maintain social order and to effectively support the rule of law. The primary public behavior that authorities have sought to shape is compliance with the law. It is widely recognized by legal authorities that the law is of little importance if it does not shape how people act in their everyday lives.1 This instrumental view of people’s relationship to the law has a long history within legal scholarship and criminology.2 An alternative model views legal authority as being based on the consent of the governed. Such consent could occur for many 257

258

Tom R. Tyler

reasons. The focus here is on consent based on the belief that legal authorities are entitled to be obeyed because they are regarded as appropriate and proper authorities. They are authorized to make decisions about how to best maintain social order in the community. This concept, typically referred to as legitimacy, provides an alternative value-based model of legal authority. It suggests that people can and will consent to appropriately enacted and exercised legal and political authority. Such consent has the benefit of motivating willing deference. People take the responsibility to obey upon themselves and do so because they think it is their responsibility to defer to appropriate authorities, not because they expect to receive services or because they fear being punished. The coercive vision of law has a long history, but is perhaps best illustrated by discussions of the writings of Jeremy Bentham.3 In particular, there has been a general acceptance of the idea that a key component of legal authority is the ability to motivate compliance using sanctions. This view has been recently reinforced by Schauer, who argues that force has always been and is today the central factor viewed as shaping law-related behavior.4 This perspective is supported by a large and ongoing literature on deterrence. The perceived deterrence literature treats the instrumental model as an individual-level dynamic: Those who anticipate receiving benefits or who fear sanctions convert those judgments into a subjective calculus of costs/benefits and act based upon those judgments. This model is concerned with the issue of why people obey laws. Behavioral economics researchers directly examine the connection between perceived risk and behavior in the large literature on deterrence.5 A parallel macro-level literature focuses on observed connections between objective risk and observed patterns of behavior by people as reflected in aggregate statistics such as the crime rate. This approach is often based on the assumption that deterrence is the underlying micro-level mechanism shaping macro-level effects, but that assumption is not typically empirically tested. The legitimacy tradition focuses on personal values and their capacity to elicit voluntary consent to regulation.6 This is an inherently psychological model that postulates that a person’s own values shape their actions. That influence is distinct from the impact

Evaluating Consensual Models of Governance

259

of costs and benefits. It is particularly important as an explanation for behavior when people are not under surveillance or where offenses are low level so the likelihood of being punished is small. It is also important when people are not concerned about the provision of services such as community safety. The classic example is a person who arrives at a stop sign in the dead of night with no one around. Do they stop? If so, one explanation for their behavior is that they have internalized the obligation to follow rules. The argument that leaders benefit from having value-based legitimacy is an old one and can be found in the writings of Aristotle and Plato. More recently Cicero describes the motivation of “glory,” which is associated with cooperation with others, being loved, being the recipient of goodwill, and being someone in whom others have faith.7 He contrasts this model to one based on “fear.” Similarly, Montesquieu talks about the human passions that activate governments: “virtue” in democracies; “honor” in monarchies; and “fear” in despotic states.8 In the case of democracy, Montesquieu argues that society cannot flourish without virtue, “the love of one’s country, and the preference of the interest of one’s country to one’s self-interest. It involves the love of what one’s country stands for: equality and frugality in the democratic state.”9 Essentially, virtue means the spirit of moral goodness, patriotism, and altruism among citizens. At the risk of oversimplifying history, these ideas are precursors of the idea of responsibility and obligation that we now refer to as legitimacy. What has been missing from these earlier efforts is empirical research on legitimacy that could demonstrate that legitimacy has distinct importance because it affects important public behaviors. Studying why people obey laws empirically is emphasized by social scientists ranging from Max Weber to Kurt Lewin.10 The goal of these efforts is to examine the influence of legitimacy through research and empirical evidence. There are two conditions that are necessary for a consensual model based on legitimacy to be effective.11 The first is that people have a framework for relating to authority in terms of obligation and responsibility. People have to believe in an idea of an appropriate authority and accept their responsibility to defer to it. Second, people have to evaluate that existing legal or political authorities are acting in ways that are consistent with the preconditions

260

Tom R. Tyler

for being legitimate authorities.12 Research suggests that those conditions are: they make decisions fairly; they respect people and their rights; and they acknowledge boundary conditions for their authority. It is important to emphasize that people consent to authority in the sense that they decide if the authority meets the conditions they have for viewing someone as legitimate and, only then, deferring to them. The argument that fair procedures legitimize political authority is reflected in recent discussions of the rule of law. The idea of the “rule of law” has a long history within philosophies of governance and jurisprudence, is central to early discussions about American law, and is now widely articulated as an important goal for reform in law throughout the world.13 As would be expected, such a basic concept is explained in a variety of ways by different authors. I will focus on three core ideas associated with the rule of law that track the ideas of procedural justice noted above. The first aspect of the rule of law is rule-based decision-making. There are four elements to the idea of rule-based decision-making. First, affected parties can present evidence before decisions are made or are otherwise allowed to participate or be represented in decision-making. Second, decisions are made impartially based on facts, not personal opinions or biases. Third, rules are consistently applied. Fourth, procedures are transparent, so people know how decisions are being made. In describing the rule of law, the US State Department Bureau of International Information Program (IIP) description of democracy points to the need for the government to treat all persons equally, making decisions without regard for “race, religion, gender or minority status.” Decisions need to be made based on clearly defined laws and by people of integrity, acting on legal principles. A similar argument contends that a key element is “reasoned decisions articulated in terms of the governing rules.”14 These elements also speak to the manner in which leaders are chosen, e.g., through neutral and transparent procedures that allow citizen involvement and input. A second aspect of the rule of law is respect for people’s rights and for the principles embodied in the law. Two dimensions of respect for rights and for the law are important. First, whether the decisions made are consistent with the law or with relevant rules

Evaluating Consensual Models of Governance

261

or standards. Second, whether legal authorities act in ways that respect people’s legal rights. In describing the rule of law, the IIP points to the government being bound to act in accordance with the law and not violating the rights of the people. The idea of acting in terms of governing rules has already been noted, while legal rights are a concept central to Western democratic governments as is the idea that laws “enshrine and uphold the political and civil liberties that have gained status as universal human rights over the last half-century.”15 This respect for people’s rights is reflected in their right to participate in choosing leaders. A third idea underlying the rule of law is respect for people. Here the focus will be on the degree to which people are accorded dignity and treated with courtesy, i.e., on respect for people as people, who are entitled to treatment with human dignity. In describing the rule of law, the IIP points to the need to recognize every person’s “dignity and humanity.” It has been recognized that adherence to formal decision-making rules is not enough, since rules may themselves be inconsistent with principles of decency and fairness. Hence, the content of rules must be evaluated for consistency with ideas of human rights.16 These three elements underlying the rule of law are similar to core ideas already outlined and used to define fair procedures.17 Hence, by upholding the principles of the rule of law, legal authorities are creating the conditions of perceived legitimacy, since many studies link popular legitimacy to the perception that authorities exercise their authority through fair procedures. Legitimacy first refers to the political leaders already discussed and to the extension of legitimacy to the laws that they create. The core idea of the rule of law is that fair governmental procedures, such as elections, legitimate the rule of law.18 This is consistent with the psychological literature on legitimacy, which further links legitimacy to the fair exercise of authority. And, again this is a psychological model linking the perception of fairness to judgments about subjective legitimacy. Fortunately, empirical research has advanced considerably in recent years, and in particular there is greater willingness to support broad efforts at data collection.19 We can take advantage of those efforts to compare these models. Of particular importance is the European Social Survey, round five, which collected data that

262

Tom R. Tyler

can be used to compare coercive and consensual models of legal authority across Europe. This includes a number of societies, some traditionally democratic and others emerging from the aftermath of the cold war. In the study, twenty-seven countries were surveyed and more than 50,000 people were interviewed. We can compare the results of this broad study to a recent 2012 national survey of 1,603 Americans.20 Empirical Research Empirical research involves three tasks: concretizing concepts in measurable ways that allow quantification of theoretical variables; collecting data using those operationalizations to index relevant concepts; and connecting the concepts empirically through statistical analysis. This approach has the advantage of making it clear what abstract concepts such as legitimacy are understood to mean in concrete measurable terms. And it provides a mechanism for falsifying models that specify connections between variables. But it also makes weaknesses clear, in particular by making clear any assumptions about how abstract concepts ought to be operationalized. Ideas like legitimacy have been discussed for centuries, frequently in a vague way that did not make clear exactly how we might know how legitimate an authority or institution actually was. Empirical research requires a clear specification of how concepts will be defined. The first issue in any empirical analysis is establishing goals. There must be some objective that other concepts are expected to influence (a “dependent” or impacted variable). The classic empirical approach is that a variety of concepts are tested as possible antecedents of an objective or goal. One reason that management science has emerged as an arena for empirical research on organizations that is important is that researchers have had a clear goal against which to measure influence: profitability. This lack of ambiguity concerning the objective of a business has led to a great deal of research once it was established that the design of an organization influenced whether it was profitable. There has been less clarity concerning the desired impact of the legal system. However, in recent decades there has been considerable convergence on the goal of lowering the crime rate by

Evaluating Consensual Models of Governance

263

maximizing compliance with the law. And there have been empirical efforts to understand how to do that by varying the nature of the policies and practices of the police and the courts. The beginning point of this analysis is accepting the focus on achieving compliance, although I will argue that a broader set of goals is also desirable. It is equally important to operationalize the categories of factors expected to impact the goal (i.e., the influencing or “independent” variables). In this arena there are three classes of variable: The first are measures of the classic perceptual deterrence issues of risk of being caught and punished. In addition, the legal system (and in particular the police) provides benefits of safety through crime control, and the degree of those benefits can be quantified. A second set of variables are sociolegal and include social norms and moral values. These are societal factors that shape compliance but are rooted in society, not legal authority. Finally, there is the legal system variable of legitimacy, which has traditionally been operationalized as the perceived obligation to obey; to convey trust and confidence; and to reflect normative alignment. The focus is on local legal authorities, i.e., the police and courts. In particular, the police are the visible face of legal authority for most people. This includes dealing with the police because people ask for help and because the police regulate public behavior over minor crimes. However, people also deal with everyday courts over tickets and misdemeanor offenses, such as shoplifting or other everyday crimes. The focus on local authority is natural in the United States, since an emphasis on local legal authority is the focus of the American legal system. In Europe it is more typical to have national police forces and court systems, although the point of contact with law through legal authorities is still local. The primary benefit of the data used in these analyses is their scope. There have been a number of studies of particular cities, including Chicago and New York, but the national sample collected in 2012 in the United States is unusual. Prior to this sample, studies beyond general public opinion polls like the Gallup or NORC polls have not sought to establish the national mood toward the police and courts, and have not systematically considered why people do or do not view the police and courts as legitimate. Similarly, the European Social Survey is the first cross-Europe effort to assess national levels of popular legitimacy.

264

Tom R. Tyler

Legitimacy versus Risk-Based Deterrence If we focus on a comparison of these two theoretical models, we need to operationalize two variables: legitimacy and risk. There is a large literature on risk that operationalizes it as subjective risk perception, i.e., the perceived likelihood of being caught and punished for wrongdoing. More recent instrumental models also emphasize the provision of benefits as indexed by the ability to manage community problems such as crime. Tyler and Jackson argue that the concept of legitimacy is composed of three measurable indexes: perceived obligation to obey; trust and confidence; and normative alignment.21 The argument is that each of these factors makes a unique empirical contribution to predicting the behaviors of concern in models of law. It is important to note, as both Lenowitz and Viehoff do in their comments on this chapter, that there are a variety of ways to decide what an appropriate concept is to reflect legitimacy, and changes in concept influence which variables should be measured empirically.22 The model outlined here justifies a concept by its distinct contribution to explaining behavior. Other models might be defined in terms of categories that political or moral philosophers think should matter. There are also a variety of ways to provide evidence that some category is important. Other researchers have adopted the argument that showing an empirical connection to behavior is important, but philosophers such as David Beetham argue for the importance of different issues on philosophical terms.23 Three behaviors are outlined in this presentation: compliance, cooperation, and engagement. While Tyler and Jackson argue for the appropriateness of these indexes, an additional benefit is that these three elements of legitimacy are measured in both the European Social Survey (Round 5, conducted in 2010) and an American national survey (conducted in 2012).24 Hence, it is possible to examine their influence empirically. It is equally true, however, that other factors that might be included in conceptual grounds cannot be tested here because they are not measured.

Evaluating Consensual Models of Governance

265

Legitimacy In both of the surveys, legitimacy was assessed using the same theoretical framework: perceived obligation to obey; trust and confidence; and normative alignment. American sample. Twenty-six items are used, each with an agreedisagree format. The items are: “All laws should be strictly obeyed”; “People should do what the law says”; “It is hard to break the law and keep your self-respect”; “A person who disobeys laws is a danger to others in the community”; “Obeying the law ultimately benefits everyone in the community”; “Some laws are made to be broken (reverse coded)”; “Sometimes doing the right thing means breaking the law (reverse coded)”; “There are times when it is ok to ignore the law (reverse coded)”; “Sometimes you have to bend the law for things to come out right (reverse coded)”; “The law represents the values of the people in power, rather than the values of people like yourself (reverse coded)”; “People in power use the law to try to control people like you (reverse coded)”; “The law does not protect your interests (reverse coded)”; “Your own feelings about what is right and wrong usually agree with the laws that are enforced by the police and courts”; “The laws in your community are consistent with your own intuitions about what is right and just”; “You should support the decisions made by police officers even when you disagree with them”; “You should do what the police tell you even if you do not understand or agree with the reasons”; “You should do what the police tell you even if you do not like how they treat you”; “The police in your community are legitimate authorities and you should do what they tell you to do”; “You trust the police in your community”; “You generally support how the police act in your community”; “The police generally have the same sense of right and wrong that you do”; “The police stand up for values that are important to you”; “The police usually act in ways consistent with your own ideas about what is right and wrong”; “You and the police want the same things for your community; “The values of most of the police officers who work in your community are very similar to your own”; and “The law represents the moral values of people like yourself.”

266

Tom R. Tyler

European sample. A nine-item scale involving agreeing or disagreeing that: “The police have the same sense of right and wrong as me”; “The police stand up for values that are important to people like me”; “I trust the police”; “I trust the legal system”; “I generally support how the police act”; “The police seldom take bribes”; “It is your duty to back decisions made by the police even if you disagree with them”; “It is your duty to do what the police say, even if you do not understand or agree”; and “It is your duty to do what the police say even if they treat you badly.” Risk-Based Deterrence American sample. If you committed the following crimes, how likely is it that you would be caught and punished: “Made an exaggerated or false insurance claim”; “Bought something you thought might be stolen”; “Took something from a store without paying for it.” European sample. If you committed the following crimes how likely is it that you would be caught and punished: “Made an exaggerated or false insurance claim”; and “Bought something that might be stolen.” Other Factors American sample Disorder. How often do you see: “Teenagers hanging around the streets”; “Rubbish or litter lying around”; “Vandalism, graffiti, or other property damage”; and “People being drunk or rowdy in public places.” Police performance. “How successful do you think the police are at preventing crimes where violence is used?”; “How successful are the police at catching burglars?”; and “If a violent crime occurred and the police were called, how quickly would they respond?” European sample Police performance. “How successful are the police at preventing crimes in your country?”; “How successful are they at catching home burglars?”; “How quickly would the police arrive at a violent crime/burglary scene near where you live?”

Evaluating Consensual Models of Governance

267

Safety. “How safe do you feel walking alone in your neighborhood after dark?” Fear. “How often do you worry about your home being burglarized?”; “How often do you worry about being the victim of a violent crime?” Country-level factors The European Social Survey included respondents from twenty-seven countries. In an effort to take some of the basic features of these different countries into account, three country-level variables were included in the European analysis. The first was the average family income, as reported by the World Bank. The second was an aggregate political freedom score from Freedom House. And the third was the nature of the legal system (common law; mixed; civil). This coding is based on data from the World Factbook. Why Do People Comply with the Law? Measuring Compliance American sample. “How often in the last five years have you (never; once; twice; 3–4 times; 5 or more times)”: “Made an exaggerated or false insurance claim (3%, 1 or more)?”; “Bought something that you think might be stolen (6%, 1 or more)?”; or “Taken something from a store without paying for it (6%, 1 or more)?” (alpha = 0.79). European sample. “How often in the last five years have you (never; once; twice; 3–4 times; 5 or more times)”: “Made an exaggerated or false insurance claim (3%, 1 or more)?”; or “Bought something that might be stolen (7%, 1 or more)?” (alpha = 0.36). Results: Why Comply? What factors are shaping compliance? In this case it is possible to justify either a coercive or a consensual model of authority. The variables reflecting either model shape compliance, as is indicated by the significant influence of both factors, shown in table 9.1. In both cases, however, even with compliance as the dependent variable, the influence of legitimacy is stronger than is that of risk. In the United States, the legitimacy factor is beta = 0.12; risk is beta =

268

Tom R. Tyler

0.06; performance is beta = 0.08; in Europe, the numbers are 0.08, 0.03, and 0.01. All of these coefficients are statistically significant. Hence it is possible to argue that either model is supported, since each significantly influences compliance. However, legitimacy emerges as a stronger predictor when the two are directly compared. On the other hand, the two influences are not strikingly different in magnitude, so this difference should not be overstated. Table 9.1. Legitimacy and Compliance United States

Europe

Legitimacy

0.12***

0.08***

Risk of punishment

0.06*

0.03***

Police effectiveness

.08**

0.01***

Safety

—

−.02

Disorder

0.06*

—

Fear

−.13***

0.04***

Hispanic

0.05*

—

African American

0.01

—

Age

.10***

0.13***

Education

0.13***

0.00

Gender

0.10***

0.09***

Income

0.04

0.04***

Economic level

—

0.01*

Political freedom

—

−.01

Legal system

—

0.00

10%***

4%***

1,603

52,250

Individual-level variables

Country-level variables

*p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

One additional way to differentiate the models is to consider long-term consequences. The consensual model has the possibility of creating trust and building cooperation in the long run, while the coercive model does not. This is an argument based on the

Evaluating Consensual Models of Governance

269

findings in prior studies that authority backed up by threats does not promote the development of later supportive values. It may on the contrary undermine them, something referred to as “crowding out.” Hence, in the larger context there may be reasons to prefer consent, even when consent and coercion are equally powerful in shaping behavior toward the law in the short term. A Broader Image of Authority Why Do People Cooperate? As noted, the traditional focus of law has been on compliance. The issue has been determining why people obey laws. These results suggest that both legitimacy and risk shape compliance with everyday laws. Of the two, legitimacy is as influential as or more so than risk, suggesting that legitimacy can be a viable basis for law and governance. These results suggest that either coercion or consent could be a basis for a system focused on compliance. However, compliance is only one focus for the legal system. Many of the problems associated with the operation of the legal system flow from a lack of willing and active cooperation with legal authorities. The police seek cooperation in reporting crimes and identifying criminals. The courts seek people who will be witnesses and jurors. These are actions that are hard to motivate through promises of reward or threats of punishment. They are found, as in this study, to be better motivated by values, such as feelings of responsibility and obligation to cooperate to manage social order. Measuring Cooperation American sample. “If you saw someone push a person to the ground and steal their purse or wallet, how likely would you be to call the police?”; “How willing would you be to identify the person who had committed the crime?”; “How willing would you be to give evidence in court against the accused?” European sample. “How likely would you be to call police if you see a man get his wallet stolen?”; “How willing would you be to identify the person who did it?”; “How willing would you be to give evidence in court against the accused?”

270

Tom R. Tyler

Results: Why Cooperate? Although either coercion or consent can be the basis for compliance, the benefits of a consensual system become clearer when the focus is on willing cooperation. The analysis shown in table 9.2 supports the argument that legitimacy is the primary antecedent of cooperation. This effect is very strong in the United States, where legitimacy (beta = 0.31) is much stronger than either effectiveness (beta = 0.07) or risk (beta = 0.11). In Europe, legitimacy is also the strongest factor (beta = 0.12) and is stronger than either effectiveness (beta = 0.05) or risk (beta = 0.01), although the influence of legitimacy is weaker than it is in the United States. Table 9.2. Legitimacy and Cooperation United States

Europe

Legitimacy

0.31***

0.12***

Police effectiveness

0.07**

0.05***

Risk

0.11***

0.01

Safety

—

−.08***

Disorder

0.03

—

Fear

−.05

−.05***

Hispanic

−.06*

—

African American

0.00

—

Age

0.15***

0.02***

Education

0.01

0.08***

Gender

0.03

−.03***

Income

0.09***

0.04***

Economic level

—

0.31***

Political freedom

—

0.02***

Legal system

—

0.02***

22%***

19%***

1,603

52,250

Individual-level variables

Country-level variables

*p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

Evaluating Consensual Models of Governance

271

Why Do People Engage in Communities? Beyond the question of cooperation is the potential role of law and legal authorities in encouraging engagement in communities. If legal policies and practices can motivate community development, there exists the long-term possibility of helping communities to develop their way out of crime. Measuring Engagement American sample Collective efficacy. Collective efficacy reflects the degree to which people believe that others in their community will enforce norms. Respondents were asked to estimate how likely it is that their neighbors would intervene if: “Children were skipping school”; “Children were spray-painting graffiti on a local school”; “A fight broke out in front of your house”; “Children were showing disrespect to an adult”; or “The fire station in your neighborhood was going to be closed.” Identification/engagement. A prelude to engagement in the community is feeling psychologically connected with it, so identification and engagement were treated as a single idea, reflecting a psychological and behavioral orientation toward the community. Responders were first asked to agree-disagree with items about identification: “You are proud to live in your community”; “When someone praises the achievements of others in your community, it feels like a personal compliment to you”; “The things your community stands for are important to you”; “Being a part of the community you live in is important to the way you think of yourself as a person”; “Others in your community respect how you live your life”; “Others in your community respect what you contribute to the community”; “Others in your community respect your values.” They were then asked about the frequency of their behaviors reflecting community engagement: “How often do you attend meetings involving local officials to discuss community problems”; “How often do you

272

Tom R. Tyler

vote in local elections”; “How often do you communicate your views about community issues to elected officials”; and “How often do you talk to your neighbors about community problems.” European sample Trust in other people. Would you agree or disagree that: “Most people can be trusted, or you can’t be too careful when dealing with others”; “Most people try to take advantage of you or try to be fair”; and “Most of the time people are helpful or mostly looking out for themselves.” Engagement. There was no measure of identification in the European sample, so the index is only behavioral. Respondents were asked: “Have you”: “contacted a government official”; “worked for a community organization”; “signed a petition”; and/or “boycotted products.” Results: Why Engage? The indicators of engagement differ across the two surveys. First, in the United States, the survey asks about collective efficacy; in Europe, about trust in people. Second, in the United States, the survey asks about identification with and engagement in the community. In Europe, people are only asked about engagement. Hence, comparisons across surveys need to be made carefully. The results of both surveys support the argument that legitimacy motivates engagement. In particular in the United States, legitimacy promotes both collective efficacy (beta = 0.24; see table 9.3) and community identification/engagement (beta = 0.17). That influence is stronger than is the impact of effectiveness or risk. In Europe, legitimacy is strongly associated with trust in people (beta = 0.17; see table 9.4) and is linked to community engagement (beta = 0.06). In Europe, legitimacy motivates engagement, but it is not the only important factor. People are also equally likely to engage because they feel safe and unafraid. It is also clear that in both studies demographics shape engagement. Those who are older, better educated, and have more resources are more likely to engage.

Table 9.3. Legitimacy and Engagement—United States United States Collective efficacy

Community identification/engagement

Legitimacy

0.24***

0.17***

Police effectiveness

0.15***

0.05*

Risk

0.09***

0.07***

Safety

0.05

—-

Disorder

−.03

−.04

Fear

0.06*

−.02

Hispanic

0.05

0.06*

African American

0.07**

0.06*

Age

−.02

0.25***

Education

0.04

0.10***

Gender

0.10***

−.02

Income

0.24***

0.11***

Economic level

—

—

Political freedom

—

—

Legal system

—

—

17%***

16%***

Individual-level variables

Country-level variables

n = 1,603. *p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

273

Table 9.4. Legitimacy and Engagement—Europe Europe Trust others

Community engagement

Legitimacy

0.17***

0.06***

Police effectiveness

0.13***

−.04

Risk

−.02***

−.01

Safety

−.09***

−.07***

Disorder

—

—

Fear

0.10***

−.05***

Hispanic

—

—

African American

—

—

Age

0.04***

0.17***

Education

0.07***

0.24***

Gender

0.07***

0.17***

Income

0.06***

0.09***

Economic level

0.19***

0.19***

Political freedom

0.10***

0.03***

Legal system

−.08***

0.06***

18%***

15%***

Individual-level variables

Country-level variables

n = 52,250. *p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

274

Evaluating Consensual Models of Governance

275

Summary These findings emphasize the importance of including a consideration of legitimacy when examining law-related behavior. Across three types of behavior—compliance, cooperation, and engagement— legitimacy is found to have consistent and distinct influences above and beyond any impact of instrumental judgments about risk or performance. And, the legitimacy influences are typically stronger. The Elements of Legitimacy The operationalization of legitimacy draws on prior theories which identify three key elements: obligation; trust and confidence; and normative alignment.25 Do all three of these elements matter? To address this question a model was created connecting these three elements of legitimacy to behavior. That model was then tested. In both cases (see figures 9.1 and 9.2), each of the three aspects of legitimacy separately shaped behavior. Those distinct influences were sometimes small, as would be expected given that the elements are correlated to one another. Despite this limitation, distinct influences were found. Perhaps most interesting is that trust was central in Europe and normative alignment was central in the United States. Of course, as has been emphasized throughout this chapter, the different elements were not measured in identical ways in the two studies, so comparisons must be considered with caution.

Obligation

.08***

Compliance

68% .12***

.39 Trust

.35

.76***

.33*** Cooperation

Behavior

.58 Alignment

.08***

.24*** Engagement

Figure 9.1. Antecedents of behavior in the European sample.

276

Tom R. Tyler

Obligation

.18**

Compliance

43% .14***

.39 Trust

.45

.14*

.59*** Cooperation

Behavior

.67 Alignment

.45***

.43*** Engagement

Figure 9.2. Antecedents of behavior in the American sample.

Legitimacy, Social Norms, Moral Values, and Risk Assessments Another way to evaluate the importance of legitimacy is relative to the influence of two other types of social values: social norms and moral values. These factors are social in nature but, unlike legitimacy and risk, are not under the control of legal authorities. Nonetheless, social theorists argue that they shape law-related behavior. Durkheim, among others, made the argument that social norms are important, while Freud and other psychologists have emphasized the centrality of moral values.26 Empirical research can be used to test the relative influence of different factors in shaping behavior. What other factors? As noted, the literature on compliance identifies three bodies of non-instrumental influence upon behavior: norms, moral values, and legitimacy. Norms refer to shared values in a community. People do not violate norms when they fear disapproval from others. Moral values are judgments about right and wrong. People are motivated to adhere to their own moral values. Legitimacy is responsibility and obligation to obey authorities. People defer to authorities when they think they are entitled to make decisions. All three of these factors reflect social values. Many legal scholars contrast legitimacy to these two other types of values. Tyler, as an example, used a random sample of

Evaluating Consensual Models of Governance

277

a population (in this case, of Chicago) which was sampled and interviewed about their law-related attitudes and values, as well as their degree of compliance with the law.27 Compliance with the law was reflected by choosing a set of everyday laws that people frequently have the opportunity to comply or not comply with. This study focused on behaviors such as driving a car while intoxicated; speeding; taking inexpensive items from stores; parking illegally; and littering illegally. The attitudes and values measured included social norms (Would others disapprove?); moral values (Is the behavior morally wrong?); risk (How likely is it that you would be caught and punished if you broke a particular law?); and legitimacy (Do you feel that you ought to obey the law?). The study found that several factors shape whether people report complying with laws. The most important is whether they believe the behavior in question is moral (beta = 0.33) and whether they feel an obligation to obey the law (beta = 0.11).28 Other factors have a smaller, non-significant influence (norms, beta = 0.08; sanction risk, beta = 0.02). These factors are correlated with one another, so there are overlapping reasons to obey everyday laws, i.e., behavior that is immoral is typically expected to elicit both disapproval and risk of punishment. The influences shown by the regression equation numbers (i.e., beta weights) above are the distinct strength of the influence of each factor, controlling for (i.e., removing) the joint influence of other factors. What about in this study? Table 9.5 replicates the analysis using the two samples under consideration here. The results suggest that legitimacy continues to be important in the US sample. In particular, it dominates predictions of cooperation and engagement. As Tyler demonstrated, moral values are the most important factor shaping compliance, followed by legitimacy.29 Socials norms are also important in this equation, supporting the argument that both legitimacy and social norms can be a basis for a compliancebased strategy. As shown in table 9.6, results in Europe are similar, although the dominance of moral values for compliance is more striking and moral values are more central to cooperation and social norms to engagement. These results replicate those in the United States but suggest that, in comparison to the United States, other social values are more broadly important in Europe.

Table 9.5. Comparing Legitimacy to Social Norms, Moral Values, and Risk Assessments—United States Comply

Cooperate

Engage

Legitimacy

0.10***

0.27***

0.19***

Risk

−.01

0.03

0.05

Social norms

0.07*

0.10***

0.00

Moral values

0.19***

0.20***

0.13***

Police performance

0.02

0.08***

0.07**

Economic level

—

—

—

Political freedom

—

—

—

Legal system

—

—

—

8%

24%

10%

Individual-level variables

Country-level variables

1,603 *p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

Table 9.6. Comparing Legitimacy to Social Norms, Moral Values, and Risk Assessments—Europe Comply

Cooperate

Engage

Legitimacy

0.06***

0.12***

0.09***

Risk

0.01*

−.01*

−.02***

Social norms

−.03***

0.07***

0.10***

Moral values

0.23***

0.10***

0.04***

Police performance

0.01

0.04***

0.04***

Economic level

0.01

0.32***

0.22***

Political freedom

0.03***

0.03***

0.04***

Legal system

0.03**

−.01

0.05***

6%

18%

7%

Individual-level variables

Country-level variables

52,250 *p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

278

Evaluating Consensual Models of Governance

279

Mechanisms for Building Legitimacy These findings point to the broad importance of legitimacy, especially to an expanded model of law and legal authority that includes the goals of motivating cooperation and encouraging engagement. The second element to building a legitimacy-based legal order is having a clear mechanism for building and maintaining legitimacy. Here the study draws upon psychological models to suggest that procedural justice is the central antecedent of legitimacy. Measuring Procedural Justice American sample. How often do the police: “Make fair and impartial decisions?”; “Give people a chance to tell their story?”; “Make decisions based upon the law and not their personal opinions or biases?”; “Treat people with dignity and respect?”; “Respect people’s rights?”; “Try to do what is best for the people they are dealing with?”; “Explain their decisions and actions?”; “Make decisions that are good for everyone in the community?” European sample. “How often do the police treat people with respect?”; “How often do they make fair impartial decisions?”: and “How often do the police explain their decisions and actions when asked?” Results An analysis of the factors associated with legitimacy is shown in table 9.7. The results indicate that in both surveys procedural justice is the central antecedent of legitimacy. This is more clearly true in the United States than in Europe, where effectiveness is relatively more strongly linked to legitimacy, but it is true in both cases. So a clear message for legal authorities is that they should attend to public views about the fairness of the procedures through which they create and implement laws and exercise legal authority.

280

Tom R. Tyler Table 9.7. Antecedents of Legitimacy United States

Europe

Procedural justice

0.49***

0.37***

Police effectiveness

0.11***

0.30***

Risk

0.07***

0.02***

Hispanic

0.02

—

African American

−.00

—

Age

0.15***

0.06***

Education

0.10***

0.01***

Gender

−.09***

0.01*

Income

−.02

0.03***

Economic level

—

0.14***

Political freedom

—

−.15***

Legal system

—

0.01**

41%***

54%***

1,603

52,250

Individual-level variables

Country-level variables

*p < .05; **p < .01; ***p < .001. Note: Legal (1 = common law; 3 = civil). Political freedom (1–7; high is free); GNN is the adjusted average income (high is more).

Law and Legal Authority in the Twenty- First Century The findings reported here are not unique to the datasets examined.30 But these studies are important because they more adequately examine large samples of the public than has been the case in prior research. They support the arguments of consensual models linked to legitimacy. Legitimacy is a viable basis for compliance. These findings provide empirical support for the value of legitimacy. Beyond compliance, a focus on legitimacy provides a model for transforming the vision of law from a focus on coercion toward consent. This includes as a beginning cooperation in co-policing social order via cooperation. But beyond that the findings point toward a vision of legal authority as promoting community identification and engagement as well as positive social relations among

Evaluating Consensual Models of Governance

281

neighbors. This is a new way of thinking about legal authority in the twenty-first century, and one supported by the empirical findings outlined here. Finally, these results provide evidence for a mechanism for creating and maintaining legitimacy. It would be of very little benefit to point to the potential value of being legitimate and of legitimating policing and practices without a model for creating and maintaining legitimacy. Fortunately, such a model exists and is rooted in procedural justice. My suggestion is that a central benefit of psychological theory is to broaden the theoretical framework through which we understand the human actor and to provide a mechanism, via empirical research, to demonstrate that this broader framework can provide a viable basis for a system of law based on human values. Such a framework is not simply a better way to achieve the traditional goal of compliance, although I will show that it is such a better way; I will also suggest that it also allows the nature of the relationship between people and law to be reconceptualized in ways that better facilitate the achievement of social order in twenty-first-century democratic societies. This argument can first be made in relationship to the traditional goal of law: to obtain compliance with laws and the decisions of legal authorities. Why do people obey the law? Economic models place their emphasis on the risk of punishment. Sanction risks do shape compliance; but research suggests that values, in particular the value of legitimacy, play a similar and perhaps even more important role in shaping compliance. Second, it is concerned with how legitimacy is created and maintained. Again, an economic model focuses on control over outcomes, while psychological models articulate a broader set of antecedents, including respect and trust. And, again, research suggests support for this broader model of procedural justice. Hence, psychology posits and psychological research supports a broader conception of the psychology underlying compliance. Why Is It So Hard to Accept This Argument? The idea of legitimacy has been around for thousands of years, as have the insights and opinions of authorities about what it is

282

Tom R. Tyler

and whether it matters. The advance that has occurred in recent years involves the development of better approaches to empirical research methods and through the more widespread availability of large-scale social surveys. This has allowed for the emergence of much broader field tests of core theoretical models of law and governance. Of course, field experiments would be ideal for this purpose, and those are not yet widely available in this arena. But there is a growth in non-experimental findings, as illustrated by those outlined in this chapter. What is interesting about these findings is that they paint a very different picture than the dominant narrative in law. As noted, since the time of Bentham there has been a striking instrumentality in legal authorities’ views about how they shape public actions. In tandem with a focus on the goal of compliance, these models have argued for the concentration of resources in the hands of legal authorities for use in the surveillance, apprehension, and when needed, punishment of those who break rules. Of course, the limits of this model have been noted. The deterrence approach produces minor influences on behavior through a costly system that requires widespread and ongoing monitoring of people. Is the influence important? In the area of everyday compliance, reviews of deterrence research have concluded that the relationship between risk judgments and crime is modest to negligible and that the perceived certainty of punishment plays virtually no role in explaining deviant/criminal conduct.31 According to Piquero, Paternoster, Pogarsky, and Loughran, a review of the literature results in “some studies finding that punishment weakens compliance, some finding that sanctions have no effect on compliance, and some finding that the effect of sanctions depends on moderating factors.”32 The effects of deterrence, in other words, exist, but are weak. As an example, MacCoun reviewed the literature on deterrence in the case of drug use and suggested that only about 5 percent of the variance in drug use was explained by variations in the certainty and severity of punishment.33 As this review notes, direct comparisons suggest the importance of a value-based focus. How has such evidence been minimized? One approach is to frame the question as one of whether deterrence works. The influence of deterrence factors is often significantly different from zero, so deterrence does work. Compare the

Evaluating Consensual Models of Governance

283

message of statistical significance (something has an impact) with that of the strength of influence. It is not enough to say deterrence has a statistically significant influence on behavior. Even when there are statistically significant effects, deterrence explains at best a very small proportion of the variance in law-related behavior. In other words, a relationship between two variables can be found to be statistically significant but can nonetheless explain so little of the variance between them that it is of little consequence. Empirical research can be used to test the relative influence of different factors in shaping behavior. In contrast to that approach, the analysis reported here compares the unique influence of different factors. Typically, several influences are significant, but a further question is how much of the variance in behavior is explained. At least some of this conventional wisdom is likely based on behavior that regulatory authorities see occurring every day. People faced with authorities may comply in the moment, when the threat of those authorities is palpable (although evidence suggests that they often resist). Then, later (and, more importantly, outside the scope of plausible surveillance), when their behavior is not motivated by the immediate threat of punishment, they renege. As a consequence, those in positions of authority have the continual experience of seeing power work to influence behavior, when in fact that influence is extremely short-lived. People in power therefore come to think that the threat and use of force is a more viable and effective strategy than it actually is. There is also a self-interested component to this interpretation. On a theoretical level, it is possible to recognize that immediate compliance that requires the continued ability to observe behavior is not a desirable public policy. But, from the perspective of a power holder, it is a very desirable policy. It is essentially a life-long employment arrangement. Prison guards and police officers have mixed feelings about promoting consensual models. Improvements that might allow society to better allocate its collective resources away from command and control models are not to the institutional advantage of existing authorities. Hence, focusing on immediate compliance in the presence of authorities is a highly desirable policy for legal authorities to promote. On a cognitive level, beliefs in the power of coercion reflect a “myth of self-interest.” Studies show that people expect others

284

Tom R. Tyler

to be more strongly influenced by self-interested judgments, for example about potential rewards and punishments, than they actually are. This is especially true of people in positions of authority. In reality, as I have detailed above, reviews of the literature consistently show that the threat and use of force is often ineffective and generally not as strong in its influence on behavior as is typically assumed. Variations in the likelihood of being caught and punished have a minor impact on criminal behavior at best. Whether people are punished is not reliably related to lower levels of future criminal conduct, and more severe punishment is not linked to less criminality in the future. When this widespread lack of empirical support for force-based approaches is compared to the general belief in the effectiveness of such models, the question is why this widely held, but mistaken, belief persists. To some extent this continuation in a flawed belief reflects the self-fulfilling nature of theories and models. Ferraro, Pfeffer, and Sutton discuss the dominant role of economic models in organizational theory and the similar lack of strong empirical support for such models in that context.34 They argue that whether a theory appears self-evident is more strongly related to whether it is consistent with cultural myths than whether it is supported by empirical evidence. This leads to the question of why people find it so compelling to think of people as utilitarian. Does utilitarianism in some way comport with our understanding of everyday experience? At least in the case of authorities, these beliefs persist because power produces compliance in the everyday personal experience of legal authorities. Once people have such a conception of their own and others’ motivational nature, it is difficult to let go of those beliefs, even when the evidence for them is discredited. As Baron notes: “We tend to hold to our beliefs without sufficient regard to the evidence against them or the lack of evidence in their favor.”35 Psychologists refer to this as “belief perseverance.”36 Instead of openly considering evidence questioning their beliefs, research suggests that people engage in psychological strategies to soften the impact of discordant information upon their beliefs. One approach that people use is to look primarily or exclusively for confirming information that allows them to maintain their beliefs. For example, people shape the way they frame their

Evaluating Consensual Models of Governance

285

study of problems in ways that support their prior views. I have noted a general tendency to frame deterrence studies as questions of whether deterrence works, in the sense that its effects can be shown to be significantly different from zero. Against this relatively low standard, deterrence effects are often found and researchers can conclude that deterrence is effective. The question is whether people are drawn to this approach because it supports their suppositions, or the other way around. Because of this tendency to focus on arguments or evidence consistent with one’s prior views, a key question is what type of evidence can minimize these motivations and encourage people to be open to new information. Kahan has discussed this issue in the context of messages about climate change.37 People have a natural tendency to interpret messages through the framework of their prior ideology and to engage in identity protective cognitions. On the other hand, this does not mean that people are always insensitive to information that is counter to their prior beliefs. Both the crafting of messages and general training and experience can shape openness. Kahan, for example, shows that judges are more open to discrepant messages than the general public— something that he believes may reflect their professional training and experience. The import of all of these perspectives can be summarized simply. When people have preconceptions about what is important, they have many ways to minimize or ignore evidence that undermines their point of view. Hence, a key virtue of empirical research is that it can be falsified. It can disconfirm people’s views, and when it does it is powerful. That does not mean that people cannot undermine or discredit data—they can. But as a form of evidence it has strength and value and can cause people to recognize the validity of perspectives that they have initially dismissed. The question of what persuades is always relative, and empirical research has been shown to be one of the more powerful ways to change the legal system. Psychology and Philosophy: Allies or Antagonists? I have a background in social science and have always studied the issue of legitimacy empirically. However, throughout my career my

286

Tom R. Tyler

research has been influenced by philosophy. When I first studied legitimacy, the legal landscape was dominated by economic models. My own efforts to broaden this framework were aided by a rich history of psychological research on legitimacy. However, in the area of political and legal philosophy, the work of David Beetham provided an example of addressing authority from the social theory perspective, and I relied heavily on this work. In the case of justice, philosophers have provided a variety of important suggestions concerning the categories that might define justice. I have drawn upon those in doing my studies by asking people about different elements of their experience and then testing whether those different elements affect perceived fairness. In particular, philosophies of justice have expanded the envelope of categories beyond the instrumental view of justice presented by psychologists.38 The initial idea put forward by psychologists was that people view justice instrumentally, i.e., as a way to obtain desired outcomes. Not all psychologists have held such a view, and issues of self and identity have been important, especially in European psychology. But philosophy has consistently emphasized questions of self and dignity, which turn out to be very important to everyday people when they are dealing with legal and political authority. So, again, philosophy helped me to broaden my own framework. Despite drawing upon philosophy, I have always tried to keep normative and empirical issues distinct. In this respect, I have been influenced by Weber, who separated two questions. The first is whether people ought to obey the law, which is a normative question. The second is why people will obey the law, which is an empirical question. I have focused on the second of these questions and have not tried to speak to philosophy. I think, however, that philosophers ought to be very interested in the results I outline. People relate to legal authority in very normative ways. They care about legitimacy, distinct from performance or outcomes. And, they have sophisticated models of distributive and procedural justice. For both, people distinguish among different principles of justice and have a situational template for knowing when to deploy which principles. Further, since justice cannot be helpful as a social coordination device unless it is shared among people, it is striking that within a given culture, like our own, conceptions of fairness in a given type of situation

Evaluating Consensual Models of Governance

287

are widely shared. Across cultures, conceptions of fairness vary, but within a given culture they are similar. This is what is needed for justice to be socially useful, and it is striking that evidence suggests that such agreement exists.39 Overall, the research in this area suggests that people are normative in their relationship to law on three levels. First, they base their behavior on legitimacy, distinct from risk or performance evaluations. Second, they evaluate legitimacy through the lens of procedural justice. Their version of procedural justice is heavily infused with issues of respect and dignity. And, third, people’s primary concerns in their social relationships are frequently dominated by issues of identity and status, not about material gains or losses.40 The Research Presented in This Chapter In analyzing these studies, I have acknowledged that there are weaknesses in the operationalization of the variables. There are also challenges associated with assessing both ongoing compliance and (hypothetical) willingness to cooperate if the situation arose. These problems are typical of the types of issues that arise when an effort is made to operationalize and measure philosophical ideas such as legitimacy and risk. On the other hand, these findings are equally important for illustrating the advantages of trying to operationalize, test, and confirm or falsify theories about human behavior. These problems illustrate a key issue in empirical research. Variables have to be operationalized. Abstract ideas such as legitimacy have to be translated into concrete questions that have to be asked in particular ways to particular people. Or they can be inferred from the observation of behavior. In either case, it is necessary to associate an abstract idea with a particular instantiation of that idea in a particular indicator or metric that can be identified, observed, and quantified. The path from abstraction to empirical operationalization has occurred in several other arenas, and those provide illustrations of the benefits and problems of this transformation. One is economics, which has moved from theoretical to behavioral economics. It would be hard to overstate the influence of the law and economics

288

Tom R. Tyler

model on law in the post–World War II period. That influence began on a theoretical level. People were assumed to act in particular ways and the law was designed based on that presumption. This movement led to a movement on behavioral economics, which sought to build the law based on experiments about how people behave. That research-based approach led to the important judgments and decision making–based recognition that people act in many ways that depart from theories of rationality. So, empiricism identified a different reality, the world of actual human behavior, when it sought to operationalize and measure rationality. More recently, philosophy has also moved in an empirical direction through research on issues like the trolley problem. Again, the point is that this effort to operationalize abstract concepts requires their operationalization in particular experimental tasks that are believed to reflect abstract concepts. Empirical research requires that abstract ideas be made concrete as they are operationalized. Why operationalize? Empirical realizations of abstract concepts are inviting because they suggest the possibility of verification or falsification. Consider the case of legitimacy, which is an ancient concept; empiricism provides the possibility of testing whether legitimacy matters, something that political theorists and philosophers have disagreed about for centuries. Certainly, at the time I conducted my initial research in this area, the general presumption was that legitimacy was not central to the study of law. Can that proposition be supported or falsified? Yes, it can, but only when there is an effort to operationalize it. Another way of putting it is that a focus on empiricism shifts the discussion from one about theoretical categories to appropriate operationalization. What is legitimacy and what questions reflect it? My own approach to this question has been to look at the factors that shape relevant behaviors. I have argued that three aspects of legitimacy—obligation, trust and confidence, and normative alignment—are important because they shape law-related behavior. This is the approach that shapes the analysis reported in this chapter. It is important to recognize that philosophers identify other dimensions that might be relevant to understanding legitimacy. An example is Beetham, who mentions other dimensions such as legality.41 The broader issue is how to decide if these dimensions should

Evaluating Consensual Models of Governance

289

also be included in discussions about legitimacy. It is this interplay between theory and operationalization that is central to understanding how to think about an abstract like legitimacy. And it is the project for the future in empiricism. Essentially it is the question of what model of empirical model development is the most reasonable. Conclusion Research suggests that consent-based legal systems are not only possible but have clear advantages over coercive approaches. This result emerges from studies of both the United States and Europe. At a time when legitimacy, or at least the lack of it, is a preoccupation across both America and Europe, these results both provide clear evidence for the importance of addressing the legitimacy “crisis” and support the general usefulness of evidence-informed law. Notes 1. Tom R. Tyler, Why People Obey the Law (Princeton, NJ: Princeton University Press, 2006). 2. Philip Schofield, “The Legal and Political Legacy of Jeremy Bentham,” Annual Review of Law and Social Science 9, no. 1 (2013): 51–70. 3. Ibid. 4. Frederick F. Schauer, The Force of Law (Cambridge, MA: Harvard University Press, 2015). 5. Aaron Chalfin and Justin McCrary, “Criminal Deterrence: A Review of the Literature,” Journal of Economic Literature 55, no. 1 (March 2017): 5–48. 6. Tyler, Why People Obey the Law; Tom R. Tyler, “Psychological Perspectives on Legitimacy and Legitimation,” Annual Review of Psychology 57, no. 1 (2006): 375–400. 7. See the chapter by Jeffrey Lenowitz in this volume. 8. Robert Shackleton, Montesquieu: A Critical Biography (London: Oxford University Press, 1961). 9. Ibid., 273. 10. Max Weber, Economy and Society: An Outline of Interpretative Sociology, ed. Günther Roth and Claus Wittich (Berkeley: University of California Press, 1978); Kurt Lewin, Ronald Lippitt, and Ralph K. White, “Patterns of Aggressive Behavior in Experimentally Created ‘Social Climates,’” Journal of Social Psychology 10, no. 2 (2010): 269–299.

290

Tom R. Tyler

11. Tom R. Tyler and Rick Trinkner, Why Children Follow Rules: Legal Socialization and the Development of Legitimacy (New York: Oxford University Press, 2018). 12. Tyler, “Legitimacy and Legitimation.” 13. Ian Shapiro, The Rule of Law (New York: New York University Press, 1994). 14. Harry Jones, “The Rule of Law and the Welfare State,” Columbia Law Review 58, no. 2 (February 1958): 143–156. 15. R. C. van Caenegem, A Historical Introduction to Western Constitutional Law (Cambridge: Cambridge University Press, 1995); Thomas Carothers, “The Rule of Law Revival,” Foreign Affairs 77, no. 2 (March–April 1998): 95–106. 16. Blandine Kriegel, Jeffrey C. Cohen, and Marc A. LePain, The State and the Rule of Law (Princeton, NJ: Princeton University Press, 1995). 17. E. Allan Lind and Tom R. Tyler, The Social Psychology of Procedural Justice (New York: Plenum Press, 1988). 18. Bruce Gilley, The Right to Rule: How States Win and Lose Legitimacy (New York: Columbia University Press, 2009). 19. Lind and Tyler, The Social Psychology of Procedural Justice; Tyler, Why People Obey the Law; Tyler, “Legitimacy and Legitimation”; Tom R. Tyler, Why People Cooperate: The Role of Social Motivations (Princeton, NJ: Princeton University Press, 2011); Tom R. Tyler, “Justice and Effective Cooperation,” Social Justice Research 25, no. 4 (December 2012): 355– 375; Tom R. Tyler, Phillip Atiba Goff, and Robert J. MacCoun, “The Impact of Psychological Science on Policing in the United States: Procedural Justice, Legitimacy, and Effective Law Enforcement,” Psychological Science in the Public Interest 16, no. 3 (December 2015): 75–109; Tom R. Tyler and Yuen J. Huo, Trust in the Law: Encouraging Public Cooperation with the Police and Courts (New York: Russell Sage Foundation, 2002). 20. For details of the study, see Tom R. Tyler and Jon Jackson, “Popular Legitimacy and the Exercise of Legal Authority: Motivating Compliance, Cooperation and Engagement,” Psychology, Public Policy, and Law 20, no. 1 (2014): 78–95. 21. Tyler and Jackson, “Popular Legitimacy.” 22. The reference here is to Viehoff’s commentary on this chapter at the ASPLP conference. Also see David Beetham, The Legitimation of Power, 2nd ed. (Basingstoke: Palgrave Macmillan, 2013); Anthony Bottoms and Justice Tankebe, “‘Voice Within’: Power-holders’ Perspectives on Authority and Legitimacy,” in Legitimacy and Criminal Justice: An International Exploration, ed. Justice Tankebe and Alison Liebling (Oxford: Oxford University Press, 2013), 60–82.

Evaluating Consensual Models of Governance

291

23. David Beetham, The Legitimation of Power (New York: Palgrave, 1991). 24. Tyler and Jackson, “Popular Legitimacy.” 25. Ibid. 26. Emile Durkheim, Moral Education: A Study in the Theory and Application of the Sociology of Education (New York: Free Press, 1973). Sigmund Freud, Civilization and Its Discontents, trans. James Strachey (New York: Norton, 1961); Martin L. Hoffman, “Moral Internalization: Current Theory and Research,” Advances in Experimental Social Psychology 10 (1977): 85– 133. 27. Tyler, Why People Obey the Law. 28. Ibid. 29. Tyler, “Psychological Perspective.” 30. Ibid., 59. Also, Tom R. Tyler and Jeff Fagan, “Legitimacy and Cooperation: Why Do People Help the Police Fight Crime in Their Communities?” Ohio State Journal of Criminal Law 6 (2008): 231–275; Tom R. Tyler, Jeff Fagan, and Amanda Geller, “Street Stops and Police Legitimacy: Teachable Moments in Young Urban Men’s Legal Socialization,” Journal of Empirical Legal Studies 11, no. 4 (2014): 751–785. 31. Travis C. Pratt, Francis T. Cullen, Kristie R. Blevens, Leah E. Daigle, and Tamara D. Madensen, “The Empirical Status of Deterrence Theory: A Meta-Analysis,” in Taking Stock: The Status of Criminological Theory, ed. Francis T. Cullen, John Paul Wright, and Kristie R. Blevins (New Brunswick, NJ: Transaction, 2006), 383. 32. Raymond Paternoster, “The Deterrent Effect of the Perceived Certainty and Severity of Punishment: A Review of the Evidence and Issues,” Justice Quarterly 4, no. 2 (1987): 191; “How Much Do We Really Know About Criminal Deterrence?” Journal of Criminal Law and Criminology 100 (2010): 765–824; Alex R. Piquero, Raymond Paternoster, Greg Pogarsky, and Thomas Loughran, “Elaborating the Individual Difference Component in Deterrence Theory,” Annual Review of Law and Social Science 7, no. 1 (2011): 335–360. 33. Robert MacCoun, “Drugs and the Law: A Psychological Analysis of Drug Prohibition,” Psychological Bulletin 113, no. 3 (1993): 497–512. 34. Fabrizio Ferraro, Jeffrey Pfeffer, and Robert I. Sutton, “Economics Language and Assumptions: How Theories Can Become Self-Fulfilling,” Academy of Management Review 30, no. 1 (January 2005): 8–24. 35. John Baron, Thinking and Deciding, 3rd ed. (New York: Cambridge University Press, 2000). 36. John W. Thibaut and Laurens Walker, Procedural Justice: A Psychological Analysis (Hillsdale, NJ: L. Erlbaum Associates, 1975). Also see Craig A. Anderson, Mark R. Lepper, and Lee Ross, “Perseverance of Social Theo-

292

Tom R. Tyler

ries,” Journal of Personality and Social Psychology 39, no. 6 (1980): 1037–1049; Lee Ross, Mark R. Lepper, and Michael Hubbard, “Perseverance in SelfPerception and Social Perception: Biased Attributional Processes in the Debriefing Paradigm,” Journal of Personality and Social Psychology 32, no. 5 (1975): 880–892. 37. Dan M. Kahan, “Ideology, Motivated Reasoning, and Cognitive Reflection,” Judgment and Decision Making 8, no. 4 (2013): 407–424. 38. Thibaut and Walker, Procedural Justice. 39. Tyler, “Justice and Effective Cooperation.” 40. Tom R. Tyler and Steve Blader, Cooperation in Groups: Procedural Justice, Social Identity, and Behavioral Engagement (Philadelphia, PA: Psychology Press, 2000). 41. David Beetham, The Legitimation of Power (Atlantic Highlands, NJ: Humanities Press International, 1991).

10 ON THE EMPIRICAL MEASUREMENT OF LEGITIMACY JEFFREY A. LENOWITZ

Legal and political theorists have discussed the concept of political legitimacy, in the sense of a morally justified right to rule, since the ancient world. From Socrates’ famous discussions in the Crito about the political obligations of citizens, to Weber’s discussion of the three primary sources of legitimacy, to late twentieth-century debates about legitimacy and the possibility of political obligation, scholars have grappled with what it means and requires for an authority to be legitimate, and how citizens should or do react to such an authority.1 Despite this long theoretical debate, the concept of legitimacy rarely appeared in empirical social-scientific work dealing with when and why citizens actually follow the dictates of their legal and political authorities. However, things have begun to change over the last several decades. One of the scholars leading this change is Tom R. Tyler, whose work has brought the concept of legitimacy into the socialscientific mainstream, especially in the areas of criminology and social psychology. Due to his and his colleagues’ efforts, recent empirical studies not only include legitimacy as an explanatory variable for citizen behavior, but also routinely demonstrate its significance. For everything from employee rule compliance in financial services institutions, to cooperation with local police, to tax avoidance, to interaction with courts, to respect for national parks, these studies not only show that legitimacy needs to be taken into account when seeking to understand how and why individuals defer to and follow rules, but also reveal how we might reform our 293

294

Jeffrey A. Lenowitz

institutions to make them more effective, efficient, and compatible with the human beings that inhabit them.2 At the risk of oversimplification, we can state Tyler’s central argument and findings as providing support for two causal relationships. First is what I call legitimacy leads to compliance (LC), the idea that beliefs in the legitimacy of legal and political authorities can motivate people to comply, cooperate, and positively engage with them. Second is procedural justice leads to legitimacy (PJL), the idea that people whose authorities treat them in a procedurally just manner come to see these authorities as legitimate. Tyler and his co-authors use the results of their research, which support both of these relationships, to push for what they call a consensual approach to law and authority, one that deemphasizes coercion to focus on reforming institutional procedures to be more transparent, fair, and participatory, with the aim of inculcating lifelong habits of compliance.3 The fruits of Tyler’s work can be seen across the globe, especially when it comes to policing reform proposals. In the United States, for example, both the President’s Task Force on 21st Century Policing and the Department of Justice’s National Initiative for Building Community Trust & Justice emphasize the importance of procedural justice on the grounds of increasing a community’s felt obligation to obey the law.4 In his contribution to this volume, Tyler gives a snapshot of his work over the years on legitimacy and the effects of procedural justice, as well as presenting new evidence in support of his main claims. In this chapter, I use Tyler’s contribution as a jumping off point to challenge several features of his own work, and the literature that it gave rise to. Specifically, I show that Tyler’s findings, while undeniably valuable, do not tell us much about legitimacy as it is commonly understood, for his operationalization of the concept distorts its meaning. I then demonstrate that on its own his work does not support the policy implications he draws from it— that political and legal institutions should reform themselves to be more procedurally just—and instead only supports efforts to make institutions look just. In the end I argue that while some of the appeal of Tyler’s work is its seeming ability to offer a way around tough moral argumentation, this is illusionary. Moral argument is still needed. Before making these critiques, I sketch a conceptual scheme of legitimacy types and the two ways that procedures can

On the Empirical Measurement of Legitimacy

295

aid in legitimating institutions. I then explain LC and PJL using this conceptual scheme, clarifying the nature of Tyler’s intervention and orienting his work amidst the larger discussions of political legitimacy in the fields of legal and political theory. Tyler has long championed, quite rightly, the contribution of empirical methods to theoretical discussions of normative concepts such as legitimacy and justice. Here, I hope to show the reverse, the contribution of philosophical methods to empirical discussions of these very same concepts. Understanding Legitimacy and Procedures Tyler characterizes his work as falling into the “legitimacy tradition.” Thus, from the beginning, it is important to understand what this concept means. I argue that in the political and legal context there are three types of legitimacy: moral legitimacy, sociological legitimacy, and legal legitimacy.5 Much of the confusion and disagreement that regularly characterizes debates on the concept stem from failing to differentiate between these types.6 Legitimacy Types Moral legitimacy relates to “moral justifiability or respect worthiness,” the latter of which is usually interpreted as necessitating some sort of duty of compliance toward that which is morally legitimate.7 More precisely: x is morally legitimate if and only if x’s claim to authority is morally justified. Here, a claim to authority makes recourse to the concept of practical authority, which a person, group, or institution possesses if its demands or instructions in a particular arena provide reasons for action meant to preempt or override alternative reasons directly related to the action in question.8 Thus, a state claims to be authoritative when it asserts a unique right to rule and exercise political power and demands some sort of deserved and dutiful, non-selfish compliance from its citizenry as a result.9 A state claiming authority expects citizens to comply with its edicts because it makes them, rather than because of their specific content. Citizens should follow the demand that they pay their taxes because this is what the state demands, not because tax

296

Jeffrey A. Lenowitz

payments contribute to national security, failing to do so results in punishment, or doing anything else wrongs one’s neighbors. In claiming authority, a state is asserting that there are sufficient content-independent moral reasons to treat its edicts as preemptive in this fashion. When such reasons exist, the state is morally legitimate. The concept of morally legitimate authority becomes clearer when contrasted with de facto authority. A state or regime is de facto authoritative when its claim to authority is successful rather than morally justified. In other words, a state that claims a right to rule, acts as if it has such a right, and succeeds in establishing and maintaining its rule is de facto authoritative, but may or may not be morally legitimate.10 Thus, a state can claim to be authoritative without being a de facto or morally legitimate authority, can be a de facto authority without being a legitimate authority, and can be a legitimate authority without being a de facto authority. Sociological legitimacy piggybacks on moral legitimacy. It indicates that citizens believe that their regime’s claim to authority is morally justified and hold content-independent reasons for obedience as a result.11 This legitimacy type descends from Max Weber, who described it as one of the keys for the survival of social orders. For Weber, to say that an authority is legitimate is to say that its citizens, or a bulk of its citizens, believe it to be legitimate. As he explains, the “basis of every system of authority, and correspondingly of every kind of willingness to obey, is a belief, a belief by virtue of which persons exercising authority are lent prestige.”12 In other words, a regime is sociologically legitimate to the degree that its citizens see it as morally justified according to their own beliefs.13 What is purported to follow from this is what Tyler calls “legitimate normative obedience,” i.e., obedience due to a belief in the rightful authority of officials, rather than due to instrumental reasons (obeying in this instance will benefit me) or general moral beliefs about proper behavior (obeying in this instance is right according to my non-authority-related moral beliefs).14 We can formally define sociological legitimacy as follows: x is sociologically legitimate if and only if a significant portion of the population that x claims authority over believe x to be morally legitimate. Note that sociological legitimacy is, as Friedrich notes, “a question of facts.”15 It is thus fundamentally a descriptive concept, i.e., it describes the

On the Empirical Measurement of Legitimacy

297

sorts of beliefs that citizens have about their regime. Such beliefs, of course, can be false, and thus moral legitimacy does not necessarily correspond with sociological legitimacy. The final legitimacy type is also the simplest: x is legally legitimate if x is lawful according to the legal system of which it is a part. Legal legitimacy is equivalent to the positivist conception of legal validity, according to which something is legally valid if and only if it belongs to a legal system in force in a certain country or is enforceable within it.16 Asserting legal legitimacy is to make a legal statement, i.e., the truth conditions are legal facts rather than moral properties or actually existing beliefs.17 Creating Legitimacy Tyler is interested in how one might reform existing regimes by changing their procedures in order to make them more legitimate. I argue that procedures can do so in two distinct ways.18 First, a procedure might have a procedural legitimating effect, in that a procedure serves as a legitimating criterion itself and legitimates a regime by its presence and use. Second, a procedure might have an outcome legitimating effect, meaning that it leads to certain outcomes such that the regime or its decisions come closer to meeting some independent legitimacy criterion. This distinction applies regardless of what type of legitimacy one is interested in. For instance, procedural legitimation might produce actual moral reasons for an authority’s claim to authority or increase the likelihood that a given population believes that it does. Similarly, outcome legitimation might lead a regime to meet real criteria of moral legitimacy, or criteria endorsed by a significant number of people. The effectiveness or appropriateness of procedural or outcome moral legitimation will depend on the content of one’s particular theory of moral legitimacy (what actually produces good reasons for claims to authority), while the effectiveness or appropriateness of procedural or outcome sociological legitimation will depend on what produces legitimacy beliefs. Similarly, the nature of the relevant legal order will determine whether procedural or outcome legitimation is most appropriate for ensuring that an institution and its actions are lawful. To illustrate the distinction between outcome and procedural legitimation, consider a town choosing rules to select their sheriff,

298

Jeffrey A. Lenowitz

with the aim of ensuring that the office is deemed legitimate. One citizen could claim that the town should use sortition to select a sheriff from all those willing and able to perform the job, for this ensures that all interested parties have an equal chance to obtain the desirable position. This would be an argument for procedural moral legitimation, in that the resultant sheriff would be morally legitimate, at least partially, because a process that treats citizens in an important moral way selected them, rather than because they are a certain type of person or took certain types of actions. Alternatively, another townsperson could argue that all candidates should take a series of exams testing for hidden biases, quick thinking, knowledge of the law, etc., with the candidate achieving the highest scores being declared sheriff. This is an argument for outcome moral legitimation, in that it assumes that there is an independent criterion of moral legitimacy (being able to make the fairest, fastest, and generally best decisions) and puts forward a decision-making procedure designed to maximize the possibility that the outcome meets it. The procedure still plays a role, but the legitimation flows from the outcome fitting the standard. Tyler’s Theory For my purposes, the centerpiece of Tyler’s work on legitimacy is legitimacy leads to compliance (LC) and procedural justice leads to legitimacy (PJL). In this section, with the help of the conceptual scheme outlined in the previous section, I explain each of these arguments. In the process, I seek to emphasize the relevance of Tyler’s contribution, and to situate it within political and legal theory. Legitimacy Leads to Compliance Tyler’s first set of arguments, which I refer to as LC, consists of three claims. The first two replicate Weber’s claim that the beliefs of subjects in the moral legitimacy of their authorities can be a real basis for the stability of a social order. Specifically, Tyler claims that (1) some people possess beliefs in the moral legitimacy of their authorities, “such that they [the authorities] are entitled to be obeyed because they are regarded as appropriate and proper,” and (2) that these beliefs compel individuals to comply, cooperate, and

On the Empirical Measurement of Legitimacy

299

positively engage with their legal authorities.19 Put more simply, Tyler claims that the phenomenon of sociological legitimacy is real and exists in certain contexts, and that it motivates several forms of obedience. When people believe that authorities have a right to rule and demand obedience, they are more likely to obey. In addition, Tyler makes a comparative claim, arguing (3) that the effects of legitimacy on compliance, cooperation, and engagement are greater than those from the threat of coercion. Thus, a belief in legitimacy is not only supposed to motivate individuals to comply and cooperate on its own, i.e., apart from other possible motivations such as fear of punishment, social norms, or a belief in the moral principles that align with or underlie a law, but also to do so better than the promise of retaliatory harm. Numerous works by Tyler and his co-authors report findings that give empirical support to these claims, the following of which is just a small sample. In a 1990 study, Tyler used a twowave panel survey of Chicago residents and found that legitimacy is positively associated with compliance, with increases in the latter following increases in the former.20 In 2003, Sunshine and Tyler found that legitimacy independently and positively shapes cooperation, compliance, and support for policies that empower the police, and the effect was greater than fear of getting caught. These results stemmed from a survey about the New York Police Department given to two samples of residents in New York City.21 In 2012, Jackson and colleagues extended Tyler’s research to England and Wales, once again finding evidence that people are more likely to comply with the law when they see their legal authorities as legitimate.22 In a 2014 study, Tyler, Fagan, and Geller used panel data from two surveys filled out by 1,261 young New Yorkers before and after their interactions with the police to investigate the effects of stop and frisk policing. Among other things, they found that perceived legitimacy significantly predicted legal obedience and cooperation with the police and the legal system.23 Finally, in the chapter included in this volume, Tyler replicates these findings using data from round five of the European Social Survey and a national survey of 1,603 Americans.24 He finds that legitimacy is a stronger determinant of compliance, cooperation, and positive engagement than coercion in both the United States and Europe.

300

Jeffrey A. Lenowitz

These findings are not minor. For one thing, many normative beliefs or judgments, even deeply held ones, frequently fail to motivate behavior. Many more people believe that they should eat a healthy diet and exercise regularly than actually do. Most people believe they should always treat their parents politely, yet somehow the average family holiday dinner often has moments that lack much in the way of mere civility. Tyler claims to show that beliefs in legitimate authority both exist and at the least are correlated with an increase in obedience and compliance. His work suggests that something like Hart’s “puzzled man” might actually exist, and that Weber’s assertions from more than one hundred years ago are supported by more robust evidence than mere observation.25 In addition, these findings make a persuasive intervention in the debate between the comparative merits of fear and respect for the law when it comes to motivating citizens. This is an old debate. One form of it can be seen in Machiavelli’s attack on Cicero’s remarks in De Officiis.26 In a series of provocative passages, Cicero describes glory as something akin to legitimacy—it is a state, associated with the cooperation of others, of being loved, being the recipient of goodwill, and being someone whom others’ have faith in—and argues that it far outranks being feared when it comes to inspiring obedience and maintaining command and control.27 In The Prince, Machiavelli mocks such a position, famously concluding that if one has to choose, it is much safer for a prince to be feared rather than loved due to the innate ungratefulness and fickleness of men. He insists that “love is held together by a chain of obligation that, since men are a wretched lot, is broken on every occasion for their own self-interest; but fear is sustained by a dread of punishment that will never abandon you.”28 Tyler not only supports Cicero and rejects Machiavelli, but also uses the inverse logic of the latter to criticize coercion. While Machiavelli argues in favor of fear because it is ever present while positive feelings are easily forgotten or overridden by self-interest, Tyler argues that people may comply in the moment when faced with authorities, but “when their behavior is not motivated by the immediate threat of punishment, they renege.”29 He backs up this assertion by citing studies showing that variance in the severity and likelihood of punishment has far less of an effect on crime and recidivism than most people assume.30

On the Empirical Measurement of Legitimacy

301

Of course, Tyler is not interested in sparring with Machiavelli, but is instead involved in the modern form of this debate. He seeks to oppose what, at least at the time of Why People Obey the Law, he saw as the dominant model of approaching social control: deterrence. This model, broadly inspired by Bentham and more fully developed by social scientists such as Gary Becker, treats individuals as rational actors perpetually calculating the costs and benefits of legal compliance.31 To secure obedience, political and legal authorities ensure that the benefits achieved through noncompliance are much lower than those following compliance. To do so, they utilize the levers of spending on surveillance, which determines its effectiveness, and coercion, which breaks down into the form and degree of punishment.32 A more recent philosophically inclined version of this argument can be seen in parts of Schauer’s The Force of Law, where contra Hart and Tyler, he insists that we have good reasons to believe that coercion is a more effective means of social control than legitimacy.33 As we have seen, Tyler’s response is to put forward and provide evidence for his psychological model, grounded in the idea that values, of which a belief in legitimacy is one, shape action apart from and sometimes better than the calculation of costs and benefits. Procedural Justice Leads to Legitimacy Tyler’s second central argument, PJL, consists of two claims. First, Tyler asserts that procedural justice is “a mechanism for creating and maintaining legitimacy.”34 Here, authorities are procedurally just to the extent that they “wield their power in fair, just and neutral ways,” do so with transparency, and give citizens ample opportunity for input when possible.35 Procedural justice is therefore distinct from effectiveness, valence, or even fairness of outcome. Tyler argues that procedurally just treatment strengthens the social bonds between individuals and institutions, causing the former to identify with the latter and eventually internalize the belief that they should obey and cooperate.36 This is an example of procedural rather than substantive sociological legitimation because the legitimacy of the authorities stems from the nature of the procedures they use, rather than from the particular substance of their decisions. People come to see authorities as legitimate not because

302

Jeffrey A. Lenowitz

fair transparent procedures are more likely to lead to decisions or outcomes they deem as legitimate, but because the procedures themselves have an independent effect. Second, Tyler claims that procedural justice is more effective at stimulating sociological legitimacy than providing good or fair outcomes. So, for example, the degree to which a person sees a court’s proceedings as fair will play a larger role in their legitimacy beliefs than the court’s ruling. This is a claim for the superiority of procedural legitimation based on just procedures over outcome legitimation of two types. Again, Tyler supports these claims with multiple findings. Initially, Tyler found that the fairness of procedures was about five times more effective at making people willing to accept the outcome of a police interaction when compared to the nature of the outcome.37 In later studies, he found evidence that legitimacy was the cause. For example, Sunshine and Tyler also found that procedural justice was the primary antecedent of legitimacy, while “judgments of police performance in fighting crime and estimates of the risk of being caught and punished  .  .  . were weak predictors.”38 In another 2003 study, this time utilizing mail-based surveys in New York City, Tyler found that police seen as engaging in racial profiling, a practice perceived as violating procedural justice, were less likely to be seen as legitimate.39 Tyler, Fagan, and Geller, in their 2014 study on stop and frisk, identified “the justice of public policies and practices as key factors shaping political legitimacy.”40 Similarly, also in 2014, using data from 1,603 respondents, Tyler and Jackson found that while the justice of decision-making and the fairness of interpersonal treatment mattered when it came to legitimacy, neither decision accuracy nor outcome favorability has a significant influence.41 In a departure from survey research, Mazzerole et al. conducted a randomized experiment involving police stops, and found that “the more ‘procedurally just’ the police strive to make even a short encounter, the more likely citizens are to perceive the police as legitimate.”42 Finally, in a study about patients’ responses to government health-care experts, Tyler, Mentovich, and Satyavada found that beliefs in the fairness and justice of experts correlated with beliefs in their legitimacy.43 PJL is, for many, a strikingly counterintuitive finding. Even those not wedded to an economic model of purely rational calculating individuals are likely surprised to find that people, when it

On the Empirical Measurement of Legitimacy

303

comes to accepting legal decisions and forming legitimacy judgments, care more about how they are treated by authority figures than the performance and decisions of these authorities. This, in a way, explains what Tyler refers to as “the paradox of policing” in the United States today, how the objective performance of the police continues to increase nationwide while support for the police remains stagnant.44 Perhaps, Tyler suggests, the problem is a failure to adopt procedural justice criteria. Moreover, in showing PJL, Tyler has once again provided empirics and social psychological explanation, rather than a few choice historical examples, for an argument stemming from the early days of political thought. We can once more look to Cicero for an example. After explaining the ultimate importance of glory for command and control, Cicero claims that all three of its dimensions—getting the masses to love you, have faith in you, and to think you worthy of honor— “are all achieved by justice.”45 People do not worry about deceit and misbehavior when it comes to the just person, and thus see it as right to “entrust to them our safety, our fortunes, and our children.”46 Moreover, this effect is greater and more reliable than the attribute of good sense or competence/intelligence, in other words, the demonstrable ability to obtain good outcomes, because only a just ruler inspires long-term faith. Tyler produces evidence and argument that purport to back just such a conclusion. In addition, PJL is of particular significance because, in combination with LC, it seems to offer a way to avoid some tricky moral argumentation about the procedures of the criminal justice system. In liberal democratic societies, citizens are thought to require, and often demand, treatment from their authorities that recognizes and respects their basic rights and equal status. This imposes limitations on the criminal justice system that sometimes conflict with its attempt at accomplishing one of its central tasks: to efficiently provide a stable political and legal order. Thus, it continues to be a live debate as to how the police and other authorities should treat the communities and persons they govern and what actions are and are not permissible. For instance, is it acceptable for the police to lie to suspects in order to extract confessions?47 Can police use ethnicity or race to decide whom to investigate or search?48 Can the courts give disproportionate sentences if the practice maximizes

304

Jeffrey A. Lenowitz

deterrence? Should juries be informed of mandatory sentencing laws? In general, should the police and courts emphasize crime reduction and decreased recidivism at all costs, or should they constrain their actions by taking seriously the importance of fair and just treatment of those they purport to protect? Answering these and similar questions requires making moral arguments for and in the public sphere. It requires setting out moral values and weighing them against one another. This is hard and contentious work, made all the more difficult in today’s pluralist states. An undeniable appeal of Tyler’s work is that it seems to provide a way out of such debates. It gives the reformer reason to shift focus away from new forms of coercion and surveillance, cease obsessing over crime statistics, and instead emphasize transparency, respect, and fair treatment. This is not because such changes are required by a particular moral position demanding just procedures, but rather because such changes are effective in securing greater compliance and cooperation. Rather than having to weigh efficiency and effectiveness against considerations of justice, Tyler’s research tells us the two are not in conflict. The criminal justice system is simply more efficient and effective if it adopts so-called procedural justice features, so the (possibly contested) moral foundations upon which procedural justice rests need not be unearthed or examined. Concerns with Operationalization I now turn to two aspects of Tyler’s work that I wish to challenge. In this section, I will focus on LC and argue that, despite decades of research, Tyler has not convincingly shown this relationship to be true. He has not demonstrated sufficiently that perceptions of legitimacy cause people to be more compliant, cooperative, and positively engaged than they otherwise would be. In making this argument, I am not claiming that it is unlikely that beliefs in legitimacy make people more likely to demonstrate desirable compliance behavior, but rather that certain conceptual mistakes and operationalizing decisions make the evidence given less persuasive then it initially seems. Of course, others have sought to question LC. The most persuasive arguments claim that Tyler and his colleagues cannot

On the Empirical Measurement of Legitimacy

305

sufficiently control for the possibility of unobserved heterogeneity or reverse causality.49 The former refers to the possibility of unobserved variables that have a similar effect on both perceptions of legitimacy and compliance-related behavior. For example, individuals with a vested interest in the power relations codified by an existing legal regime might be more likely to see the authority as legitimate and do what they can to support it. Or, perhaps some people are simply more prone to question authority, such that they see their authorities as illegitimate and are more likely to disobey. Reverse causality refers to the possibility that how someone responds to their legal authorities might cause changes in their legitimacy perceptions, rather than the other way around. For instance, law-breakers might convince themselves that the laws are illegitimate in order to justify their actions internally.50 Or, forced to explain their routinized behavior of compliance, some people might self-rationalize that legal authorities are legitimate and thus deserving of obedience. The existence of plausible unmeasured confounding variables and the possibilities of reverse causality are worrisome, because either might explain the found correlation between legitimacy and behavior instead of LC. And, thus far, Tyler and his colleagues have been unable to produce empirical work that sufficiently wards off both possibilities. My concern is different, and instead focuses on the operationalization of legitimacy. Rather than posit rival causes of the relationship between legitimacy and compliance-related behavior, I question whether Tyler’s studies actually measure legitimacy at all. As mentioned, Tyler consistently endorses a standard account of sociological legitimacy when describing the concept. At the beginning of his landmark book, Why People Obey the Law, he describes legitimacy as the feeling “that the authority has the right to dictate behavior,” and in a later article he notes that “when one recognizes the legitimacy of an institution, one believes that the institution has the right to prescribe and enforce appropriate behavior, and that one has a corresponding duty to bring one’s behavior in line with that which is expected.” Thus, when confronted with a particular “decision made or rule created by these authorities,” one sees it as “valid in the sense that it is entitled to be obeyed by virtue of who made the decision or how it was made.”51 These and many other examples, as well as Tyler’s contrasting of legitimacy

306

Jeffrey A. Lenowitz

with alternative explanations of compliance such as fear of punishment or community norms, make it clear that when discussing legitimacy he is referring to beliefs about the existence of real content-independent moral reasons that support a given authority’s claim to authority. Nonetheless, Tyler deviates from this conceptualization when he operationalizes legitimacy. Initially, in Why People Obey the Law, Tyler broke down the concept into two dimensions: (1) a perceived obligation to obey the law and (2) trust/support for authorities. In more recent work, which I will concentrate on here, Tyler and his co-authors add a third dimension, (3) moral alignment/solidarity.52 Questions are given to research subjects that correspond with each dimension, and their answers are combined to create a legitimacy score for each individual. What “dimensions” mean unfortunately is never clarified. Two possibilities seem likely: The three dimensions are constituent components of legitimacy, such that what it means to believe that an authority is fully legitimate is just to feel an obligation to obey, to trust the authorities, and to view them as having similar values. Alternatively, the three dimensions might simply be proxies for legitimacy, i.e., holding an authority legitimate is known to (uniquely) elicit obligation, trust, and feelings of normative alignment, such that measuring their presence allows one to surmise the presence of legitimacy beliefs. Unfortunately, as I argue below, each of these dimensions is an imperfect component or proxy of legitimacy. Thus, responses to survey questions designed to measure them and the analysis of the resultant data are less telling than they might otherwise seem, for we cannot confidently say that what is being measured is legitimacy. Obligation to Obey Tyler’s first dimension of legitimacy is a felt obligation to obey the law, which he describes as “reflecting the internalization of the value that it is appropriate to obey the police and the law.”53 In the philosophical literature on political obligation and moral legitimacy, there is push-back on just this point; some argue that an authority can be legitimate for a person without there being a corresponding obligation to obey.54 For the purposes of this analysis, however, let us assume with Tyler that when it comes to sociological

On the Empirical Measurement of Legitimacy

307

legitimacy, a person who sees the police, courts, and/or the state as legitimate also believes they have an obligation to obey, either because such obligations are a constituent component of legitimacy beliefs or because a belief in legitimacy always produces a belief in such an obligation. The question then is whether we can, as Tyler does, assume the opposite, i.e., assume the presence of sociological legitimacy beliefs from a person’s positive responses to questions designed to measure her felt obligation to obey. Unfortunately, there are two basic problems with such an assumption. On the one hand, a person might feel or actually have an obligation to obey the law or legal authorities for reasons unrelated to legitimacy.55 For instance, a person might believe they have a natural duty to do justice, that obeying the laws of their existing authorities is the best means to meet it, and that therefore they have a duty to obey the law.56 Or, a person might feel obligated to obey the laws because they think not doing so might bring harm to their loves ones, or because they see law-breaking as being unfair to those who already sacrificed by complying. All of these possible sources for a felt obligation to obey are compatible with simultaneously believing that there is nothing about the law itself, absent its content, that justifies its claim to authority, i.e., with failing to see the law and its enforcers as legitimate. This means that, whether conceptualized as a component or a proxy of legitimacy, measuring a felt obligation to obey might produce false positives.57 People might be reported as seeing their authority as more legitimate then they actually do just because, for reasons unrelated to legitimacy, they think they are obliged to obey. On the other hand, many of the standard questions used to measure a felt obligation do so poorly, and thus inferring such an obligation from positive responses might yield bogus results. Here are some statements, common to numerous studies in the last decade, for which respondents are supposed to indicate their level of agreement or disagreement: Q1. People should do what the law says. Q2. You should do what the police tell you to do, even when you think they are wrong. Q3. You should do what the police tell you even if you do not understand or agree with the reasons.

308

Jeffrey A. Lenowitz

Q4. Obeying the law ultimately benefits everyone in the community. Q5. Communities work best when people follow the directives of the police.58 These and similar questions fall short for two reasons. First, some might pick up motivating reasons for legal compliance that are unrelated to any obligation at all. Consider a woman named Lara who not only considers her local government illegitimate, but also spends a considerable amount of her time trying to challenge and change it. Nonetheless, Lara might answer Q1–Q3 in the affirmative because she is scared of the police, and/or because she believes that the best means of resisting illegitimacy is through the courts or campaign work, not by civil disobedience or lawbreaking, which she sees as ineffective and often unnecessarily costly. Moreover, Lara might, for similar reasons, answer yes to Q4 and Q5. Even though she is dedicated to opposing her local authorities, Lara might think everyone in her community benefits from legal obedience, because law-breaking leads to community reprisals, serves as ammunition for her political enemies, or is simply futile. Now, one might disagree with Lara’s reasoning, but that is beside the point. What she shows is that it is plausible to imagine individuals answering questions in a way that Tyler and his co-authors take as indicating a felt obligation to obey, and thus legitimacy, when in fact this is not the case. Second, some of Tyler’s survey questions treat a felt political obligation as indefeasible, i.e., they assume that the relevant obligation is absolute and cannot be overridden. This is inconsistent with both any reasonable theory of political obligations and our everyday experience with moral reasoning. Like nearly all moral obligations, real or felt, an obligation to obey legal authorities means that there are or there seem to be very powerful moral reasons to obey them. On some accounts, these reasons are particularly weighty. On others, they are protected reasons, meaning that they give us reason not only to do a particular thing (what the authority says) but also to disregard other reasons against doing it.59 Regardless, these reasons are pro tanto reasons, in that they are inconclusive all things considered reasons.60 This means that the reasons produced by political obligations can be overridden

On the Empirical Measurement of Legitimacy

309

by weightier or unexcluded competing reasons. Thus, the presence of a political obligation does not always mean one should obey, just like claiming that a person has an obligation to fulfill their promises does not mean that they should always fulfill their promises. Conflicts of moral obligations are a part of our everyday moral experience for just this reason. Tyler asks respondents to agree or disagree with the following, all of which are reverse coded when determining legitimacy beliefs: Q6. Sometimes doing the right thing means breaking the law. Q7. There are times when it is ok to ignore the law. Q8. Sometimes you have to bend the law for things to come out right. Q9. There are times when it is ok for you to ignore what the police tell you to do.61 Consider Anthony, a person who, unlike Lara, is strongly committed to his local authorities and believes wholeheartedly that they are legitimate and that he has a resultant duty to obey. Nonetheless, Anthony can think of a variety of situations in which he would break or ignore the law. This could be because of context. Anthony would quickly overrule his obligation to obey jaywalking rules if he saw a small child wandering on a busy intersection and that child required saving.62 Or, this could be because of the content of a particular law. Anthony might view a law requiring him to report a felony drug offense as remarkably unjust, given the legal consequences of the likely resultant arrest, and may believe in this instance that his duty to do justice overrides the moral reasons to obey legitimate authority. Thinking of these and other circumstances, Anthony, and any similarly situated citizen, might plausibly express strong agreement to the above questions, and the survey instrument would produce a false negative, inaccurately seeing these responses as a strike against the presence of a legitimacy belief rather than the result of normal moral reasoning. Moral Alignment, Trust, and Confidence The second and third dimensions of legitimacy are trust and confidence, and normative alignment. Trust and confidence refers

310

Jeffrey A. Lenowitz

to the belief that authorities have the people’s best interests at heart.63 As Tyler and Jackson explain, “if people feel that the authorities are sincere, benevolent, and concerned about their welfare, then they trust them to act in ways that benefit the people over whom they exercise authority.”64 Moral alignment is the citizens’ “belief that power-holders have values, goals and purposes that align with their own.”65 In this section I treat these dimensions together because they face the same two problems: (1) they are not constituent parts of legitimacy; and (2) there is no necessary relationship between them and legitimacy. Trust and moral alignment are not reliable proxies or constituent components of legitimacy, and we should therefore be skeptical of research that relies on a legitimacy index that includes them. First, and most simply, it makes little theoretical sense to state that sociological legitimacy simply means trust and normative alignment. Sociological legitimacy, as Tyler, Weber, and many others note, is the belief that moral reasons to obey the demands of an authority exist and are independent of the contents of such demands. This belief might frequently correspond, cause, or flow from trusting legal officials, or seeing an authority as having similar moral values as yourself, but the three concepts and their corresponding psychological states are separable. If Tyler wants to insist that trust and normative alignment are constituent components of what he is measuring, then he must stop calling it legitimacy. Second, legitimacy beliefs do not always co-exist with, produce, or follow feelings of trust or beliefs in normative alignment. In other words, there is no necessary relationship between these dimensions and legitimacy itself, and thus trust and moral alignment are mistaken proxies for legitimacy. To illustrate this, let us return to Anthony and Lara. Now, it is easy to imagine a scenario in which Anthony trusts his legal authorities, and his trust helps lead to his legitimacy beliefs. He knows that the police have his best interests in mind and thus comes to take their commands as being authoritative and worthy of being obeyed generally without examining their content. Or, Anthony might come to believe that his legal authorities share his personal moral convictions and are guided by similar values, and thus instill their commands with the authority of the moral reasons he thinks drives them both. In doing so, Anthony is motivated by his belief in legitimacy, which

On the Empirical Measurement of Legitimacy

311

has taken over for or made irrelevant his moral values. In both scenarios, Anthony holds something like Raz’s service conception of authority, in that he takes as authoritative and obeys the directives of his legal authorities because he believes that, by doing so, he best conforms to reasons that already apply to him.66 We can also imagine a situation in which the causality flows the other way: Anthony might self-rationalize that his legal authorities have his best interests at heart or share his moral values in an effort to come to grips with his beliefs in legitimacy. In all of these scenarios, Anthony’s responses to questions measuring trust and moral alignment will be positive, which would accurately push his legitimacy score higher as a result. Learning that Anthony has trust in his legal authorities, and that he sees them as sharing his values, accurately indicates that he sees his authorities as legitimate. However, we can imagine a quite different but nonetheless plausible scenario. Now, Anthony is a committed procedural democrat who believes that the laws, officials, and employees of his state are legitimate because they all come from and are structured by a democratic process that gives equal consideration to all citizens under its jurisdiction. However, Anthony is unhappy with the most recent round of elections. He believes that most of those elected are untrustworthy and do not have his best interests at heart. He is confident that his values and theirs are quite different. Recent laws and enforcement methods are repugnant to him. If he is asked whether he thinks his legal authorities are trustworthy or share normative values with him, he would very quickly answer in the negative. However, Anthony still thinks his legal authorities are legitimate, he still thinks he has good moral reasons to follow the commands of the authorities because his conception of legitimacy is grounded in the institution of democracy, not the most recent actions by legal officials.67 This scenario is plausible because the dimensions of trust and moral alignment are not necessarily in a causal relationship with, and do not necessarily correspond with, beliefs in the legitimacy of legal authorities. They can be in such relationships, but they also cannot be. This means that Tyler and his co-authors might score people with high legitimacy beliefs as having low-level legitimacy beliefs simply because they don’t also see their authorities as particularly trustworthy or sharing similar values.

312

Jeffrey A. Lenowitz

Now consider Lara, who trusts that her legal authorities have her best interests at heart and share her moral values. She responds positively to statements such as “the police care about the well-being of everyone they deal with” and “the laws in your community are consistent with your own intuitions about what is right and just.”68 However, she also does not believe her authorities are legitimate; she does not see herself as having contentindependent moral reasons to obey their commands. This might be for a variety of reasons. Perhaps she lacks the framework for or concept of legitimacy. Perhaps she is a democratic proceduralist who sees the local democratic process as corrupt. Regardless, as long as someone like Lara is possible and plausible, then there is another problem with using trust and normative alignment as indicators for legitimacy. People like Lara might score highly in at least two dimensions of legitimacy, and be treated as seeing their authority as legitimate, but nonetheless deny the legitimacy of their institutions and authorities. In this way, measuring legitimacy through the dimensions of moral alignment and trust, like with felt obligation, is liable to produce both type 1 and type 2 errors. The Independent Force of Moral Value Finally, moral alignment creates an additional problem for Tyler’s index. For, whenever it occurs, whenever an individual believes they share a common morality with authorities and the law, moral values might directly lead them to comply for reasons unrelated to legitimacy. This means that including moral alignment in a legitimacy index might make legitimacy seem more closely correlated with the desired compliance behavior than it actually is, and thus provides further reason to be wary of treating such alignment as a legitimacy indicator. Imagine that Anthony and Lara see moral alignment between themselves and their legal authorities. Here are some of the questions used to measure moral alignment: Q10. My own feelings about what is right and wrong generally agree with what the law says.69 Q11. The police generally have the same sense of right and wrong that you do. Q12. Judges stand up to values that are important to you.70

On the Empirical Measurement of Legitimacy

313

Anthony and Lara answer strongly in the affirmative to these questions, as they see the law and its officials as broadly aligned with their personal set of moral values. This is the very definition of moral alignment, and it also means that Anthony and Lara likely believe that, generally, following the law, listening to the police, and obeying the edicts of a judge is compliant with their given moral values. Without wading into debates about norms, morality, and practical reasons, we can nonetheless say that moral or normative values are reasons for action.71 If I morally value animal life, this is a reason to be vegetarian; I have a reason to stop eating animals because I value animal life. Put this way, it is easy to see why Anthony and Lara have reasons to obey the law or follow the dictates of their legal authorities independent of legitimacy—this is what their own values tell them to do. Anthony and Lara see the law and their legal authorities as tracking their own moral values. In each instance this is true, they have reasons to comply, cooperate, or positively engage. These reasons are their moral values, and the reasons act upon them directly. The fact that Anthony believes his authorities are legitimate, and that Lara believes the opposite, has no necessary bearing on this observation, for the reasons generated by the drive to live morally and enact one’s sense of value on the world can operate separately from legitimacy. Tyler himself acknowledges that moral values are one of the rival non-coercive sources of legal compliance to legitimacy, describing them as “personal standards to which people attempt to align their behavior.”72 In his contribution to this volume, he lists moral values as one of “three bodies of non-instrumental influence upon behavior” along with norms and legitimacy.73 It is thus striking that he simultaneously includes moral alignment as a dimension of moral legitimacy itself. If a person sees their personal standards in the law and those that enforce it, then the drive to align her behavior with those standards will also be a drive to obey the law and respect authorities. Legitimacy beliefs do not necessarily need to enter into the requisite moral reasoning. This means that part of the found correlation between legitimacy and compliance and cooperation might be from the independent effect of moral values.

314

Jeffrey A. Lenowitz

Tyler’s Response As I have argued, there are a variety of reasons to doubt the accuracy of Tyler’s operationalizing of legitimacy and the strength of LC. In his contribution to this volume, Tyler rightly notes that “there are a variety of ways to decide what an appropriate concept is to reflect legitimacy” and that “changes in concept influence which variables should be measured empirically.” In defense of his own conceptualization of legitimacy, specifically the breakdown of the concept into three dimensions, he explains that “each of these factors make a unique empirical contribution to predicting the behaviors of concern in models of law,” and that “the model outlined here justifies a concept by its distinct contribution to explaining behavior.”74 This defense is striking in its inadequacy. To be clear, the behavior in question is compliance, cooperation, and positive engagement. These are the dependent variables in Tyler’s model. The independent variables include legitimacy, coercion/risk, social norms, and moral values. Tyler’s explanation for conceptualizing and measuring one of his independent variables in a particular way is because doing so ensures that the independent variable correlates with the dependent variable. A felt obligation to obey, trust and confidence, and normative alignment, Tyler thus argues, are acceptable dimensions of legitimacy because each of them seems to explain variability in cooperation, compliance, and engagement. This, of course, is no justification at all. It is closer to an admission of manipulating the measure of a concept in order to get one’s desired results. Level of education, past criminal behavior, local temperature, each of these might “make a unique empirical contribution to predicting” lawabiding behavior, but that is no argument for declaring them dimensions of legitimacy and including them in the construction of a legitimacy index. Questioning Procedural Justice Leads to Legitimacy I now turn to procedural justice leads to legitimacy (PJL), the claim that authorities are more likely to be seen as legitimate if they are procedurally just, and that this effect is greater than simply giving

On the Empirical Measurement of Legitimacy

315

individuals good outcomes. Combined with LC, this means that procedurally just authorities are most likely to elicit cooperation, compliance, and positive engagement. As Tyler summarizes: “A core finding of that literature is that authorities and institutions are viewed as more legitimate and, therefore, their decisions and rules are more willingly accepted when they exercise their authority through procedures that people experience as being fair.”75 In this section I question two aspects of PJL. The first concerns the concept of trust, which appears in Tyler’s measures for both procedural justice and legitimacy. The second concerns the policy implications of PJL. Trust Tyler breaks down procedural justice into four measurable criteria split into two dimensions. The first dimension is the fairness of decision-making, consisting of voice and neutrality, and the second is the fairness of interpersonal treatment, consisting of trust and respect.76 Consider a routine traffic stop by a policewoman. The stop meets the criteria of fairness in decision-making if the policewoman initiated and carried out the stop in accordance with settled law and policy rather than because of personal opinion or bias (neutrality), and the driver was able to explain his actions before any formal decision was made (voice). The stop consisted in fair interpersonal treatment if the policewoman appeared motivated to treat the driver fairly and helpfully (trust) and acted in a polite and dignified manner, respectful of the driver’s legal rights (respect). Thus, procedural justice involves respectful and unbiased treatment with trustworthy motives that take an individual’s concerns into account before reaching a decision. A problem emerges once one considers the category of trust and how it is similar to the dimension of legitimacy Tyler refers to as “trust and confidence.” For instance, Tyler and Jackson define trust, a component of procedural justice, as the belief that “the authorities are . . . benevolent and caring, and are sincerely trying to do what is best for the people with whom they are dealing.” The authors measure the concept by asking respondents how often the police or courts “make decisions that are good for everyone in the community” or “try to do what is best for the people they are dealing with.” In the same article, they define legitimacy’s trust

316

Jeffrey A. Lenowitz

and confidence as the belief that authorities are “sincere, benevolent, and concerned about their [citizens’] welfare,” such that they will “act in ways that benefit the people over whom they exercise authority.” Questions used to measure this dimension of legitimacy include asking respondents to give their level of agreement or disagreement to statements such as: “you generally support how the police act in your community,” “the police only care about the views of some of the people in your community” (reverse scored).77 As this shows, the definition of and the questions used to measure trust, a component of procedural justice, are similar to the definition of and the questions used to measure trust and confidence, a component of legitimacy. The worry, of course, is that some of the found correlation between procedural justice and legitimacy stems not from the former leading to the latter, but rather from including the same indicator within the measures of the two concepts. In other words, some of the reason that procedural justice is found to lead to legitimacy could be because Tyler is measuring the same psychological state for both concepts; the predictor and outcome variables are not entirely independent of one another. This casts doubts on the strength of PJL. Policy Implications The policy implications of PJL seem relatively straightforward: Transform legal authorities and their respective institutions so that they embody procedural justice norms rather than simply emphasizing specific positive outcomes and coercive force. “The message means that in this arena,” Tyler claims, “doing what is right (acting ethically) and doing what works converge.”78 In this section, I argue that even if we overlook all of the other concerns with LC and PJL that I mention above, we still cannot draw the conclusion that acting ethically and effectively converge. Taken at face value, Tyler’s empirical research has not produced definitive reasons for police departments, courts, and other legal authorities to act ethically. Instead, he has simply provided empirical support for Machiavelli’s claim that the appearance of virtue, rather than its possession, is all a prince needs. To measure procedural justice, Tyler and his co-authors rely again on a self-report survey methodology, asking individuals to

On the Empirical Measurement of Legitimacy

317

answer questions about their previous interactions with authorities that correspond to the four criteria listed above. Thus, for example, respondents are asked how often the authorities in question “treat people with dignity and respect,” “make fair and impartial decisions,” “try to do what is best for the people they are dealing with,” and “give people a chance to tell their side of the story.”79 Such questions gather data about subjective procedural justice, the degree to which the police appear procedurally just to those interacting with them. Tyler and his co-authors thus do not actually measure or analyze the effects of procedural justice. Instead, they look at perceptions of procedural justice. Thus, at best, Tyler shows the effects of the appearance of procedural justice on individual perceptions of legitimacy. The distinction between being procedurally just and being perceived as procedurally just matters because the two are different and can diverge. It is entirely possible for an individual to perceive an unfair procedure as fair, to view a biased authority as neutral, and to believe they had an opportunity to explain their actions and affect an outcome when in fact they did not. And, the reverse of all of these is possible as well. Thus, it is a mistake when researchers make procedural recommendations upon the presumption that an individual’s subjective experience mirrors what authority figures actually did. The one prominent study testing the relationship between procedural justice and its perception gives some support for the possibility of divergence, finding a correlation between objective procedural justice and perceived procedural justice of only 0.15.80 Even Tyler acknowledges that among studies that attempt to find a strong connection between objective measures of police procedural justice and perceived procedural justice, “none find a strong connection” between the two.81 Thus, even if we accept LC, the policy implications of Tyler’s work are thus different than how he presents them. Rather than giving reasons for police departments and other authorities to be more ethical and procedurally just, Tyler gives reasons in favor of making whatever reforms are necessary to amplify perceptions of procedural justice. This is a slightly different goal. It might be the case that changing an institution such that it is fair and neutral requires the same alterations as changing it so that people will see the institution as fair and neutral, but one cannot assume this at the outset.

318

Jeffrey A. Lenowitz

Despite what was said earlier, this gives a Machiavellian gloss to Tyler’s research. I bring up Machiavelli because the conflation of these two goals—appearing just and being just—is perhaps his most central critique of his contemporaries, who he sees as unwisely following the Ciceronian tradition of blindly urging rulers to cultivate moral virtues, especially justice, in pursuit of the goals of honor, glory, and fame.82 Machiavelli accepts the importance of these values (which somewhat overlap with what we now call legitimacy), but argues that to achieve honor, glory, and fame “it is not necessary for a prince to possess all of the above mentioned qualities [virtues], but it is very necessary for him to appear to possess them.”83 His point was that with regard to maintaining control of the state, the benefits of being just and virtuous stem solely from appearing to be just and virtuous, and thus a ruler only needs to keep up such appearances. This will likely involve being just and virtuous most of the time, but it also means actions in opposition to such values are never out of bounds. Thus, nothing in Tyler’s findings rule out institutional reforms that have no effect or a negative effect on objective procedural justice, just so long as they improve perceived procedural justice and have a minimal risk of being discovered as fraudulent. These can vary from the wasteful, like public money being used for public relations campaigns designed to make criminal justice institutions appear more procedurally just, to reforms more sinister or manipulative. For example, Tyler mentions the introduction and advertisement of transparent means for citizens to issue police complaints as a way of giving citizens voice in local policing practice, something that is central to promoting perceptions of procedural justice.84 Of course, the procedurally just reform would be to have these complaints read, analyzed, and distributed to officers when necessary. But, considering that these are costly behind-closeddoors sorts of processes, it might be the case that collecting complaints but ignoring them will have the same effect when it comes to perceptions of procedural justice. Similarly, racial profiling is an unjust practice correlated with high perceptions of illegitimacy.85 Rather than stop it, police might simply occasionally question people from a non-targeted race in order to obscure their real methods. All things being equal, Tyler’s results cannot recommend the real procedural justice reform over the artificial one.

On the Empirical Measurement of Legitimacy

319

Conclusions Tyler profoundly shaped, and continues to shape, research on rule compliance and cooperation across the social sciences. His work has not only led to positive changes in the practices of policing and the courts, his primary focus, but also has influenced policy formation in the areas of dispute resolution processes, immigration control, workplace governance, and more. His contribution to this volume summarizes his research and adds further empirical support to what I argue are his two central arguments: (LC) that sociological legitimacy exists, is an effective means of securing desirable legal behavior, and does so better than coercion; and (PJL) that procedural justice not only helps foster legitimacy beliefs, but is more effective at it than giving citizens desirable or fair outcomes. This, as I explained, is an argument for procedural sociological legitimation. Tyler’s work, its implications, and the research agenda it has led to, are valuable. However, I argue that its contributions are significantly different than usually characterized. Most important, the research tells us little about legitimacy. Tyler’s operationalization of the concept distorts it beyond common meaning, including the meaning Tyler ascribes to it himself. Multiple sources cause this distortion. First, the three dimensions used to measure legitimacy are unreliable, in that they can produce both false positives and false negatives. I can believe my authorities lack legitimacy, yet also feel obliged to obey the law, or have trust in my authorities, or believe they share a common normative outlook with me. Holding these beliefs together is not contradictory. At the same time, I can distrust my authorities, consider my normative outlook different than their own, yet nonetheless believe they are legitimate and that I therefore have an obligation to obey. Thus, Tyler’s legitimacy index has the potential to count beliefs as indicative of legitimacy when they are not, and to count beliefs as signaling a lack of legitimacy when the opposite is the case. Second, the questions used to gauge a felt obligation to obey the law are especially problematic. By treating such an obligation as indefeasible, the measure scores people as lacking legitimacy beliefs simply because they hold that, in certain circumstances, they can be overridden. At the same time, the questions are not

320

Jeffrey A. Lenowitz

fine-grained enough to separate beliefs that one ought to obey the law that stem from fear, or strategic calculations, from those that come from a genuine legitimacy belief. Again, these discrepancies introduce further potential errors in Tyler’s measurement of legitimacy. Third, the fact that moral values can directly lead a person to comply, as Tyler himself admits, and that moral alignment occurs when a person sees her legal authorities as having compatible moral values, means that Tyler includes a confounding variable in the measure of his independent variable itself. This casts the resulting analysis even more in doubt. Given that Tyler’s chosen means of measuring legitimacy is unreliable—it does not necessarily track legitimacy beliefs—one cannot conclude anything about legitimacy from his findings. Thus, LC and PJL are unproven. We cannot interpret the empirical findings as telling us anything about legitimacy’s effect on compliance, cooperation, and engagement, nor its correlation with perceived procedural justice. At best, the version of LC that Tyler proves is something like the following: Individuals are more likely to comply, cooperate, and positively engage with the law when they feel obligated to obey it, when they trust their legal authorities, and when they believe they share moral values with the law and its enforcers, and this effect is greater than when individuals are simply worried about getting punished for law-breaking. This is still an important finding, one that can be useful for the purposes of reform and for understanding when and why people obey, but it is far from a finding about legitimacy itself. It tells us little about the effects of a genuine belief in the rightful authority of the law on legal compliance, and thus has little to say to those preoccupied with that question. In addition, Tyler’s work, regardless of his statements to the contrary, does not give reasons to make the institutions of state power more procedurally just. What he finds is that if people perceive their authorities to be procedurally just, then they are more likely to trust their authorities, see normative alignment, and feel an obligation to obey, all of which make legal compliance more likely. However, this is not the same thing as providing evidence that authorities should become more just, because what it takes to seem just will include different reforms than what is needed to be just. This, of course, undercuts the idea that Tyler offers us a way

On the Empirical Measurement of Legitimacy

321

to avoid moral argument when it comes to balancing the goal of securing legal stability and compliance with the need to respect the rights and equality of all citizens. In order to successfully argue that the institutions of the criminal justice system, and the state more generally, need to move closer toward the ideals of justice, moral argument is needed. This, of course, is a task that philosophers and normative theorists are more than happy to take up. Notes 1. Plato, “Crito,” in Plato: Five Dialogues (Indianapolis, IN: Hackett, 2002); Max Weber, The Theory of Social and Economic Organization (New York: Free Press, 1997), 215–216; George Klosko, Political Obligation (Oxford: Oxford University Press, 2005); A. John Simmons, Moral Principles and Political Obligations (Princeton, NJ: Princeton University Press, 1979); Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986). 2. Tom R. Tyler and Steven Blader, “Can Businesses Effectively Regulate Employee Conduct? The Antecedents of Rule Following in Work Settings,” Academy of Management Journal 48, no. 6 (2005): 1143–1158; Tom R. Tyler and Yuen J. Huo, Trust in the Law: Encouraging Public Cooperation with the Police and Courts (New York: Russell-Sage Foundation, 2002); Kristina Murphy, “Regulating More Effectively: The Relationship Between Procedural Justice, Legitimacy and Tax Compliance,” Journal of Law and Society 32 (2005): 562–589; Valerie Braithwaite, “Resistant and Dismissive Defiance Towards Tax Authorities,” in Legitimacy and Compliance in Criminal Justice, ed. Adam Crawford and Anthea Hucklesby (New York: Routledge, 2013); Doreen McBarnet, “Questioning the Legitimacy of Compliance: A Case Study of the Banking Crisis,” in Legitimacy and Compliance in Criminal Justice, ed. Adam Crawford and Anthea Hucklesby (New York: Routledge, 2013); Marc J. Stern, “Coercion, Voluntary Compliance and Protest: The Role of Trust and Legitimacy in Combating Local Opposition to Protected Areas,” Environmental Conservation 35, no. 3 (2008): 200–210. 3. Tom R. Tyler, “Legitimacy and Criminal Justice: The Benefits of SelfRegulation,” Ohio State Journal of Criminal Law 7 (2009): 307–359. 4. COPS Office, “President’s Task Force on 21st Century Policing Implementation Guide: Moving from Recommendations to Action,” (Washington, DC: Office of Community Oriented Policing Services, 2015); “Executive Summary: About the National Initiative,” National Initiative for Building Community Trust & Justice, http://trustandjustice.org. 5. I take this conceptual taxonomy from Richard Fallon’s wonderful article on constitutional legitimacy, though I alter, add to, and depart from

322

Jeffrey A. Lenowitz

his substantive understanding of each legitimacy type. Richard H. Fallon, Jr., “Legitimacy and the Constitution,” Harvard Law Review 118, no. 6 (2005): 1787–1853. For somewhat similar conceptions, see: Bruce Gilley, The Right to Rule: How States Win and Lose Legitimacy (New York: Columbia University Press, 2009), 7; David Beetham, The Legitimation of Power (New York: MacMillan, 1991); Leslie Green, “Law, Legitimacy, and Consent,” Southern California Law Review (1998): 797; Wilfried Hinsch, “Justice, Legitimacy, and Constitutional Rights,” Critical Review of International Social and Political Philosophy 13, no. 1 (2010): 40. 6. For example, Simmons’s criticism of the Weberian conception of legitimacy loses much of its purchase once one realizes that while Simmons is concerned with moral legitimacy, the social scientists deploying the Weberian conception are solely concerned with sociological legitimacy. A. John Simmons, Justification and Legitimacy: Essays on Rights and Obligations (Cambridge: Cambridge University Press, 2001), 133–134. 7. Fallon, “Legitimacy and the Constitution,” 1786. 8. Raz, Morality of Freedom, 23–99; Joseph Raz, Between Authority and Interpretation: On the Theory of Law and Practical Reason (Oxford: Oxford University Press, 2009), 126–66; Joseph Raz, The Authority of Law, 2nd ed. (Oxford: Oxford University Press, 2009), 3–33, 233–291. Note that I add “override” to Raz’s description of authoritative reasons as preemptive in order to accommodate the possibility that reasons are extremely weighty, rather than fully preemptive. For such an account, see Scott Shapiro, “Authority,” in The Oxford Handbook of Jurisprudence and Philosophy of Law, ed. Jules L. Coleman and Scott Shapiro (Oxford: Oxford University Press, 2002). 9. One fundamental component of a state’s assertion that it has a right to rule is its right to coerce. Understood in this way, the alleged split between understanding legitimacy as relating to authority or justified coercion seems to disappear or at least be minimized. For a discussion of this debate, see Arthur Ripstein, “Authority and Coercion,” Philosophy & Public Affairs 32, no. 1 (2004): 2–35. 10. Raz, Morality of Freedom, 26; Christopher McMahon, Collective Rationality and Collective Reasoning (Cambridge: Cambridge University Press, 2001), 50. 11. This is the sense of legitimacy most popular in the political science literature on legitimacy, where David Easton’s description of legitimacy as “the conviction ‘That is it right and proper . . . to obey the authorities and to abide the requirements of the regime. It reflects the fact that in some vague or explicit way [a person] sees these objects as conforming to his own moral principles, his own sense of what is right and proper in the political sphere.’” David Easton, “A Re-Assessment of the Concept of

On the Empirical Measurement of Legitimacy

323

Political Support,” British Journal of Political Science 5, no. 4 (1975): 435– 457, at 451. 12. Weber, The Theory of Social and Economic Organization, 382. See also Robert Grafstein, “The Failure of Weber’s Conception of Legitimacy: Its Causes and Implications,” Journal of Politics 43, no. 2 (May 1981): 456–472. 13. Beetham helpfully points out that “a given power relationship is not [sociologically] legitimate because people believe in its legitimacy, but because it can be justified in terms of their beliefs.” Beetham, The Legitimation of Power, 11. 14. Tom R. Tyler, Why People Obey the Law (New Haven, CT: Yale University Press, 1990); Anthony Bottoms and Justice Tankebe, “Beyond Procedural Justice: A Dialogic Approach to Legitimacy in Criminal Justice,” Journal of Criminal Law & Criminology 102, no. 1 (2012): 120; Carl Friedrich, Man and His Government (New York: McGraw-Hill, 1963), 234. 15. Friedrich, Man and His Government, 234. 16. Raz, Between Authority and Interpretation, 153; Hans Kelsen, General Theory of Law and State (New Brunswick, NJ: Transaction, 2006), 117. 17. Joseph Raz, “On Lawful Governments,” Ethics 80, no. 4 (1970): 300; Green, “Law, Legitimacy, and Consent,” 797; H.L.A. Hart, Essays in Jurisprudence and Philosophy (Oxford: Clarendon Press, 1983), 21–48. 18. The following is inspired by Rawls’s famous discussion of perfect, imperfect, and pure procedural justice. John Rawls, A Theory of Justice, Revised ed. (Cambridge, MA: Harvard University Press, 1999), 74–76. 19. Tom R. Tyler, “Evaluating Consensual Models of Governance: Legitimacy Based Law,” in Annual Meeting of the American Society for Political and Legal Philosophy (Kansas City, MO, 2017), 1, 22. For Tyler, compliance refers to doing what the law or political authorities demand, cooperation means proactively assisting legal authorities, and engagement refers to the generation of supportive attitudes and behavior toward one’s community. See Tom R. Tyler and Jonathan Jackson, “Popular Legitimacy and the Exercise of Legal Authority: Motivating Compliance, Cooperation and Engagement,” Psychology, Public Policy, and the Law 20, no. 1 (2014): 79–80. 20. Tyler, Why People Obey the Law. 21. Jason Sunshine and Tom R. Tyler, “The Role of Procedural Justice and Legitimacy in Shaping Public Support for Policing,” Law & Society Review 37, no. 3 (2003): 513–548. 22. Johnathan Jackson et al., “Why Do People Comply with the Law?,” British Journal of Criminology 52 (2012): 1051–1071. 23. Tom R. Tyler, Jeffrey Fagan, and Amanda Geller, “Street Stops and Police Legitimacy: Teachable Moments in Young Urban Men’s Legal Socialization,” Journal of Empirical Legal Studies 11, no. 4 (2014): 751–785. 24. Tyler, “Evaluating Consensual Models of Governance.”

324

Jeffrey A. Lenowitz

25. H.L.A. Hart, The Concept of Law, 2nd ed. (Oxford: Oxford University Press, 1994), 40. 26. Machiavelli was attacking Cicero indirectly, in that his true targets were his more Ciceronian contemporaries. See Quentin Skinner, The Foundations of Modern Political Thought: The Age of Reformation, vol. 1 (Cambridge: Cambridge University Press, 1978), 128–138. 27. Marcus Tullius Cicero, On Duties, ed. M. T. Griffin and E. M. Atkins, trans. E. M. Atkins (Cambridge: Cambridge University Press, 1991), II.21–40. 28. Niccoló Machiavelli, The Prince, trans. Peter Bondanella (Oxford: Oxford University Press, 2005), 58. 29. Tyler, “Evaluating Consensual Models of Governance,” 26. 30. Ibid. 31. Jeremy Bentham, The Principles of Morals and Legislation (Amherst, NY: Prometheus Books, 1988). 32. Gordon Tullock, “Does Punishment Deter Crime?,” Public Interest 36 (1974): 103–111; Gary S. Becker, “Crime and Punishment: An Economic Approach,” Journal of Political Economy 76, no. 2 (1968): 169–217. 33. Frederick Schauer, The Force of Law (Cambridge, MA: Harvard University Press, 2015). 34. Tyler, “Evaluating Consensual Models of Governance,” 21. 35. Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 81. 36. Tom R. Tyler and Johnathan Jackson, “Future Challenges in the Study of Legitimacy and Criminal Justice,” in Legitimacy and Criminal Justice: An International Exploration, ed. Justice Tankebe and Alison Liebling (Oxford: Oxford University Press, 2013), 88. 37. Tyler and Huo, Trust in the Law; Tyler, Why People Obey the Law. 38. Sunshine and Tyler, “The Role of Procedural Justice and Legitimacy.” 39. Tom R. Tyler, “Procedural Justice, Legitimacy, and the Effective Rule of Law,” Crime and Justice 30 (2003): 283–357. 40. Tyler, Fagan, and Geller, “Street Stops and Police Legitimacy,” 264. 41. Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority.” 42. Lorraine Mazerolle et al., “Shaping Citizen Perceptions of Police Legitimacy: A Randomized Field Trial of Procedural Justice,” Criminology 51, no. 1 (2012): 33–64. 43. Tom R. Tyler, Avital Mentovich, and Sagarika Satyvada, “What Motivates Adherence to Medical Recommendations? The Procedural Justice Approach to Gaining Deference in the Medical Arena,” Regulation & Governance 8, no. 3 (2014): 350–370.

On the Empirical Measurement of Legitimacy

325

44. Tom R. Tyler, “Legitimacy and Compliance: The Virtue of SelfRegulation,” in Legitimacy and Compliance in Criminal Justice, ed. Adam Crawford and Anthea Hucklesby (New York: Routledge, 2012), 8. 45. Cicero, On Duties: II.37. 46. Ibid., II.33. 47. Oregon v. Mathiason (1977) 429 U.S. 492; Frazier v. Cupp (1969) 394 U.S. 731. 48. Whren v. United States (1996) 517 U.S. 806; United States v. BrignoniPonce (1975) 422 U.S. 873. 49. See the following two excellent articles: Amy Nivette and Manuel Eisner, “Does Low Legitimacy Cause Crime? A Review of the Evidence,” in Legitimacy and Criminal Justice: An International Exploration, ed. Justice Tankebe and Alison Liebling (Oxford: Oxford University Press, 2013); Daniel S. Nagin and Cody W. Telep, “Procedural Justice and Legal Compliance,” Annual Review of Law and Social Science 13 (2017): 5–28. 50. Albert Bandura, “Mechanisms of Moral Disengagement,” in Origins of Terrorism: Psychologies, Ideologies, Theologies, States of Mind, ed. Walter Reich (Cambridge: Cambridge University Press, 1990). 51. Tom R. Tyler, “Psychological Perspectives on Legitimacy and Legitimation,” Annual Review of Psychology 57 (2006): 377; Tyler, Why People Obey the Law, 4. 52. Tyler provides an excellent overview of the history of using trust and obligation as measures of legitimacy in the literature that preceded his own work. See Tyler, Why People Obey the Law, 27–36. Moral solidarity is used in place of moral alignment in Jason Sunshine and Tom R. Tyler, “Moral Solidarity, Identification with the Community, and the Importance of Procedural Justice: The Police as Prototypical Representatives of a Group’s Moral Value,” Social Psychology Quarterly 66, no. 2 (2003): 153–165. 53. Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 81. 54. M.B.E. Smith, “Is There a Prima Facie Obligation to Obey the Law?,” Yale Law Journal 82 (1973): 950–976; Rolf Sartorius, “Political Authority and Political Obligation,” Virginia Law Review 67, no. 1 (1981): 3– 17; Allan Buchanan, “Political Legitimacy and Democracy,” Ethics 112, no. 4 (2002): 689–719; Kent Greenawalt, Conflicts of Law and Morality (Oxford: Clarendon Press, 1987), 47–59. 55. Raz, Between Authority and Interpretation, 127. 56. Rawls, A Theory of Justice, 100. 57. Tankebe also criticizes the inclusion of obligation into the measure of legitimacy, but does so by mistaking any decision that one should obey as indicative of a belief in an obligation to obey. Justice Tankebe, “View-

326

Jeffrey A. Lenowitz

ing Things Differently: The Dimensions of Public Perceptions of Police Legitimacy,” Criminology 51, no. 1 (2013): 105. 58. Tyler, “Evaluating Consensual Models of Governance,” 11; Sunshine and Tyler, “The Role of Procedural Justice and Legitimacy,” 539. 59. Raz, The Authority of Law, 18, 22. 60. S. L. Hurley, Natural Reasons: Personality and Polity (Oxford: Oxford University Press, 1989), 133–134. 61. Tyler, “Evaluating Consensual Models of Governance,” 11; Sunshine and Tyler, “The Role of Procedural Justice and Legitimacy,” 539; Tom R. Tyler and Jeffrey Fagan, “Legitimacy and Cooperation: Why Do People Help the Police Fight Crime in Their Communities?,” Ohio State Journal of Criminal Law 6 (2008): 270. 62. On a weighty reason account, the weight of the reason in favor of saving a life would outweigh the reason in favor of obeying. On a protected reason account, saving a life would be an unexcluded reason, given that context-specific reasons on such matters are not the sort of reasons that legal authorities can take into account when making general laws. Raz, Between Authority and Interpretation, 144. 63. I simply call this dimension “trust” going forward. 64. Tyler and Jackson, “Future Challenges in the Study of Legitimacy and Criminal Justice,” 88. 65. Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 81. 66. Raz, The Morality of Freedom, ch. 3; Raz, Between Authority and Interpretation, 136–137. 67. Perhaps, over time, if things don’t change, Anthony’s legitimacy beliefs will decrease. This is entirely possible but does not refute my point that it is possible to imagine a plausible scenario in which a lack of trust corresponds with a belief in legitimacy. 68. Tyler and Fagan, “Legitimacy and Cooperation,” 270; Tyler, “Evaluating Consensual Models of Governance,” 12. 69. Sunshine and Tyler, “Moral Solidarity,” 156–157. 70. Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 84. 71. Joseph Raz, Practical Reason and Norms (Oxford: Oxford University Press, 1975), 25. 72. Tyler, “Psychological Perspectives on Legitimacy and Legitimation,” 390. 73. Tyler, “Evaluating Consensual Models of Governance,” 20. 74. Ibid., 11. 75. Tyler, “Psychological Perspectives on Legitimacy and Legitimation,” 379.

On the Empirical Measurement of Legitimacy

327

76. Tom R. Tyler and Avital Mentovich, “Mechanism of Legal Effect: Theories of Procedural Justice” (Temple University Beasley School of Law: Public Health Law Research Program), 2; Kristina Murphy, “Procedural Justice and Its Role in Promoting Voluntary Compliance,” in Regulatory Theory: Foundations and Applications, ed. Peter Drahos (Canberra: Australian National University Press, 2017), 46; Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 81–82. 77. Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 81–84. 78. Tom R. Tyler, “Legitimacy and Compliance: The Virtue of SelfRegulation,” in Legitimacy and Compliance in Criminal Justice, ed. Adam Crawford and Anthea Hucklesby (New York: Routledge, 2012), 10. 79. Tyler, “Evaluating Consensual Models of Governance,” 22; Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 84. 80. Robert E. Worden and Sarah J. McLean, Mirage of Police Reform: Procedural Justice and Police Legitimacy (Berkeley: University of California Press, 2017), 134. Objective procedural justice was measured by researchers’ coding of audio and visual recordings of citizen/police interactions. 81. Tom R. Tyler, “Procedural Justice and Policing: A Rush to Judgment?,” Annual Review of Law and Social Science 13 (2017): 29–53. 82. Skinner, The Foundations of Modern Political Thought: The Age of Reformation, 1: 131–132. See also Maurizio Viroli, Machiavelli (Oxford: Oxford University Press, 1998), 50–55. 83. Machiavelli, The Prince, 61. 84. Tyler, “Legitimacy and Compliance: The Virtue of Self-Regulation,” 17; Tyler and Jackson, “Popular Legitimacy and the Exercise of Legal Authority,” 81–82. 85. Tyler, Fagan, and Geller, “Street Stops and Police Legitimacy.”

11 THE EMPIRICAL STUDY OF LEGITIMATE AUTHORITY NORMATIVE GUIDANCE FOR POSITIVE ANALYSIS SANFORD C. GORDON AND GREGORY A. HUBER

In this chapter we examine three related questions of broad interest across the social sciences. First is the legitimacy of institutions and individuals in positions of authority a meaningful, analytically distinct empirical phenomenon? Second, if it is, is a rigorous, causally identified empirical analysis of its origins and consequences possible? And third, what can empirical scholars interested in the study of legitimacy learn from the rich normative literature on the subject? Since the seminal writings of Max Weber in the early twentieth century, the idea of legitimacy has emerged as a core concept in the positive analysis of human interaction across the social sciences. What might be called the “legitimacy hypothesis”—that beliefs about the legitimacy of authorities and institutions are essential to long-term cooperation and compliance—enjoys prominence in numerous fields of relevance to political science. In the domestic context, this is particularly the case in the analysis of courts and law enforcement; in the comparative context, in discussions of statebuilding, counterinsurgency, and pro- and anti-regime sentiment and activity; and in the international context, in examinations of norms and cooperation with supranational organizations. Generally 328

The Empirical Study of Legitimate Authority

329

speaking, a common view that emerges across these subject areas is that material incentives alone are inadequate to facilitate social cooperation. At the heart of the legitimacy hypothesis is the idea that citizens may voluntarily comply for reasons over and above the likelihood of reward or punishment, but which are tied instead to the characteristics or behavior of authorities and institutions.1 We examine the normative political theory literature on legitimacy and argue that it provides useful interpretive guidance on how to put the positive literature on firmer analytical foundations. Two issues are of particular note: the close association of legitimacy with the justifiability of an authority’s actions and the question of whether citizens, as moral agents, have an obligation to comply with authority. The normative literature’s focus on these two issues in particular is helpful to us in constructing a definition of legitimate authority as an empirical phenomenon. Specifically, obligation implies a motivation to comply apart from extrinsic, material motivations; while justification implicates citizens’ beliefs about authorities and institutions. Understanding the structure of legitimacy in terms of motivations and beliefs allows us, as positive analysts, to draw on game-theoretic intuitions and equilibrium logic. Specifically, we can think of legitimacy as a feature of an equilibrium in which citizens’ intrinsic motivations are enhanced by those beliefs about authorities, and the actions of governing institutions and their incumbents are consistent with those beliefs.2 (Throughout, we use the term intrinsic motivations to refer to internally generated reasons for behavior [e.g., moral values] and the term extrinsic motivations to refer to external pressures for behavior [e.g., financial rewards, the threat of coercion, etc.].) We want to make it clear at the outset that we do not mean to engage in definitional trickery, merely redefining legitimacy in such a way as to find fault with the extant empirical literature. Instead, our aim is to isolate the features of explanations for deference to authorities specific to the idea of their legitimacy that make them analytically distinct from other categories of explanation—particularly those based on extrinsic material rewards and punishments. It is our belief that definitional clarity facilitates improved empirical work and will also allow that empirical work to speak more clearly to the concerns undergirding the normative scholarship.

330

Sanford C. Gordon and Gregory A. Huber

Next, we review the empirical political science research on legitimacy appearing in selected key journals since 1970 in three subject areas: courts, regimes in general, and international organizations. Our discussion of these literatures suggests the existence of three fundamental challenges to studying legitimacy empirically. First, the term “legitimacy” as defined in the literature tends to conflate a number of analytically distinct factors affecting attitudes about, and cooperation and/or compliance with, authorities and institutions. Attempts to measure beliefs about legitimacy frequently are confounded by this same theoretical ambiguity. Second, legitimacy is often studied in institutional environments where its behavioral consequences are likely to be obscured. And third, studies of legitimate authority are, with a few exceptions, insufficiently attentive to issues of causal inference and identification. Having explored the empirical literature, we then elaborate on the challenges to inference while providing some guidelines to researchers about research designs with which scholars can uncover legitimating effects of institutions and authority actions, should those effects exist. Our analysis points to situations in which the identification of “institutional” and “personal” legitimacy is and is not feasible; to conditions under which material factors are likely to yield upward versus downward bias in estimates of legitimacy effects; and to conditions under which variation in perceptions of an institution’s legitimacy are most likely to yield observable behavioral implications. The Research Environment Insights from Normative Theory Though a full account of the development of the concept of legitimacy in normative political theory and philosophy is beyond the scope of the current inquiry, several key points of discussion stand out as relevant to our treatment of legitimacy as an empirical phenomenon.3 The first of these is the problem of justification. Raz, for example, notes that a key feature of authority (e.g., of the state) is the right to issue exclusionary reasons to perform actions, that is, reasons that categorically override or exclude other reasons (for

The Empirical Study of Legitimate Authority

331

acting differently).4 Insofar as the coercive component of authority may conflict with the autonomy of the individual citizen— itself a primary good—the state must have a compelling rationale to exercise this authority. In a similar vein, Habermas states that legitimacy “means that there are good arguments for a political order’s claim to be recognized as right and just,” referring to an order’s “worthiness to be recognized.”5 Anscombe associates legitimate authority explicitly with the administration of justice.6 And Beetham directly ties justification to legitimacy, arguing that in a given power relationship an authority gains its legitimacy from the fact that it can be justified in terms of the beliefs of its participants.7 Overall, justification is therefore linked to being able to express reasons, often based in values, for acting as an authority. What constitutes adequate justification is itself likely to be contingent on the historical peculiarities and moral traditions of the polity in question, a point noted by Habermas in his discussion of the historical evolution of the legitimation activities of different regimes. From the perspective of liberal political theory, however, justification entails an alignment of government’s exercise of power with commonly (or universally) accepted principles. This is the position articulated by Rawls, who associates the legitimate exercise of state power with “a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason.”8 Relatedly, Nagel associates legitimate liberal governance with being impersonal or impartial.9 The second key point of discussion in this literature is the relationship between consent on the one hand and the obligation to obey on the other. Consent, of course, forms the basis of all contractarian theories of governance. A substantial range of opinion exists, however, concerning the duty of the citizen to comply with the edicts of a duly constituted authority. At one extreme lies the Hobbesian view, which holds that the social contract between sovereign and citizen entails an absolute duty of the citizen to comply with the sovereign’s commands so long as the commands do not threaten the subject’s inviolable right to (physical) selfpreservation. At the other extreme, philosophical anarchists articulate a view of individual autonomy as sacrosanct, to the point of rejecting any notion of a citizen owing a debt of obedience to an

332

Sanford C. Gordon and Gregory A. Huber

authority (although citizens are certainly entitled to give weight to the instructions of authorities as part of a constellation of reasons).10 Dworkin offers a middle ground perspective, endeavoring to articulate an account of legitimacy grounded in the obligation of the citizen to the community rather than to the state itself.11 The linkages among legitimacy, obligation, and consent suggest a behavioral consequence to more legitimate authority, because greater legitimacy will translate into consent and (at least weakly) a greater tendency to obey or cooperate with an authority. For example, Beetham notes that legitimacy provides a “moral grounds” for cooperation and obedience that exists in a complex set of motivations for compliance that may include both immediate self-interest as well as moral or normative reasons.12 Importantly, this means that legitimacy may affect behavior, and also that consent alone is not evidence of legitimacy because there are many other reasons for cooperation. Similarly, Hampton distinguishes between what she refers to as “convention consent” and “endorsement consent.”13 The former is behavior that is de facto supportive of, or at the very least not undermining of, the regime. By contrast, endorsement consent is behavior that conveys active approval of the regime. Endorsement consent, she writes, “is a decision to support [the regime] because of one’s determination that it is a good thing to support. By giving this form of consent, the subject conveys her respect for the state, her loyalty to it, her identification with it, and her trust in it.” Consent can therefore vary in degree with legitimacy, and all consent is not evidence of legitimacy. What are the valuable insights from this literature for empirical scholars, and what are some pitfalls to avoid? The first concerns the notion of obligation. Irrespective of where one comes down on the duty to obey an authority, it is clear that normative theorists are articulating a notion of motivation to comply distinct from extrinsic motivations. When we speak of obligation, then, we are speaking of intrinsic motivations to cooperate with or support an authority. And, similarly, cooperation alone is not evidence of that intrinsic motivation. Most bluntly, this means that empirical researchers need to isolate intrinsic motivations for behavior in order to instantiate the concept of legitimacy as articulated by normative theorists. The second insight concerns justification. An authority’s actions implicate the beliefs of agents about the fairness or justice

The Empirical Study of Legitimate Authority

333

of institutions or individuals, and it is only fair and just institutions that are “worthy” (to use Habermas’s term) of deference. That is, perceptions of fairness (among other values) enhance or diminish the intrinsic motivation to comply. Together, obligation and justification suggest a normative core to legitimacy in the motivation of citizens to comply with authority. Critically, and borrowing from the language of game theory (a tool we employ in greater depth below), this normative core is an equilibrium: The beliefs associated with justification, and the actions associated with obligation, must be mutually consistent. If the authority acts in an unjustifiable way, obligation is undermined. And if citizens feel no obligation, there is little incentive for an authority to justify its actions. The normative literature accordingly points us in the direction of an equilibrium definition of legitimacy potentially amenable to empirical analysis: Legitimacy is a feature of an equilibrium in which (a) subjects’ favorable beliefs about an authority enhance their intrinsic motivations to comply with its directives and (b) the authority’s actions are consistent with those beliefs.

Before explicating this notion of legitimacy and the attendant problems for isolating its causal effects, we offer a caveat: Any behavioral model that aims at prediction rather than prescription must depart from the normative language of absolute (or near-absolute) rights and obligations and consider the concepts outlined above in terms of marginality and conditionality. Rather than speak of a general obligation to comply with an authority (or absolute autonomy from such an obligation), we will speak instead of legitimation “on the margin”: conditions in which the enhanced intrinsic motivations associated with a legitimate authority incrementally affect the likelihood of compliance, other things being equal. Given this treatment, the legitimacy of the authority is just one of many factors affecting compliance. Others include extrinsic reasons (the apprehension of punishment or reward), informal norms in the polity, or altruism, a perspective found in some normative work (e.g., Beetham). Given individual-level heterogeneity in other non-legitimacy-related beliefs that affect compliance (or probabilistic responsiveness to legitimacy), it is therefore also

334

Sanford C. Gordon and Gregory A. Huber

likely that the behavioral consequences of an authority’s legitimacy will vary across individuals (or over time). This approach, which makes the empirical study of legitimacy feasible, is also, of course, what makes the empirical identification of legitimacy so vexing. Extant Empirical Research We examined empirical research on legitimacy in six major political science journals (American Political Science Review, American Journal of Political Science, Journal of Politics, Comparative Political Studies, International Organization, and World Politics) for the period 1970–2016.14 These papers can be divided roughly into three sets: research on public support for courts; research on regime legitimacy; and research on the legitimacy of supranational authorities. Before reviewing this research, we briefly examine its roots in the seminal work of Max Weber and David Easton. Weberian and Eastonian Roots. The literature referencing legitimate authority as an empirically distinct and important phenomenon is voluminous. From a political science perspective, its roots can be traced to two critical sources. Most familiar is the treatment of legitimate forms of domination by Weber, who defines legitimacy as a belief in, or orientation to, the validity of a social order.15 For Weber, the idea of legitimacy is itself closely connected to the apparent need of an authority to justify itself. Critically, for Weber the belief in the legitimacy of an authority motivates compliance with authority, over and above other motivations, among which Weber includes custom, personal advantage/material motives, and “purely affectual or ideal motives of solidarity.”16 And empirically, evidence of the legitimacy of a system is manifested in “the probability that to a relevant degree the appropriate attitudes will exist and the corresponding practical conduct ensue.”17 This Weberian notion of legitimacy is expanded upon in Easton’s seminal Systems Analysis of Political Life.18 Like Weber, Easton ties the notion of legitimacy to compliance rooted in the beliefs of citizens about the normative appropriateness of a regime’s actions: “the most stable support [for a regime] will derive from the conviction on the part of the member that it is right and proper for him to accept and obey the authorities and to abide by the requirements of the regime.”19 There are two points

The Empirical Study of Legitimate Authority

335

of elaboration that are critical for understanding the importance of Easton’s contribution to subsequent political science research on legitimacy. The first is that Easton views legitimacy as a potential determinant of a broader concept that he terms diffuse support, defined as “a reservoir of favorable attitudes or goodwill that helps members to accept or tolerate outputs to which they are opposed or the effects of which they see as damaging to their wants.”20 The second point is that, although Easton views belief in the legitimacy of a regime as a “major source of diffuse support in most systems we know about,” he is quite clear in stating that systems can enjoy diffuse support even in the absence of legitimacy. This last point is critical because it suggests that subjects’ beliefs in the legitimacy of a regime are not a necessary condition for diffuse support of that regime.21 Accepting Easton’s premise, then, we cannot infer the existence of subjects’ beliefs in the legitimacy of an institution or authority from evidence of diffuse support. Put bluntly, there are many reasons apart from legitimacy that individuals support existing political arrangements. Inadequately accounting for these other reasons is a key deficiency in the extant political science literature on the subject. On the Legitimacy of Courts. In Federalist No. 78, Hamilton famously argued that the federal courts would have “neither force nor will, but merely judgment.” It is unsurprising, therefore, that scholars have suggested that a widespread belief in the legitimacy of the courts can explain how unelected judges maintain support for themselves and the institution of judicial review more generally despite their potentially unpopular decisions and the absence of a direct accountability linkage with the public. Drawing on earlier work by Caldeira, Caldeira and Gibson construct a battery of survey questions designed to measure diffuse support.22 The battery includes questions about the respondent’s attitudes toward curtailing the power of the Supreme Court or even abolishing the institution outright. Gibson, Caldeira, and Spence operationalize “institutional loyalty” toward the Supreme Court explicitly to capture Easton’s “reservoir of goodwill” account of diffuse support.23 In addition to questions concerning the desirability of limiting the Court’s power, the survey battery they employ includes questions about trust in the Court, its propensity to favor certain groups over others, and its involvement in politics. In this

336

Sanford C. Gordon and Gregory A. Huber

paper, the authors elide Easton’s distinction between diffuse support and legitimacy, even stating that “Easton and many others use ‘diffuse support’ as a synonym for legitimacy.”24 With survey-based measures of support, scholars have also examined how support for courts is shaped by individual factors (e.g., personal ideology and political awareness). Gibson and Caldeira, for example, examine the antecedents of support for the US Supreme Court and the European Court of Justice.25 Other research considers how specific court decisions and their alignment with respondent ideology, along with justifications offered for those decisions, affect support for courts.26 A related literature, building on the idea that diffuse support can shield the courts from the consequences of unpopular decisions, examines the importance of a court’s popularity in explaining patterns of conflict between it and legislatures or other bodies.27 This literature therefore similarly equates support for an institution with its legitimacy. On the Legitimacy of Regimes Generally. A second, related area of research concerns the concept of regime legitimacy. This work, like its counterpart in the courts literature, frequently draws on the Weber/Easton understanding of the bases of support for a regime. An early example of this is found in work by Gurr and McClelland, who define legitimacy as “the extent that a polity is regarded by its members as worthy of support.”28 Research in this tradition tends to compare regime support (1) across political systems, (2) over time, and (3) across individuals (within regimes). Cross-country studies have found that higher levels of income inequality are associated with reduced trust and lower evaluations of government performance;29 that the legitimacy of communist rule in Eastern Europe was based on economic performance, but also that deference arose because of the absence of credible alternatives to the communist regime (arguably a concept independent of legitimacy);30 and that challenges to Soviet and Sovietproxy rule were more frequent in countries where pre-communist education systems inculcated nationalist conceptions of the right to govern that undercut the legitimacy of communist claims about the basis of authority.31 Studies exploiting over-time variation have found that elections followed by changes in control of government enhance legitimacy among those previously excluded from

The Empirical Study of Legitimate Authority

337

power.32 Others have found that legal reforms in China increased trust in the (national) government;33 and that the determinants of regime support evolved in post-Franco Spain.34 Scholarship focusing on individual-level variation has shown that individual-level reported corruption experience is associated with reduced diffuse support in four Latin American countries;35 that trust in government in the Dominican Republic is correlated with perceptions of economic and political performance rather than democratic values or civic engagement;36 that support for the post-apartheid democracy in South Africa is strongly correlated with race;37 and that among those found guilty of a felony, those who judge the procedures to be fair exhibit more support for law and the broader political system after accounting for their assigned sentence.38 While the research described above focuses mainly on attitudes, one important strand of work focusing on individual-level differences concerns behavior: specifically, which individuals support a regime and comply with its edicts. Mueller and Jukam, for example, show that general dissatisfaction with a governance system (trust, performance, etc., rather than just its particular elected leaders) is correlated with reported anti-system behavior.39 And Giles and Gatlin show that individuals are more likely to exit racially desegregated schools for private schools, a behavior the authors classify as defiance, if they perceive court intervention to mitigate school segregation as less legitimate, which is operationalized by items assessing beliefs about desegregation moving too quickly and viewing it as undesirable to have courts involved in desegregation policy.40 A more recent strand of related research takes an experimental approach: In a laboratory setting, Grossman and Baldassari randomly assign an “authority” who has the ability to punish participants in a public goods game to be either selected at random or by voting from among participants. They find that the elected authority is associated with more frequent contributions and interpret this effect as arising due to legitimacy after showing that elected and appointed authorities do not behave differently.41 Olken reports results from a field experiment in Indonesia showing that when local citizens can vote on development projects, they are more knowledgeable, perceive larger benefits from those projects, and are more satisfied than when citizens do not have

338

Sanford C. Gordon and Gregory A. Huber

the opportunity to vote on selected projects.42 Further evidence shows relatively minor effects on which projects are chosen, which Olken argues that the process of democratic project selection may enhance the legitimacy of those projects. Legitimacy of International Organizations. The final area of concern for political scientists seeking to document legitimacy perceptions and their effects is that of supranational organizations and institutions.43 This literature is largely qualitative in nature: Grant and Keohane, for example, discuss ways in which supranational authorities, which lack direct ties to those they govern, are accountable and therefore obtain elements of legitimacy.44 Barnett discusses the way in which international political actors, and in particular the UN, gain legitimacy, acknowledging the difficultly of distinguishing the strategic logic of supporting the UN from accounts based in the normative legitimacy of the body.45 Finnemore considers the difficulty a unipole faces in becoming legitimate—that is, able to garner support from others absent threats of force, and suggests obtaining legitimacy often requires reducing one’s short-term power.46 Borzel et al. conduct a quantitative analysis of EU member states’ violations of European law as a function of legitimacy, which they operationalize as citizen support for the rule of law and public support for the EU. They find these factors do not predict violations, while state capacity and material interests do.47 Observations Viewed through the lens of our analysis of the normative-theoretic literature on legitimate authority, our brief tour through the recent journal-based empirical literature suggests three preliminary observations. The first of these concerns the behavioral relevance of diffuse support for, or beliefs in the legitimacy of, specific institutions. Take, for example, the extensive literature on public support for national high courts. In environments where the legal system is largely enshrined, it is generally costless to the individual to criticize or defend it. Moreover, for the typical survey respondent, there is little chance that the absence of support for an institution entails a significant opportunity or motive to disobey it. Thus, the

The Empirical Study of Legitimate Authority

339

behavioral implications of an absence of legitimacy perceptions may, in many cases, be either unclear or miniscule. Indeed, research by Caldeira and Gibson on South Africa finds that despite apparently low legitimacy, individuals are willing to defer to legal institutions even in that relatively nascent political system, suggesting that (low) diffuse support need not have substantial behavioral consequences. Below, we discuss the importance for research designs investigating legitimacy to consider cases in which citizens are closer to indifferent between compliance and non-compliance, and where, as a consequence, the behavioral implications of changes or differences in legitimacy perceptions are more likely to arise. Second, even in situations where the behavioral implications of support for or opposition to institutions are clearer, a more pervasive issue of construct validity comes into focus. As alluded to above, support for a regime or institution is not synonymous with beliefs in the justness or rightness of the regime, its occupants, or its actions. Citizens may (diffusely) support a regime for many reasons: because of its past performance, custom and norms, convenience, or immediate self-interest.48 Past performance seems particularly salient as an antecedent of Easton’s notion of a reservoir of goodwill: Why abandon your support for an institution that regularly produces desirable results when, on occasion, it generates an undesirable one? In this regard, batteries of survey questions that are interpreted as measures of perceptions of legitimacy often implicate attitudes about trust in and satisfaction with an institution, for which a respondent, when answering, might easily consult his or her attitudes about a regime’s performance rather than its normative appropriateness. Indeed, part of the reason that legitimacy has been dismissed as a useful theoretical concept by some scholars is the multiplicity of other, performance-related factors that give rise to support.49 Evidence of trust in institutions (by which we mean confidence in their capabilities, decisions, and directives) and norms of compliance with those institutions (by which we mean the tendency of individuals to comply because they perceive doing so as standard, proper, or typical) can also arise from numerous sources aside from perceptions of legitimacy. Citizens may decline to resist an authoritarian regime that delivers substantial GDP growth and threatens to torture

340

Sanford C. Gordon and Gregory A. Huber

subversives: In such cases, a fear of material punishment coupled with possible trust in the regime’s ability to deliver economically (or mistrust of the opposition) may confound our efforts to assess perceptions of its legitimacy. A further reason to express caution about the legitimacy-revealing nature of measures of diffuse support is that norms and trust can arise through patterns of equilibrium selection in repeat play. For example, individuals may coordinate on a set of behaviors and beliefs that produce desirable outcomes and involve widespread compliance and regime support. In and of itself, this provides no evidence of a belief in the regime’s legitimacy. A final, closely related set of concerns, which we develop more fully in a broader discussion in the next section, is about the issues of (1) causal inference and (2) attribution. To begin with, as a matter of inference, a key challenge to persuasive causal identification is coming up with a source of variation (in legitimacy) that is unrelated to other causes of the outcome or the outcome itself. Next we take up this attribution problem, highlighting how changes in legitimacy often implicate other motivations for compliance. The causal inference issue is particularly thorny in observational analysis because it raises the essential question about what causes differences in perceptions of legitimacy. For example, if one group of individuals perceive their treatment by the courts as less legitimate and report they are less likely to comply with the law, what explains this individual-level variation in perceptions and reported behavior? Suppose people who are inclined to disobey (for reasons unrelated to legitimacy) justify their behavior by reporting low legitimacy, or more perniciously, suppose some (omitted from a statistical model) third factor explains both legitimacy perceptions and state support. It is precisely these concerns that have led many scholars to embrace randomized experiments as a tool for persuasive causal identification, because by design they break the correlation between all characteristics of subjects (and therefore outcomes) and their treatment status. At the same time, it is important to note that most extant experiments rely on survey measures of outcomes, which itself returns us to the points raised above about the behavioral relevance and meaning of those responses. Moreover, suppose we were to empirically detect a relationship between some feature of an institution generally perceived as just or appropriate and either expressed attitudes toward the

The Empirical Study of Legitimate Authority

341

institution or behavior supportive of it. Could we conclude that the institutional feature under study increased support or compliance by enhancing legitimacy? Even setting aside the issue of nonrandom assignment of treatment, it is not clear that this is feasible absent very strong assumptions. This is because variation in the institutional feature that affects perceptions of its legitimacy may simultaneously affect the performance of, or trust in, the institution.50 In the next section, we consider this attribution problem in greater detail and show that it can arise for a number of different reasons. We then provide guidelines that should aid researchers working in the field and hoping to overcome it. A Positive Model of Legitimate Authority Baseline: The State of Nature To clarify intuition, we present a very simple model of social interaction. Imagine a stateless polity consisting of two citizens, 1 and 2. Each citizen has an endowment normalized to one, and both simultaneously choose to contribute the endowment to a common pot or keep it for themselves. For each contribution to the common pot, both citizens receive a return of ρ, which lies somewhere between 1/2 and 1. Absent further elaboration, this structure corresponds to a standard prisoners’ dilemma (PD). We modify the PD by adding idiosyncratic shocks to the payoff from contributing for each citizen, labeled w1 and w2, which are drawn from a commonly known probability distribution. The realization of each shock wi is the private information of citizen i. The payoff matrix, then, is as shown in table 11.1. Note that the modified game is only a PD for certain draws of w1 and w2; otherwise, any of the four possible outcomes could be an equilibrium. Table 11.1. Baseline Game: A Randomly Perturbed Prisoners’ Dilemma Citizen 2

Citizen 1

Contribute

Keep

Contribute

2ρ + w1 , 2ρ + w2

ρ + w1 , 1 + ρ

Keep

1 + ρ , ρ + w2

1,1

Payoffs are listed for Citizen 1 and then Citizen 2.

342

Sanford C. Gordon and Gregory A. Huber

Enforcement Complementarity Suppose now that there is an authority with some resources to punish citizens who are caught keeping their endowment. For this section, we treat the authority as a non-strategic actor—in particular, assume that the authority endeavors to punish a citizen if and only if she observes that citizen to have kept. The authority has two relevant characteristics: capacity and legitimacy. Capacity represents the resources available to the authority to implement punishment. For example, if only one citizen keeps his endowment, that citizen is punished with probability c, but if both keep, each is punished with some probability cʹ < c. Taking our cue from the normative and social psychological literatures, we model legitimacy as an enhancement to the intrinsic benefit to a citizen of contributing, labeled λi for each citizen i. For now, we keep legitimacy exogenous, allow it to vary across the two citizens, and assume that each citizen’s λ is common knowledge.51 Three insights emerge after solving for the equilibrium strategies of the two citizens (which can be described by cutpoints in w), three insights emerge. First, citizens’ choices are strategic complements: that is, the more likely citizen i is to contribute, the higher the expected sanction to j from not contributing, and thus, the stronger the incentive of j to contribute. Interestingly, complementarity is enhanced by an aspect of the authority’s weakness: the extent to which widespread noncompliance taxes their ability to maintain order. An increase in the state’s ability to reliably punish all detected non-compliers reduces the extent of the complementarity.52 Even in the absence of legitimacy concerns, the complementarities between the citizens’ strategic choices implies the potential existence of multiple equilibria: If Citizen 1 (2) believes that Citizen 2 (1) will contribute, then she will also believe that the likelihood of punishment for keeping is relatively high, which may drive her to contribute (a “good” equilibrium). But if Citizen 1 (2) believes that Citizen 2 (1) will not contribute, then she will also believe that the likelihood of punishment for keeping is lower, which may drive her to keep (a “bad” equilibrium). The potential existence of multiple equilibria means that complementarities introduce a coordinative aspect into the baseline compliance game.

The Empirical Study of Legitimate Authority

343

What, then, is the effect of the authority’s legitimacy on the citizens’ choices? Most obviously, citizen i is more likely to contribute if he or she views the authority as legitimate (i.e., the probability that i contributes is increasing in λi). More interestingly, though, citizen i is more likely to contribute if the other citizen views the authority as legitimate (i.e., the probability that i contributes is also increasing in citizen j’s legitimacy perception jλ). This is owing to the strategic complementarity between the citizens discussed in the previous paragraph: High λj increases the likelihood that j contributes, which increases the expected sanction to i from keeping, which increases the likelihood that i contributes. Even if i does not perceive the authority as legitimate at all, the legitimacy of the authority from j’s perspective still enhances i’s compliance for purely extrinsic reasons. More generally, the authority’s legitimacy in the eyes of either citizen may dramatically reduce the range of parameter values supporting the “bad” equilibrium. These effects are roughly equivalent to what Levi terms “quasivoluntary compliance.”53 As an aside, note that legitimate authority is a sufficient but not necessary attribute of Levi’s notion of quasivoluntary compliance. For example, suppose the distributions from which the wi’s are drawn differ by citizen. If, for example, j is known to be more altruistic on average, this fact will also increase the probability that i contributes, and also be a form of quasivoluntary compliance. More generally, spillovers from one citizen to another can arise due to both extrinsic and intrinsic (e.g., legitimacy) motivations for the other citizen to comply. The issue of strategic complementarity implies a violation of the Stable Unit Treatment Value Assumption (SUTVA) that underlies estimation of average treatment effects. Consider a plausibly exogenous increase to legitimacy of the authority in the eyes of one set of citizens that, via strategic complementarity, affects the behavior of another set of citizens who do not recognize that increase. A researcher who ignored this complementarity and compared the behavior of the citizen whose legitimacy perceptions changed to the unaffected citizens would underestimate the effect of the change in legitimacy, because it had behavioral effects on both groups. Likewise, a researcher documenting an association between the shock and a widespread increase in compliance (or other form of cooperation with the institution) would be estimating the total

344

Sanford C. Gordon and Gregory A. Huber

effect of the legitimating shock. The total effect would encompass both the direct legitimating effect on the first group of citizens and the purely material effect emerging through the complementarity mechanism. Absent additional assumptions, we would not be in a position to apportion causation across the two mechanisms. Note that an analyst can overcome the SUTVA issue either via research design (e.g., by changing the unit of analysis) or statistically (e.g., via clustering) to estimate a valid estimate of the total causal effect. This total effect, while arising due to a change in legitimacy perceptions, does not isolate individual-level changes in legitimacy perceptions (i.e., document the psychological mechanism) necessary to show that any behavioral change arises due to that individual’s personal legitimacy perceptions. To summarize, when authority capacity is limited, strategic complementarities mean that variation in the legitimacy and other motivations for compliance of one citizen affect the incentives of other citizens to comply. These complementarities make isolating which actors’ perceptions of legitimacy affect their (versus others’) behavior difficult. As such, situations in which such complementarities are minimized will make it easier to isolate the psychological mechanism underlying legitimate authority. Institutional and Personal Legitimacy We next consider a broader class of inference problems that may emerge even in the absence of strategic complementarities between the citizens.54 To do so, we consider what might occur in situations where an authority can take some (possibly costly) action to enhance the quality of a governing institution. Before specifying what we mean by quality, consider that there are at least three channels through which this enhancement to quality might cause changes in the behavior of citizens. 1. The quality of the institution may affect the citizens’ extrinsic incentives by changing the mapping between citizen action and expected punishment and/or rewards. 2a. The quality of the institution may engender greater affinity toward it, and hence toward the act of compliance itself. We refer to this as the institutional legitimacy channel.

The Empirical Study of Legitimate Authority

345

2b. The authority’s action can affect the citizens’ beliefs about the authority’s type (i.e., worthiness), which may in turn affect the intrinsic benefit of compliance with the authority. We refer to this as the personal legitimacy channel. There are many ways to operationalize quality, but taking our cue from the extant literature, we proceed from the premise that the perceived fairness of the institution plays a role in enhancing its legitimacy in the eyes of citizens.55 We follow Leventhal, who draws a useful distinction between distributive and procedural fairness, and focus for the moment on the latter, as it highlights the identification issues most clearly.56 One critical aspect of procedural fairness is accuracy: the extent to which an institution accurately assigns sanction to non-compliers while exculpating compliers. We can permute our randomly perturbed PD from above as follows. First, we shut down the complementarities described in the preceding section by making the probability of punishment independent of the number of contributors. This will have the effect of making the relevant strategic interaction the one between the authority and the individual citizen, rather than the one between the citizens. Second, we suppose the authority is a strategic player. Prior to the contribution stage, the authority chooses an observable level of effort e, which affects the accuracy/quality of the institution q. We assume that accuracy is strictly increasing in effort. Authorities are characterized by a type, ß, which scales the extent to which she cares about procedural fairness primitively, that is, independent of its effects on contributions. Letting Y denote the number of contributions to the common pot, the authority seeks to maximize ρY + ßq − e. Finally, we modify the citizen’s intrinsic benefit by assuming that it is responsive to three factors: the idiosyncratic term wi; institutional legitimacy from the perspective of i, λi(q); and personal legitimacy, captured by the citizen’s posterior belief concerning the extent to which the authority cares primitively about fairness, ß̂ . Citizens value the extrinsic and intrinsic benefits from contributing, as well as the extrinsic cost of being punished (which, given inaccurate institutions, may happen even if the citizen contributes). We now move to a consideration of what makes identifying personal or institutional legitimacy separately from extrinsic incentives so challenging.

346

Sanford C. Gordon and Gregory A. Huber

Figure 11.1. Graphical Representation of Identification Issues in the Study of Institutional and Personal Legitimacy. Oval nodes represent observables, and rectangular nodes, unobservables.

Identifying Institutional Legitimacy. Holding constant the authority’s effort level (and therefore the citizen’s beliefs about the leader’s personal legitimacy), consider the effect of a change in institutional quality on the behavior of the citizen. There are two channels through which a change in institutional quality can affect the probability the citizen contributes. The first is the extrinsic channel: By improving the mapping between anti-social behavior and sanctioning, an increase in quality corresponds to an increase in marginal deterrence. The second is the intrinsic, institutional legitimacy channel: An increase in enforcement accuracy enhances the citizen’s positive affect toward the institution and thus the tendency toward compliance. These two channels are captured in the portion of figure 11.1 in the inner dashed rectangle labeled “Institutional.” The multiplicity of channels of influence complicates efforts to isolate the legitimating effect of an increase in institutional quality empirically. Note that randomization of the institutional environment does not alleviate the problem. While it would permit a valid causal estimate of the total effect of institutional quality on citizen behavior, it would not permit the analyst to isolate the legitimating effect of institutional quality. This observation, of course, does not imply that the legitimacy channel is inoperative; indeed, a standard observation concerning compliance with legal norms is that citizens often comply even in the absence of strong material incentives to do so. That said,

The Empirical Study of Legitimate Authority

347

any observed association between institutional quality and citizen behavior cannot be attributed solely to the legitimacy or extrinsic motivation channels absent additional assumptions. What is the overall takeaway from these considerations? To estimate the legitimating effects of institutions, it is advisable to consider institutions that have minimal effects on the material incentives of individuals to comply. In many cases, this can present a challenge for the researcher. For example, consider the recommendation of the community policing movement that police legitimacy will be improved if police officers are residents of the neighborhoods they patrol. The residence of the officers may seem materially irrelevant. But police officers who live in proximity to their beats are more likely to possess dense knowledge about the residents of those neighborhoods. This may affect the accuracy with which they assign sanctions, and thus the material incentives of citizens. Personal Legitimacy. To examine the personal legitimacy effect, note that in a separating equilibrium to the game, an authority with higher ß exerts greater effort, and citizens infer the authority’s type from the observed effort level and act accordingly. Can we identify the legitimating effect of the authority’s effort choice from the observed relationship between the authority’s action and the citizen’s response? Not in the current context, as it turns out. The causal effect of the authority’s effort choice propagates through three distinct channels. The first is the personal legitimacy channel: By informing the citizen of the authority’s type, the authority’s effort affects the intrinsic motivations of the citizen, which in turn affects behavior. Second, the authority’s effort affects the character of the institution, which in turn affects the extrinsic motivations of the citizen. Finally, the effect of the authority on the institution also affects the intrinsic motivations of the citizen through the institutional legitimacy channel. Figure 11.1 depicts these channels graphically. Clearly, in light of this complex casual pathway, one cannot simply regress the citizen’s behavior on the authority’s action and claim to isolate a legitimacy effect. And controlling for the institution is not feasible, as it is, in this model, deterministically tied to the authority’s effort choice. It would seem, then, that identification requires severing the deterministic linkage between the

348

Sanford C. Gordon and Gregory A. Huber

authority’s choice and the institution. This, however, introduces two additional subtleties in light of the posited causal mechanism. First, the researcher might be tempted to randomize the authority’s effort choice, or find an instrumental variable that affects effort but neither institutional quality nor citizen behavior. The challenge here is that if effort is known to be randomly or quasirandomly assigned, the citizen should draw no inferences about the authority’s type from the observed level of effort, just the same way that a coerced confession is uninformative with respect to actual guilt or innocence. Second, suppose the researcher decides instead to randomize the mediating variable q (institutional quality). A necessary and sufficient condition for a separating equilibrium (in which higher types take costlier actions) is that type and effort are complements. This is true by construction above, given the relationship between quality/accuracy and the authority’s effort level. If accuracy is completely randomized, however, then the separating equilibrium will collapse: Higher types will gain nothing from their effort, and all types will choose no effort. In this situation, it is no longer feasible to detect personal legitimacy, and further, if accuracy shapes both material incentives and intrinsic motivations (via the institutional legitimacy channel), then those effects remain intertwined as we discussed above. Dickson, Gordon, and Huber describe sufficient conditions for identifying personal legitimacy effects given these issues.57 Essentially, what is required is that the authority’s level of effort induces overlapping lotteries over levels of accuracy that are orderable by stochastic dominance. Under these conditions, the separating equilibrium will be preserved, and the personal legitimacy effect can be identified from the relationship between the authority’s and citizen’s behavior, controlling for the realized level of quality.58 That paper instantiates this strategy in a laboratory experiment as follows: If the authority takes a costly action, with 50 percent probability the accuracy is “high” and with 50 percent probability it is medium; if the authority foregoes the costly action, with 50 percent probability the accuracy is “medium” and with 50 percent probability it is low. In this design, the high type still has an incentive to take the costly action (thereby revealing her type), but one can obtain medium accuracy regardless of whether the costly

The Empirical Study of Legitimate Authority

349

action was taken. The paper then compares citizen behavior when the authority took the action with that behavior when she did not, restricting attention to cases in which the realized accuracy level was medium. In summary, identifying a causal relationship between the action of an authority and response of a citizen that can accurately be described as “legitimacy” or “legitimation” is difficult owing to the relationship between the authority’s legitimating action and the institutional environment of the subject, which activates extrinsic motivations as well as institutional legitimacy considerations. This is a particular kind of causal mediation problem, but for the reasons detailed above, the answer is not as simple as “randomize the mediator.” Looking for situations in which the institutional environment is related probabilistically to the authority’s actions is one way to identify the personal legitimacy effect. Another would be to look for a valid instrumental variable that, along with the authority’s action, affected the institutional environment without separately affecting the authority’s effort choice or citizen behavior. While each strategy provides purchase to identify the personal legitimacy effect, these strategies do not seem feasible in the realm of institutional legitimacy (unless the institution itself has no effect on material incentives). Other Considerations The Case for Indifference An important research design challenge is identifying cases in which it is possible to ascertain whether a theoretically relevant construct operates as hypothesized. Note that this is analytically distinct from asking whether a concept is relevant or operative in a given domain, but is instead about whether one could detect the importance of a given concept in a given situation. Put differently, we wish to look for instances in which a marginal change in some theoretically relevant concept would produce an observable behavioral effect. The general question of “where to look” is particularly important in the study of legitimacy because, as we articulated above, a belief in legitimacy is just one among many reasons individuals may comply with an authority or institution. For example, even in

350

Sanford C. Gordon and Gregory A. Huber

the version of the model described above that assumes away complementarities, an individual’s decision to comply is characterized by a cutpoint w*, and compliance occurs if and only if the sum of idiosyncratic individual-level motivations to comply, material/ deterrence motivations, and legitimacy effects exceed this cutpoint. Explicitly modeling compliance in this way makes clear the importance of choosing the right cases for study. Suppose for the moment that the legitimating effect of an institution or authority choice could be isolated from material compliance motivations using one of the strategies described above. Then the average treatment effect of an incremental change in the institution will be proportional to the height of the shock density around the cutpoint. If most citizens are already inclined to comply (w* in the left tail of the distribution) or not comply (w* in the right tail), then even those reforms that yield a large change in the cutpoint may have only negligible behavioral consequences.59 What does this mean for research design? We note that many cases where legitimacy is studied are precisely those where compliance seems unlikely to change given posited changes in legitimacy. Take, for example, the question of how Americans respond to Supreme Court decisions they perceive as illegitimate or to legal verdicts they perceive as unfair. While prior work provides extensive survey evidence about perceptions of these outcomes, evidence about their behavioral consequences is scant because nearly all individuals will comply with those decisions regardless of marginal perturbations to their legitimacy perceptions for other reasons. In parallel fashion, in regimes where individuals routinely ignore legal verdicts, even substantial increases in legitimacy are unlikely to manifest in detectable behavior changes. Would this mean legitimacy is inconsequential? No, because the lack of a behavioral consequence may arise precisely because individuals are far from indifferent with regard to compliance. Of course, cumulative and large changes in legitimacy may matter for things like Supreme Court decisions, but as a matter of objects of study, the local changes in legitimacy induced by most interventions for these sorts of policies are unlikely to be those with substantial behavioral consequences. This means we may wish to look elsewhere to understand how much and for whom different policies induce behaviorally relevant changes in legitimacy.

The Empirical Study of Legitimate Authority

351

By contrast, in cases where more individuals are close to the compliance threshold, it is much more likely that legitimacyenhancing or -degrading policy changes or choices will be consequential, and studying those circumstances will help us to instantiate the behavioral importance of legitimacy (subject to the identification concerns detailed above). Thus, the key takeaway point from this analysis is that if one wants to provide behavioral evidence about legitimacy, one of two conditions should hold. First, one should look for situations in which, either as a product of researcher intervention or experimental design, individuals are close to indifferent with regard to compliance. Second, and alternatively, one should focus study on those who are otherwise likely to be close to indifferent with regard to compliance, even if this is only a subset of the larger population affected by a regime. Prior research on regime legitimacy provides suggestive evidence of the fruitfulness of this targeted approach. Among West German college students, it was leftists whose perceptions of the right-leaning German regime’s legitimacy that were correlated with self-reports of anti-regime activity.60 By contrast, for rightists, for whom the regime’s legitimacy was arguably secondary to its ideological attractiveness, no such effects were observed. The Case for Hard Cases We have, at this point, identified several research strategies for grappling with the deep entanglement between material and legitimacy-based reasons for compliance with an authority. But a deeper understanding of how this entanglement operates in different environments points to another approach to identification: the examination of “hard cases” for legitimacy. These are instances in which, unlike in the cases described above, institutions, policies, and choices that potentially enhance legitimacy diminish material compliance motivations. In such situations, the detection of legitimacy effects will be in spite of, rather than owing to, the extrinsic motivation channel. Or, to use the language of statistics, estimated legitimacy effects will be downward- rather than upward-biased. A prominent example of such a case is group- or neighborhoodbased profiling in law enforcement: for example, hot spot policing, racial profiling in motor vehicle and pedestrian stops, and

352

Sanford C. Gordon and Gregory A. Huber

ethno-religious profiling in airport security screening. Suppose a local police department allocates more intense scrutiny to a neighborhood where it anticipates more crime. From the perspective of the police, this may be efficient: The marginal deterrent effect of this reallocation should improve compliance with the law (or detection of law-breakers). A law-abiding citizen in the heavily scrutinized neighborhood may view the allocation as illegitimate, however: From her perspective, it may be motivated by bias rather than efficiency. (We set aside for the moment the possibility of mistreatment by the police.) Suppose a law enforcement agency moved toward a more equitable allocation of resources, and we simultaneously observed an increase in compliance in the previously targeted neighborhood. The increase in compliance would not, in this circumstance, be attributable to enhanced material incentives to comply, but rather, to enhanced perceptions of the fairness of policing (increased legitimacy) net of diminished deterrence.61 Conclusion Legitimacy is a foundational concept in both normative political theory and positive political science and is widely seen as being an important explanation for observed patterns of citizen behavior and institutional support. Indeed, the frequency with which the concept of legitimacy is invoked by academics and policymakers alike as an explanation for the success or failure of policy reveals our prior beliefs about its importance. Despite the frequency with which it is invoked, however, there remains substantial theoretical and empirical imprecision about its meaning, which may account in part for the fact that some scholars are highly skeptical of its practical importance. In this chapter, we have discussed the normative underpinnings of the concept of legitimacy and linked those theoretical concepts to the problem of conducting empirical research that provides informative evidence about its importance. This review highlights several core issues for empirical research that invokes legitimacy, including problems of conceptual clarity, measurement, and causal identification. To more fully elucidate these claims, we present a simple model that helps us to understand the causal and theoretical challenges facing empirical researchers.

The Empirical Study of Legitimate Authority

353

In this conclusion, we briefly recount what we see as the key points we have identified based on this review and their implications for future work: 1. The concept of legitimacy, as understood in normative work and foundational empirical work, is distinct from extrinsic (material) motivations for consent. Legitimacy should be understood as affecting the intrinsic motivations of citizens to comply because they care about the values an institution or authority embraces or instantiates. 2. In equilibrium, legitimacy obtains when citizens hold beliefs about an authority that enhance these intrinsic motivations to comply, and authorities take (justifiable) actions that are consistent with those beliefs. Legitimacy therefore has to be understood in terms of the interaction between what authorities and citizens do. 3. In moving from normative absolutes to empirical prediction, we must recognize that legitimacy is one motivation among many others for differences in support for an authority or state. Legitimacy is therefore likely to matter on the margin in generating compliance. a. It is important not to conflate evidence for diffuse support with evidence for legitimacy. Individuals may value an authority or institution over feasible alternatives for a variety of reasons, but diffuse support (and proxy measures of trust, confidence, and support for political institutions or authorities) is not itself evidence that such diffuse support arises due to legitimacy. b. If legitimacy is to matter decisively in explaining behavior, it will only be in cases where, ceteris paribus, individuals are otherwise close to indifferent about their decision to comply. Research that invokes legitimacy as causally important must therefore anticipate not only that legitimacy considerations are operative, but also that perturbations to those considerations will have behaviorally relevant implications. 4. Because there are many different motivations for compliance, empirical work must grapple with the fact that institutions, choices, and actions by an authority may simulta-

354

Sanford C. Gordon and Gregory A. Huber neously affect the intrinsic and extrinsic motivations for compliance. This has three key implications for existing work. a. First, at one extreme, if decisions by a governing authority enhance both material and non-material motivations for compliance (thereby implicating material incentives, institutional legitimacy, and potentially an authority’s personal legitimacy), then empirical research will not be able to isolate the causal effect of legitimacy absent strong assumptions. This problem is likely to be exacerbated in situations where state capacity is limited, because effects on one citizen’s willingness to comply (whether due to legitimacy or other factors) can indirectly affect other citizens’ incentives to comply. Only in cases where legitimacy-enhancing institutions have no effect on the other (material) motivations to comply is it possible to isolate institutional legitimacy effects, and only in situations where strategic complementarities among citizens are minimal can an individual citizen’s responses to a posited institutional change be isolated as arising due to its effect on perceptions of legitimacy. For researchers beginning empirical work, a primary question of research design is therefore to assess whether factors thought to affect legitimacy have any effect on extrinsic motivations for behavior. b. Second, if factors that improve legitimacy also affect extrinsic rewards, it is precisely hard cases—those in which enhanced legitimacy and improved material incentives have opposite effects on behavior—that provide the most promising opportunities for understanding the consequences of improved institutional legitimacy. This is because it is these situations that estimates of institutional legitimacy effects will be downwardly biased, and so finding evidence for legitimacy in these situations would provide persuasive evidence not only about the existence of, but also the empirical importance of, enhanced legitimacy. These are also the key “real-world” situations in which arguments about

The Empirical Study of Legitimate Authority

355

the trade-offs between legitimacy and material incentives are the most pressing for public policy. c. Third, simple randomization of an institution cannot solve this problem of attribution. This is true first because even randomized institutions alter both intrinsic (institutional legitimacy) and extrinsic motivations for behavior. It is also true because if it is an authority’s choices that communicate her concern for values the citizen cares about—thereby enhancing her legitimacy—simple randomization makes her choice meaningless or undercuts the reasons for that action altogether. We specify the conditions under which an authority’s choice to pursue a superior institution can, through conditional randomization with overlapping support in realized institutions, create situations where estimating a personal legitimacy effect is not conflated by material concerns (or institutional legitimacy). Stepping back somewhat, our view is that legitimacy remains an important area for political science research. With renewed attention to a conceptually delimited and rigorously defined definition of legitimacy, many of the complaints about the theoretical imprecisions surrounding claims of legitimacy may be addressed. At the same time, this more circumscribed understanding of legitimacy itself reveals the thorny empirical challenges that make isolating legitimacy so hard. All of this begs the question of “why bother?” That some reform “works,” whether for legitimacy or other reasons, may be all that we care about. We close with two observations. First, as academic researchers, we believe one of our key roles is to isolate theoretical mechanisms. Prediction without understanding, particularly given the frequency with which legitimacy is invoked causally, risks misunderstanding. This misunderstanding may itself be undesirable, but if it in turn leads to bad policy advice, the consequences may also be deleterious from a public welfare perspective. Nor is this problem limited to empirical research. Many important normative arguments draw on inferences and intuitions we make about “what causes what,” and as such, making sure we understand whether

356

Sanford C. Gordon and Gregory A. Huber

legitimacy is operative helps us understand whether individuals are in fact motivated by these concerns. Second, there are many situations in which policymakers confront choices about which (costly) reform to implement. While it is often the case that enhanced legitimacy and improved material motivations for compliance move hand in hand, they do not always do so, and sometimes interventions that affect one causal pathway more than another are differentially costly. Fully isolating and understanding the empirical consequences of legitimacy for compliance allows us to better make predictions and recommendations in situations in which the contours of policy involve choices along these lines. Acknowledgments We thank Eric Dickson for his contributions to other work in our collaborative research agenda, and gratefully acknowledge the helpful suggestions of Leonie Huddy, Macartan Humphreys, and Margaret Levi. Jennifer Wu provided superlative research assistance. This research is supported in part by NSF grants 1556047 and 1555494. Notes 1. See, for example, Tom R. Tyler, “The Psychology of Legitimacy: A Relational Perspective on Voluntary Deference to Authorities,” Personality and Social Psychology Review 1, no. 4 (November 1997): 323–345. 2. For an approach similarly grounded in the importance of obligation, see Margaret Levi, Audrey Sacks, and Tom R. Tyler, “Conceptualizing Legitimacy, Measuring Legitimating Beliefs,” American Behavioral Scientist 53, no. 3 (2009): 354–375. 3. For a helpful overview, see Morris Zelditch, “Theories of Legitimacy,” in The Psychology of Legitimacy: Emerging Perspectives on Ideology, Justice, and Intergroup Relations, ed. John T. Jost and Brenda Major (Cambridge: Cambridge University Press, 2001), 33–53. 4. Joseph Raz, The Authority of Law, 2nd ed. (Oxford: Oxford University Press, 2009). 5. Jürgen Habermas, Communication and the Evolution of Society (Boston: Beacon Press, 1979).

The Empirical Study of Legitimate Authority

357

6. Gertrude Elizabeth Margaret Anscombe, Ethics, Religion and Politics: Collected Philosophical Papers, Volume 3 (Oxford: Basil Blackwell, 1981). 7. David Beetham, The Legitimation of Power (New York: Palgrave MacMillan, 1991). 8. John Rawls, Political Liberalism (New York: Columbia University Press, 1993). 9. Thomas Nagel, “Moral Conflict and Political Legitimacy,” Philosophy and Public Affairs 16, no. 3 (1987): 215–240. 10. See, especially, Robert Paul Wolff, In Defense of Anarchism (New York: Harper & Row, 1970). 11. Ronald Dworkin, Law’s Empire (Cambridge, MA: Belknap Press, 1986). 12. Beetham, The Legitimation of Power. 13. Jean Hampton, Political Philosophy (Boulder, CO: Westview Press, 1997). 14. Specifically, a research assistant obtained all articles using the word legitimacy from JSTOR (for all included journals except CPS) or SAGE (for CPS) for this period. From this initial set, all articles where legitimacy was either an outcome or explanatory concept were retained. We read this initial set of sixty-four articles and discuss here those we deemed most relevant. 15. Max Weber, Economy and Society: An Outline of Interpretative Sociology (Berkeley: University of California Press, 1978). 16. Ibid. 17. Ibid. 18. David Easton, A Systems Analysis of Political Life (New York: Wiley, 1965). 19. Ibid. 20. Ibid. 21. It need not be a sufficient condition either, but that is not of particular concern at present. 22. Gregory A. Caldeira, “Neither the Purse Nor the Sword: Dynamics of Public Confidence in the Supreme Court,” American Political Science Review 80, no. 4 (December 1986): 1209–1226; Gregory A. Caldeira and James L. Gibson, “The Etiology of Public Support for the Supreme Court,” American Journal of Political Science 36, no. 3 (August 1992): 635–664. 23. James L. Gibson, Gregory A. Caldeira, and Lester Kenyatta Spence, “Measuring Attitudes toward the United States Supreme Court,” American Journal of Political Science 47, no. 2 (April 2003): 354–367. 24. Ibid. 25. For example, James L. Gibson and Gregory A. Caldeira, “Confirmation Politics and the Legitimacy of the U.S. Supreme Court: Institutional Loyalty, Positivity Bias, and the Alito Nomination,” American Journal

358

Sanford C. Gordon and Gregory A. Huber

of Political Science 53, no. 1 (January 2009): 139–155; James L. Gibson and Gregory A. Caldera, “The Legitimacy of Transnational Legal Institutions: Compliance, Support, and the European Court of Justice.” American Journal Science 39, no. 2 (May 1995), 459–489. 26. For a representative sample, see Brandon L. Bartels and Christopher D. Johnston, “On the Ideological Foundations of Supreme Court Legitimacy in the American Public,” American Journal of Political Science 57, no. 1 (January 2013): 184–199; Dino P. Christenson and David M. Glick, “Chief Justice Roberts’s Health Care Decision Disrobed: The Microfoundations of the Supreme Court’s Legitimacy,” American Journal of Political Science 59, no. 2 (April 2015): 403–418; James L. Gibson, “Challenges to the Impartiality of State Supreme Courts: Legitimacy Theory and ‘New-Style’ Judicial Campaigns,” American Political Science Review 102, no. 1 (February 2008): 59–75; James L. Gibson, “‘New-Style’ Judicial Campaigns and the Legitimacy of State High Courts,” Journal of Politics 71, no. 4 (October 2009): 1285–1304; James L. Gibson and Michael J. Nelson, “Is the U.S. Supreme Court’s Legitimacy Grounded in Performance Satisfaction and Ideology?” American Journal of Political Science 59, no. 1 (January 2015): 162–174; James L. Gibson, Gregory A. Caldeira, and Vanessa A. Baird, “On the Legitimacy of National High Courts,” American Political Science Review 92, no. 2 (June 1998): 343– 358; Stephen P. Nicholson and Thomas G. Hansford, “Partisans in Robes: Party Cues and Public Acceptance of Supreme Court Decisions,” American Journal of Political Science 58, no. 3 (July 2014): 620–636; James L. Gibson and Gregory A. Caldeira, “Campaign Support, Conflicts of Interest, and Judicial Impartiality: Can Recusals Rescue the Legitimacy of Courts?” Journal of Politics 74, no. 1 (January 2012): 18–34; James R. Zink, James F. Spriggs, and John T. Scott, “Courting the Public: The Influence of Decision Attributes on Individuals’ Views of Court Opinions,” Journal of Politics 71, no. 3 (July 2009): 909–925; James L. Gibson and Gregory A. Caldeira, “Defenders of Democracy? Legitimacy, Popular Acceptance, and the South African Constitutional Court,” Journal of Politics 65, no. 1 (February 2003): 1–30. 27. For example, Tom S. Clark, “The Separation of Powers, Court Curbing, and Judicial Legitimacy,” American Journal of Political Science 53, no. 4 (October 2009): 971–989; Gretchen Helmke, “The Origins of Institutional Crises in Latin America,” American Journal of Political Science 54, no. 3 (July 2010): 737–750. 28. Ted Robert Gurr and Muriel McClelland, Political Performance: A Twelve Nation Study (London: Sage Publications, 1971). 29. Christopher J. Anderson and Matthew M. Singer, “The Sensitive Left and the Impervious Right: Multilevel Models and the Politics of Inequality, Ideology, and Legitimacy in Europe,” Comparative Political Studies 41, no. 4–5 (April 2008): 564–599.

The Empirical Study of Legitimate Authority

359

30. Stephen White, “Economic Performance and Communist Legitimacy,” World Politics 38, no. 3 (April 1986): 462–482. 31. Keith Darden and Anna Grzymala-Busse, “The Great Divide: Literacy, Nationalism, and the Communist Collapse,” World Politics 59, no. 1 (October 2006): 83–115. 32. Devra C. Moehler and Staffan I. Lindberg, “Narrowing the Legitimacy Gap: Turnovers as a Cause of Democratic Consolidation,” Journal of Politics 71, no. 4 (October 2009): 1448–1466. The authors operationalize legitimacy using indexes of institutional trust, perceptions of the accountability of political leaders, deference to be governed, support for constitutions, and satisfaction with democratic performance. 33. Xiaobo Lü, “Social Policy and Regime Legitimacy: The Effects of Education Reform in China,” American Political Science Review 108, no. 2 (May 2014): 423–437; Susan H. Whiting, “Authoritarian ‘Rule of Law’ and Regime Legitimacy,” Comparative Political Studies 50, no. 14 (January 2017): 1907–1940. 34. Peter McDonough, Samuel H. Barnes, and Antonio López Pina, “The Growth of Democratic Legitimacy in Spain,” American Political Science Review 80, no. 3 (September 1986): 735–760; “The Nature of Political Support and Legitimacy in Spain,” Comparative Political Studies 27, no. 3 (October 1994): 349–380. 35. Mitchell A. Seligson, “The Impact of Corruption on Regime Legitimacy: A Comparative Study of Four Latin American Countries,” Journal of Politics 64, no. 2 (May 2002): 408–433. 36. Rosario Espinal, Jonathan Hartlyn, and Jana Morgan Kelly, “Performance Still Matters: Explaining Trust in Government in the Dominican Republic,” Comparative Political Studies 39, no. 2 (March 2006): 200–223. 37. James L. Gibson, “The Legacy of Apartheid: Racial Differences in the Legitimacy of Democratic Institutions and Processes in the New South Africa,” Comparative Political Studies 36, no. 7 (September 2003): 772–800. 38. Tom R. Tyler, Jonathan D. Casper, and Bonnie Fisher, “Maintaining Allegiance toward Political Authorities: The Role of Prior Attitudes and the Use of Fair Procedures,” American Journal of Political Science 33, no. 3 (August 1989): 629–652. 39. Edward N. Muller and Thomas O. Jukam, “On the Meaning of Political Support,” American Political Science Review 71, no. 4 (December 1977): 1561–1595. 40. Michael W. Giles and Douglas S. Gatlin, “Mass-Level Compliance with Public Policy: The Case of School Desegregation,” Journal of Politics 42, no. 3 (August 1980): 722–746. 41. Guy Grossman and Delia Baldassarri, “The Impact of Elections on Cooperation: Evidence from a Lab-in-the-Field Experiment in Uganda,” American Journal of Political Science 56, no. 4 (October 2012): 964–985.

360

Sanford C. Gordon and Gregory A. Huber

42. Benjamin A. Olken, “Direct Democracy and Local Public Goods: Evidence from a Field Experiment in Indonesia,” American Political Science Review 104, no. 2 (May 2010): 243–267. 43. Ian Hurd, “Legitimacy and Authority in International Politics,” International Organization 53, no. 2 (Spring 1999): 379–408. 44. Ruth W. Grant and Robert O. Keohane, “Accountability and Abuses of Power in World Politics,” American Political Science Review 99, no. 1 (February 2005): 29–43. 45. Michael N. Barnett, “Bringing in the New World Order: Liberalism, Legitimacy, and the United Nations,” World Politics 49, no. 4 (July 1997): 526–551. 46. Martha Finnemore, “Legitimacy, Hypocrisy, and the Social Structure of Unipolarity: Why Being a Unipole Isn’t All It’s Cracked Up to Be,” World Politics 61, no. 1 (January 2009): 58–85. 47. Tanja A. Börzel, Tobias Hofmann, Diana Panke, and Carina Sprungk, “Obstinate and Inefficient: Why Member States Do Not Comply with European Law,” Comparative Political Studies 43, no. 11 (July 2010): 1363–1390. 48. A related concern involves relevant counterfactuals: Specifically, what implicit comparison is a citizen making when announcing his or her support for an institution (as against, e.g., a political party)? More broadly, calling into question a political system or institution may be a “cheap talk” survey response, in which individuals are not communicating sincere beliefs or intentions. On the broad problem of “cheerleading” in survey response, see John G. Bullock, Alan S. Gerber, Seth J. Hill, and Gregory A. Huber, “Partisan Bias in Factual Beliefs about Politics,” Quarterly Journal of Political Science 10, no. 4 (2015): 519–578 [in the context of partisan bias]. Such concerns also arise if one attempts to use survey responses to “control for” different explanations for outcomes, particularly if individuals misreport those behaviors or intentions in ways that are correlated with the treatment or outcome variables of interest. 49. For example, see Theda Skocpol, States and Social Revolutions: A Comparative Analysis of France, Russia, and China (Cambridge, MA: Cambridge University Press, 1979); Alan Hyde, “The Concept of Legitimation in the Sociology of Law,” Wisconsin Law Review 1983 (1983): 379–426. 50. And, in a regression context, if two factors are correlated but only one causes the outcome of interest, regression will tend to give weight to both factors, a pattern that can be exacerbated by (differential) measurement error. This means that efforts to “control for” extrinsic rewards (material incentives) is a complete solution only under specific restrictive conditions. 51. Introducing uncertainty over λ would leave the intuition qualitatively unchanged.

The Empirical Study of Legitimate Authority

361

52. Note that this does not mean that enhanced state capacity reduces compliance overall: The dominant, first-order effect of an increase in capacity is to increase the probability of compliance. 53. Margaret Levi, Of Rule and Revenue (Berkeley: University of California Press, 1988). 54. This section builds on Eric S. Dickson, Sanford C. Gordon, and Gregory A. Huber, “Identifying Legitimacy: Experimental Evidence on Compliance with Authority,” (typescript, New York University, 2017). 55. Tom R. Tyler, “Procedural Justice, Legitimacy, and the Effective Rule of Law,” Crime and Justice 30 (2003): 283–357; Jason Sunshine and Tom R. Tyler, “The Role of Procedural Justice and Legitimacy in Shaping Public Support for Policing,” Law & Society Review 37, no. 3 (September 2003): 513–548. 56. Gerald S. Leventhal, “What Should Be Done with Equity Theory?” in Social Exchange: Advances in Theory and Research, ed. Kenneth J. Gergen, Martin S. Greenberg, and Richard H. Willis (Boston: Springer, 1980), 27– 55. 57. Dickson, Gordon, and Huber, “Identifying Legitimacy.” 58. For a related identification strategy—termed “selective trials”— involving probabilistic treatment assignment following a choice by the subject in an experiment, see Sylvain Chassang, Gerard Padró i Miquel. and Erik Snowberg, “Selective Trials: A Principal-Agent Approach to Randomized Controlled Experiments,” American Economic Review 102, no. 4 (June 2012): 1279–1309. 59. In environments with complementarities and heterogeneous legitimacy effects, the marginal effect of an increase in the legitimacy perceptions of a subset of citizens can be magnified, via the mechanism described above. For example, thinking through the logic of the game above, suppose both players did not contribute under policy p1. If the change in legitimacy associated with a move to policy p2 induced one player to contribute, then the second player would need a smaller change in legitimacy to contribute than in the absence of complementarities. 60. Muller and Jukam, “On the Meaning of Political Support.” 61. This, of course, assumes the availability of measures of compliance independent of the police themselves, a nontrivial issue in itself, but one beyond the scope of the current discussion. See Eric S. Dickson, Sanford C. Gordon, and Gregory A. Huber, “Profiling in the Lab: Experimental Evidence from a Novel Subject Population,” (typescript, New York University, 2017).

12 TRUSTWORTHY GOVERNMENT AND LEGITIMATING BELIEFS MARGARET LEVI

Government trustworthiness accounts for considerable variance in why and when citizens (and subjects more generally) comply with the extractive demands of governments.1 A trustworthy government is one that keeps its promises (or has exceptionally good reasons why it fails to), is relatively fair in its decision-making and enforcement processes, and delivers goods and services. A legitimate government is one that appeals to widely accepted justifications for its selection, maintenance, and policies. Investigations across history and countries reveal that the more trustworthy the government, the more likely it is to evoke observation of its laws and acquiescence to policies. Less clear is the link between perceptions that government is trustworthy and beliefs that it is legitimate, at least in countries claiming or trying to be democratic.2 Being trustworthy in practices and outcomes may contribute to perceptions of government legitimacy. However, trustworthiness is, at best, a necessary but not sufficient condition for legitimating beliefs. Some important contemporary circumstances reveal the distinction between trustworthy provision of goods and services, on the one hand, and the legitimacy of those in power, on the other hand. It is that distinction I want to explore.3 After using current instances in advanced democratic countries to illuminate what is at issue, I then turn to attempts to establish legitimate government where there are partial states, alien rule, post-conflict, or newly established states attempting to demonstrate governmental 362

Trustworthy Government and Legitimating Beliefs

363

capacity while also achieving some semblance of democracy.4 I do this by reexamining my own work on historical state-building and the growing literature on contemporary state-building, particularly research exploring the empirical link among government service delivery, perceptions of the trustworthiness of government, and attributions of legitimacy. Variation in the two major attributes of trustworthy government, provision of promised goods and services, and procedural justice in the determination and implementation of policy are not always sufficient to account for the variation in legitimating beliefs. What is also necessary is a set of accepted justifications for government and its actions, justifications based on widely shared moral principles and beliefs. This is hardly news, but we may, for the first time, have empirical means for figuring out how much such justifications matter and under what conditions—and how they are affected by actual government performance. Reassessing the Link Between Trustworthy Government and Its Legitimacy The United States of Trump and the Britain of Brexit exemplify national states that have penetrated deeply into their populations and, by objective standards, act within and are accountable to long-established democratic and bureaucratic procedures in the delivery of goods and services. Yet, questions are arising about the legitimacy of their governments. In both countries, the disruptions caused by technological change and the indifferent responses of policymakers mean that these governments are perceived as failing to deliver what many people need and believe they deserve. In both countries, albeit more extremely in the United States, there are those who reject almost any interventions of the state in their lives—be it via gun control or immunization vaccines. Then there are those whose civil rights have actually been denied through problematic policies and practices, such as American mass incarceration and questionable police actions so prevalent among communities of color, the poor, and immigrants. Claims that government is acting arbitrarily and discriminatorily, particularly when reinforced by actual and observable government actions, feed and intensify beliefs that government

364

Margaret Levi

is untrustworthy. They produce legitimacy concerns when combined with beliefs that the governments or their policies were not selected according to democratic justifications. Brexit was a legal vote, but there are issues about whether this was the appropriate way to reach such a complex and consequential decision. Trump is legally the president, but he came to office despite losing the popular vote and under the shadow of possible Russian meddling in the election process. Trump’s loss of the popular vote does not represent a challenge to legitimacy. It is written into the constitutionally established electoral college, the result of a compromise that allowed the United States to form, but has long raised concerns about its effect on equality of votes and majority determination of outcome. Russian meddling does raise a legitimacy challenge by violating widely shared values about democratic elections as free from illegal manipulation and from foreign interference. The president’s legitimacy is further undermined, at least in the eyes of some, by his breaking of long-established norms of presidential behavior, words, and probity. While the president is not the whole government, the tolerance of the US Congress for his conduct smears that institution as well. Despite such concerns, democracy continues to exist and so does (largely) trustworthy government by objective standards, but their existence is far from enough to legitimate government. Legitimacy rests on more than effective governance; it requires popularly acceptable justifications for who holds the reins of power, who the leadership is, and for the policies promoted. Doubts about government effectiveness and fairness in providing services and security undermine the popular belief in its trustworthiness and certainly weaken its claim for legitimacy. Loss of confidence in the very rules of the game—be it democracy, autocracy, or theocracy—for ensuring appropriate succession, accountability, or policies undermines the justificatory basis of legitimacy. It is arguable that we are seeing a fundamental, if often inchoate, challenge to the principles that made our long-standing democracies legitimate in the eyes of their populations. We are most certainly witnessing the fraying of the political economic framework that guided action for decades, and with that comes contestation of the values that undergird the framework. Periodically, it is necessary to update the prevailing moral economy of any

Trustworthy Government and Legitimating Beliefs

365

governmental system: the extra-market reciprocal rights and obligations that link populations, governments, corporations, and all the other various organizations that make up the society. Economies are a moral and political choice. A polity can and does make decisions about the nature of its moral economy, and those decisions—and the ethical premises that underlie them— vary across time and place. The Thatcher-Reagan brand of neoliberalism replaced the Keynesian approach that dominated the post–World War II developed West, but the Scandinavian countries resisted this movement, at least to some extent. Russia, China, and India made different decisions than the West in the 1950s; they continue to have different conceptions of who receives social protection and how it is provided. These are the principles that define who is in and who is out of the web of services and protections and guide their definition and delivery. Most of the world lives in polities in which the industrial and green revolutions—be it in this past century or the one before— obliterated the long-standing responsibilities of landlords toward the peasants who worked their lands. States now take on the burden of care, but according to different principles of who is deserving, who is included, and who pays, principles that were often long in coming and full of conflict. Once established, however, they became part of the justification for the legitimacy of government. Let us take the example of protection of workers. In the United States and much of Europe, free markets brutally drove workers’ lives from the mid-nineteenth through the mid-twentieth centuries, when most industrial countries established an interconnected framework of labor rights and citizen benefits to ensure that rising productivity was more equitably shared. This transformation involved changing laws but, more important, changing expectations about what capitalism ought to be, who it should serve, and what it could become. Not unrelatedly, these countries enjoyed the greatest growth in human history. As the moral economy of the post–World War II era, embodied in Keynesianism, was overtaken by the moral economy embodied in the neoliberalism of Friedrich von Hayek and Milton Friedman, the prevailing prescription that employers have obligations to their employees began to disintegrate. Further decimation of the New Deal labor framework is reinforced by the automation

366

Margaret Levi

and outsourcing of our manufacturing sectors and by gig-labor business models that evade the employer-employee relationship assumed in our laws. Productivity gains are no longer shared. Firms no longer feel responsible for ensuring employment security, health insurance, or other benefits. In the Keynesian era, despite the 1947 Taft-Hartley Act constraints on unions, union power in America increased. So did the role of unions in popular culture: The meme was Big Labor as an equal with Big Business and Big Government. Major newspapers of the time regularly covered union actions. By the 1970s, more than 20 million Americans were in unions, which remained a vital part of the Democratic Party’s coalition and were major advocates of social insurance and other citizen protections. Then things began to change—and union power began being rolled back, first gently and then with increasing ferocity. Several factors led to the devastation in the house of labor. Even in its heyday, the labor movement failed to represent that part of the labor force working in the agricultural, domestic, and low-end service sectors. With the neoliberal onslaught, led in the United States by President Ronald Reagan, the employers regained the upper hand, knowing that government would either look the other way or actively assist in the employer campaign to undermine unions organizing more workers and effectively enforcing collective bargaining rights. Moreover, enough time had gone by since the acme of labor organizing that young workers no longer credited unions with the improvements in their standard of living. The combined loss of power and the increased focus of many unions on particularistic gains made it difficult for unions, even the large confederations, to block the policies that amplify inequality in the United States. One result in recent decades has been real wage declines even as US productivity has grown.5 At the turn into the twenty-first century, survey evidence reveals that workers still wanted unions.6 A 2002 study, confirmed with additional data in 2005, found that more than 50 percent of nonunion, non-managerial workers preferred union representation. The percentages were particularly high among 18- to 34-year-olds (58 percent), those with incomes under $40,000 (59 percent), and minorities (74 percent). Fifteen years on, it would be worth doing such an analysis today, particularly among younger workers.

Trustworthy Government and Legitimating Beliefs

367

My own research on and experience with those employed in the new gig economy suggest that interest in having a voice no longer translates into interest in being represented by a union, an organization perceived as constraining individual prerogatives while taking a cut for doing so. These workers have bought into the prevailing moral political economy, with its overriding emphasis on freedom of choice and opposition to governmental intervention. In 2016, the total union membership in the combined public and private sectors was only 10.7 percent of all non-agricultural wage and salary workers, or 14.6 million people.7 Unions have always been weak in the South and in the non-coastal western states. For the industrial heartlands, the decline in union membership in recent years is notable. South Carolina has the lowest membership, at 1.6 percent. The highest Southern state is Alabama, at 8.1 percent, placing it in a dead heat with Wisconsin, which used to be a union stronghold. These aggregate figures hide the real story, however. Only 6.4 percent of private sector workers now belong to unions, down from the high of over 35 percent in 1954. By and large, what is keeping unions alive is government employment: 34.4 percent of public sector workers belong to unions. And yet the future of public sector unions is, arguably, tenuous.8 Note the reactions against them in once union-proud states such as Wisconsin. There is, in these states, growing antagonism among the public to the demands and strikes of government employees. At the same time, government austerity measures have further undermined the strength of the public sector. All evidence suggests that Trump and the courts will perpetuate the undermining of workers’ rights and wages. In 1947, the Taft-Hartley Act made it possible to create what became known as “right-to-work” laws, which enabled workers to opt out of dues-paying even when gaining the benefits of representation. Eleven states, most of them in the South, immediately adopted “right-to-work” regulations. Another five became “right to work” by 1955. Today, twenty-eight states and Guam have this status. As a result, union power nationally has been decimated. In the late 1990s and again in the Obama era, there was some guarded optimism that these trends could be reversed, that unions could regain some of their leverage and that a new generation of labor activists might even take leadership on questions of equality

368

Margaret Levi

and equity, both economic and racial.9 There is no such hope with Donald Trump, a man with a long history of hostility to unions and of engaging in unfair labor practices in his own businesses. His appeals to workers play to their basest interests and their current fears and add to the delegitimizing of unions. Without question, manufacturing—the basis of the grand industrial unions—has been disappearing from the United States for several decades. And whatever Trump’s rhetoric around protectionism and manufacturing, it is most unlikely that he will be able to re-industrialize the heartlands. People are hurting, and they no longer believe that their children will be better off than they were. They are looking for a scapegoat, and they demand a silver bullet to restore what they once had. Yet many hang on to the moral economy in the neoliberal variant of the political economy.10 Instead of advocating for programs to assist those in need and those striving to prepare themselves for the transformation of the economy, Trump pushes his protectionist trade policies, a hard line on immigration, and tax reform, none of which will ultimately benefit those left behind by recent economic trends. He does this by making exaggerated and often decidedly false claims that American jobs have gone overseas or to new immigrants; and he ignores the role of technological change. This strategy, while unlikely to actually generate manufacturing jobs in the United States, has the added benefit of demonizing unions—such as the National Domestic Workers Alliance and Service Employees International Union (SEIU)—that are growing in prominence and that disproportionately represent immigrant labor. Trump’s analysis may be largely inaccurate, but it finds resonance among white workers, whose jobs are precarious and whose standard of living is falling. And it supports the efforts to weaken and delegitimize the labor movement further. This extended example reveals two different kinds of legitimacy issues. The first has to do with the extent to which an organization—in this instance a union—is compatible with prevailing values about appropriate influences on individual action and intervention in the economy. Within the prevailing moral political economy, there is a reasonable case for undermining union power. There are many who disagree, however, raising questions about the appropriateness of the moral political economy itself. The

Trustworthy Government and Legitimating Beliefs

369

second issue is the legitimacy of a state that is upholding values that are currently being widely contested, due to questions about their moral rightness and about their consistency with prevailing ideas about what constitutes a state that is just, equitable, and democratic. Fashioning a new “moral political economy” will require shifting popular ideas about work, framing new labor laws, fashioning a safety net with new options, including possibly a universal basic income, and, most important, reestablishing the justifications for assessing a legitimate government. We cannot return to the postwar manufacturing boom nor to the moral political economy that governed that era. But we have choices regarding where we will go next. Similarly, we have decisions in other spheres as well: the mitigation of and adaptation to climate change; the determination of immigrant rights; the design and effects of bioengineering on our societies; how artificial intelligence evolves and with what impacts. With such shifts in moral reasoning also come new choices about how to fashion our lives and societies, posing challenges to our ethical systems, notions of equity, conceptions of who is deserving and who is not, and what forms of governance are best suited to a changed world. Britain and the United States alert us to the fragility of the relationship between trustworthy government and its legitimacy. However, these cases are so complex that it is difficult to use them to disentangle the links among quality of service delivery, processes by which they are determined and delivered, the selection and character of those who hold decision-making positions, perceptions of government trustworthiness, and assessments of legitimacy. To better understand what the necessary links are and how they are built and secured may require us to consider cases where government penetration is underway or incomplete. Several recent papers indicate that beliefs about the trustworthiness of government may reflect expectations—not just delivery—of government services; the higher the expectations, the more brittle the assessment of trustworthiness.11 But what role does trustworthiness play in promoting legitimacy, and is trustworthiness—even if secured—enough?

370

Margaret Levi A Quick Frolic through the Literature

The literature on legitimacy is vast but not totally satisfying. Max Weber’s writing suggests that the concept is best understood as an adjective characterizing the type of power: legitimate or illegitimate power, with a typology of legitimate power as traditional, charismatic, and rational-legal.12 What makes power legitimate is that it is “rightful” according to prevailing standards of morality; those who hold it came to their positions according to widely recognized and acceptable processes and then act according to given norms and values. David Beetham interprets Weber as arguing that it is the beliefs of people in a government’s legitimacy that make it legitimate.13 He goes on to list what, in his view, constitutes legitimate government and governmental practices: (i) it conforms to established rules; (ii) the rules can be justified by reference to beliefs shared by both dominant and subordinate; and (iii) there is evidence of consent by the subordinate to the particular power relation.14 The question is then how best to determine empirically when legitimacy exists and where and why it varies. We have made some progress, if incomplete, on this front. To give some prominent examples: Fritz Scharpf pioneered an approach to legitimacy that emphasizes “input” and “output” legitimacy.15 The inputs are the voices and decision processes that produce government policy; the outputs are the implementation and delivery of those policies, that is, how government performs. More recently, John Patty and Maggie Penn formalized a testable model of legitimacy that emphasizes the justificatory principles and the extent to which decision processes are consistent with and a means for realizing those principles.16 Combining these approaches with our takeaways from political and sociological theorists such as Weber, Beetham, and Habermas, we have most of the pieces we need to begin to link principles, their justification, the rules that embody them, the processes that realize them, and the consent of the governed.17

Trustworthy Government and Legitimating Beliefs

371

My Research on Building and Maintaining Trustworthy Governments My own work largely focuses on the conditions that promote popular perceptions of and belief in the trustworthiness of both policies and public officials and how those perceptions and beliefs affect the willingness of citizens to comply with costly demands and extractions. More precisely, I have attempted to: (1) find behavioral indicators of a population’s beliefs that governing arrangements are trustworthy; (2) understand the reasons for variation in perceptions of trustworthiness; and (3) learn how variations in perceptions affect variations in responses to government demands. I posit that individuals, if they assess government as trustworthy, will feel obligated to obey it. The markers of trustworthy government are its promise-keeping (which reflects both its commitments and its capacity), the fairness of its processes (given the norms of place and time), and evidence that the government will and can punish free-riders. When individuals are legally required to comply but do so out of a combination of confidence in the government and a sense of obligation, I conceive of their compliance as quasi-voluntary.18 When individuals are being asked by government to engage in voluntary activity—be it for military service and perhaps as voters—I conceive of their volunteering behavior as contingent consent.19 With these concepts, I was then able to derive hypotheses about when a government was objectively trustworthy (or not), whether it was believed to be trustworthy, and how and to what extent that should affect observable and measurable behavioral compliance. I was also able to account for the variation in the response of different parts of the population to the extractive demands of a government. In Consent, Dissent and Patriotism, I considered why young men chose to volunteer (or not) for military service in the twentieth-century wars in the United States, Australia, New Zealand, Canada, Britain, and France.20 They were being asked to pay a very high price to show their allegiance to the government. The decision was individual but informed by social networks and communities, producing a huge difference in how distinctive populations responded. For example, anglophone Canadians volunteered in high numbers and francophone Canadians resisted. They felt the federal Canadian government was untrustworthy,

372

Margaret Levi

failing to keep its promises of bilingual education and general respect for their language. Francophones were also worried—and reasonably so—that military orders would only be given in English, which not all of them spoke. Perhaps even more important, they questioned the legitimacy of the war and of the Canadian government’s insistence they serve in it; from their perspective the federal government was violating the constitutional justification that conscription could be considered only if Canada was invaded. Francophone Canadians during both world wars, working-class Australians in World War I, and dissidents throughout history have used avoidance of and outright refusal to serve in the military as a way to proclaim their opposition to particular policies. Additional indicators reveal oppositional stances: Shirking and property destruction are among the “weapons of the weak”; disobedience to the law, tax evasion, inoculation resistance, and even non-voting can represent active non-compliance.21 Of course, how to read the meaning of these actions depends on the motivations of the actors. Sometimes non-compliance is simply a reflection of venality or laziness or ignorance. But there are ways to study what the likely motivations are. In all of these instances, qualitative information often provides clues that can be followed with systematic analysis of hypothesized variations. I have long recognized the importance of the principles of fairness, due process, and proper selection of authorities in affecting legitimating beliefs. However, I have not always paid adequate attention to the content and effect of justifications—from both governing actors and their opposition—in influencing the perception of how well goods and services are being delivered and, perhaps more important, in prompting acceptance of governing authority as legitimate. More recent work attempts to fill that lacuna. Using Afrobarometer data, Audrey Sacks, Tom Tyler, and I attempt to test the link between aspects of an objectively trustworthy government as determined by measurable government activity, the creation of value-based legitimacy, and the consequent behavioral legitimacy (or quasi-voluntary compliance).22 We use survey questions about willingness to comply with tax requirements as an indicator of value-based legitimacy but lack measures of actual compliance with those requirements. Even so, we are able to find evidence that a “virtuous circle” (see figure 12.1) can exist in which

Trustworthy Government and Legitimating Beliefs

Administrative competence

373

Government performance

Government effectiveness

Procedural justice

Value-based legitimacy: Legitimating beliefs

Behavioral-based legitimacy: compliance

Figure 12.1. An illustration of the theoretical framework in Levi, Sacks, and Tyler, “Conceptualizing Legitimacy.” This models legitimacy as a sense of obligation or willingness to obey authorities (value-based legitimacy) that then translates into actual compliance with governmental regulations and laws (behavioral-based legitimacy).

a trustworthy government spawns legitimacy beliefs, which in turn leads to greater compliance, which feeds back into government whose increase in capacities enables it to provide more services and garner greater legitimacy. Others have since made advances beyond our initial effort; I discuss their work below. Suggestive findings about legitimacy also emerged in my coauthored book with John Ahlquist, In the Interests of Others.23 We investigate unions in the United States and Australia as a way to investigate how variations in democratic governance arrangements, on the one hand, and provision of promised goods and services, on the other hand, influence the development of both legitimating beliefs about the organization and behaviors that reflect those beliefs. The justification for unions is the delivery of higher wages, better benefits, improvements in health and safety, and relative job security, and no union leadership will retain legitimacy if it fails to deliver on these dimensions. Such performance

374

Margaret Levi

or output legitimacy facilitates compliance, but some unions wish to evoke actions from their members that serve an extended “community of fate,” that is, the group of people—some of whom may be distant and unknown—with whom one’s fate is linked and who one protects. For example, when dockworkers refuse to load cargo on ships because of some larger objective such as opposition to the Japanese invasion of Manchuria in the 1930s or to South African apartheid in the 1980s, they are acting in the interests of others but against their own material interests. Indeed, they are engaging in costly actions, risking loss of pay, loss of jobs, and even jail time. Union and leadership requests for such actions are more likely to produce value-based and behavioral legitimacy when there is greater input legitimacy, reflected in accountable union leaders, direct democracy (universal votes on officers and contracts), and inclusive governance arrangements that permit widespread voice and access. Our detailed dissection of a particular set of governance apparatuses clarifies the means by which the leadership and members/ citizens of an organization construct consent. Discussions and debates about the proper course of action ensured the creation of shared and consensual values about the principles underlying union activities. Information—and its contestation—meant that decisions of the collective were based on the best available evidence about various actions. Further, the deliberations legitimated the process itself; not everyone agreed with the choice, but virtually all went along with it once made. Establishing Legitimacy What we have learned from exploration of contemporary issues of democratic government, from the literature, and from my own work, is that there is indeed a link between trustworthy government and legitimacy, but that they are not conceptually the same. Legitimacy requires trustworthy government but also requires a set of shared values or principles that guide the actions of government actors and the responses of those they are governing. In practice, this means that input legitimacy is as important as output legitimacy. However, input processes are more than following procedures. First, they require adherence to higher principles,

Trustworthy Government and Legitimating Beliefs

375

justified by appeal to the constitutional history and ethics of the polity. Second, there should be good reason to believe that input processes will actually lead to outputs consistent with what is promised and expected as a result of the input processes. The first is a sine qua non of legitimacy and the second, an indicator of the trustworthiness of government that is also a necessary (but not sufficient) condition of legitimacy.24 My conclusions, based on the research to date, is that we would still use compliance with various government policies as an indicator of popular beliefs about trustworthiness and legitimacy but recognize three conditions for what appears to be the same behavioral choice: 1. If an individual believes government is untrustworthy and illegitimate, she will comply only if coerced. 2. If an individual believes government is trustworthy, she will comply if she believes government will ensure freeriders comply. 3. If an individual believes government is trustworthy and legitimate, she will comply without coercion and without assurance of others complying. The empirical difficulty of altering these conditions to determine likely citizen responses is overwhelming, although perhaps a clever researcher might eventually find a way. An only somewhat easier approach is to look for indicators of beliefs about trustworthiness and legitimacy. Even then, the extent to which the perceived trustworthiness of government influences or sustains legitimating beliefs remains a problematic issue. Let us start by figuring out when a populace believes government is trustworthy and why. Surveys provide one means of getting at trustworthiness of government institutions, leaders, and policies. However, even when available, few record attitudes over time from the same population, and most of the widely used questions are more likely to evoke from respondents their preferences for a person or party than a judgment about trustworthiness.25 The only questions that come close to eliciting useful information are those that ask about particular institutions and authorities and, even better, about assessments of various behaviors of those authorities.26

376

Margaret Levi

That being said, some recent researchers use surveys to good advantage in ascertaining the degree to which developing democratic states have achieved trustworthiness. A common mantra of aid agencies is: Provide decent goods and services to the population, and they will come to trust the government. Understanding when that presumption is true, if it ever is, has become a subject of interesting new research. The case of post-conflict Iraq, for example, reveals that the creation of a trustworthy government depends on more than objectively measured reasonable service delivery.27 The perception and reality of satisfactory service delivery do not always coincide. How people perceive what they are getting from government turns out to be complex everywhere, but particularly where the state is perceived as alien, where it has not fully penetrated society, and when there are good reasons to be wary of those bearing gifts. One particularly compelling recent finding is that the key intervening variable between objective service delivery and its perception as such is the expectation of the recipient.28 There is often greater trust of government in badly served rural populations and far less trust among those who have received relatively decent service. The level of expectations appears to be far more explanatory than the actual quality of services. Work in Tanzania also reveals variation in expectations around local services that have consequences for perceptions of trustworthiness.29 Why people hold such expectations is another still unanswered question deserving exploration, particularly for those keen to construct trustworthy and legitimate governments. Field experiments are another means for assessing the link between perceptions of trustworthy government and compliance. Interestingly, research based solely in surveys and those that use field experiments can have very distinct findings. Where government lacks capacity, often non-profits and foreign aid step in, creating questions about who deserves credit for provision. Using the Afrobarometer and her own survey in Zambia, Sacks found that aid from donors actually improved citizen-state relationships.30 This is contrary to what others, using field experiments instead of surveys, find in Uganda.31 In Bangladesh, Dietrich, Mahmud, and Winters, also using field experiments, find that information about foreign aid contributed to positive attitudes toward local government but had little effect on national government.32 Indeed, it appears that

Trustworthy Government and Legitimating Beliefs

377

one means for improving government trustworthiness in the presence of foreign aid and low state penetration is to decentralize services, as research on post-conflict Sierra Leone reveals.33 There is a long-standing literature on the benefits of decentralization and federalism in state-building to which these arguments contribute.34 Populations are more likely to trust local than national governments, in part because they can see how the sausage is made and who is getting what and why. Services, including security, often go hand-in-hand with extractive demands, such as taxation or military service. Citizens may want the first but not necessarily the second, particularly if they fear the processes of extraction will be unequally distributed and if they worry whether government is honest. Of equal import may be concerns about government sensitivity to their particular needs and circumstances. This suggests that unequal and palpably unfair distribution of services can actually undermine the virtuous circle by undermining perceptions of government trustworthiness.35 To understand the relationship between perceptions of how trustworthy government is and compliance with it also requires consideration of cases where government is not serving its citizens. Again, surveys and field experiments help us get at this, but so, too, do case studies of instances of alternative service provision. State incapacity, incompetence, disregard, or worse are unfortunately commonplace throughout much of the world, historically and today. Organized groups often arise directly to fill the gaps states neglect. The grandmothers of Yuendemu, an indigenous community in the central desert of Australia, are illustrative. They formed a night watch to keep the children from glue-sniffing, which government—despite some money and intention—was failing to do. Their actions did not seem to reflect a belief that the state was illegitimate, but certainly reflected lack of confidence in its ability to act on their behalf in this important area of concern for them. In other cases, e.g., organized groups, existing for a very different purpose, the population turns to alternatives to the state. This is the situation in some of the favelas of Brazil, slums of India, and indeed among populations throughout the world; the police are trusted less than the gangs.36 Again, is the state thought to be illegitimate or the gangs legitimate? That is far from clear. What is

378

Margaret Levi

evident is that government has failed in the important domain of trustworthiness, and others have picked up the slack. Even if perceptions of the trustworthiness of government can be improved, this does not necessarily lead to its achieving legitimacy in the eyes of the population. Legitimacy requires shared values being upheld and evidence that the processes of governance are justifiable according to commonly accepted principles. Legitimacy grounded only in provision of goods and services, including some modicum of security not otherwise available, is an extremely limited form of performance legitimacy. Arguably, it is not legitimacy at all. Although there may be some element of voluntary compliance flowing from the quid pro quo between the government and the populace it oversees, it is not likely to evoke compliance in the absence of assurances free-riders are being made to obey. Because of the conceptual muddiness in the relationship between the trustworthiness of the state and its legitimacy, there are few studies that parse this question using surveys or field experiments.37 Still, we can consider other actions that reveal questions about the legitimacy of governance institutions and officials. If we believe these are indicators, then we may be able to infer that any non-compliance we observe among those populations reflects their assessments of legitimacy. Observation and documentation of factors outside the individual can offer a sense of shared beliefs. In francophone and anglophone Canada, newspaper editorials, sermons, and other such public statements offered guidance to public sentiment. But there are many circumstances in which such guidance is not possible. As Kuran’s work makes clear, the perceived legitimacy of a regime is apparent only at the moment at which it is being actively overturned.38 Even so, public debate, voting, protest, and other actions provide a means for assessing which parts of the populace support government and which do not, even in autocratic situations.39 Another kind of example is the post–Civil War American South. The seemingly legitimate national state violated local values and was perceived as an alien ruler. The Ku Klux Klan represents organized vigilantism that uses violence to impose its values on its society. The Taliban represents another form of the same general phenomenon. It is a quasi-state that not only refuses to abide by the laws of the national government but also, like the Klan, imposes

Trustworthy Government and Legitimating Beliefs

379

its view of the world on the parts of society it controls through violence. Legitimacy, I suspect, can only develop with explicit discussions during the founding moments and reinforced over time through public debates in parliaments, the courts, the media, and public fora about the principles under which decisions are made, policies enacted, and fundamental conflicts arbitrated within the polity. There must be some agreement, albeit usually implicitly, about what the moral economy is and what that implies about how citizens (and non-citizens) are treated, what they can expect from the government. The citizens, governments, and corporations are in a community of fate based on shared values and reciprocal obligations. This way of framing the issue, of course, presents an additional and major empirical challenge: How can we tell what the consensual principles are, if indeed there are any at all? Elections are one indicator; widespread acceptance of policies another. But given evidence of popular apathy and even ignorance of what is at issue, these are partial indicators at best. In ascertaining the underlying consensual principles, there is good reason to consider founding moments or periods of crises (depressions, natural disasters, wars—internal and external). These are likely to be times when the population coalesces and clarifies the values held highly and most commonly and when it establishes rules, laws, and institutions that guide behavior and define the terms of reciprocity. Such rules inevitably erode over time, either because they are less applicable when peace or calm are restored or because special interests actively undermine or change them or because the world in which they were developed is no longer the world in which people live.40 Equally inevitably, it seems, popular divisions appear or reappear: Even when many of us share a core set of values, we may disagree on their order of importance, which can make all the difference not only in elections and policies but also in what each of us believes is legitimate.41 I have argued elsewhere that critical to eliciting common principles and beliefs in non-crisis situations is the existence of mediating associations.42 In the United States and Europe, these include such organizations as labor unions, political parties, and PTAs (Parent-Teacher Associations), groups that might create what

380

Margaret Levi

Putnam labeled “bridging social capital.”43 At the least they form a constraint on actions and words that are outside the ethical pale. At best, as in the case of the dockworker unions Ahlquist and I studied, they provide a basis for learning, discourse, and deliberation. The construction of a shared ethos and of guiding principles is the by-product of joining together for a common purpose and, through working together, developing mutual respect and a broader community of fate. We appear to be in a hiatus between the moral political economy created by neoliberalism and whatever moral political economy emerges. In the interim, the prevailing justifications for government action seem tenuous at best. The breakdown of neoliberalism as the basis of shared social values is evident in the developed democracies, but it is also part of the issue confronting developing states in this highly globalized world. The establishment of new and acceptable justifications requires developing a new moral economy, illuminating its principles, and then deriving appropriate policies consistent with those principles. Notes 1. The claims in this paragraph draw largely on research in which I’ve been involved, including: Margaret Levi, Of Rule and Revenue (Berkeley: University of California Press, 1988); Consent, Dissent and Patriotism (New York: Cambridge University Press, 1997); Karen S. Cook, Russell Hardin, and Margaret Levi, Cooperation without Trust? (New York: Russell Sage Foundation, 2005). 2. Monarchies, theocracies, and other forms of autocratic regimes may have claims to be legitimate even when they are far from trustworthy in the provision of goods and services, but democracies generally justify their legitimacy at least partially with demonstrations of their trustworthiness. 3. Cord Schmelzle and Eric Stollenwerk are engaging in a related enterprise, but their emphasis is on effective governance, not its trustworthiness per se. They have pulled together several conferences, including the Conference on Legitimacy and Governance in Areas of Limited Statehood: Theoretical and Empirical Perspectives, Freie Universität Berlin, June 10–11, 2017, and published a special issue of the Journal of Intervention and Statebuilding. The lead paper is their “Virtuous or Vicious Circle? Governance Effectiveness and Legitimacy in Areas of Lim-

Trustworthy Government and Legitimating Beliefs

381

ited Statehood,” Journal of Intervention and Statebuilding 12, no. 4 (2018): 449–467. 4. Michael Hechter, “Alien Rule and Its Discontents,” American Behavioral Scientist 53, no. 3 (October 2009): 289–310. 5. For a review of the literature and analysis of the role of unions, see John S. Ahlquist, “Labor Unions, Political Representation, and Economic Inequality,” Annual Review of Political Science 20, no. 1 (May 2017): 409–432. 6. Richard Freeman, “Do Workers Still Want Unions? More than Ever!” Briefing Paper No. 182, Economic Policy Institute, Washington, DC (February 2007); Jake Rosenfeld, What Unions No Longer Do (Cambridge, MA: Harvard University Press, 2014). 7. “Union Membership (Annual) News Release,” Bureau of Labor Statistics, January 19, 2018, https://www.bls.gov. 8. John S. Ahlquist, “Public Sector Unions Need the Private Sector (Or Why the Wisconsin Protests Were Not Labor’s Lazarus Moment),” Forum 10, no. 1 (2012), doi:10.1515/1540–8884.1499. 9. Bill Fletcher, Jr., and Fernando Gapasin, Solidarity Divided (Berkeley: University of California Press, 2008); Dorian T. Warren, “‘Labor in American Politics’: Continuities, Changes, and Challenges for the Twenty-FirstCentury Labor Movement,” Polity 42, no. 3 (July 2010): 286–292. 10. Katherine J. Cramer, The Politics of Resentment: Rural Consciousness in Wisconsin and the Rise of Scott Walker (Chicago: University of Chicago Press, 2016); Arlie Russell Hochschild, Strangers in Their Own Land: Anger and Mourning on the American Right (New York: New Press, 2016). 11. Derick W. Brinkerhoff, Anna Wetterberg, and Erik Wibbels, “Distance, Services, and Citizen Perceptions of the State in Rural Africa,” Governance 31, no. 1 (February 2018): 103–124; Claire Mcloughlin, “When Does Service Delivery Improve the Legitimacy of a Fragile or ConflictAffected State?,” Governance 28, no. 3 (July 2015): 341–356; Tanja A. Börzel and Thomas Risse, “Dysfunctional State Institutions, Trust, and Governance in Areas of Limited Statehood,” Regulation & Governance 10, no. 2 (October 2016): 149–160. Also see the set of papers in Journal of Intervention and Statebuilding 12, no. 4. 12. Max Weber, Economy and Society (Berkeley: University of California Press, 1978). 13. David Beetham, The Legitimation of Power, 2nd ed. (New York: Palgrave Macmillan, 2013). 14. Ibid., 37. 15. Fritz W. Scharpf, Games Real Actors Play: Actor-Centered Institutionalism in Policy Research (New York: Westview Press, 1997); “The Joint Decision Trap: Lessons from German Federalism and European Integration,” Public Administration 66, no. 3 (September 1988): 239–278.

382

Margaret Levi

16. John W. Patty and Elizabeth Maggie Penn, Social Choice and Legitimacy: The Possibilities of Impossibility (New York: Cambridge University Press, 2014). 17. Jürgen Habermas, Legitimation Crisis (London: Heinemann, 1976). 18. Levi, Of Rule and Revenue. 19. Levi, Consent, Dissent and Patriotism. 20. Ibid. 21. James C. Scott, Weapons of the Weak (New Haven, CT: Yale University Press, 1985). 22. Margaret Levi and Audrey Sacks, “Legitimating Beliefs: Sources and Indicators,” Regulation & Governance 3, no. 4 (December 2009): 311– 333; Margaret Levi, Audrey Sacks, and Tom R. Tyler, “Conceptualizing Legitimacy, Measuring Legitimating Beliefs,” American Behavioral Scientist 53, no. 3 (October 2009): 354–375. 23. John S. Ahlquist and Margaret Levi, In the Interests of Others: Organizations and Social Activism (Princeton, NJ: Princeton University Press, 2013). 24. Beliefs in the legitimacy of a government or a public official may also affect the perception of their trustworthiness. Eric Stollenwerk makes this point well; see Eric Stollenwork, “Securing Legitimacy? Perceptions of Security and ISAF’s Legitimacy in Northeast Afghanistan,” Journal of Intervention and Statebuilding 12, no. 4 (2018): 506–526. Cognitive dissonance and other psychological mechanisms can positively bias the construal of the actions of those the observer already approves. See, e.g., Herbert C. Kelman, “Reflections on Social and Psychological Processes of Legitimization and Delegitimization,” in The Psychology of Legitimacy: Emerging Perspectives on Ideology, Justice, and Intergroup Relations, ed. John T. Jost and Brenda Major (New York: Cambridge University Press, 2001), 54–73. For a compendium of biases in judgment, see Daniel Kahneman, Thinking, Fast and Slow (New York: Farrar, Straus, and Giroux, 2011). 25. Margaret Levi and Laura Stoker, “Political Trust and Trustworthiness,” Annual Review of Political Science 3, no. 1 (2000): 475–507; Karen S. Cook, Russell Hardin, and Margaret Levi, Cooperation without Trust? (New York: Russell Sage Foundation, 2005). There is, of course, debate about the usefulness of these indicators. For efforts to use these measures to good effect, see Bruce Gilley, The Right to Rule: How States Win and Lose Legitimacy (New York: Columbia University Press, 2009); Marc J. Hetherington, Why Trust Matters: Declining Political Trust and the Demise of American Liberalism (Princeton, NJ: Princeton University Press, 2004). 26. Soren Holmberg, Bo Rothstein, and Naghmeh Nasiritousi, “Quality of Government: What You Get,” Annual Review of Political Science 12, no. 1 (2009): 135–161; Bo Rothstein, “Social Trust and Honesty in Government:

Trustworthy Government and Legitimating Beliefs

383

A Causal Mechanisms Approach,” in Creating Social Trust in Post-Socialist Transition, ed. Janos Kornai, Susan Rose-Ackerman, and Bo Rothstein (New York: Palgrave Macmillan, 2004), 13–30; Peter Nannestad, “What Have We Learned About Generalized Trust, If Anything?,” Annual Review of Political Science 11, no. 1 (2008): 413–436. 27. Derick W. Brinkerhoff, Anna Wetterberg, and Stephen Dunn, “Service Delivery and Legitimacy in Fragile and Conflict-Affected States,” Public Management Review 14, no. 2 (March 2012): 273–293. Also see Mcloughlin, “When Does Service Delivery Improve.” 28. Brinkerhoff, Wetterberg, and Wibbels, “Distance, Services, and Citizen Perceptions.” 29. Odd-Helge Fjeldstad, Lucas Katera, and Erasto Ngelewa, “Disparities Exist in Citizens’ Perceptions of Service Delivery by Local Government Authority Authorities in Tanzania,” REPOA Brief No. 13, Research on Poverty Alleviation, Dar es Salaam (November 2008). 30. Audrey Sacks, “Can Donors and Non-State Actors Undermine Citizens’ Legitimating Beliefs?” Policy Research Working Paper No. 6158, World Bank, Washington, DC (August 2012). 31. Helen Milner, Daniel Nielson, and Michael Findley, “Which Devil in Development? A Randomized Study of Citizen Actions Supporting Foreign Aid in Uganda” (May 12, 2013), https://papers.ssrn.com. 32. Simone Dietrich, Minhaj Mahmud, and Matthew S. Winters, “Foreign Aid, Foreign Policy, and Domestic Government Legitimacy: Experimental Evidence from Bangladesh,” Journal of Politics 80, no. 1 (January 2018): 133–148; Simone Dietrich and Matthew S. Winters, “Foreign Aid and Government Legitimacy,” Journal of Experimental Political Science 2, no. 2 (Winter 2015): 164–171. 33. Audrey Sacks and Marco Larizza, “Why Quality Matters: A Multilevel Analysis of Decentralization, Local Government Performance and Legitimating Beliefs in Postconflict Sierra Leone,” Policy Research Working Paper No. 6021, World Bank, Washington, DC (April 2012). 34. Erik Wibbels, “Madison in Baghdad?: Decentralization and Federalism in Comparative Politics,” Annual Review of Political Science 9, no. 1 (2006): 165–188. 35. Levi, Consent, Dissent and Patriotism; Levi and Sacks, “Legitimating Beliefs”; Claire Mcloughlin, “When the Virtuous Circle Unravels: Unfair Service Provision and State De-Legitimation in Divided Societies,” Journal of Intervention and Statebuilding 12, no. 4 (2018): 527–544. 36. Tariq Thachil, Elite Parties, Poor Voters: How Social Services Win Votes in India (New York: Cambridge University Press, 2014); Beatriz Magaloni, Edgar Franco, and Vanessa Melo, “Killing in the Slums: An Impact Evaluation of Police Reform in Rio de Janeiro,” CDDRL Working Paper No. 556,

384

Margaret Levi

Center on Democracy, Development, and the Rule of Law, Stanford University (December 2015); Rachel Kleinfeld and Elena Barham, “Complicit States and the Governing Strategy of Privilege Violence,” Annual Review of Political Science 21, no.1 (2018): 215–238; Rachel Kleinfeld, A Savage Order: How the World’s Deadliest Countries Can Forge a Path to Security (New York: Random House, 2018). 37. One notable exception is Eric Stollenwerk’s study of Afghanistan in “Securing Legitimacy?.” 38. Timur Kuran, “Sparks and Prairie Fires: A Theory of Unanticipated Political Revolution,” Public Choice 61, no. 1 (April 1989): 41–74; “Now Out of Never: The Element of Surprise in the East European Revolution of 1989,” World Politics 44, no. 1 (October 1991): 7–48; Private Truths, Public Lies (Cambridge, MA: Harvard University Press, 1995). 39. Support of government is one commonly used indicator of legitimacy. See David Easton, A Framework for Political Analysis (Englewood Cliffs, NJ: Prentice-Hall, 1965). Federico Ferrara, “Why Regimes Create Disorder: Hobbes’s Dilemma During a Rangoon Summer,” Journal of Conflict Resolution. 47, no. 3 (June 2003): 302–325. 40. For a detailed account of the role of special interests in eroding the New Deal and World War II consensus, see Kim Phillips-Fein, Invisible Hands: The Making of the Conservative Movement from the New Deal to Reagan, 1st ed. (New York: Norton, 2009). 41. This is the argument made popular by Jonathan Haidt, The Righteous Mind: Why Good People Are Divided by Politics and Religion, 1st ed. (New York: Pantheon Books, 2012). 42. Margaret Levi, “The Devastated House of Labor,” Public Books, November 1, 2017, http://www.publicbooks.org. 43. Robert D. Putnam, Bowling Alone (New York: Simon & Schuster, 2000).

INDEX

Africa, 10 Afrobarometer, 372, 376 agency. See joint intentions; self-directed agency; shared will Ahlquist, John, 373, 380 Alabama, 367 alienation: and colonialism, 24–25; and quality assent, 78. See also coercion; political legitimacy; self-directed agency alignment: allegiance and political alignment, 80, 93; legitimacy and normative alignment, 264, 275–76, 306, 309–13; quality assent and political alignment, 80–81 allegiance: Burke on, 79; and military service, 371; and political alignment, 80, 93 American South, the. See South, the anarchism, 331–32; Kant on, 39; and state legitimacy, 62 annexation, 10, 14, 15, 26–27, 41, 47–48, 50–55. See also colonialism Anscombe, Elizabeth: on legitimate authority, 331 apartheid, 208–9, 337, 374 Aristotle, 105–12, 115, 118–19, 120, 121, 123, 128, 130, 135, 259; on civic life, 106–8; on the good life, 105–7, 112; on human nature, 105–112; qualifications for citizenship, 113–14; and republicanism, 104–5 assent, 70–71; versus consent, 76. See also quality assent Atack, Iain, 239 Australia, 371–73 authority, 180–82; and coordination, 181–82; de facto, 74, 296; of democracy, 21; justification of, 330–31; and

legitimacy, 68–69; and normative power, 180–85; and obligation to obey, 295–96, 333; of political subjects, 73–74; Raz on, 68, 330–31; and the right to err, 181–85; value of, 77, 81–83; Weber on sources of, 75 autonomy, 24, 28–30; and colonialism, 24–28; and correspondence, 31, 33–34; Kant on, 103; and selfdetermination, 24; and unilateral coercion, 24, 27–29; value of, 30. See also self-directed agency, political autonomy Baldassarri, Delia, 337 Bangladesh, 376 Barnett, Michael, 338 Baron, John, 284 Becker, Gary, 301 Beetham, David, 264, 333, 370; on legitimacy, 67, 71, 79, 288, 332; on legitimate authority, 286, 331; on Weber, 370 Bentham, Jeremy, 258, 282, 301 Blake, Michael, 36 Brazil, 377 Brexit, 363–64 Britain, 39, 363, 369, 371. See also England; United Kingdom Buchanan, Allen, 91–92, 101n62, 237– 38, 245 Burke, Edmund: on allegiance, 79. See also allegiance Calain, Philippe, 240 Caldeira, Gregory, 335, 336, 339 Canada, 371–72, 378 capitalism: and employers’ obligations, 365

385

386 CARE, 231, 243 causal inference, and legitimacy, 305, 340, 346–49 Chicago, 263, 277, 299 China, 337, 365 Christiano, Thomas, 187 Cicero, 259, 300, 303 civic life: Aristotle on, 106–8; and equality, 114–15; and ‘the good life,’ 110–11; and law-abidingness, 122–24 citizenship, 53–54, 116–17; Aristotle on qualifications for, 113–14; changes of citizenship without consent, 53–56, 62; and collective political autonomy, 59–60; equality and republican, 114– 15, 134–36; and franchise, 116. See also colonialism; functionalism Coakley, Mathew, 92 coercion: alien, 31, 33, 41; autonomy and unilateral, 24, 27–29; colonialism and unilateral, 25–28; cooperation harmed by, 22; and correspondence, 31; exceptions to wrongness of unilateral, 22–24, 37–40; and freedom-as-independence, 22–23; inequality in unilateral, 20–21; justice and unilateral, 20–22; not necessary to analyze legitimacy, 68–69; political autonomy and omnilateral, 14; and political legitimacy, 86–87, 206–7; self-directed agency and omnilateral, 33; and voluntarism, 13 coercive model, 257–58, 283–84; and compliance, 267–69; and cooperation, 269–70, 329; and deterrence, 258, 282–85, 301; and engagement, 271–74. See also consensual model Collingwood, Vivien, 249 colonialism: and alienation, 24–25; and autonomy, 24–28; and functionalism, 10–12, 53–54, 62–63; and inequality, 25–26; and political autonomy, 15, 48; and political autonomy view, 49– 52; and unilateral coercion, 25–28 common good: and contestation, 126– 28; and franchise, 118–21; and limits

Index to republican authority, 129–30; and political obligation, 127–28; and rule of law, 122–23. See also republicanism compensation: legitimate authority and compensatory demands, 174–75, 178–79; and right to err, 187 compliance: and the coercive model, 267–69; and the consensual model, 267–69; government trustworthiness and quasi-voluntary, 343, 371–73; indicating beliefs about government trustworthiness, 375; and legitimacy, 275–76, 299–300, 328–29, 343–44, 350–51, 353; and legitimate authority, 333; measurement of, 267 Congress, 364 consensual model, 257–60, 294, 298–99; and compliance, 267–69; and cooperation, 3, 267–70; and engagement, 271–74; and legitimacy, 259–60 consent: versus assent, 76; changes of citizenship without, 53–56, 62; functionalism versus consent-based views, 60–62; Hampton on, 332; and input legitimacy, 374; and legal authority, 257–58; and legitimacy, 42–43, 63n1, 332, 333; and the obligation to obey, 331–32; and political autonomy view, 42–43, 47–48; republicanism and hypothetical, 128, 133; and voluntary activity and contingent, 371 contestation: and the common good, 126–28; and political obligation, 127–28 cooperation, 269–70; and the coercive model, 269–70; 329; and the consensual model, 3, 267–70; from intrinsic motivations, 332; harmed by unilateral coercion, 22; with legal authorities, 269, 275, 299; and legitimacy on, 269–70, 290, 294, 298–99, 304, 312–23, 320; measurement of, 269 cooperators, 15, 20, 34; and joint intentions, 15, 20, 48; and political autonomy, 35–36

Index correspondence, 31, 33–34, 48–49; and autonomy, 33–34; and coercion, 31; and functionalism, 49–52; interest in, 31 cultural nations: and selfdetermination, 13. See also nationalist theory culture: identification with civic life not requiring shared, 43; similar conceptions of fairness within one, 286–87. See also political autonomy democracy, authority of, 21 Democratic Party, 366 deterrence, and the coercive model, 258, 282–85, 301 Dietrich, Simone, 376 diffuse support, 335, 338–39. See also Easton, David disagreement: justification under, 154, 158–61; and Objective Reason Conception, 148–49; and political legitimacy, 161–66; practical, 158–59; and Public Reason Conception, 148– 49, 153–54 discrimination: discretionary, 6; harmless, 218–20; political legitimacy and intentional, 209–10, 217–18; relevance of past, 134; religious, 201 Doctors Without Borders, 231, 240, 242–43 Doctrine of Double Effect, 213–14. See also intentions domination: Pettit on nondomination, 68, 104; Weber on legitimate, 79, 334 Dominican Republic, 337 Donini, Anthony, 232 Durkheim, Emile, 276 duty to enter the state, 16, 18; as a requirement of justice, 16, 18–19 Dworkin, Ronald, on legitimacy, 332 Eastern Europe, 336 Easton, David: on legitimacy, 334–35; on ‘diffuse support,’ 335

387 Edwards, Michael, 240 employers’ obligations: and capitalism, 365; under Keynesianism, 365; after New Deal, 366. See also unions; workers’ rights engagement, 271–4 England, 299. See also Britain; United Kingdom Enoch, David, 151. See also Objective Reason Conception equal concern, and official intentions, 208–9 equality: authority of democracy and, 21; in civic life, 114–15; colonialism and inequality, 25–26; and franchise, 115–16; and republican citizenship, 114–15, 134–36; inequality in unilateral coercion, 20–21; unions and inequality, 366. See also equal concern Establishment Clause, 201, 202, 206, 210 Europe, 262, 263, 268, 270, 272, 274, 275, 277, 278, 279, 280, 289, 299, 365, 379 European Court of Justice, 336 European Social Survey, 261, 263, 264, 299 European Union (EU), 89, 338 Evaluating Transnational NGOs: Legitimacy, Accountability, Representation, 241 Fagan, Jeffrey, 299, 302 fairness: legitimacy and perceptions of, 261; obligation to obey and perceptions of, 333; procedural, 8, 345; and right to err, 195 Fallon, Richard, 214–15, 216 Federalist Paper No. 78, 335 Ferraro, Fabrizio, 284 Finnemore, Martha, 338 France, 25, 26, 371 franchise, 115–21; and citizenship, 116; and common good, 118–21; and equality, 115–16; and republicanism, 133–36

388 freedom-as-independence, 15–16, 41; and natural duty of justice, 15–16, 20, 41; and personal self-determination, 16; and unilateral coercion, 22–23. See also justice, rights Freud, Sigmund, 276 Friedman, Milton, 365 Friedrich, Carl, 296 functionalism, 12–13: versus consentbased views, 60–62; and colonialism, 10–12, 53–54, 62–63; and correspondence, 49–52; and legitimacy, 12–13; and political autonomy view, 49–52, 55; and political legitimacy, 53–56, 60–62 Gatlin, Douglas S., 337 Gaus, Gerald, 154 Geller, Amanda, 299, 302 Germany, 11; Nazi Germany, 123; West Germany, 351 Gibson, James, 335, 336, 339 Giles, Michael W., 337 good life, the: in Aristotle, 105–7, 112; civic life and, 110–11. See also republicanism government legitimacy: and conflict with local values, 378–79; and government trustworthiness, 362, 364, 369 government trustworthiness, 362; beliefs in, 371, 372–73; compliance indicating beliefs about, 375; and expectation of government services, 369, 375–76; and foreign aid, 376– 77; and government legitimacy, 362, 364, 369; markers of, 371; and quasi-voluntary compliance, 343, 371–73; perceptions of, 370; and state alternatives, 377–78; surveys and field experiments limited at indicating, 375–76, 378; and unequal distribution of services, 377 Grant, Ruth, 338 Grossman, Guy, 337 Gurr, Ted Robert, 336

Index Habermas, Jürgen, 67, 370; on legitimacy, 331 Hamilton, Alexander, 335 Hampton, Jean, on consent, 332 Hart, H. L. A., 67, 108, 300, 301 Hobbes, Thomas, 88, 93, 114; and the obligation to obey, 331; on political legitimacy, 75 humanitarian intervention: and selfdetermination, 11. See also annexation; colonialism humanitarian organizations. See INGOs Hurrell, Andrew, 237 India, 39, 365, 377 Indonesia, 337 inequality. See equality intentions: collective intentionality, 204; distinguished from motivations, 205; and the Doctrine of Double Effect, 213–14; equal concern and official, 208–9; evidence for, 204–5; existence of, 204; and harmless discrimination, 218–19; of officials, 203–5; intention-sensitive theories of political legitimacy, 203, 205, 206–11, 216–20; irrelevance of intentions for political legitimacy, 211–16; in Palmer v Thompson, 220–21; and pure proceduralism, 207–8; Rawls and official, 209–10; in reflective equilibrium, 212–13, 220; role of intentions in proceduralism with substantive constraints, 208–9 International Non-Governmental Organizations (INGOs), 231; charitybased approaches to, 244–45; political legitimacy of, 232–51; resistance against, 246 Jackson, Jon, 264, 299, 302, 310, 315 Jackson, MS, 220–21 Jim Crow, 208. See also South, the joint intentions: and cooperators, 15, 20, 48; omnilateral coercion based on, 33. See also shared will

Index Jukam, Thomas O., 337 justice: and duty to enter the state, 16, 18–19; freedom-as-independence and natural duty of, 15–16, 20, 41; Kant on, 15–16; legitimacy and procedural, 279–80, 301–304, 315–18; obligation to obey and procedural, 301; and quality assent, 83–84; rights and the duty of justice, 16; rule of law and procedural, 260–61; and unilateral coercion, 20–22 justification: of authority, 330–31; under disagreement, 154, 158–61; and objective reasons, 151–54, 160, 165–66; practical reasoning versus practical, 154, 157–58, 160–61; Rawls on, 331; republican justification of sovereignty, 104, 124–25; and subjective reasons, 153–54, 161–62 Kahan, Dan, 285 Kant, Immanuel, 18, 19, 20, 39, 103, 105, 130, 131; on anarchism, 39; on autonomy, 103; on duty to enter the state, 9, 14, 15, 16, 18, 19; on freedom-as-independence, 15–16; on justice, 15–16; on liberalism, 130–31; on omnilateralism, 14, 19–20; on selfdetermination, 13–14; on unilateral coercion, 16–18 Keohane, Robert, 237, 238, 245, 338 Keynesianism, 365; unions in the Keynesian era, 366 Kolodny, Niko, 31 Ku Klux Klan, 378 Kuran, Timur: on perceived legitimacy, 378 law-abidingness: and civic life, 122–24; and republicanism, 122–24. See also compliance; obligation to obey; political obligation legitimacy, 12–15, 73, 75–76, 84–91, 234, 239, 241, 305, 333, 364, 379; and authority, 68–69; Beetham on, 67, 71, 79, 288, 332; behavioral, 373–34;

389 behavioral implications of, 338– 39; and causal inference, 305, 340, 346–49; challenges to democratic, 364–65; and compliance, 275–76, 299–300, 328–29, 343–44, 350–51, 353; and consensual model, 259–60; and consent, 42–43, 63n1, 332, 353; consent and input, 374; and cooperation, 269–70, 290, 294, 298–99, 304, 312–13, 320; of courts, 335–36; distinguished from trust, 339–40; Dworkin on, 332; Easton on, 334–35; empirical, 177; as an empirical phenomenon, 329; and empirical research, 259, 264, 287–89; 305, 334, 343–44, 354; as an equilibrium, 333; and functionalism, 12–13; Habermas on, 331; input and output, 370, 373–75; institutional, 345–47; of international organizations, 338; justifications for, 373; Kuran on perceived, 378; legal, 297; and loss of confidence, 364; and mediating associations, 379–80; and mistakes, 175–85; moral, 295–96; and normative alignment, 264, 275–76, 306, 309–13; and obligation to obey, 264, 275–76, 306–9; and perceptions of fairness, 261; personal, 347–49; and political obligation, 91–92; and procedural justice, 279–80, 301–304, 315–18; and quality assent, 73, 75, 83–86; Rawls on, 209; of regimes, 336–38; as a right to err, 174–75, 180–85; and right to rule, 12; rule of law and perceived, 261; requiring shared values, 259, 282–83, 378–80; and self-determination, 13; sociological, 296–97; and support for institutions, 335–36, 339, 344–45; and trust, 264, 275–76, 306, 309–12; of unions, 373–74; Walzer on, 35; Weber on sociological, 296–97, 370. See also legitimate authority; political legitimacy legitimate authority, 181–82, 186–88; Anscombe on, 331; Beetham on,

390 legitimate authority (cont.) 286, 331; behavioral and compensatory demands, 174–75, 178–79, 187; and compliance, 333; conditions for, 260; consequences to, 332; and normative power, 177, 180–85; and the right to err, 177, 180–85; and the right to rule, 185; Weber on, 334. See also legitimacy; political legitimacy legitimate power, Weber on, 328, 370 Leventhal, Gerald, 345 Lewin, Kurt, 259 Liao, Matthew, 216 liberal theories: Kantian liberalism, 130–31; and political obligation, 103, 108; Rawlsian liberalism, 131–32; and sovereignty, 102–3, 105 Libya, 25–27 Locke, John, 90, 108, 114 Loughran, Thomas, 282 MacCoun, Robert, 282 Macdonald, Terry, 237 Machiavelli, Niccolo, 104, 121, 300, 301, 316, 318 Mahmud, Minhaj, 376 Manchuria, Japanese invasion of, 374 May, Simon, 208 McClelland, Muriel, 336 Médecins Sans Frontières (MSF). See Doctors Without Borders Mentovich, Avital, 302 military service: allegiance and, 371; refusal to take part in, 372 Montesquieu, Baron de, 259 moral political economy, 365; in the neoliberal variant, 368; new, 368, 380; and support for unions, 366–67 motivations: cooperation from intrinsic, 332; distinguished from intentions, 205; intrinsic versus extrinsic, 329; obligation to obey and intrinsic, 306, 332, 346–47. See also compliance, intentions

Index Nagel, Thomas, 331 National Domestic Workers Alliance, 368 nationalist theory, 13. See also cultural nations neoliberalism, 365, 366, 368 New Deal, 365–66 New York, 263, 299, 302 normative power, 177; and authority, 180–85; and the right to err, 177, 180–85; and the right to rule, 185 North Korea, 202 Nozick, Robert, 108, 109, 121 Obama, Barack, 367 Ober, Josiah, 119–20 Objective Reason Conception, 148–49, 150–51; and disagreement, 148–49; and political legitimacy, 150–51. See also objective reasons objective reasons, 150–51; and justification, 151–54, 160, 165–66; and political legitimacy, 151; and practical reasoning, 157. See also Objective Reason Conception obligation to obey: and authority, 295–96, 333; and consent, 331– 32; Hobbes on, 331; and intrinsic motivation, 306, 332, 346–47; and legitimacy, 264, 275–76, 306–9; and perceptions of fairness, 333; and political legitimacy, 245–47; and procedural justice, 301 occupation, military, 10, 11, 24. See also colonialism Olken, Benjamin A., 337–38 omnilateralism, 15, 19; omnilateral coercion, 33; omnilateral will, 14; and shared will, 33–34. See also coercion; unilateralism Orbinski, James, 242 Oxfam, 231, 243 Pakistan, 243 Palmer v Thompson, 220–23 Parent-Teacher Associations: as mediating associations, 379–80

Index Parr, Tom, 218–19 parties, political: as mediating associations, 379–80 Paternoster, Raymond, 282 Patty, John, 370 Penn, Maggie, 370 Pettit, Phillip, 114, 128; on nondomination, 68, 104 Pfeffer, Jeffrey, 284 Philippines, the, 10 Piquero, Alex, 282 Plato, 120, 259 Pogarsky, Greg, 282 political alignment, 80–81 political autonomy, 13–14, 28–36; citizenship and collective, 59–60; claims to, 28, 36, 39; and colonialism, 15, 48; colonialism and political autonomy view, 49–52; consent and political autonomy view, 42–43, 47– 48; and cooperators, 35–36; functionalism and political autonomy view, 49–52, 55; interest in, 14, 28; and omnilateral coercion, 14; possible when cooperators have joint intentions, 15, 20, 48; and rights, 37–38; and self-determination, 13–15 political legitimacy, 66–69, 73, 75, 206, 235, 239; accountability as a criterion of, 248; in cases of mistake, 37, 207; and coercion, 86–87, 206–7; criteria for, 232, 235, 239, 249–51; and disagreement, 161–66; and functionalism, 53–56, 60–62; of INGOs, 232–51; Hobbes on, 75; intention-sensitive theories of, 203, 205, 206–11, 216– 20; and intentional discrimination, 209–10, 217–18; irrelevance of intentions for, 211–16; and moral taint, 220–23; normative versus empirical, 67–68, 239–41, 243, 245; and Objective Reason Conception, 150–51; and objective reasons, 151; and obligation to obey, 245–47; and organizational autonomy, 242; and republicanism, 114; and subjective reasons,

391 148–49; and the right to rule, 68–69, 206–7, 232, 235–36, 239–40, 242 political obligation: and the common good, 127–28; and contestation, 127– 28; and legitimacy, 91–92; and liberal theories, 103, 108; and republicanism, 104–5, 124–25 political order, 69; and quality assent, 72–73 practical reasoning, 152, 155–57; and justification, 154, 157–58, 160–61; and objective reasons, 157; and subjective reasons, 157 proceduralism: intentions and pure, 207–8; role of intentions in proceduralism with substantive constraints, 208–9 Public Reason Conception, 148, 162; and disagreement, 148–49, 153–54. See also public reasons; subjective reasons Putnam, Robert, 380 quality assent, 70–71; and basic security, 72–73; and justice, 83–84; and legitimacy, 73, 75, 83–86; and political alignment, 80–81; and political order, 72–73; and stability, 78–80; value of, 76–78, 81–83. See also assent Rawls, John, 67, 108, 164, 331; on justification, 331; on legitimacy, 209; and official intentions, 209–10; Rawlsian liberalism, 131–32 Raz, Joseph, on authority, 68, 330–31 Reagan, Ronald, 365, 366 reasons. See objective reasons; public reasons; subjective reasons reflective equilibrium, intentions in, 213–14, 220 republicanism: in Aristotle, 104–5; common good and limits to republican authority, 129–30; equality and republican citizenship, 116, 134–38; and franchise, 133–36; and hypothetical consent, 128, 133;

392 republicanism (cont.) justification of sovereignty, 104; and law-abidingness, 122–24; and political legitimacy, 114; and political obligation, 104–5, 124–25 rights: acquired, 16, 18; and autonomy, 24; and the duty of justice, 16; moral, 17; and political autonomy, 37–38; private, 17–18; property, 9, 17 right to err, 185–88; and authority, 181–85; as a claim right, 185–86; and compensation, 187; and fairness, 195; and legitimacy, 174–75, 180–85; and legitimate authority, 177, 180– 85; as a liberty right, 186; and normative power, 177, 180–85 right to rule: and legitimate authority, 185; and normative power, 185; and political legitimacy, 68–69, 206–7, 232, 235–36, 239–40 right to vote. See franchise right-to-work, 367; and union power, 367 Ripstein, Arthur, 103 Rousseau, Jean-Jacques, 107; on political autonomy, 29–30 rule of law: and the common good, 122–23; and perceived legitimacy, 261; and procedural justice, 260–61 Russia, 364–65 Sacks, Audrey, 372, 376 Satyvada, Sagarika, 302 Scanlon, Thomas, 214 Scharpf, Fritz, 370 Schauer, Frederick, 258, 301 Scheffler, Sam, 110 self-determination, 11–12, 13–14, 42; and autonomy, 24; and cultural nations, 13; freedom-asindependence and personal, 16; and legitimacy, 13; and nationalist theory, 13; and political autonomy, 13–15 self-directed agency, 28–29; and autonomy, 28; and omnilateral coercion and, 33

Index self-rule, 11. See also self-determination Service Employees International Union, 368 shared will, 32–33; and correspondence, 33–34; and omnilateralism, 33–34. See also joint intentions Sierra Leone, 377 Skinner, Quentin, 104 Slavny, Adam, 218–19 Slim, Hugo, 241 Socrates, 293 South, the: American South, 124; antebellum South, 123; post-Civil War American South, 378; unions in the South, 367. See also Jim Crow South Africa, 337, 339, 374 South Carolina, 367 sovereignty: and liberal theories, 102–3, 105; political obligation and republican, 104–5, 124–25; republican justification of, 104 Soviet Union, 336 Spain, 337 Spence, Lester Kenyatta, 335 stability, 298; and quality assent, 78–80 state legitimacy: and anarchism, 62; moral, 295–96; and moral political economy, 368–69; and neoliberalism, 368–69 state of nature: acquired rights unenforceable in, 18; justice unenforceable in, 17–18 Steffek, Jens, 241 subjective reasons, 152–53; and justification, 153–54, 161–62; and the Objective Reason Conception, 152– 54; and political legitimacy, 148–49; and practical reasoning, 157; and the Public Reason Conception, 153–54 Sunshine, Jason, 299, 302 Supreme Court, 202, 220–21, 224, 335, 336, 350 Sutton, Robert, 284 Taft-Hartley Act, 366, 367 Taliban, 378–79

Index Tanzania, 376 Thatcher, Margaret, 365 travel ban, 201–204, 206, 209, 211, 214, 222 Trump, Donald, 201–203, 204, 206, 214, 215, 222, 363–64, 367, 368 Trump v Hawaii, 224 trust: and legitimacy, 264, 275–76, 306; distinguished from legitimacy, 339–40. See also government trustworthiness Uganda, 376 Unilateralism, 14, 20–25; autonomy and unilateral coercion, 24, 27–29; colonialism and unilateral, 25–28; exceptions to wrongness of unilateral coercion, 22–24, 37–40; freedom-asindependence and unilateral coercion, 22–23; justice and unilateral coercion, 20–22. See also coercion; omnilateralism unions, 366; and behavioral legitimacy, 373–74; decrease in the power of, 366, 368; and inequality, 366; in the Keynesian era, 366; legitimacy of, 373–74; as mediating associations, 379; membership, 367; and moral political economy, 366–67, 368; in the Obama era, 367; public sector, 367; and right-to-work, 367; in the South, 367; support for, 366–67; in the Trump era, 368

393 United Kingdom, 26. See also Britain; England United States, 11, 27, 132, 136, 201, 202, 223, 263, 267, 268, 270, 272, 273, 275, 277, 278, 279, 280, 289, 294, 299, 303, 363–64, 365, 366, 368, 369, 371, 373, 379 Venezuela, 202 voluntarism, 13, 42–43 voluntary activity: and contingent consent, 371; and military service, 371 von Hayek, Friedrich, 365 Waldron, Jeremy, 117, 186, 187 Wales, 299 Walzer, Michael, 52; on legitimacy, 35 Weber, Max, 67, 259, 286, 293, 298, 300; Beetham on, 370; on legitimate authority, 334; on legitimate domination, 79, 334; on legitimate power, 328, 370; on sociological legitimacy, 296–97; on sources of authority, 75 Williams, Bernard, 86–87 Williamson, Timothy, 155 Winters, Matthew, 376 Wisconsin, 367 workers’ rights: and capitalism, 365; under Keynesianism, 365; after New Deal, 366. See also employers’ obligations; unions Zambia, 376